Re: Unable to upload screenshots
Yet more info, please excuse the top post. We hit this again (reviewboard 1.6.11 silently does not create thumbnails now, rather than raising an error). I think I have a better handle on why thesimple test works and yet Django/Reviewboard fails. Image.open() doesn't do anything interesting. It isn't until you try and read the file that anything is done. The real test is: try: import Image except ImportError: from PIL import Image x = Image.open(/home/ingres/patchdoc_new.png) x.tostring()[0] The to string fails for me with: IOError: decoder zip not available Due to some weird inability to find zlib. There are some notes online about Ubuntu (and suggesting soft linking) but they didn't work for me. I ended up trying Pillow (they've done some cleanup for easy install). My new fix is now: sudo rm -rf /usr/local/lib/python2.7/dist-packages/PIL-1.1.7-py2.7-linux-x86_64.egg/ # or what ever version is being used sudo easy_install Pillow In the course of this research I ended up installing a few dev libs but using PIL with them did NOT help, I'm not sure if they helped with Pillow or not :-) sudo apt-get install libpng-dev sudo apt-get install libfreetype6-dev Using the base OS provided library is usually the better option though (as per my last update). On Monday, October 24, 2011 2:29:59 PM UTC-7, Chris Clark wrote: Thilo-Alexander Ginkel wrote: On Mon, Feb 14, 2011 at 23:01, Christian Hammond wrote: I don't believe they're stored temporarily anywhere. If so, it's internal to Django and I don't know it off-hand. The only thing I've ever seen permission-wise is when there's a directory within media/uploaded that doesn't have the proper write permissions for the server. Hm... Actually, after doing some more testing I think that permissions aren't an issue as GIF files can be uploaded properly. I also can't imaging why libpng should be loadable from the command line, but not from the Apache process... Arise zombie thread, live again! ;-) OK, so it turns out after years of happily running with RB we've hit this too. I KNOW this used to work (as my image testing when we first deployed was with PNG files). Our problem appears to be identical to this, regular RB install (i.e. easy_install based) which always seems to pull down PIL even though it is already installed (we're running Ubuntu server and I always ensure the debian PIL package is installed that has full image support). We could only attach new GIF screen shots to reviews. Using PIL in a Python terminal always worked (i.e. open the supposedly bad image via Image.open), but RB/Django would always fail with: -- Cut here -- Upload Screenshot One or more fields had errors Caption: Path: * Upload a valid image. The file you uploaded was either not an image or a corrupted image. -- Cut here -- On a hunch I thought I'd have a look at what PIL instances are on the file system (e.g. all the eggs that easy_install pulled down), after upgrading a few times over the years we had 2. reviews:~$ find / -name \*PIL\* 2/dev/null /usr/share/pyshared/PIL.pth /usr/share/pyshared/PIL /usr/share/pyshared/PIL/PIL-1.1.6.egg-info /usr/lib/python2.5/site-packages/PIL.pth /usr/lib/python2.5/site-packages/PIL-1.1.6-py2.5-linux-i686.egg /usr/lib/python2.5/site-packages/PIL-1.1.6-py2.5-linux-i686.egg/PIL /usr/lib/python2.5/site-packages/PIL /usr/lib/python2.5/site-packages/PIL/PIL-1.1.6.egg-info /usr/lib/python2.4/site-packages/PIL ... So I removed all the PIL eggs (remember I already have PIL installed): sudo rm -rf /usr/lib/python2.5/site-packages/PIL-1.1.6-py2.5-linux-i686.egg /usr/lib/python2.5/site-packages/PIL/PIL-1.1.6.egg-info Followed by a reviewboard restart: sudo /etc/init.d/apache2 restart And we can now attach PNGs and not just GIFs. So what out for easy_install and PIL. I know pip purports to support removal (and would probably be a better solution) so this is not a recommendation of resolution! I'm just sharing what I found on my system, don't go deleting files if you don't know the impact/dependencies! Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: reviewboard 1.6.11 and reviewing problems with FireFox 3.6 and IE 9
The FF user who had problems has Firebug installed ready for the next time, nothing has happened so far. In a similar vein, I've Windows XP user with IE8 (the latest IE for that platform) that had a single occurrence of writing a massive comment on a review (along with diff/code comments) and the review being lost. He (luckily copy/pasted the comment text into an email) hit publish, went away for an hour for lunch and came back it was still publishing and so he ended up closing the session down and lost his review (he did NOT any point use the save option). I've just got Chrome installed on their machine but I wanted to share this experience in case anyone else experiences this. The other weird behavior (and I've seen this in FF too) is occasionally hitting Publish will show the do you want to navigate away from this page dialog shows up. I'm a pretty compulsive ctrl-s saver so I know this is benign (and I'm hitting publish too) but again sharing in case anyone else is seeing this. We're planning an upgrade to 1.6.12 soon, so maybe these will go away? Chris On Thursday, September 27, 2012 2:38:14 PM UTC-7, Chris Clark wrote: On Thursday, September 27, 2012 2:12:08 PM UTC-7, Christian Hammond wrote: The IE thing is weird, because I don't think anything should have changed as far as text entry goes in a while... Which IE? The IE behavior may not be new, for the version see the title (9) but recall it was seen with 8 too. I can get access to someones machine where this seems to occur on a regular basis. For Firefox, is there anything in the JavaScript console? Installing Firebug (you may need an older one for FF3.6) would give you a lot more useful information. That might be tricky, I'm personally not seeing the problem but I'll mail out to the upgrade hold outs :-) and make sure they have it installed. I'll keep you posted on what we find. -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
reviewboard 1.6.11 and reviewing problems with FireFox 3.6 and IE 9
We upgraded a few weeks ago to 1.6.11 and we've hit a few minor problems. For IE users when they edit the description the newlines disappear and they get one massive line of text. Some times this occurs when viewing the description too and not editing. The IE compatibility mode usually helps with the display issue but not editing. I'm promoting Chrome and FF (15) internally to avoid both issues :-) I've had a few people with older FF version (3.6) where they can't edit/publish/discard reviews they started (e.g. comments on other peoples or their own review). Using FF 15 or Chrome has worked in most cases. I just now had a user tell me the are using FF 15.0.1 and they've had issues hitting the Publish button too. We've hit http://code.google.com/p/reviewboard/issues/detail?id=2728 too but I think I have a fix for that (without migrating to 1.7 beta) What kind of information can I get to help diagnose what is going on? For the FF 3.6 case I have a dump of what the web browser showed (which showed no www errors), see end of message for that. I've seen the IE newline issue (over someones shoulder) so I think I can probably get more information about that one. Thanks! --- FF 3.6 error: Error Code: 200 Error Text: OK Request URL: http://MYREVIEWBOARDSERVERADDR/api/review-requests/16046/reviews/draft/ Request Data: (none) Response Data: There may be useful error details below. The following error page may be useful to your system administrator or when reporting a bug. To save the page, right-click the error below and choose Save Page As, if available, or View Source and save the result as a .html file. Warning: Be sure to remove any sensitive material that may exist in the error page before reporting a bug! ok 0 http://MYREVIEWBOARDSERVERADDR/api/review-requests/16046/reviews/33072/diff-comments/ GET http://MYREVIEWBOARDSERVERADDR/api/review-requests/16046/reviews/33072/file-attachment-comments/ GET http://MYREVIEWBOARDSERVERADDR/api/review-requests/16046/reviews/33072/ GET http://MYREVIEWBOARDSERVERADDR/api/review-requests/16046/reviews/33072/ PUT http://MYREVIEWBOARDSERVERADDR/api/review-requests/16046/reviews/33072/screenshot-comments/ GET http://MYREVIEWBOARDSERVERADDR/api/users/Gosia/ GET http://MYREVIEWBOARDSERVERADDR/api/review-requests/16046/reviews/33072/replies/ GET http://MYREVIEWBOARDSERVERADDR/api/review-requests/16046/reviews/33072/ DELETE 2012-09-24 06:18:30 33072 0 -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Screenshots for version 1.7?
On Wednesday, September 26, 2012 11:26:36 AM UTC-7, Robert Dailey wrote: Are there any screenshots I can look at for various UI improvements for version 1.7? What do code review pages look like? Have you seen the live site http://reviews.reviewboard.org/r/3347/ ? -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: reviewboard 1.6.11 and reviewing problems with FireFox 3.6 and IE 9
On Thursday, September 27, 2012 2:12:08 PM UTC-7, Christian Hammond wrote: The IE thing is weird, because I don't think anything should have changed as far as text entry goes in a while... Which IE? The IE behavior may not be new, for the version see the title (9) but recall it was seen with 8 too. I can get access to someones machine where this seems to occur on a regular basis. For Firefox, is there anything in the JavaScript console? Installing Firebug (you may need an older one for FF3.6) would give you a lot more useful information. That might be tricky, I'm personally not seeing the problem but I'll mail out to the upgrade hold outs :-) and make sure they have it installed. I'll keep you posted on what we find. -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Unable to upload screenshots
Thilo-Alexander Ginkel wrote: On Mon, Feb 14, 2011 at 23:01, Christian Hammond chip...@chipx86.com wrote: I don't believe they're stored temporarily anywhere. If so, it's internal to Django and I don't know it off-hand. The only thing I've ever seen permission-wise is when there's a directory within media/uploaded that doesn't have the proper write permissions for the server. Hm... Actually, after doing some more testing I think that permissions aren't an issue as GIF files can be uploaded properly. I also can't imaging why libpng should be loadable from the command line, but not from the Apache process... Arise zombie thread, live again! ;-) OK, so it turns out after years of happily running with RB we've hit this too. I KNOW this used to work (as my image testing when we first deployed was with PNG files). Our problem appears to be identical to this, regular RB install (i.e. easy_install based) which always seems to pull down PIL even though it is already installed (we're running Ubuntu server and I always ensure the debian PIL package is installed that has full image support). We could only attach new GIF screen shots to reviews. Using PIL in a Python terminal always worked (i.e. open the supposedly bad image via Image.open), but RB/Django would always fail with: -- Cut here -- Upload Screenshot One or more fields had errors Caption: Path: * Upload a valid image. The file you uploaded was either not an image or a corrupted image. -- Cut here -- On a hunch I thought I'd have a look at what PIL instances are on the file system (e.g. all the eggs that easy_install pulled down), after upgrading a few times over the years we had 2. reviews:~$ find / -name \*PIL\* 2/dev/null /usr/share/pyshared/PIL.pth /usr/share/pyshared/PIL /usr/share/pyshared/PIL/PIL-1.1.6.egg-info /usr/lib/python2.5/site-packages/PIL.pth /usr/lib/python2.5/site-packages/PIL-1.1.6-py2.5-linux-i686.egg /usr/lib/python2.5/site-packages/PIL-1.1.6-py2.5-linux-i686.egg/PIL /usr/lib/python2.5/site-packages/PIL /usr/lib/python2.5/site-packages/PIL/PIL-1.1.6.egg-info /usr/lib/python2.4/site-packages/PIL ... So I removed all the PIL eggs (remember I already have PIL installed): sudo rm -rf /usr/lib/python2.5/site-packages/PIL-1.1.6-py2.5-linux-i686.egg /usr/lib/python2.5/site-packages/PIL/PIL-1.1.6.egg-info Followed by a reviewboard restart: sudo /etc/init.d/apache2 restart And we can now attach PNGs and not just GIFs. So what out for easy_install and PIL. I know pip purports to support removal (and would probably be a better solution) so this is not a recommendation of resolution! I'm just sharing what I found on my system, don't go deleting files if you don't know the impact/dependencies! Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Diff viewer vs. control-L
Bill Fenner wrote: We've been having trouble with the diff viewer highlighting the wrong part of the file (e.g., it highlights 6 lines as being added, but those 6 lines are 12 lines above the lines that were actually added). I finally root-caused this to embedded control characters inside comments. This source, for better or for worse, has comments like /* ^L */ where ^L is literally a control-L. In vi with :number set, this appears as: 889 890 /* ^L */ 891 In the reviewboard diff viewer, this appears as: 889 890 /* 891 */ 892 ..snip My 2 cents, having anything on the same line after a FormFeed is kinda hard to grok. I have an old code base (presumably like yours) that has FF (CTRL-L, 0x0c, etc.) characters all over. Reviewboard (and pygments) haven't had any problems BUT the use of ^L in our the code has been consistently followed by a new line (most of these are not in comments but that is a by-the-by). Options appear to be: 1. fix your code base, e.g. write a small python script to replace all '\x0c' occurrences with '\x0c\n' and just submit the updated code. Kinda dirty and a pain if you have multiple branches 2. modify pygments 3. modify Reviewboard For #2 and #3 this would be something like convert all '\x0c' occurrences with '^L' so that the display shows something. You may want to mail the pygments mailing list and see what they say about displaying ^L. I'd suggest generating an ascii table with each character taking one line (with some padding) and see what pygments does for all the characters from 0-127, there may be other control characters that an alternate visual replacement would make sense for (e.g. BEL). Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Is there a print function in review board?
One option is to print the email sent from reviewboard. If you want to print the diffs you are likely to find sizing issues with the web page approach with the wide side-by-side diff view. We've modified our RB server to include the diff in the email that is sent out for review (we've not yet worked out how to NOT include the diffs in replies/comments). That can help a little bit but the main advantage we've found is using the browser to see the diffs and then immediately expand code and add comments. You could do this with face-to-face reviews if the computer is used instead of print. Chris On 9/30/2011 9:06 AM, Nelson Jones wrote: I see. Thanks Christian! On Sep 30, 12:00 pm, Christian Hammondchip...@chipx86.com wrote: Hi Nelson, There's just the standard browser print capabilities. We don't do any special formatting. Christian On Friday, September 30, 2011, Nelson Joneshuawen@gmail.com wrote: Hi, Just wonder is there a print function in review board? In case, a face to face review is needed? Thanks, -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Is there a print function in review board?
On 9/30/2011 11:34 AM, Stephen Gallagher wrote:
On Fri, 2011-09-30 at 11:32 -0700, Chris Clark wrote:
One option is to print the email sent from reviewboard. If you want to
print the diffs you are likely to find sizing issues with the web page
approach with the wide side-by-side diff view.
We've modified our RB server to include the diff in the email that is
sent out for review (we've not yet worked out how to NOT include the
diffs in replies/comments). That can help a little bit but the main
advantage we've found is using the browser to see the diffs and then
immediately expand code and add comments. You could do this with
face-to-face reviews if the computer is used instead of print.
Does it send the diff in the email contents or as an attachment?
It depends :-)
It is semi-configurable, i.e. there is code that has a parameter but it
is hard coded at the moment.
We default to inline (this matches our old review standards that
predated re) but switch to an attachment if we have no idea what the
encoding of the diff is.
We need to add a size filter, e.g. we had a monster review the other day
and what should have happened is the diff should have been compressed
and attached.
