Hi Christian,

I want to thank you for your help last night with all of these problems, you 
were super responsive.

However, I later realized that the reason I decided to try reviewboard in the 
first place was silly.

I needed to do a code review on a very large commit in GitLab, and it wouldn't 
show me the diff because of its size, so I couldn't add any comments.

Last night I realized that all I had to do (in addition to a squash-merge), was 
unstage and revert the files that were irrelevant to the code review. This 
significantly reduced the size of the diff and GitLab finally showed it. So for 
now I'll stick with GitLab, for a few reasons:

1. ReviewBoard doesn't work with UNIX domain sockets for memcached.
2. ReviewBoard doesn't use the system's sendmail/exim4/mail command to send 
email, and it failed to send email with the SMTP settings I gave it.
3. This problem with GitLab integration and fastcgi.
4. ReviewBoard's interface is too complicated for my taste. Basic things, like 
being able to delete a comment on a review, are either impossible or too 
difficult to figure out.
5. I hate Python, and ReviewBoard just confirmed this for me.


So thanks again for your help, but for now I'll stick to GitLab. Maybe I'll 
revisit it in the future though, because while there are a lot of thing that I 
don't like about it, there are also a lot of things that ReviewBoard does that 
nothing else seems to do, for example, it's super nice to be able to comment on 
arbitrary diffs.

Cheers,
Greg

--
Please do not email me anything that you are not comfortable also sharing with 
the NSA.

On Feb 13, 2014, at 2:11 AM, Christian Hammond <chip...@chipx86.com> wrote:

> I don't mean to run it in place of fastcgi. Run it as its own port, separate 
> from the fastcgi server, and connect to it directly. We're trying to rule out 
> the interference of any layers between the web browser and the Django/Review 
> Board code.
> 
> Christian
> 
> -- 
> Christian Hammond - chip...@chipx86.com
> Review Board - http://www.reviewboard.org
> Beanbag, Inc. - http://www.beanbaginc.com
> 
> 
> On Thu, Feb 13, 2014 at 12:06 AM, Greg <g...@kinostudios.com> wrote:
> well, if a server isn't expecting fastcgi, it won't be happy: 
> http://pastebin.com/04P7fsgk
> 
> 
> --
> Please do not email me anything that you are not comfortable also sharing 
> with the NSA.
> 
> On Feb 13, 2014, at 1:53 AM, Christian Hammond <chip...@chipx86.com> wrote:
> 
>> Okay. There's something I want to have you try to rule out fastcgi. We're 
>> going to run Review Board's single-threaded devserver and see if it's 
>> susceptible to the same problem. That will help narrow down if it could be 
>> fastcgi or if it's definitely not.
>> 
>> Run:
>> 
>>     rb-site manage /path/to/site runserver 0.0.0.0:8080
>> 
>> Then connect and repeat the attempt. If you get the same error, it's most 
>> likely a Paramiko bug somewhere. If it's gone, then it's yet another bad 
>> interaction with fastcgi.
>> 
>> Christian
>> 
>> -- 
>> Christian Hammond - chip...@chipx86.com
>> Review Board - http://www.reviewboard.org
>> Beanbag, Inc. - http://www.beanbaginc.com
>> 
>> 
>> On Wed, Feb 12, 2014 at 11:36 PM, Greg <g...@kinostudios.com> wrote:
>>> So.. wait, let's take a step back. Are you using a GitLab repository here, 
>>> or are you specifying a standard Git repository and pointing to the local 
>>> path, or are you specifying an SSH or git protocol path?
>> 
>> The first two (and potentially third, but that seems like the least 
>> convenient of the three).
>> 
>> I have a gitlab instance running on the same machine, and so reviewboard can 
>> also access its repo via the filesystem. 
>> 
>> Oddly enough, after I turned on logging, while the log file was created, it 
>> contained nothing in it, even after setting it to 'debug' and trying again 
>> to add the repo.
>> 
>> However, my other logs had this:
>> 
>> Feb 13 02:07:39 host rb-site[7497]: 
>> /usr/local/lib/python2.7/dist-packages/Django-1.4.10-py2.7.egg/django/views/generic/simple.py:8:
>>  DeprecationWarning: Function-based generic views have been deprecated; use 
>> cla...ed views instead.
>> Feb 13 02:07:39 host rb-site[7497]: DeprecationWarning
>> Feb 13 02:07:39 host rb-site[7497]: 
>> /usr/local/lib/python2.7/dist-packages/Django-1.4.10-py2.7.egg/django/template/defaulttags.py:1237:
>>  DeprecationWarning: The syntax for the url template tag is changing. Load 
>> t...the new behavior.
>> Feb 13 02:07:39 host rb-site[7497]: category=DeprecationWarning)
>> Feb 13 02:24:58 host rb-site[7497]: ERROR:root:