If the
latter, could you share your patch with the rest of the class? :)
Right now the code is in a private git repo :-( This is an excellent
reminder that we should try and make this public even if no one will
want all of it. Right now what we have is NOT ready for prime time
As a taster, here is the code, it is not pep8 compliant and it isn't a
diff (we're on an older version), but it should allow you to make a
similar (more robust!) change:
in reviewboard/templates/notifications/review_request_email.html:
{% if raw_diff %}
table width=100% bgcolor=#ff cellspacing=0 cellpadding=10
style=border: 1px solid #b8b5a0
tr
td
pre style=margin: 0; padding: 0; {{precss}}{{ raw_diff }}/pre
/td
/tr
/table
{% endif %}
--- this is the inline diff option
In /reviewboard/notifications/email.py:
context['MEDIA_SERIAL'] = settings.MEDIA_SERIAL
# should really (add to) repository and add include raw diffs option
# for now check scm type or always do it!
include_raw_diff=True
#include_raw_diff=False ## DEBUG FIXME REMOVE!
## NOTE some of our mailing list do not retain (diff) attachments.
attachment_message='Diffs attached'
emaildiffs_inline=True ## if False make an attachment
context['raw_diff']=''
#context['raw_diff']='debug raw diff FIXME REMOVE'
if include_raw_diff:
## Other option, instead of a flag include_raw_diff could
lazy evaluate
## http://pyds.muensterland.org/ has a nice short one (only
needs minor cleanup, and I already have a cleaned up copy)
diffset = review_request.diffset_history.diffsets.latest()
raw_diff =
review_request.repository.get_scmtool().get_parser('').raw_diff(diffset)
if emaildiffs_inline and isinstance(raw_diff, str):
try:
_ = raw_diff.decode('us-ascii') # should check
repository encoding type..
except UnicodeDecodeError:
# non-ASCII data in diff, email is (probably) utf8 and
# so will not display inline. Switch to an attachment
emaildiffs_inline=False ## if False make an attachment
attachment_message='Diffs contain non-ASCII
characters/bytes, unable to determine encoding. See attachment.'
if emaildiffs_inline:
context['raw_diff']=raw_diff
else:
# attach
context['raw_diff']=attachment_message
message = SpiffyEmailMessage(subject.strip(), text_body, html_body,
from_email, list(to_field),
list(cc_field),
in_reply_to, headers)
if include_raw_diff and not emaildiffs_inline:
message.attach('diff.diff', raw_diff, 'text/x-diff')
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
Re: Change email of users
Martin wrote: Hi, I'm using RB 1.6.1 with LDAP authentication. I want to each user can modifiy his email address. I tried add the users a group with auth|user|Can change user, but it still not working... NOTE I'm not using LDAP with RB :-) Usually when using LDAP it is a read only resource, updates are usually performed via ldap (or rather via the team/person administering LDAP). How are users in LDAP handled? Do you have a way for users to update their own details in LDAP? My guess is you can't do this via RB. If you're not using LDAP you can update the email addresses for the users via the management shell, I had to do this recently to automate a domain change for users in batch. I suspect this won't work for LDAP. $ cd /var/www/reviewboard/conf ; django-admin.py shell --settings=settings_local --pythonpath=/var/www/reviewboard/conf from reviewboard.accounts.models import User for x in User.objects.all(): print x.email x.email = gen_new_email(x.email) # defined somewhere else. x.save() # commit -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: upgrade from 1.0.9 tp 1.6.1 failed
premod dev wrote: I have been trying to upgrade my current reviewboard instance to 1.6.1 from the older 1.0.9. But when I try to upgrade the site, its giving me the following error and failing the upgrade. Any idea why? ... No fixtures found. Registering new SCM Tool Plastic SCM (reviewboard.scmtools.plastic.PlasticTool) in database Evolution could not be simulated, possibly due to raw SQL mutations Error: Error applying evolution: (1050, Table 'reviews_reviewrequest_changedescs' already exists) This is kinda of a sucky answer but have you tried an older release, e.g. 1.5.x? I think Plastic support was added in 1.6 so this would get you the latests final/stable release until the next RC (or final) release of 1.6. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
changing backends Re: Upgrading database/site from 1.0 to 1.6.1
Christian Hammond wrote: [RE changing backend databases] Unofficially, you can do the following: On the old site: rb-site manage /path/to/site dumpdata dbdump.json On the new: . loaddata dbdump.json There are no guarantees it'll work, but give it a try. Note: They will both need to be updated to the same version of RB first. I know this is off topic but we've had success at Actian (we changed our name just last week from Ingres) with a json dump transfer. I didn't perform the migration myself, one of my colleagues did, so I can't offer any thing other than this is worth trying, it worked for us. Of course we moved from version X to version X (not version X to Y). We have previously moved from version X to Y but by using the same database backend, see old thread titled Successful upgrade from Reviewboard 0.9 to 1.0.5.1. For the upgrade from 1.0 to 1.5 series the evolve just worked. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: How can I run ReviewBoard 1.5.5 and 1.6rc1 in same time?
Christian Hammond wrote: ...A couple important notes: 1) I would strongly recommend that, if you want to test new versions, that you do it on a non-production system, such as your own machine or on a VM. You can always copy over your database if you need to. That's the best way of making sure things work without all the mess below, and that you don't inadvertently break your production install. My 2 cents, even in the test machine, I take backups[1] ahead of time so that if things go badly I can re-run (and document failures for bug reports, etc.). Going backwards without a backup is just not worth the hassle (and doesn't usually work). [1] Where for me backups might be as simple as a VM snapshot, or a database backup along with www file system tar up. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Easy_install post-review issue with 64-bit windows
Eric Johnson wrote: I've got some clients of the server I set up that wish to use 64-bit Windows. Unfortunately, it appears that Python easy_install doesn't work properly in that environment. Consequently, those users cannot get RBTools (post-review) to work. Is there a work-around, or instructions I can point them to for manually installing RBTools? I'm just giving my users a binary, seems to work well for them. They don't have to install anything, they just but the binary in the path along with the source code control client binary. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Minimum version of Python for RBTools?
What is the minimum supported version of Python for RBtools? The Windows specific docs mention 2.5 http://www.reviewboard.org/docs/manual/dev/admin/installation/windows/ I just stumbled across some code that fails in 2.4 in the ClearCase code (and hacked it to work with 2.4): # fails with python 2.4 self.viewtype = 'dynamic' if 'dynamic' in properties else \ 'snapshot' if 'dynamic' in properties: self.viewtype = 'dynamic' else: self.viewtype = 'snapshot' I'm wondering if I should consider an upgrade of my postreview clients. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: About the submit workflow
Hongbin Lu wrote: My name is Hongbin Lu. I am currently working on the issue below. http://code.google.com/p/reviewboard/issues/detail?can=2q=693colspec=ID%20Type%20Status%20Priority%20Component%20Owner%20Summary%20Milestoneid=693 http://code.google.com/p/reviewboard/issues/detail?can=2q=693colspec=ID%20Type%20Status%20Priority%20Component%20Owner%20Summary%20Milestoneid=693 Basically, several users like to attach comment when clicking submit. The solution we can think of is clicking submit one time, popup an windows to prompt for comment, and click submit second time. In other words, there will be two click instead of one click(current version). We want to know whether we should go this way or there is a better solution you can think of. Having comment on close would be ideal. What we do here is use postreview to post a comment with the change details, then postreview to close it. I.e. mostly automated client side at commit time. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: 1.5.5 diff viewer not showing files with a single whitespace change?
Bruce Cran wrote: I recently upgraded to 1.5.5 and found that with one review which had several files changed, Review Board starts out listing all the files under Files Changed, but quickly decides not to show one of them, which has a 1-line change to the amount of leading whitespace. Other files with whitespace changes were shown, so I'm wondering if this is a bug? Probably not a bug but a feature. There are some settings related to white space and file types. See topic ReviewBoard not displaying content of diff file http://groups.google.com/group/reviewboard/browse_thread/thread/fad1e1741c19d606/28ccf483f4fd81e1?lnk=gstq=chris+white+space#28ccf483f4fd81e1 Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: I have some problems about Languages.
Christian Hammond wrote:
I think we need a better way to factor this all out, though. Otherwise
you'll have to maintain a custom patch for every upgrade.
Can you tell me where you placed the first bit of code (encoding
changeset0['desc'])?
We probably can reuse the repository's Encoding field and insert the
lines into perforce.py to respect it, in much the same way that you
did. So instead of unicode(something, 'cp949'), it'd be
unicode(something, repository.encoding). Right now, that field only
appears to affect the patches.
And in your case, what does p4.charset show?
This is a good example of why we need support for named encoding in
postreview, doing a look up is certainly the right thing to do; by
checking the source code control system (where possible) and the
ReviewBoard repository encoding. However we also need an override/force
option.
In the first instance a command line parameter would be useful.
RE where the decode() call (rather than unicode) takes places, I'd
suggest around line 3877. Either before or after:
if options.output_diff_only:
# The comma here isn't a typo, but rather suppresses the extra newline
print diff,
sys.exit(0)
The print would probably need a repr() around it if done before.
Line 3854 also needs a change to open options.diff_filename in binary
('rb') mode.
Example code:
diff_encoding = 'utf-8'
diff_encoding = 'cp949' # ... from command line, config file, lookup in
RB repo, etc.
..
diff = diff.decode(diff_encoding)
Chris
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
Re: I have some problems about Languages.
jh5774@samsung.com wrote:
Well, I found a little clue about encoding.
/usr/lib/python2.6/json/__init__.py
in this codes, they said python only decode 'ASCII' and 'Unicode'.
So, python couldn't decode when encounter like a 'Korean word'
I find some solution about decoding.
For example,
a= '테스트'
unicodeA =a.decode('euc-kr')
utf8A = unicodeA.encode('utf-8')
or
u = unicode(a, 'euc-kr').encode('utf-8')
Now, I must find the where reviewboard receive response message
which their request to perforce using 'p4 describe' command.
Anyone who knows about that, please tell me.
I'd be tempted to take a look at postreview and look at adding an
encoding flag. I haven't yet really looked at the transport mechanism,
there is a string implication that Unicode (specifically utf8) is sent
for the diffs, really bytes should be sent. This may be because json is
supposed to be utf8. Basically you would do the first part of your
example above in postreview. The other option (and this is a hack) is to
change your site packages to change the default system encoding for
Python from ASCII (default for windows) to euc-kr. This normally isn't
recommended BUT it would be a good diagnostic step to take to
prove/disprove the theory. See
http://blog.ianbicking.org/illusive-setdefaultencoding.html
Let us know how you get on.
The extreme solution would be to use utf8 encoding for all source code,
which is fine for new projects but more tricky for older projects.
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
Re: AttributeError: 'DashboardDataGrid' object has no attribute 'id_list'
Leonel Togniolli wrote: Every week or so I get one of these below in the mail from my RB server. I wasn't able to determine what triggers it, unfortunately. Me too. See http://code.google.com/p/reviewboard/issues/detail?id=1526 I suggest you log a comment on the issue/bug so that there is a record on how many people are seeing this. I'm now seeing this every day, I think with different users. I know one user has seen this a couple of times (some of the VPN users makes it tricky to know for sure who hits it). I didn't see it with: Django-1.1.1, Djblets-0.5.5, Reviewboard 1.0.5.1 Your seeing it with: Django-1.2.3, Djblets-0.6.6, Reviewboard-1.5.1 I'm seeing it with: Django-1.3, Djblets-0.6.7, Reviewboard-1.5.5 Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: AttributeError: 'DashboardDataGrid' object has no attribute 'id_list'
Leonel, I forgot to ask, are you seeing this after an upgrade? Chris Clark wrote: Leonel Togniolli wrote: Every week or so I get one of these below in the mail from my RB server. I wasn't able to determine what triggers it, unfortunately. Me too. See http://code.google.com/p/reviewboard/issues/detail?id=1526 I suggest you log a comment on the issue/bug so that there is a record on how many people are seeing this. I'm now seeing this every day, I think with different users. I know one user has seen this a couple of times (some of the VPN users makes it tricky to know for sure who hits it). I didn't see it with: Django-1.1.1, Djblets-0.5.5, Reviewboard 1.0.5.1 Your seeing it with: Django-1.2.3, Djblets-0.6.6, Reviewboard-1.5.1 I'm seeing it with: Django-1.3, Djblets-0.6.7, Reviewboard-1.5.5 Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Error with post-review. I get Failed to execute command: ['git', 'symbolic-ref', '-q', 'HEAD']
Martin wrote: Failed to execute command: ['git', 'symbolic-ref', '-q', 'HEAD'] What happens when you manually run: git symbolic-ref -q HEAD -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: UnicodeDecodeError with Visual Studio files, RB 1.5.4, RBTools-0.3.2 (Python 2.7.1)
Craig A wrote: This may well be fixed by trying python 2.6 (or even 2.5) as you suggested. I tried Python 2.6.6 (could not find an msi installer for 2.5.5) and voila! It works now, no more problems with BOM characters. Thanks for your help Chris, I am glad this is working now. Whilst this impacted a Windows user (Windows Python defaults to 7 bit US-ASCII encoding) with a utf8 encoded (diff) file. This could easily impact any user on any platform where the encoding for the file is not valid for the (default) Python site string encoding. For example this is likely to impact users of Python 2.7 if: * diff contains a single latin1 character (e.g. copyright symbol, an e with an accent, etc.) and the site encoding is not latin1. This would impact all Unix/Linux platforms where the Python installation tends to default to utf8 It looks like the Python 3.x byte/string difference that is in 2.7 will need to be looked at. Possibly via an encoding flag to postreview. I'm snowed under at the moment otherwise I'd look at this myself :-( I have a number of Windows users (with non ascii diffs) so I'm at risk for this problem BUT this isn't an issue for me as I supply them with an exe generated with Python 2.4 and py2exe. Craig This may be your best option. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Successful upgrade from Reviewboard 1.0.5.1 to 1.5.5
We're still running fine with the new release, we've had a minor issue (which is a known issue, but without a consistent reproducing test case). More information at http://code.google.com/p/reviewboard/issues/detail?id=1526#c8 We're still pleased we've upgraded :-) Chris Christian Hammond wrote: Awesome :) Thanks for sharing. It's nice to hear on this list when things actually go smoothly ;) Christian -- Christian Hammond - chip...@chipx86.com mailto:chip...@chipx86.com Review Board - http://www.reviewboard.org VMware, Inc. - http://www.vmware.com On Wed, Mar 30, 2011 at 6:29 PM, Chris Clark chris.cl...@ingres.com mailto:chris.cl...@ingres.com wrote: We had a customized version of 1.0.5.1 and we now have a (less) customized version 1.5.5 (custom SCM back end and minor email/template tweaks). We're using Ubuntu 8.04.3 LTS (hardy) on x86 (same as the last time we upgraded, we've left the OS alone) . We had no problems at all with this upgrade! So unlike my previous upgrade reports this really is an uninteresting one :-) The only minor oddity was that setup tools needed to be upgraded. In our haste to upgrade we simply used easy_install for that (it may have been that apt-get would have found a new version). When the test upgrades had been done, a different OS version was used so this was our only surprise. We also went from Django-1.1.1 (we're behind a firewall so this old version has been fine for us) to Django-1.3 (other stuff was upgraded too, like Djblets-0.5.5 to Djblets-0.6.7, but this is one of the significant ones). The main reason for the upgrade was that we needed auth support for svn, we've wanted the other new features (drag-n-drop images, improved username lookup, ) for a while but we *needed* auth. Thanks again to everyone for making the upgrade path so easy, with great/clear documentation. We maybe brave soon and enable the Lucene support. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com mailto:reviewboard%2bunsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Successful upgrade from Reviewboard 1.0.5.1 to 1.5.5
We had a customized version of 1.0.5.1 and we now have a (less) customized version 1.5.5 (custom SCM back end and minor email/template tweaks). We're using Ubuntu 8.04.3 LTS (hardy) on x86 (same as the last time we upgraded, we've left the OS alone) . We had no problems at all with this upgrade! So unlike my previous upgrade reports this really is an uninteresting one :-) The only minor oddity was that setup tools needed to be upgraded. In our haste to upgrade we simply used easy_install for that (it may have been that apt-get would have found a new version). When the test upgrades had been done, a different OS version was used so this was our only surprise. We also went from Django-1.1.1 (we're behind a firewall so this old version has been fine for us) to Django-1.3 (other stuff was upgraded too, like Djblets-0.5.5 to Djblets-0.6.7, but this is one of the significant ones). The main reason for the upgrade was that we needed auth support for svn, we've wanted the other new features (drag-n-drop images, improved username lookup, ) for a while but we *needed* auth. Thanks again to everyone for making the upgrade path so easy, with great/clear documentation. We maybe brave soon and enable the Lucene support. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: UnicodeDecodeError with Visual Studio files, RB 1.5.4, RBTools-0.3.2 (Python 2.7.1)
Craig A wrote: I am getting an error similar to what was reported some time ago in this post: http://groups.google.com/group/reviewboard/browse_thread/thread/56fb450ceaef45c1 UnicodeDecodeError: 'ascii' codec can't decode byte 0xef in position 1285: ordinal not in range(128) Specifically you are hitting: http://groups.google.com/group/reviewboard/msg/13225461e83d7743 The BOM mark at the start of the file. Thanks for making the file available so this could be confirmed. If this is a new file the diff will include the BOM. The error above usually occurs when there is an assumption that there is a a Unicode string, and file diff contents should be bytes (i.e. it shouldn't go through ANY string/unicode processing). Can you post the error, the complete traceback? Along with the version you are using. You mention you've customized postreview (pretty common, my company does too) but it is possible that your new code has the Unicode assumption in it rather than the original RBtools. It is worth your while doing a quick test with a virgin version of the latest RBTools (that traceback would be the most helpful to people on the list). Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: UnicodeDecodeError with Visual Studio files, RB 1.5.4, RBTools-0.3.2 (Python 2.7.1)
Craig A wrote:
You mention you've customized postreview ...
is it possible that your new code has the Unicode assumption in
it rather than the original RBtools.
I don't think so, the relevant line of code looks like this:
return content_type, content
The trace back does not appear to show that line of code. There appears
to be a concatenation of unicode and (2.x) str (bytes) type.
This may well be fixed by trying python 2.6 (or even 2.5) as you
suggested. You'd need to debug this through to see where the Unicode
string is coming from, 2.7 is the cross over version for 2.x and 3 and
it does try to deal with Unicode types when possible. Of course json
data is supposed to be utf8 encoded so this maybe where this is failing
down.
I've left the traceback below for reference.
Traceback (most recent call last):
File C:\apps\python\2.7.1\scripts\post-review-script.py, line 8,
in module
load_entry_point('RBTools==0.3.2', 'console_scripts', 'post-
review')()
File C:\apps\python\2.7.1\lib\site-packages\rbtools-0.3.2-py2.7.egg
\rbtools\postreview.py, line 3800, in main
submit_as=options.submit_as)
File C:\apps\python\2.7.1\lib\site-packages\rbtools-0.3.2-py2.7.egg
\rbtools\postreview.py, line 3466, in tempt_fate
parent_diff_content)
File C:\apps\python\2.7.1\lib\site-packages\rbtools-0.3.2-py2.7.egg
\rbtools\postreview.py, line 769, in upload_diff
fields, files)
File C:\apps\python\2.7.1\lib\site-packages\rbtools-0.3.2-py2.7.egg
\rbtools\postreview.py, line 982, in api_post
return self.process_json(self.http_post(path, fields, files))
File C:\apps\python\2.7.1\lib\site-packages\rbtools-0.3.2-py2.7.egg
\rbtools\postreview.py, line 903, in http_post
data = urllib2.urlopen(r).read()
File C:\apps\python\2.7.1\lib\urllib2.py, line 126, in urlopen
return _opener.open(url, data, timeout)
File C:\apps\python\2.7.1\lib\urllib2.py, line 392, in open
response = self._open(req, data)
File C:\apps\python\2.7.1\lib\urllib2.py, line 410, in _open
'_open', req)
File C:\apps\python\2.7.1\lib\urllib2.py, line 370, in _call_chain
result = func(*args)
File C:\apps\python\2.7.1\lib\urllib2.py, line 1186, in http_open
return self.do_open(httplib.HTTPConnection, req)
File C:\apps\python\2.7.1\lib\urllib2.py, line 1155, in do_open
h.request(req.get_method(), req.get_selector(), req.data, headers)
File C:\apps\python\2.7.1\lib\httplib.py, line 941, in request
self._send_request(method, url, body, headers)
File C:\apps\python\2.7.1\lib\httplib.py, line 975, in
_send_request
self.endheaders(body)
File C:\apps\python\2.7.1\lib\httplib.py, line 937, in endheaders
self._send_output(message_body)
File C:\apps\python\2.7.1\lib\httplib.py, line 795, in
_send_output
msg += message_body
UnicodeDecodeError: 'ascii' codec can't decode byte 0xef in position
689: ordinal not in range(128)
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
Re: Users are automatically created if authenticated through ldap
Christian, Is this a Reviewboard feature or a Django one? I would have guessed this was a Django behavior. I'm not even a Django (auth) novice, let alone expert, so sorry if this is a dumb question! Chris Christian Hammond wrote: Hi Pablo, No, there's no way to do this out of the box. You would need to write a custom auth backend that augments the LDAP backend to so this. There are plans in 1.6 to e-mail admins for new users. It'd make sense to have an option to disable new users until the admin approves them. If you want to file a bug on this, it'll help us track this. Christian On Wednesday, March 23, 2011, Pablo G pablo...@gmail.com wrote: Hi guys. I have a RB installation validating against an LDAP server. It looks like RB creates users in the local database whenever a user is correctly authenticated with LDAP. What I need is a way to disable this automatic user creation, so the site admin can decide which users in the full LDAP repository can access the RB. Is this possible? I am on RB 1.5.2. thanks a lot. -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Help debugging a hang in RB 1.5
Alfred von Campe wrote: We are experiencing a strange hang when viewing one particular review. When we click on the View DIff we get a new web page and the Files Changed: section lists all the files that have changed in this particular request. There are 20 files, and the information for the first 6 is updated in a few seconds. But then it hangs on the 7th file until we navigate away from the page (my last attempt was about 20 minutes ago and I still have 14 spinning circles on the screen). I know I should upgrade to the 1.5.3 release from earlier today, but before I do that I want to try to debug this issue. The problem is, I don't know where to even start. BTW, when the user that stumbled across this issue reported it to me last week, I was able to load this particular review, but today it is hanging for me as well. Always on the same file. Funny thing is that there is only a very minor change to this file. (I've tried multiple browsers in case that matters). I've had 2 main issues with slow diffs: 1. database locking (for testing I was using sqlite which is single user and horrible for *large* [10Mb diffs] reviews as multi user concurrency is awful, for most reviews sqlite is actually pretty good). 2. complex diffs being slow to generate diffs, this is a feature of the batteries includes difflib in python. #1 is pretty easy to check, if you get timeouts and errors mailed to the admin about database locks you know you hit #1 :-) #2 is more sinister as it doesn't happen that often but you can get pathologically bad diffs that takes mins/hours for difflib to examine. The only way to know if this is what you are hitting is to write a python script to use difflib with the original and new file. I've attached a demo of #2 where there is a subtle difference mid-line for each line. There is no easy way around this unless a different diff module is used. NOTE attached demo uses an iterator, if you do NOT use an iterator it still takes 2 mins but you get no output during that 2 minute period and so looks like it has hung! I've no idea if you are hitting these but I've been meaning to post my experiences in this area for some time and this was a good reminder. Chris demo of built in diff lib cpu clobbering really heavy cpu intensive, but performs VERY SLOWLY (it looks like it hangs, but it does complete) import difflib file1content = ['27 mss gw001 PASS\n', '27 mss gw002 PASS\n', '27 mss gw003 PASS\n', '27 mss gw004 PASS\n', '27 mss gw005 PASS\n', '27 mss gw006 PASS\n', '27 mss gw007 PASS\n', '27 mss gw008 PASS\n', '27 mss gw009 PASS\n', '27 mss gw010 PASS\n', '27 mss gw011 PASS\n', '27 mss gw012 PASS\n', '27 mss gw013 PASS\n', '27 mss gw014 PASS\n', '27 mss gw015 PASS\n', '27 mss gw016 PASS\n', '27 mss gw017 FAIL\n', '27 mss gw018 PASS\n', '27 mss gw019 PASS\n', '27 mss gw020 PASS\n', '27 mss gw021 FAIL\n', '27 mss gw022 PASS\n', '27 mss gw023 PASS\n', '27 mss gw024 PASS\n', '27 mss gw025 PASS\n', '27 mss gw026 PASS\n', '27 mss gw027 PASS\n', '27 mss gw028 PASS\n', '27 mss gw029 PASS\n', '27 mss gw030 PASS\n', '27 mss gw031 FAIL\n', '27 mss gw032 PASS\n', '27 mss gw033 FAIL\n', '27 mss gw034 PASS\n', '27 mss gw035 PASS\n', '27 mss gw036 PASS\n', '27 mss gw037 PASS\n', '27 mss gw038 PASS\n', '27 mss gw039 FAIL\n', '27 mss gw040 FAIL\n', '27 mss gw041 PASS\n', '27 mss gw042 PASS\n', '27 mss gw043 PASS\n', '27 mss gw044 PASS\n', '27 mss gw045 PASS\n', '27 mss gw046 PASS\n', '27 mss gw047 PASS\n', '27 mss gw048 PASS\n', '27 mss gw049 PASS\n', '27 mss gw050 PASS\n', '27 mss gw051 PASS\n', '27 mss gw052 PASS\n', '27 mss gw053 PASS\n', '27 mss gw054 PASS\n', '27 mss gw055 PASS\n', '27 mss gw056 PASS\n', '27 mss gw057 PASS\n', '27 mss gw058 PASS\n', '27 mss gw059 PASS\n', '27 mss gw060 PASS\n', '27 mss gw061 PASS\n', '27 mss gw062 PASS\n', '27 mss gw063 PASS\n', '27 mss gw064 PASS\n', '27 mss gw065 PASS\n', '27 mss gw066 PASS\n', '27 mss gw067 PASS\n', '27 mss gw068 PASS\n', '27 mss gw071 PASS\n', '27 mss gw075 PASS\n', '27 mss gw076 PASS\n', '27 mss gw077 PASS\n', '27 mss gw078 PASS\n', '27 mss gw079 PASS\n', '27 mss gw081 PASS\n', '27 mss gw085 PASS\n', '27 mss gw086 PASS\n', '27 mss gw088 PASS\n', '27 mss gw091 PASS\n', '27 mss gw092 PASS\n', '27 mss gw093 PASS\n', '27 mss gw094 PASS\n', '27 mss gw095 PASS\n', '27 mss gw096 PASS\n', '27 mss gw097 PASS\n', '27 mss gw098 PASS\n', '27 mss gw099 PASS\n', '27 mss gw100 PASS\n', '27 mss gw101 PASS\n', '27 mss gw102 PASS\n', '27 mss gw103 PASS\n', '27 mss gw104 PASS\n', '27 mss gw105 PASS\n', '27 mss gw106 PASS\n', '27 mss gw108 PASS\n', '27 mss gw109 PASS\n', '27 mss gw110 PASS\n', '27 mss gw111 PASS\n', '27 mss gw112 PASS\n', '27 mss gw113 PASS\n', '27 mss gw114 PASS\n', '27 mss gw115 PASS\n', '27 mss gw116 PASS\n', '27 mss gw117 PASS\n', '27 mss gw118 PASS\n', '27 mss gw120 PASS\n', '27 mss gw121 PASS\n', '27 mss gw122 PASS\n', '27 mss gw123 PASS\n', '27 mss
javascript detect/protect back button presses when content updated?
I've recently been using a laptop with forward/back keys. These are conveniently located by the arrow/cursor keys. In a web browser these keys are the equivalent of alt-left/right (or the backwards/forwards on screen button). I've just spent the last 10 mins re-entering (multiple times) in text for a review description as I fat fingered the back key. I'd not hit the SAVE button and so I lost my work (the save button was physically off screen due to the size of the description box.text, saving is a real pain with the current UI if you have lots of text). I have lost work a few times in the past by clicking on bug numbers to remind myself (or copy/paste from) what to put into the description but this is the first time fat-fingers have caused me to loose review work. I see this behavior on all versions (including 1.6 alpha on http://reviews.reviewboard.org/ ). The other (more) minor annoyance is that I sometimes hit CTRL-s thinking that this will save (I'm just used to that combo) which of course doesn't work :-) It tends to call the web browser save as html dialog. Has anyone else had similar issues? This is an user error but it is a UI annoyance that could be handled in Reviewboard. I've 2 ideas: 1) support CTRL-S for saving - Google docs does this as does gmail, it is a small but awesome thing :-) 2) if the description (or testing, etc.) has been updated and navigation is attempted send up a pop-up are you sure?, gmail does this when composing an email and it has not yet been saved. webchat on freenode does this too, it can even detect window/tab closing. Comments? This isn't meant to be a complaint, I'm trying to see if I'm the only one seeing this. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: ReviewBoard not displaying content of diff file
cogo wrote: I have just started trying out reviewboard and I have some issues with some diffs not beeing displayed when I click View diff in the review request. When i download the diff I can see the content: Index: //trunk/library/extras/global-post-commit-hook.php === --- //trunk/library/extras/global-post-commit-hook.php (revision 1901) +++ //trunk/library/extras/global-post-commit-hook.php (revision 1902) @@ -64,4 +64,4 @@ $revision)); } -// Test \ No newline at end of file +// Test Why does this not get displayed? This looks like a white-space-only-diff issue. There is a config option that _may_ help here (and there is a bug too), see : http://code.google.com/p/reviewboard/issues/detail?id=1780http://code.google.com/p/reviewboard/issues/detail?id=1780 http://code.google.com/p/reviewboard/issues/detail?id=1511 Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Timezone for London
brokenn wrote: I've just installed Reviewboard 1.5RC2 on a Centos 5 box. When selecting the timezone from the Admin-Settings-General Settings menu, it lists time zones for all parts of the world, but under Europe there are no UK cities. I'm not London-centric, but would like to select a sensible timezone! Any ideas where to look for track this issue down? The 1.5 release (not RC) has Europe/London on my machine. I'm not sure the RC is the issue but I'd use the GA release before I starting debugging further. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Error: No module named django_evolution
I was following the instructions on http://www.reviewboard.org/docs/codebase/dev/getting-started/ And at the setup Reviewboard stage: python ./contrib/internal/prepare-dev.py I got : Error: No module named django_evolution I worked around it by issuing: python setup.py develop python ./contrib/internal/prepare-dev.py As background, I'm using Python 2.5.2 under a virtualenv with no site packages, e.g: virtualenv --no-site-packages rb15 Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: post-review fails on HTTP POST
jdt141 wrote: File /usr/lib/python2.5/site-packages/RBTools-0.2-py2.5.egg/rbtools/ postreview.py, line 549, in process_json AttributeError: 'module' object has no attribute 'loads' I'm at a loss... Any and all help would be appreciated. It looks like the json/simplejson module isn't there (although one would expect a different kind of error, much sooner). The path indicates Python 2.5, so you will need a 3rd party library (like simplejson), see http://code.google.com/p/simplejson. The other option would be to try source (rather than the egg) and then trace with pdb to more information. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Accessing different repo types from post-review
Jay wrote: On Sep 21, 12:11 pm, Chris Clark chris.cl...@ingres.com wrote: It looks like p4 is claiming there is a perforce repo in the svn location. A quick hack/test would be to modify postreview to check svn first. I.e. hack the SCMCLIENTS def. Let's assume this suggestion actually works since it looks like the script first tries hg, then p4, probably then svn. It seems there needs to be a way to explicitly indicate which scm provider to use, rather than do it implicitly since it's conceivable they could all work for a given directory. Not really, it isn't good practice (IMHO) to use multiple SCMs in the same working directory. You may have a mix of SCM's under a given tree but not in the same directory. Take for instance the following structure: c:/dev - this is the root for the perforce clientspec c:/dev/myproject/vers1 - this is a perforce version of the project called out in the clientspec c:/dev/myproject/vers2 - this is the root for an svn repository for the project I could see how perforce might be picked up for myproject/vers2 even though it isn't specified in the perforce clientspec. If there was a way to explicitly declare the repo type, then this wouldn't be an issue. The way to do this would be to edit postreview.py either by adding some sort of directive as I think you are suggesting or by customizing for your site/usage. If it were me, I'd edit the SCMCLIENTS list definition (I've actually done this for our site, I have customization for VMS so svn isn't checked as VMS command spawning is not implemented in CPython or Jython). See http://www.reviewboard.org/docs/codebase/dev/getting-started/ under RBTools if you want to start with headrevs from git. E.g. I have something like: ### re-define SCMCLIENTS, this makes merging changes easier (than customizing SCMCLIENTS) :-) SCMCLIENTS = ( SVNClient(), PiccoloClient(), ) Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Accessing different repo types from post-review
Jay wrote: I have post-review installed on windows and have successfully been creating RB requests against perforce. We also have SVN projects, but when I try to use post-review against them it wants to talk to perforce instead. For instance if I enter: post-review --server=http://somehost/reviewboard --repository- url=https://some-svn-repo-url --revision-range=0:999 --debug I see: hg root p4 info repository info: Path: sdgperforce:1666, Base path: None, Supports changesets: True So how do I get post-review to speak svn instead of perforce for a given request? The reviewboard instance is aware of the svn repositories and others that only post against svn don't seem to have a problem. It looks like p4 is claiming there is a perforce repo in the svn location. A quick hack/test would be to modify postreview to check svn first. I.e. hack the SCMCLIENTS def. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Release party planning for Review Board 1.5
Christian Hammond wrote: We're finalizing the documentation and performing some final testing before releasing Review Board 1.5. As per tradition, we're looking at putting together a release party, and we'd like to invite anyone who wants to come. Details are still being worked out, but will probably be dinner and drinks sometime during the month of October in the Palo Alto/Mountain View/Menlo Park area in California, USA. At this point, we're looking to get an idea as to how many people would be seriously interested in coming, and what your availability looks like. Once we have a rough idea, we can start finalizing the venue and date/time. Please e-mail me directly if you'd be interested in going. I'd be interested. It would be good to finally meet :-) RE timing, I work in Redwood City (commuting from east bay) so after work weekdays would be good for me. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: post-review tool: How to post only *.c and *.h files
sakthi wrote: Could find i have to use post-review tool to post the code into review board. Can some one help me out how to post ONLY *.c and *.h files recursively from SVN directory. Is there any automated script available.. where i shall modify the require information to post the code that is required? Let me pose a question to you ;-) How are you going to submit just *.c and *.h files to subversion when you come to commit? There is no reviewboard involved at that stage. Postreview will accept the same filenames on command line that subversion will take at submit time. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: post-review error with svn and gnu diffutils (new review with an added binary file)
Timothy Pinet wrote: In which file does reviewboard invoke the svn diff --diff-cmd=diff command? I have been looking but can not find it. I then thought that you probably call the pysvn lib for this. I checked the pysvn documentation (http://pysvn.tigris.org/docs/ pysvn_prog_ref.html#pysvn_client_diff) and there is a flag ignore_content_type for the pysvn.Client.diff command. I assume that reviewboard calls this method. It actually isn't in ReviewBoard, it is in postreview.py :-) Jut fork postreview locally and you can then customize it for your environment. This is very common, I've done it for my team as we needed a bunch of customizations which really aren't appropriate for anyone else. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: post-review error with svn and gnu diffutils (new review with an added binary file)
Timothy Pinet wrote:
I wanted to check post-review however since I am on Windows my
Python26\Scripts\post-review.exe is a compiled executable and not a
python file. Would it be enough to grab the post-review.py from the
repo and delete the exe?
Not with the current version. There are a few dependencies (not many),
e.g. a version module. So you should take the whole thing. You can of
course make an exe from it very easily. See end of mail for the custom
setup script I use.
Chris
Issue:
NOTE using Python 2.4, this results in an exe about 4Mb in size.
NOTE using Python 2.6, this results in an exe about 5.5Mb in size.
c:\python24\python ingres_setup.py py2exe
setup.py py2exe
Quick-N-Dirty create win32 binaries and zip file script.
Zero error checking.
TODO inject 'py2exe' into sys.argv?
import os
import glob
import shutil
from distutils.core import setup
import py2exe
# Clean temp Python/Jython files
delete_list=glob.glob('simplejson/*.pyc')+glob.glob('simplejson/*$py.class')
for x in delete_list:
os.remove(x)
try:
shutil.rmtree('dist')
except WindowsError, info:
# assume directory does not exist
pass
# disable optimization- we _may_ need docs strings, specifically copyright
setup(
options = {py2exe: {
#includes: [decimal],
optimize: 1, ## 1 and NOT 2 because I use
the __doc__ string as the usage string. 2 optimises out the doc strings
'bundle_files': 1,
## options to reduce size of final exe
#~ 'ascii': True, # Exclude encodings
'excludes':[
'_ssl', # Exclude _ssl
'pyreadline', #'difflib',
'doctest', #'locale',
#'optparse',
'pickle', #'calendar',# Exclude
standard library
#'re',
],
}},
zipfile = None, ## try and make a single exe, if do not want this
loose this and the 'bundle_files' option
console=['postreview.py']
)
zipfilename='distribute_me.zip'
zipfilelist=['ingres_readme.txt', 'postreview.py', ] +
glob.glob('simplejson/*')+ glob.glob('dist/*')
import zipfile
z = zipfile.ZipFile(zipfilename, 'w')
for x in zipfilelist:
z.write(x)
z.close()
print 'Created:', zipfilename
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
Re: Facing problem installing in Windows!
sakthi v wrote: I followed the following link to install ReviewBoard in Windows http://yuanyi-wang.blogspot.com/2009/03/install-review-board-on-windows.html I tried installed twice.. and from blank page .. i got stuck in page showing Manual Server Updates Required I tried re-installation PIL 1.1.7, even Patch.exe. i can execute patch.exe from Cmd window. Can some one give me some pointer to resolve this.. It is urgent. If this genuinely is urgent (i.e. must fix, don't care how) the best advice I have is to set up a Ubuntu box (perhaps a virtual one in VMware) and install it there. The apache setup under Linux is pretty automatic and so much easier to deploy, no messing around with patch, etc. There are a few Windows users so hopefully you'll get some more windows focused advice if you are stuck with Windows. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: How to create svn diff for a new file
ms wrote: We use SVN for source control. I have a bunch of newly written code in new java files that are not yet committed in subversion. I want to review the code in these files using ReviewBoard, but I'm wondering how to create diff for these files since they do not have a prior version (these are brand new files). SVN won't let me create a diff for these files since there is no version of these files in SVN. I triesd to use GNU Diff to create a diff file (using a blank file as previous version file), but that doesn't load either in ReviewBoard. Any ideas on how I can review these newly created uncommitted files using ReviewBoard? From memory I'm sure postreview does this for you (it has been a while since I added a file in svn so I'm a little hazy). Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Custom authentication against external DB?
abrightwell wrote: I am considering integrating Review Board for my current project. However, before doing so, I need to determine whether or not it will place nicely with our current authentication process. All authentication for every tool on our project revolves around the authentication scheme for our XMPP server. The idea, essentially, is to have a single sign on for all of our tools. So, the question is this, is there a way to configure Review Board to use an external database (such as my XMPP server's database) for authentication? If so, where would I look to learn more about how to properly use this feature? This isn't something I've done but the mailing list has had a few discussions on this, related to LDAP. Basically this is a Django auth issue, try searching the list with those keywords. Also a Django resource would be a good place to check too (although I don't have any recommendations). Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Programmatic way to check review status?
Adam wrote: However what I'm not seeing is any scriptable way of checking the status of a review. Does something like this exist and I'm just not finding it? http://www.reviewboard.org/docs/manual/dev/webapi/ or you could hit the database directly (using raw SQL or the Django ORM). Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Successful upgrade from Reviewboard 0.9 to 1.0.5.1
Chris Clark wrote: Follow up after a days usage after the upgrade. Follow up a number of months later :-) It turns out I also hit a problem with missing screenshots, but no error for the user see: http://groups.google.com/group/reviewboard/browse_thread/thread/d158d97bf5344847/dc951bbe59c85974#dc951bbe59c85974 I.e. user uploads a screen shot image/picture BUT it shows up as a blank box in Reviewboard. The screenshots were in the old media directory (even recently uploaded ones). Simply updating the settings via the admin interface was all I needed to do. We didn't notice this for a long time as screen shots are rarely used (most of our reviews are of server processes with no UI). E.g. URL: http://DOMAIN.COM/admin/db/siteconfig/siteconfiguration/ then click on the site, then update the media entry (i.e. 'site_media_root'). The admin user may see errors in the apache log, related to missing dirs (and permissions), e.g.: grep -i exist /var/log/apache2/error.log # or where ever the log file is... [Tue May 25 02:21:15 2010] [error] [client xxx.xxx.xxx.xxx] File does not exist: /var/www/reviewboard/htdocs/media/uploaded/images/2010, referer: . Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: How apply patches in pre-review process?
Jan Koprowski wrote: . When review is done and submitted developer can change everything he want in working tree and commit everything he want. I'm thinking how be sure that reviewed changes is commited changes. For us, we are just using the honour/honor system, i.e. no checks :-) This mostly works BUT occasionally people do make mistakes (e.g. submit an extra file that was not in the review) so a check would be nice, for me personally I'm not sure it is worth spending the time on it. For a central based SCM (like svn) you could add a commit hook that pulls the diff from reviewboard and compare the diff with what the SCM has been given, as you suggested. I suspect this is the easiest thing to setup and would help catch accidents rather than deliberate deceptive developers ;-) The other idea similar to your manual reviewer idea would be for a script to pull the diff from RB (that is marked with Ship It!) and submit the change (assuming the patch applies cleanly). If patches don't apply cleanly the review could be marked with patch didn't apply cleanly to [rev in SCM] and the developer an resubmit. If you already have a buildbot system this is probably your best option (especially if you have tests as part of the build bot, you could add additional rejection checks). I've deliberately minimized the procedures/policies we have for RB at our site so this isn't something I plan on implementing myself but I can see the value. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: No new line at end of file crash diffviewer
Christian Hammond wrote: What version of patch are you guys running? I'm running Ubuntu 8.04.3 LTS - I have patch 2.5.9 ing...@reviews:~$ cat /etc/*elease* DISTRIB_ID=Ubuntu DISTRIB_RELEASE=8.04 DISTRIB_CODENAME=hardy DISTRIB_DESCRIPTION=Ubuntu 8.04.3 LTS ing...@reviews:~$ patch -v patch 2.5.9 Copyright (C) 1988 Larry Wall Copyright (C) 2003 Free Software Foundation, Inc. This program comes with NO WARRANTY, to the extent permitted by law. You may redistribute copies of this program under the terms of the GNU General Public License. For more information about these matters, see the file named COPYING. written by Larry Wall and Paul Eggert And do the diffs have any indication at the end saying that there's no newline at the end? Yes and no ;-) For me it depends on the diff tool used. E.g. some diff tools dump: \ No newline at end of file at the point the newline is missing to stdout, for a single file diff this is at the end. For multiple file diffs this could be in the middle of the diff. Some diff tools (e.g. Solaris) dump that information to stderr, if you redirect both stdout and stderr the position of that message can vary wildly. Some diff tools do nothing. Which means that multi-file diffs with the bad file in the middle look corrupted as the next file starts on the last line of the previous files diff. Also, what SCM? I'm using p2 which has the option to use the platform provided diff (so sort of a lottery on the newline detection) OR perform diffs server side (which gives consistent output). The default diff format that is used for p2 is the default diff output (i.e. not a unified diff). I'm not sure the SCM relevant, I think the issue is the diff and patch tools. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: No new line at end of file crash diffviewer
Jan Koprowski wrote: When I make review for file where some changes was made near last line and it have not new line at end of file diffviewer crush :/ I'm thinking about adding one more fake new line sign at the end everytime. Is this patch will be submitted if I create one? That is probably not going to work. We've seen a few of these too :-( It isn't reviewboard that is complaining, it is the patch tool. I've been suggesting to my users that they make a separate submission into SCM that simply adds a newline to the end of file of any files that have this problem without going through reviewboard. Then (re-)edit with the real change and the diff will be fine as you would expect. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Is it possible to specify a change number/revision for an SVN repository when creating a review request via json?
Stodge wrote: Thanks. I got confused between revisions and changesets, especially as I develop stuff using Mercurial, while the other developers use SVN! I'm curious why I can't create a review request via the json api with the revision numbers as parameters. I guess I assumed review board would then extract the diff from SVN. I'm trying to create an SVN post-commit hook that generates the review request. So what's the best/optimal way to do this? Should I spawn the diff tool and then hand off the resultant file to review board? Clarifying what I think you want to do; use postreview to take an existing (already submitted to svn) svn change number and post the code diffs up for review? I don't know if there are already some post commit hooks for svn around but here your idea is the way to go regardless of which SCM is used. To test it out try doing it manually first, i.e.: svn diff -rNEW:OLD my.patch where NEW and OLD are the revs to compare. There maybe a better way to do this with svn I'm no expert. then use my.patch with --diff-filename NOTE see http://reviews.reviewboard.org/r/1393/ for information on using --diff-filename with svn You could then include this as an option to postreview (I've added built in support for our SCM tool for already committed code as it makes seeing what was changed easier even if the code isn't being reviewed, it is also good for generating test data for postreview). Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Can I use anonymous mail ID as Mail Sender for for review requests and reviews ?
qhlonline wrote: Hi, all In the Admin page, There will be an Email setting. items will be as follows: Mail Server: Port: Username: Password: So there will be SMTPAuthentication. But we want to use this Email account to send email anonymously. if we had started the Send e-mails for review requests and reviews without set the password of this account, ReviewBoard will defaultly use the first superuser's mail account. We just want to send Email without authenticaition, Our mail server support anonymous. But why ReviewBoard can't ? If I want to send email anonymously, What should I do to change the Reviewboard source? Anon works for me with 1.0.5.1, where anon means no auth specified (just the server/port) and the email gets sent as the owner of the review. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: RB review Process Automation
One of the many things on my todo-but-never-get-to-it list is to add
close/submit support to postreview.
Roshan, if I was in your shoes I'd extend postreview to do what you want.
postreview already has the plumbing in place for auth and talking to the
RB server, admittedly postreview is really for posting code so this is
changing the behavior slightly. I think it does fit into what postreview
is used for, tying in working copies to reviews.
If you do this, I'd love to get a copy :-)
Chris
Christian Hammond wrote:
Are you storing and sending the cookie? We handle authentication
persistence by cookies, so you'd have to do this. If you're writing
this in Python, take a look at the CookieJar stuff in post-review.
Christian
--
Christian Hammond - chip...@chipx86.com mailto:chip...@chipx86.com
Review Board - http://www.reviewboard.org
VMware, Inc. - http://www.vmware.com
On Thu, Feb 4, 2010 at 11:21 PM, roshan pius
roshanpiustho...@gmail.com mailto:roshanpiustho...@gmail.com wrote:
Hi Christian,
I'm not able to close the review through HTTP POST, it kept
returning error to login, even though i had logged in just before
that.
I'm using a python script with simplejson to do this. I've tried
logging in as the user who submitted the review as well as the
admin. But i'm not able to close the review. Do i need to send any
parameters for this HTTP post ?
Request to server: http://IPADDRESS/api/json/accounts/login/
Response from server: {'stat': 'ok'}
Closing the review for Reviewrequest 604...
Close review URL: reviewrequests/604/close/submitted
Request to server:
http://IPADDRESS/api/json/reviewrequests/604/close/submitted/
Response from server: {'stat': 'fail', 'err': {'msg': 'You are not
logged in', 'code': 103}}
Roshan Pius
On Thu, Feb 4, 2010 at 7:37 AM, roshan pius
roshanpiustho...@gmail.com mailto:roshanpiustho...@gmail.com
wrote:
Hi Kunjal,
Perforce has the p4
trigger(http://www.perforce.com/perforce/doc.current/manuals/cmdref/triggers.html)
mechanism which can be used to run a script post-submission.
So if you want to close the review automatically you could run
a script, which would do the HTTP POST to
/api/json/reviewrequests/id/close/submitted/ as suggested by
Christian.
Also if you use change-commit or change-submit triggers
you would get the original changelist numbers itself, before
they are changed by Perforce during submission.
We currently use the change-submit trigger to check the
whether the changelist was shipped before it can be
submitted into Perforce.
Roshan Pius
On Thu, Feb 4, 2010 at 2:40 AM, Christian Hammond
chip...@chipx86.com mailto:chip...@chipx86.com wrote:
It is available today. You can do an HTTP POST to
/api/json/reviewrequests/id/close/submitted/
So Kunjal, basically you can write a post-commit hook that
automatically closes the review request. It will require
that there's some indication as to which review request
the change is associated with. In Perforce, the change
numbers will themselves change upon submit, but if you
knew the previous change number that was posted to Review
Board, you could use the
/api/json/reviewrequests/repository/repositoryid/changenum/changenum/
call to get the info for the review request, pull the ID
out of there, and then use that to call the
/close/submitted/ call above.
Christian
--
Christian Hammond - chip...@chipx86.com
mailto:chip...@chipx86.com
Review Board - http://www.reviewboard.org
VMware, Inc. - http://www.vmware.com
On Wed, Feb 3, 2010 at 11:00 AM, Chris Clark
chris.cl...@ingres.com mailto:chris.cl...@ingres.com
wrote:
Kunjal wrote:
Once the Review is done, user has to make Review
as Close-Submitted
eles reviewers dashboard will still have the review.
We enforce this process on developers and
developers are saying that
this is one extra step.
I was thinking whether it is possible to close the
review
automatically?
For example, for given Review, if the shipit_count
is more then zero
then review is approved and we can close it.
Is there any way to do it programatically?
IMHO having shiptit mark as submitted would not be
useful. Where we have RB reployed Submitted means
Re: RB review Process Automation
Kunjal wrote: Once the Review is done, user has to make Review as Close-Submitted eles reviewers dashboard will still have the review. We enforce this process on developers and developers are saying that this is one extra step. I was thinking whether it is possible to close the review automatically? For example, for given Review, if the shipit_count is more then zero then review is approved and we can close it. Is there any way to do it programatically? IMHO having shiptit mark as submitted would not be useful. Where we have RB reployed Submitted means the code was submitted into source code control (which is a semi-manual process). What we'd find useful would be a json call (possibly called from postreview) to mark the review as submitted (if this is already present please let me know). Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Does reviewboard support new review notifications other than email?
Tino Breddin wrote: I'm interested in notifying external system about new review requests where email isn't a good option. I've read in the roadmap that generic hooks are planed for version 2.0, but is there anything similar in place now which could be used for this purpose? This suggestion is likely to make you cry you could implement an smtplib module that appears to have the same API as the one that ships with Python and put that in the reviewboard/django path first. You could then do other stuff. Essentially monkey patching a module by injecting a SMTP class that doesn't do smtp but offers the same API, you would probably only need to implement sendmail() and pass for all the other methods. Then you could post to IRC, Jabber, etc. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Is there a size limit on diffs?
Dave Hocker wrote: We have a large Perforce change list that generates a 550K diff file. When we submit the review request, the request is created but there is no diff attached. .snip logs. Is there a limit on diff file size? If so, can it be changed? I can't comment on the Perforce support but I recently reviewed a 1Mb (1,096,113 bytes) diff with Reviewboard without too much problem. I'd be surprised if there is a diff limit HOWEVER the diff is stored in the database, depending on your database backend there is a chance you have a limit in there (this has been discussed on this list with reference to comments rather than diff sizes). You could try using the management shell to check, see http://reviews.reviewboard.org/r/598/diff/ (near the end) for a code snippet of how to pull out the diff attachment. Sage (UK) Limited. Registered in England at North Park, Newcastle upon Tyne, NE13 9AA. Registered number 1045967 Small world, way back when I used to work on a Point of Sale application that used Sage for stock levels tracking (MS DOS days) :-) Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Question on code review for files after they've been modified by another persons checkin
seacgmk wrote: I've recently set up ReviewBoard with Perforce configured for pre- commits. Perforce allows source files to be checked out by multiple developers at the same time. If one developer submits a file for review and then a second developer checks in their changes to the same file before the code reviewer has had a chance to look at the first developers changes, the diff will be out of date. The diff stored in ReviewBoard uses the rev of the source code in Perforce at the time the review was submitted. Is there any workaround for this situation? Is there any way for ReviewBoard to at least flag the reviewer that they're not reviewing the latest revision of the source code from source control. Or maybe have ReviewBoard notify the developer that they need to resolve changes and re-submit the file for review? Not at the moment. If it were me I would add this to postreview rather than Reviewboard. I've not used p4 but I've used p2; p2 has a need command to check for new files, I suspect it would probably be fairly easy to do and quick. Having Reviewboard check that the version being reviewed is the latest is certainly something that would be nice to have but it would need to be generic and would probably need a lot of thought to cover all the different SCMs (it may be more appropriate do this via the plug-in architecture that Christian has mentioned will be available for later releases of RB). Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: post-review failing to work on CVS because I have cleartool with a view set
Christian Hammond wrote: What you could do now is modify postreview.py so that the CVSClient instance is created before the ClearCaseClient. That would guarantee that CVS would check first. An easy technique for dealing with this is to add another definition for SCMCLIENTS, this way if the SCMCLIENTS is ever changed upstream it is still possible to automatically merge your diffs without a manual conflict resolution, e.g.immediate after SCMCLIENTS: SCMCLIENTS = ( SVNClient(), CVSClient(), GitClient(), MercurialClient(), PerforceClient(), ClearCaseClient(), ) ### re-define SCMCLIENTS, this makes merging changes easier (than customizing SCMCLIENTS) :-) SCMCLIENTS = ( SVNClient(), PiccoloClient(), ) Sort of brute force but it works :-) Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: post-review failing to work on CVS because I have cleartool with a view set
David McCloskey wrote: But Another problem is appearing now. Once it tries to do the post, I get an error that the connection was refused on port 2401. I'm guessing this is because we're using ssh tunnels only to connect to our CVS servers. Is there some special way to configure my repository through the web interface to make it work through the ssh tunnel? Or to configure my environment somehow? I don't know very much about CVS unfortunately. I'm doing this to help a colleague. I'm not a CVS user either. However you may find you need the latest RB as ssh implies authentication, authentication isn't support with RB 1.0. If you have the new version the config for auth should be obvious in the web admin - again should be I'm not using either CVS or any auth options for the SCM with ReviewBoard. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: post-review doesn't find my CVS repository
The other option is to add a command line override for the SCM
repository information.
I ended up doing this for the SCM we use . although I've not yet
needed to make use of it :-) I think this is what the --p4-port flag is
for with Perforce.
You could also have some sort of mapping in the .reviewboardrc file
(again you would need to implement this logic). These would probably be
good contributions back for post-review if you do work on this.
Chris
qhlonline wrote:
Hi,sir.
I have met this problem before. post-review's CVSClient now get the
host name of CVS Server through socket.getfqdn(host), to replace
host part of your CVSPath string. But ReviewBoard server seems not
support it. So you can edit post-review yourself. You should find the
CVSClient part and function get_repository_info(), you move the last
sentence of this function after fp.close() and ignore following
source code, That may be fine. You can have a try.
Best Regards!
??2010-01-06 13:19:31??heffe ghe...@gmail.com ??
I have been using Review Board for a 3-4 months using the Web UI to
submit review requests. I'm now trying to make use of post-review but
I am unable to get it to work with my CVS repository configuration. I
have multiple repositories configured on my Review Board server, some
SVN, some CVS. One of the SVN repositories I have tested works with
post-review, but CVS does not.
My Review Board server configuration has a CVS repository configured
with:
:pserver:u...@cvsserver.domain.com:/reporoot
(my CVS login doesn't need a password)
When I try to use post-review, I see
$ post-review -d
svn info
repository info: Path: cvsserver.domain.com:/reporoot, Base
path: None, Supports changesets: False
cvs diff -uN
Looking for 'reviewboard.mydomain.com /' cookie in /home/
heffe/.post-review-cookies.txt
Loaded valid cookie -- no login required
Attempting to create review request for None
HTTP POSTing to
http://reviewboard.mydomain.com/api/json/reviewrequests/new/:
{'repository_path': 'cvsserver.domain.com:/reporoot'}
Error creating review request: The repository path specified is not
in the list of known repositories (code 206)
I've been digging for hours to try and get around this but I can't
figure it out. I tried removing the username from the Review Board
repository configuration and then adding it to the username field on
the repository editor page, but that fails completely to talk to my
CVS repository.
Any thoughts? Why does post-review not identify the CVS repository
configuration I have in my Review Board server?
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
Re: Totals / Reports?
grimbeaver wrote: I need a simple way to get the total number of reviews in a group. We've assigned all the code reviews into groups based on project. Project management has to report the total number of code reviews completed for a project to our partner. When you look at the group view it only gives you a count of the pending reviews, no total count. Obviously in an ideal world there will be better reports some day, but for now I just need an easy way to get this number when needed. Off the top of my head there are 2 options: 1. use the Django manage shell and write simple Python queries 2. access the database directly, e.g. using native sql terminal monitor tools and query the table(s) directly I don't have queries for these but it probably wouldn't take long by poking around. The Django tutorials are pretty good if you want to try option #1 which requires basic Python skills. Option #2 requires basic SQL skills. If you create useful scripts they would be good to post back to the mailing list. Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: About Post-Review User change problem
qhlonline wrote: 在2009-12-23，Chris Clark chris.cl...@ingres.com 写道： Try the --submit-as flag instead. e.g.: post-review --submit-as=A myfile post-review --submit-as=B myfile But We need that the loged user and the submiter to be the same user. We are desiring detaied permission control on ReviewBoard, So that if the loged user has no derect relation with one ReviewRequest, He will never see it. So we should make judge whether the loged on user is the submitter or reviewer of this reviewRequest. And use --submit-as will make the judgement complex. Is there any other way to fix this bug? I don't completely understand the requirement as I thought a regular user can not impersonate another user (i.e. --submit-as is likely to require the authenticated user to be the same as the submitter). As to fixing the bug, IMHO this is a feature request rather than a bug. I know the Reviewboard team would like to have finer granularity of control as Christian as mentioned this a few times in this group but this is targeted at a later release. If this is important to you I would suggest starting to develop it yourself. One of the primary reasons I started using ReviewBoard was that the code was open and easy to grok. Adding finer control isn't likely to be quick and easy but you may be able to hack in some of your requirements in a (private) fork of the project. The web site does go into some detail on starting a private dev environment: http://www.reviewboard.org/docs/codebase/dev/ however you would need to have some Python (, GIT, and Django) skills. If one is using a code review system one probably already has some programming skills so this is hopefully a case of picking up new skills rather than having to learn to program from scratch. There is also a devs group that may be worth checking out, see http://www.reviewboard.org/mailing-lists/ Good luck! Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: post-review and
prestomation wrote: I am currently demo'ing review board to interface with our Perforce server. I seems that using post-review with a command like the following: post-review --username=USERNAME --server=reviewboard 65428 tries to access the api at a subfolder with the same name as the servers name itself, so in this case I get this... Unable to access http://reviewboard/reviewboard/api/json/accounts/login/. The host path may be invalid ...when it should be accessing http://reviewboard/api/json/accounts/login, right? Once I used rb-site to create a site with the same subfolder name, then it worked fine, but we don't necessarily want to do this. Is this a bug or am I missing something? Try specifying the full address/url. e.g: post-review --server=http://reviewboard.myinternaldomain.prv Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: About Post-Review User change problem
Try the --submit-as flag instead. e.g.: post-review --submit-as=A myfile post-review --submit-as=B myfile Christian Hammond wrote: Known bug. See http://code.google.com/p/reviewboard/issues/detail?id=762 Christian -- Christian Hammond - chip...@chipx86.com mailto:chip...@chipx86.com Review Board - http://www.reviewboard.org VMware, Inc. - http://www.vmware.com 2009/12/16 qhlonline qhlonl...@163.com mailto:qhlonl...@163.com Hi， I used post-review to create review request. in the form as: post-review --username=A --password=Apassword --target-person=C myfile; But when I want to post reviewrequest as another user, like B, I type: post-review --username=B --password=Bpassword --target-person=C myfile; This time , When I logged in the WebUI to publish my reviewrequest. I found that actually post-review didn't create reviewrequest as user B ,But create as user A again! I think it is because my first using user A account has created a cookie for it, and the second using of post-review is not authenticate B user at all. It will be nothing for practical usage because each RB client may be using a permanent account, But as a tester, He may need to log in the same machine with different account. So I think it is a problem then. Best Regards! -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com mailto:reviewboard%2bunsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: post-review with bypass proxy settings in Internet Explorer
My 2 cents.
Modifying the registry and then restoring is not a great idea. I can see
why you are doing it but I'd encourage you to NOT do this. There is a
potential here for a background web app to fail (e.g. web browser based
IM tool).
I'd be tempted to monkey patch urllib(2), presumably that is the beast
causing the issue here if you want a quick solution. Either monkey patch
urllib OR monkey patch _winreg lookup that urllib relies on.
As the real Chris said, improving postreview and getting the code into
the main git repo is the ideal solution here. You may be the first to
hit this but you are not going to be the last :-) so it it would be good
to get a more robust solution.
Chris
Christian Hammond wrote:
Yeah, modifying postreview.py isn't ideal. It was just one suggestion.
What I'd like to see, I think, is an optional configuration variable
in the user's .reviewboardrc that allows users to control
post-review's proxy settings manually.
Christian
--
Christian Hammond - chip...@chipx86.com mailto:chip...@chipx86.com
Review Board - http://www.reviewboard.org
VMware, Inc. - http://www.vmware.com
On Wed, Dec 2, 2009 at 11:42 AM, Akhilesh akhileshjo...@gmail.com
mailto:akhileshjo...@gmail.com wrote:
thanks Christian for suggestion. I thought if I modify postreview.py
then it will be difficult for me to roll out official updates to
individual users.
I did a workaround. Wrote following wrapper script to disable/enable
proxy. It's dirty but serves my purpose:
import _winreg, os, sys
if len(sys.argv) != 2:
print Usage: rr ChangeListId
sys.exit(0)
root = _winreg.HKEY_CURRENT_USER
keypath = Software\Microsoft\Windows\CurrentVersion\Internet
Settings
value_name = ProxyEnable
hKey = _winreg.OpenKey (root, keypath, 0, _winreg.KEY_READ |
_winreg.KEY_SET_VALUE)
existingValue, type = _winreg.QueryValueEx (hKey, value_name)
#Disable Proxy setting if ON.
if existingValue == 1:
_winreg.SetValueEx (hKey, value_name, 0, _winreg.REG_DWORD, 0)
# Call post-review.exe
command = post-review + sys.argv[1]
os.system(command)
# If original setting was PROXY ON then turn it ON.
if existingValue == 1:
_winreg.SetValueEx (hKey, value_name, 0, _winreg.REG_DWORD, 1)
_winreg.CloseKey(hKey)
On Nov 24, 4:45 pm, Christian Hammond chip...@chipx86.com
mailto:chip...@chipx86.com wrote:
Is the Review Board server on HTTP or HTTPS?
It sounds like custom code would need to be added to
specifically disable
the proxy server. It's also possible that a 2.6 release would
fix this
issue, but I don't know.
If you were to modify postreview.py , you could try adding some
code like:
proxy_support = urllib2.ProxyHandler({})
opener = urllib2.build_opener(proxy_support)
urllib2.install_opener(opener)
in ReviewBoardServer.__init__ probably at the end.
Christian
--
Christian Hammond - chip...@chipx86.com mailto:chip...@chipx86.com
Review Board -http://www.reviewboard.org
VMware, Inc. -http://www.vmware.com
On Tue, Nov 24, 2009 at 4:30 PM, Akhilesh
akhileshjo...@gmail.com mailto:akhileshjo...@gmail.com wrote:
Thanks Christian. I tried setting http_proxy environment
variable to
empty string but Windows wouldn't allow me. I set it to (with
space) - but as expected it didn't work.
Is there any work-around? The problem is that one of the sites
in our
organization uses proxy and all developers from that site have to
toggle Proxy settings to access Reviewboard/Internet as
by-pass proxy
option is not working for them.
I appreciate any help. I'm ready to test private code for
post-review
as well if needed.
Regards,
Akhilesh
On Nov 24, 1:57 pm, Christian Hammond chip...@chipx86.com
mailto:chip...@chipx86.com wrote:
Sadly, this is due to Python's usage of the system proxy
settings. I'd
have
to see if there's anything we can do for this. On Linux, I
know you can
set
the HTTP_PROXY variable to an empty string to work around
it, but I doubt
that works on Windows...
Christian
--
Christian Hammond - chip...@chipx86.com
mailto:chip...@chipx86.com
Review Board -http://www.reviewboard.org
VMware, Inc. -http://www.vmware.com
On Tue, Nov 24, 2009 at 8:16 AM, Akhilesh
akhileshjo...@gmail.com mailto:akhileshjo...@gmail.com
wrote:
We are facing a problem where if proxy settings are enabled in
Internet Explorer, post-review doesn't work (Proxy Error
502) but I we
disabled proxy settings, it works perfectly. so we added
Re: ValueError: No JSON object could be decoded
Have you tried using the:
--server=
flag?
RE the version of reviewboard, if this is a new install I would go ahead
and use 1.0.5.1 this shouldn't impact the error you are seeing but why
use an old version when you can use the new stable one ;-)
Chris
Christian Hammond wrote:
Hi,
Is it really reviewboard.com http://reviewboard.com that you're
trying to access? Because we don't own that domain :) What's
specifying that domain?
Christian
--
Christian Hammond - chip...@chipx86.com mailto:chip...@chipx86.com
Review Board - http://www.reviewboard.org
VMware, Inc. - http://www.vmware.com
On Wed, Dec 2, 2009 at 4:10 AM, Mahesh vsmait...@gmail.com
mailto:vsmait...@gmail.com wrote:
Hello brothers,
I have installed reviewboard 1.0.1 on a linux(fedora 8) m/c.
Everything is working fine in the GUI.
But when I try to use the post-review he's giving me some strange
error or rather strange to me!!
[r...@testvm cli]# /home/mahesh/lib/post-review
== Review Board Login Required
Enter username and password for Review Board at
http://reviewboard.com/
Username: mahesh
Password:
Traceback (most recent call last):
File /home/mahesh/lib/post-review, line 8, in module
load_entry_point('RBTools==0.2beta2', 'console_scripts', 'post-
review')()
File /home/mahesh/lib/RBTools-0.2beta2-py2.6.egg/rbtools/
postreview.py, line 2518, in main
File /home/mahesh/lib/RBTools-0.2beta2-py2.6.egg/rbtools/
postreview.py, line 314, in login
File /home/mahesh/lib/RBTools-0.2beta2-py2.6.egg/rbtools/
postreview.py, line 575, in api_post
File /home/mahesh/lib/RBTools-0.2beta2-py2.6.egg/rbtools/
postreview.py, line 486, in process_json
File /usr/lib/python2.6/json/__init__.py, line 307, in loads
return _default_decoder.decode(s)
File /usr/lib/python2.6/json/decoder.py, line 319, in decode
obj, end = self.raw_decode(s, idx=_w(s, 0).end())
File /usr/lib/python2.6/json/decoder.py, line 338, in raw_decode
raise ValueError(No JSON object could be decoded)
ValueError: No JSON object could be decoded
Can anyone pls help me to get out of this??
Don feel bad if it is a very basic one!!
Thanks,
Mahesh
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
mailto:reviewboard%2bunsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
Re: fields_changed in changedescs_changedescription
Out of curiosity what size limit is this likely to be? 80 chars, 1024,
1Mb, 1Gb... :-)
Chris
Christian Hammond wrote:
Hi Ronak,
We will be limiting the lengths of these fields in a future version.
We're aware of the problem and have encountered this with some users
too. Expect a field restriction in 1.0.6.
Christian
--
Christian Hammond - chip...@chipx86.com mailto:chip...@chipx86.com
Review Board - http://www.reviewboard.org
VMware, Inc. - http://www.vmware.com
On Fri, Nov 20, 2009 at 12:26 PM, Ronak ronakdotpa...@gmail.com
mailto:ronakdotpa...@gmail.com wrote:
Hi,
I get a trackback that looks something like this
Traceback (most recent call last):
File /usr/local/lib/python2.5/site-packages/Django-1.0.2_final-
py2.5.egg/django/core/handlers/base.py, line 86, in get_response
response = callback(request, *callback_args, **callback_kwargs)
File /usr/local/lib/python2.5/site-packages/ReviewBoard-1.0-
py2.5.egg/reviewboard/accounts/decorators.py, line 27, in _check
return view_func(*args, **kwargs)
File /usr/local/lib/python2.5/site-packages/ReviewBoard-1.0-
py2.5.egg/reviewboard/reviews/views.py, line 174, in review_detail
for changedesc in changedescs:
File /usr/local/lib/python2.5/site-packages/Django-1.0.2_final-
py2.5.egg/django/db/models/query.py, line 185, in _result_iter
self._fill_cache()
File /usr/local/lib/python2.5/site-packages/Django-1.0.2_final-
py2.5.egg/django/db/models/query.py, line 618, in _fill_cache
self._result_cache.append(self._iter.next())
File /usr/local/lib/python2.5/site-packages/Django-1.0.2_final-
py2.5.egg/django/db/models/query.py, line 280, in iterator
obj = self.model(*row[index_start:])
File /usr/local/lib/python2.5/site-packages/Django-1.0.2_final-
py2.5.egg/django/db/models/base.py, line 266, in __init__
signals.post_init.send(sender=self.__class__, instance=self)
File /usr/local/lib/python2.5/site-packages/Django-1.0.2_final-
py2.5.egg/django/dispatch/dispatcher.py, line 148, in send
response = receiver(signal=self, sender=sender, **named)
File /usr/local/lib/python2.5/site-packages/Djblets-0.5-py2.5.egg/
djblets/util/fields.py, line 174, in post_init
value = self.loads(value)
File /usr/local/lib/python2.5/site-packages/Djblets-0.5-py2.5.egg/
djblets/util/fields.py, line 206, in loads
val = eval(val)
File string, line 1
{testing_done:
09 11.17.09 !F
^
SyntaxError: EOL while scanning single-quoted string
After little bit of debugging I found out that if a user puts in a
long very long description in the Testing Done field and again goes
and changes that field the old and new comments are stored in
fields_changed columns of changedescs_changedescription but the text
is incomplete. Is there a way to disable the delta tracking for
Testing Done and Description fields or is there a way to inform
the user that his/her input is too long.
Thanks,
Ronak
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
mailto:reviewboard%2bunsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
--
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
Re: fields_changed in changedescs_changedescription
Christian Hammond wrote: I don't know off-hand what we're going to set it to. We're thinking about it, and suggestions are welcome. Here's where it gets tricky... We're running up to maximum field lengths for the change info, and that's a combination of multiple fields, including Description and Testing Done. So we need to make sure that we can stuff the max of both those fields, plus other change info (screenshots added, new diffs, etc.). The two possible solutions are to either 1) Make both Description and Testing Done have a max of just under half the maximum field size for the database with the smallest field restriction, or 2) Cut off the text before putting it in the change description. Maybe a combination of those things, in case you have both Description and Testing Done at max and have added a bunch of screenshots and changed other fields. Yikes more complicated than I thought :-( I guess in Reviewboard 3000 the model could be changed to deal with this :-) For a naive solution how I about polling this mailing list for user input, and provide some Django manage shell scripts to get the max (and maybe mean, mod, median) lengths for Description and Testing fields? For example without checking, I'd guess that most of the reviews in my system have larger Description text than Testing done (most testing is regression test and/or a new regression test addition so the text here tends to be short). So armed with proof I might ask for say, 80/20 ratio for Description /Testing length. This way you get some real world numbers on lengths before you make the call. I'm lucky I've not seen this limit yet! Chris -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
Re: Review Board 1.1 alpha 2 released
Christian Hammond wrote: We've just put out a release of Review Board 1.1 alpha 2. This is a major enhancement release, and we'd love to get some testing for it. It looks like 1.1 alpha 2 is at http://reviews.reviewboard.org. I've a question on bug or behavior, someone posted a review url recently: http://reviews.reviewboard.org/r/967/ Which has screen shots, in older releases screen shots had a thumbnail view on the main review page. I don't see thumbnails on http://reviews.reviewboard.org/r/967/ (FireFox 3.0.x, I also checked IE too) but clicking the screenshot links works fine. I don't have a strong preference (inclined to retain existing behavior) but it would be good to know if this is a bug or a behavior change. Chris --~--~-~--~~~---~--~~ Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Anyone seeing multiple emails from ReviewBoard
I've a report from _some_ users that they are seeing multiple emails being sent from Reviewboard. The scenario is this: 1. postreview used to create/publish new review 2. review comments posted by reviewer, this code sucks change it... etc 3. postreview used to post new diffs, post review _may_ be passing in the same description and testing content 4. ? 5. Review is published 6. Reviewboard mails out original diffs but marks the Change Description and Testing sections are marked as (updated) (although the section text is the same in the email) 7. Reviewboard mails new diffs I'm a little hazy on steps 3 and 4, I believe step 5 is via the web site. I've implemented a debug version of postreview to get to the bottom of exactly what is done client side (essentially a hard coded DEBUG=True). At this stage I'm inclined to believe this is NOT a Reviewboard issue but I wanted to see if anyone else was seeing this intermittently? Thanks! Chris --~--~-~--~~~---~--~~ Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Customize ReviewBoard for Email Subject
Kunjal wrote: we want to customize the subject line of the email which RB sends on submit of review. Check out reviewboard/reviews/email.py - mail_review_request() look for subject. I nice to have would be a way to customize this from the site admin page. I'm not likely to implement this myself though :-( Chris --~--~-~--~~~---~--~~ Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: How to post-commit review a first time checkin?
Christian Hammond wrote: Actually, I wouldn't mind such a parameter. (though, --diff-filename, as we don't use underscores in parameter names). There's actually a patch up for review for accepting via stdin (though it's a bit stale and isn't the design we want), so I guess there are more people who want this. If you'd like to contribute it, we'd take it, but I'd ask that you make the diff uploading part generic and not part of any SCMClient subclass, and that it allow - as a value, at which point we'd read from stdin. We should be able to just bypass the diff generation part of the SCMClient and instead read the data in ourselves and use it. The tricky thing is that one has to know which scm to use with the diff. By default a process is spawned to check if (say) svn is the scm in use in this directory. Assuming no one else wants this capability (avoiding processing spawning) I can contribute something ;-) http://reviews.reviewboard.org/r/1197/ NOTE this does not fulfill the read from stdin requirement, I do not have the time to test this out (e.g. does reading the password work if a prompt starts) so I've not attempted to implement this feature but someone else is welcome to make a stab at it. Do you have the original review for read from stdin? Maybe they would be interested in completing this. Chris --~--~-~--~~~---~--~~ Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Successful upgrade from Reviewboard 0.9 to 1.0.5.1
Just a quick note that I was able to upgrade from Reviewboard 0.9 to 1.0.5.1. Yes zero point nine to latest (stable) 0.9 pre-dates the rb-site so I simply had to: 1. Use rb-site to deploy a NEW EMPTY reviewboard server (yup, with a new secret key, I didn't have to mess with it at all) 2. Copy over my screenshots by hand 3. Update the settings file (and/or) copy across the old database to the new location. The database is out of date at this point so... 4. issue: sudo rb-site manage /var/www/reviewboard evolve -- --execute 5. issue: sudo rb-site manage /var/www/reviewboard syncdb 6. Change ownership of files (e.g. screenshots), issue: sudo chown -R www-data /var/www/reviewboard Obviuosly /var/www/reviewboard and the user name are specific to my site. I'm using Ubuntu LTS 8.04: ing...@reviews:~$ uname -a Linux reviews 2.6.24-24-server #1 SMP Fri Sep 18 17:24:10 UTC 2009 i686 GNU/Linux ing...@reviews:~$ cat /etc/*elease* DISTRIB_ID=Ubuntu DISTRIB_RELEASE=8.04 DISTRIB_CODENAME=hardy DISTRIB_DESCRIPTION=Ubuntu 8.04.3 LTS I actually tested out the upgrade earlier in the week on a copy of the live system so I knew what I was in for. Total down time on live was 50 mins (which included an upgrade of the bins for the OS via sudo apt-get update ; sudo apt-get upgrade). The last time I did an upgrade, evolve did not work (again I was using a pre 1.0 release) and I had to hack some SQL together to upgrade the database. So this upgrade was a real pleasure, rb-site is awesome! Not sure if this will be useful to everyone but thought I'd share the information just in case. I'll find out tomorrow what my users see :-) Thanks again to Chris, David, and everyone else for making ReviewBoard available. Chris --~--~-~--~~~---~--~~ Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Successful upgrade from Reviewboard 0.9 to 1.0.5.1
On Oct 29, 2:22 am, Christian Hammond chip...@chipx86.com wrote: I should point out for anyone else reading the notes that you were able to get away with not needing the old secret key because you're (I'm assuming) using something like LDAP or NIS or Active Directory. In these cases, we don't store an encrypted password in the database, so it should be fine. However, if you have *any* built-in auth accounts (like an initially created admin user), you'll probably want to use your old secret key. Or just use it anyway, it should be safe enough. We're currently using built-in auth (we're hoping to move to LDAP/AD soon as we already have AD setup for other stuff). Weird... I know! Thought it was worth mentioning as it (potentially) saves a step. I didn't purge cookies, I did have to login using the regular login form so I think the auth really is working (rather than cookie caching). I'm guessing the per-password salt is stored in the database along with the HMAC rather than using the site setting shared key (but this is just a wild guess). I should point out that I'm not storing auth for SCM's (all the SCM's we use do NOT use auth) which maybe where the shared key may be used. If I get burned by this later with other security stuff I'll report back :-) Chris --~--~-~--~~~---~--~~ Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
git and diffs against branches
Whilst I use other distributed SCM's I've only just started with git (as
ReviewBoard uses it) and I think I'm having some trouble understanding
git (and reviewboard).
What I'd like to do is post for review changes/diffs against a branch
(not master/bleeding edge).
E.g.
git clone git://github.com/reviewboard/reviewboard.git
git checkout release-1.0.5.1
## hack on stuff, e.g. git add myfile
postreview.py --server=http://reviews.reviewboard.org/ --debug
Here is my output (note headrevs of rbtools and this is actually posted
explictly with -r1177):
svn info
git rev-parse --git-dir
git svn info
git svn --version
git config --get svn-remote.svn.url
git remote show origin
repository info: Path:
git://github.com/reviewboard/reviewboard.git, Base path: , Supports
changesets: False
git diff --no-color --full-index master
Looking for 'reviews.reviewboard.org /' cookie in
/home/ingres/.post-review-cookies.txt
Loaded valid cookie -- no login required
HTTP GETting api/json/reviewrequests/1177/
Uploading diff, size: 467617
HTTP POSTing to
http://reviews.reviewboard.org/api/json/reviewrequests/1177/diff/new/: {}
Error uploading diff: One or more fields had errors (105)
{u'fields': {u'path': [u'int argument required']}, u'stat':
u'fail', u'err': {u'msg': u'One or more fields had errors', u'code': 105}}
Your review request still exists, but the diff is not attached.
From the instructions near the end of
http://www.reviewboard.org/docs/codebase/dev/getting-started/#gettingstarted
(Posting Changes for Review) I get the impression this is not allowed?
The checkout is based on master but it is a different branch,
I guess this maybe a bug report too as it fails to post but does create
an empty review without printing out the review number. I ended up
browsing http://reviews.reviewboard.org/r/ and looking at the top to
work out it had created one. Similar behavior if I just do a git diff
and try and use the web interface (i.e. get int arg required error).
Help!
Chris
--~--~-~--~~~---~--~~
Want to help the Review Board project? Donate today at
http://www.reviewboard.org/donate/
Happy user? Let us know at
http://www.reviewboard.org/users/
-~--~~~~--~~--~--~---
To unsubscribe from this group, send email to
reviewboard+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
-~--~~~~--~~--~--~---
Re: Can't enable search though PyLucene is installed
David, Which Ubuntu version did you use? I've not (yet) setup search but I've found the LTS (8.10) Server version to be REALLY good for reliable config. If you need a reliable, *production* server I wouldn't use anything other than LTS. I can't promise LTS will handle Lucene but it has the best chance. Chris David Resnick wrote: Understood. I'll see about filing that bug. Thanks all for your help! -David On Wed, Oct 21, 2009 at 9:33 PM, Christian Hammond chip...@chipx86.com mailto:chip...@chipx86.com wrote: We don't have the time to check every version of every distro to find out these things, so we basically learn about it when someone tells us. I would file a bug against the Ubuntu people on this. Christian -- Christian Hammond - chip...@chipx86.com mailto:chip...@chipx86.com Review Board - http://www.reviewboard.org VMware, Inc. - http://www.vmware.com On Wed, Oct 21, 2009 at 6:16 AM, David Resnick abune...@gmail.com mailto:abune...@gmail.com wrote: Thanks for the info, Tom. This is kind of ironic, considering that the main reason I choose Ubuntu was because I didn't want to manually build PyLucene :-). Probably the docs should mention which version of Ubuntu will make the setup least painful. -David On Wed, Oct 21, 2009 at 2:23 PM, grimbeaver tbrez...@gmail.com mailto:tbrez...@gmail.com wrote: I ran into this same problem. Ubuntu 9.04 has a package for PyLucene but it is built for Python 2.5. Apache uses libapache2-mod-python built for Python 2.6 so no changes to the path or default python are going to make any difference. The only way I got it to work was to manually build PyLucene instead of using the package. Hope this helps. -Tom On Oct 21, 2:34 am, David Resnick abune...@gmail.com mailto:abune...@gmail.com wrote: Do you have any idea how I can reinstall PyLucene with 2.6? I'm running a vanilla installation of Ubuntu 9.04. There's nothing in the browser JavaScript log. I'm running Firefox 3.5.3. Thanks, David On Tue, Oct 20, 2009 at 8:52 PM, Christian Hammond chip...@chipx86.com mailto:chip...@chipx86.comwrote: Switching Python paths will certainly break things if you're switching to a different version of Python. If you're using mod_python with 2.6 and using the 2.5 search path, you'll experience a number of issues. Instead, you should reinstall PyLucene with 2.6. Did you see anything in your browser's JavaScript debug log? And which browser are you using? Christian On Tuesday, October 20, 2009, David Resnick abune...@gmail.com mailto:abune...@gmail.com wrote: I got an ImportError when I tried doing imp.find_module(lucene). I saw that lucene is installed in python2.5 and the default python used was 2.6. So I changed the default python to 2.5 and now the import does work. But after restarting Apache the checkbox is still disabled. How does mod_python choose the installation of python to use? -David On Tue, Oct 20, 2009 at 11:55 AM, Christian Hammond chip...@chipx86.com mailto:chip...@chipx86.com wrote: Also, can you manually run Python and type: import imp imp.find_module(lucene) If that throws an exception (an ImportError), then PyLucene isn't set up properly. If it simply returns without any message, then it's installed fine and Review Board is seeing it, so something else is going wrong. Christian -- Christian Hammond - chip...@chipx86.com mailto:chip...@chipx86.com Review Board -http://www.reviewboard.org VMware, Inc. -http://www.vmware.com On Tue, Oct 20, 2009 at 2:08 AM, David Resnick abune...@gmail.com mailto:abune...@gmail.com wrote: Yes. -David On Tue, Oct 20, 2009 at 10:46 AM,
dev getting-started - missing django_evolution step
I'm setting up a new dev environment for some hacking and I'm following: http://www.reviewboard.org/docs/codebase/dev/getting-started/#gettingstarted After using git to get the reviewboard source and using ./contrib/internal/prepare-dev.py I get errors about django_evolution. django_evolution is a pre-requisite that is not handled when djblets (which auto installs almost everything else) is installed/setup. Here is what I did to deal with that: svn checkout http://django-evolution.googlecode.com/svn/trunk/ django-evolution cd django-evolution sed 's/distutils.core/setuptools/' setup.py setuptools_setup.py python setuptools_setup.py develop cd .. Could http://www.reviewboard.org/docs/codebase/dev/getting-started/ be updated with something like the above, just after the djblets section before the Review Board section? Thanks! Chris --~--~-~--~~~---~--~~ Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: dev getting-started - missing django_evolution step
Chris Clark wrote: I'm setting up a new dev environment for some hacking and I'm following: http://www.reviewboard.org/docs/codebase/dev/getting-started/#gettingstarted After using git to get the reviewboard source and using ./contrib/internal/prepare-dev.py I get errors about django_evolution. django_evolution is a pre-requisite that is not handled when djblets (which auto installs almost everything else) is installed/setup. Here is what I did to deal with that: svn checkout http://django-evolution.googlecode.com/svn/trunk/ django-evolution cd django-evolution sed 's/distutils.core/setuptools/' setup.py setuptools_setup.py python setuptools_setup.py develop cd .. Could http://www.reviewboard.org/docs/codebase/dev/getting-started/ be updated with something like the above, just after the djblets section before the Review Board section? Also I needed paramiko and pytz, e.g.: sudo apt-get install python-paramiko python-tz I had an issue where the default setup from ./contrib/internal/prepare-dev.py did NOT work. Specifically it appeared to work but when running the server (e.g. via ./contrib/internal/devserver.sh) a manually syncdb run was required. So I made this change: diff --git a/contrib/internal/prepare-dev.py b/contrib/internal/prepare-dev.py index 60fb740..be4b320 100755 --- a/contrib/internal/prepare-dev.py +++ b/contrib/internal/prepare-dev.py @@ -81,8 +81,9 @@ def main(): install_media(site) -print Synchronizing database... -site.sync_database(allow_input=True) +#print Synchronizing database... +#site.sync_database(allow_input=True) +print 'now issue reviewboard/manage.py syncdb' print print Your Review Board tree is ready for development. For some reason, even though PIL is installed (e.g. import Image works) and the version is 1.1.6, the dbjlets stage appears to download an egg for PIL (version 1.1.6 that is already installed). I also had an issue where my Pygments version (0.9) was not high enough and this prevented the SCM tools list from being populated (I did a quick easy_install pygments as a fix and installed 1.1.1). Hope this helps someone else. Chris --~--~-~--~~~---~--~~ Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Installing PyLucene on Windows
Christian Hammond wrote: At the time we wrote search, this functionality didn't exist. However, it's still too limited for our needs. . I'm hoping to find a better solution in the long term, but I don't believe that the native database full text search is it. Now, maybe if we developed a custom solution that provided tables for storing the pre-processed search indexes and then searched from that using full text search. If we go this route, what I'd like to do is split off the search functionality into its own django app in our tree (reviewboard.search) and add a backends directory with different search backends (fulltextdb.py, pylucene.py). We'd then need a nice API that wraps the backends. I'd be all for a solution like this. Have you seen Haystack? http://haystacksearch.org/ - specific to Django (models) with a few different search backends implemented. The closest/similar thing is http://swapoff.org/pyndexter which is not Django specific (but has not been updated in a while). Chris --~--~-~--~~~---~--~~ Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~--~~~~--~~--~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
pure python patch tool, maybe useful to win32 users
http://code.google.com/p/python-patch/wiki/README pure python patch tool (only handles unified diffs though). I've not tested it out but it may be useful to Windows users running a server who do not want to install binaries. Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Custom post-review implementations making them easier proposal
Christian Hammond wrote: ...I also would like an rbapi (or something) package for all the server communication code. This would be separate from RBTools, but RBTools would use it. post-review would be greatly simplified by these two changes. That would be awesome, we've had a couple of ideas for enhancements to post-review on modifying existing fields for a review (i.e. lookup followed by change) but the guys who want to work on it, are not web service people. I'd like to take all this a step further and provide a good way of specifying defaults for post-review without modifying anything. There are a few options for this: 1) User/system configuration files containing defaults. We'd need a good way of determining the path for the system config file, as it may vary (may want to keep it on network storage, for instance). Could even check the directory post-review is in. 2) Look for an optional rbtools_config.py in the PYTHON_PATH. This could be installed on the local user's system, or on a network share somewhere. RBTools wouldn't care, so long as PYTHON_PATH is set to wherever it is. 3) Both. Why not. How about: 1. Check for some environment variable that points to a config file, e.g. POST_REVIEW_CONF=/fullpath/filename.something 2. no environment variable? Default to a search path for a specifically named file; home directory, current directory, path for post-review, python path (maybe), etc. The names / order above would need to be thrashed out, it is more the idea that you can override with a search initiated if there is no setting. Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Custom post-review implementations making them easier proposal
Ritesh Nadhani wrote: For one, I would be interested. We use Bazaar and and have been slowly trying to intergrate ReviewBoard in our system. I looked into the post-review and it does not seem to support Bazaar. I have already extended it using the Mercurial example and override two or three methods that the Mercurial component did. I plan to soon post my changes for Bazaar once I can verify that my code actually works :) Thanks for the feedback. I've not heard any nays so I'm guessing no one hates the idea :-) I went ahead and created: http://reviews.review-board.org/r/882/ Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: give up with subversion
sailor...@gmail.com wrote: I tried a couple of days on FC4, but I still can not let reviewboard work for subversion repository. The main reason is pysvn.. I guess if someone can provide some simple way to install, it will help reviewboard is used by more guys.Maybe, specific versoin linux distribution with necessary install component and one script to install all dependency should be good for normal linux user. I'm not an FC user so I can't comment on it. I had great success with the LTS Server version of Ubuntu. Just a bunch of apt calls. The wiki does have some mention on this I think. However this could be an opportunity for some enterprising person to offer (fee base) ReviewBoard hosting. Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Mail features
Vesterbaek wrote: I would find it useful to to be able to configure this independently for groups and people, such that a user can choose to be member of a particular group, but not receive any emails from review requests to that group, but only get emails if a review request is directed at the user (listed in people). Usecase: - We have a number of development teams, which each map to one or more reviewboard groups. I've setup default reviewers for the different groups based on the repository paths of the files in the review requests. We do it this way to make sure that all developers can easily stay orientated about different review requests that is relevant for them, and possibly comment if they feel like doing so. They should not (be forced to) receive emails about these group review requests. Mandatory reviewers are put in the peoples field, and these should normally receive emails. The way we tackle this is with (Mailman) mailing lists. People join the mailing list and have the usual options; all emails, digest, or no emails. We then making the mailing-list-email-address the group email address and NEVER add anyone to the Reviewboard groups. Then when posting a review add the group, and the specific people. Not exactly what you asked for but can accomplish the same thing. This works for us as we had an existing mailing list already in place, reviewboard simply makes life easier for the people that want to use Reviewboard, this use case may not work for everyone. Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Custom post-review implementations making them easier proposal
If you want to deploy post-review with some site specific customizations (e.g. add support for a custom in house SCM) there are only a few options: * Privately fork post-review and try and sync all the changes that occur in trunk into your version * add a wrapper script around post-review, e.g. rename post-review to postreview.py (python does not allow hypens - in module/library names), import postreview in your tool and then monkey patch in your changes. The rename can be handled in setup.py when building Neither solution is optimal but this is the price you pay if you want private custom versions :-( I don't have a solution to the above but I have some ideas on making it suck slightly less by making either option easier to deal with. For instance take the case where one needs to add in support for a custom SCM. You need to add a new class (subclassed from SCMClient) and then modify determine_client() to add it to the list of SCM's. * With the forked approach you have to edit determine_client(). * With the wrapper approach you have to copy the existing determine_client() into your wrapper, modify it then monkey patch the original. Either way you have to be prepared to follow the trunk version in case determine_client() changes. What do people think of the following ideas: 1) move the tuple of SCM clients into a global variable. E.g. instead of: # Try to find the SCM Client we're going to be working with. for tool in (SVNClient(), CVSClient(), GitClient(), MercurialClient(), PerforceClient(), ClearCaseClient()): Have: scm_tuple=(SVNClient(), CVSClient(), GitClient(), MercurialClient(), PerforceClient(), ClearCaseClient()) def determine_client(): # Try to find the SCM Client we're going to be working with. for tool in scm_tuple: This makes monkey patching the list easier, you just update the global from outside. 2) modify the trunk version of post-review to change the formatting that is used in the tuple (list) of SCM clients to make it easier to edit the list. E.g.: scm_tuple=( SVNClient(), CVSClient(), GitClient(), MercurialClient(), PerforceClient(), ClearCaseClient(), ) This makes it easier to modify the code if you decide to fork the script. You can re-order, add, or remove very easily. Diffs show up very clearly. Does this appeal to anyone? #1 above would make my life easier. I wanted to get a feel from a wide audience before posting a for patch for review where it may not be seen by so many people. Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Newbie question: setup problems
Phil wrote: .Sqlite3 is complaining that it cannot open the database file. (sqlite3.OperationalError: unable to open database file) Has anyone else had this problem? My understanding is that Python 2.5 includes sqlite thus I wouldn't have to install it again -- is this incorrect? Have I missed installing something? If you are getting the sqlite error, this proves (py) sqlite is installed :-) You _probably_ have a file/path permission problem as sqlite databases are files. I suspect when you ran manage you did it as a different user than the one who runs the web server. Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: New SCMTools fot PVCS repository
Tee - your best bet is to just get stuck in and play :-) I've not come across any docs other than what is in the code, I ended up setting up a reviewboard server to svn and seeing what it did as well as creating an empty NEWscmtool that subclasses SCMTool and adding methods as they were called when I tried to use my new scm. I used a mix of live debugging in pdb and the occasional print statement. Chris tilek wrote: Hi All, I'd like to create a new SCMTool that is used to connent a PVCS repository. I read the python code and I think, I must create a new class eg.PVCSTool that extends the base class SCMTool. Thus, the new class must implement many method of base class but I don't know the objective of each method and I cannot find any document (developer guide). Could you please help me to explain about method of SCMTool? Thanks, Tee --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: C++ Highlighting
bart wrote: However I am having some problems regarding C/C++ highlighting. Generally the highlighting appears to be on but for some keywords that are highlighted by pygmentize program I am not able to see highlight in the review diff... I tried to turn on the guess option in diffutils.py but that did not work either. What could be the cause here? Am I missing some setting? I've a vague recollection that RB uses a custom colour/color set. Might be worth checking into that in svn (possibly removing it as a test?). Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Upgrade issues
I've just moved to rev 1623 in trunk (I was at 1519). I do like the new manual upgrade required screen that pops up (settings.py has been updated to change the location for htdocs/media/uploaded/images). Very slick and should saves lots of admin hair pulling :-) However when I first loaded the main page (and review page) it appeared to be messed up in Google Chrome and IE6 (to the point that diffs are not usable). This turned out to be because the whole of the htdocs directory had moved (changes to settings.py), not just the upload dir. REM this is new with heads revs as of 5th dec 2008 REM rev 1623 mkdir C:\svn\django\htdocs\media\uploaded\images mkdir C:\svn\django\htdocs\media\rb\images xcopy /S /Y C:\svn\django\reviewboard\htdocs C:\svn\django\htdocs I'm assuming I can still run from SVN without the deployment tool pre RB 1.0 - is this a bad assumption? Before the change RB was using (on this Windows box) C:\svn\django\reviewboard\htdocs. The good thing is that the manual upgrade required web page helped me work out what was going in. I also had some problems with the evolve, it worked but accessing (old) reviews (against svn) I get errors about bad change description/numbers :-(. Oddly the old reviews with my work in progress SCMTool are fine. I'm planning on starting a fresh db but I have copies (sqlitedb) if there is interest in debugging this. It isn't bothering me but I thought I would mention it (rather than logging a bug/issue). The pull from trunk also updated dbjlets, etc. I don't need fixes but wanted to share this in case other people hit this too (and in case I'm deploying incorrectly). Thanks! Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Where is the post-review tool
AndyP wrote: I installed the latest stuff (blew everything away first) and can't find the latest version of the post-review tool. Where do I go to get it? You can pull it down from svn (and you can even use wget if you don't have svn to hand): http://code.google.com/p/reviewboard/source/browse/trunk/reviewboard/contrib/tools/post-review Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Using ignore file extensions on post-review
Russell Bryant wrote: On Wed, Nov 5, 2008 at 6:54 AM, Dave Hocker [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: You can use the Perforce p4 files command to determine if Perforce thinks the file is text or binary. The last token of the output will be something like (text) or (binary+l). Note that for a binary file the character after the + is the letter L in lower case. We've had this issue and I plan to modify our post-review script to make this check. You could do something with subversion, as well, by checking the svn:mime-type property on files to see if they are text/plain or not. Subversion won't try and diff binary files so there should be no need to put this logic in post-review. Where should maybe famous last words, I'm not using subversion with reviewboard so I've not actually sat down and tested this. Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Limitations with diff files - any suggestions?
fts1050 wrote: To remedy this..any chance Post Review can be modified to handle an external file that contain a list of specific files that should be included in the review? I can do this myself but then when I upgrade I have to manage my own version of post review and that's no fun. Just a suggestion. I have a modification that implements this with a -l flag, I'm going to be out for a few weeks so I'm not in a position to share the change at the moment as I made the change to a subclass of SCMClient rather than SCMClient (so everyone can benefit) I plan on doing this at some point. Probably quicker to implement what you need now rather than waiting for me. it looks like this though: if not files_list and options.piccolo_flist: filteptr = open(options.piccolo_flist) files_list = filteptr.readlines() filteptr.close() . parser.add_option(-l, --filelist_filename, dest=piccolo_flist, default=None, help='File containing list of files in change.') Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Publishing and target group error
You won't like the answer use apache.
I reported a similar problem in the faq (with solution)
http://code.google.com/p/reviewboard/wiki/FAQ
I think this is another manifestation. If this is the case do you could
you add a comment to the faq too?
Also see
http://groups.google.com/group/reviewboard/browse_thread/thread/490a3281fd7867a7/64c910b24e3e0790?lnk=gstq=10054#64c910b24e3e0790
Chris
13Strider wrote:
when i try to run post-review with publish enabled i get this error
also when updating a file or setting a target group or target people,
i looks like the error is coming from urllib2.request, any idea how to
fix this
or what is causing it
Thank you
Publishing
HTTP POSTing to
http://192.168.100.4:8000/api/json/reviewrequests/12/publish/: {}
Unable to access
http://192.168.100.4:8000/api/json/reviewrequests/12/publish/.
The host path may be invalid
urlopen error (10054, 'Connection reset by peer')
Review request already exists. Updating it...
HTTP POSTing to
http://192.168.100.4:8000/api/json/reviewrequests/12/update_from_changenum/:
{}
Unable to access
http://192.168.100.4:8000/api/json/reviewrequests/12/update_from_changenum/.
The host path may be invalid
urlopen error (10054, 'Connection reset by peer')
HTTP POSTing to
http://192.168.100.4:8000/api/json/reviewrequests/12/save/: {}
Unable to access http://192.168.100.4:8000/api/json/reviewrequests/12/save/.
The host path may be invalid
urlopen error (10054, 'Connection reset by peer')
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
reviewboard group.
To post to this group, send email to reviewboard@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at
http://groups.google.com/group/reviewboard?hl=en
-~--~~~~--~~--~--~---
Re: How to add another scmtool support?
On 10/13/2008 8:38 AM, Jay wrote: If I want to add another scmtool support, is it enough to add a new file in scmtools folder like anotherscmtool.py which will implements several classes and methold? Or any change in other module? I can confirm that this is pretty much what you have to do as I'm going through that process myself right now. The best tip I have is to setup another svm (e.v. svn) so you can compare a working scm with your work in progress one. You also need to add the scmtool in the web admin repository page. I did find that with the diff parser I ended up sub classing a few more methods, my scmtool diff output is not unified :-( With any luck you won't have to do much. I have made a hack to the shared code for making raw diffs via a url - I'm not really in a good position to talk about that just yet as it was a quick hack to get it working. I'm hoping to do away with it but I've not spent any time working out how that will be done :-) The reason for this was that I wanted the raw diffs that come out to match the diffs that are uploaded and I needed to put back part of the special header. Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Using ignore file extensions on post-review
On 10/8/2008 1:38 AM, Marcos wrote: I am using Perforce. The error is with tga files. I'm not a Perforce user so dumb question coming. Does Perforce have the concept of binary files? Can this information be picked up from Perforce (I notice a describe is done before the diffs are started)? This would be more robust as you wouldn't have to maintain a list of binary extensions. Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: weird error in review-board and post-review in windows
On 10/8/2008 3:45 PM, Paulo Eduardo Neves wrote: .. my problem with cygwin is that I'll have to deploy the client in dozens of developer Windows machines. I don't want to force everybody to install cygwin just to have diff. Don't forget that the patch requirement is only on the SERVER, it might be worth trying cygwin on the server only and using the standalone diffutils for the clients. I have to confess I've not tried using post-review with svn clients on Windows (so I'm not clear what the requirements are). BUT I am using cygwin patch on a Windows server (the scm tool I use on the client's has it's own diff built in). One thing I did come across a while back is that Tortoise SVN does support command line arguments, you may be able to simply generate the patch(es) using that without any other tools on the clients. I've tried unxutils links but it still didn't work. I've tried cygwin and the problem persists! Then I tried to delete the file where the problem is happening... and it works! I don't know if it will be a common error or just really bad luck. Doh! I'd be interested to hear if it occurs again. Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
Re: Nothing happens when I upload a patch
On 10/8/2008 4:21 PM, Paulo Eduardo Neves wrote: I've already did some patch uploads through the web interface, but now when I try to upload a diff generated by subversion nothing happens. My steps: 1) generate diff with this command: svn diff -r 15854:15855 e:\revision15855.patch 2) go to the web interface, click in new review request. 3) there's just one repository. It is in http://subversion/svn/. My work copy is associated the the url http://subversion/svn/sac/trunk/revisado. I add my patch file, and as base diff path I put sac/trunk/revisado Now nothing happens. I see the same upload interface with Base diff path filled and the upload box empty. There's no new reviews in the all reviews request tab. The server log shows a post request: [08/Oct/2008 20:05:29] POST /r/new/ HTTP/1.1 200 4221 I've already managed to upload patches using post-review. I'm using the review-board from the HEAD of the trunk. I'm running it with sqllite in a windows 2003 machine. How do I debug this problem? When I was debugging my SCM diffs, I added print statement to the diff parser and import pdb ; pdb.set_trace() to diffutils.py so I could step through (I'm using the built in web server). Turned out in my case I had a mix of bad diffs and bad code in my diff parser. Probably not relevant to you but I had issues with tabs and spaces getting switched (due to copy/paste in a terminal). Good luck! Chris --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups reviewboard group. To post to this group, send email to reviewboard@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en -~--~~~~--~~--~--~---
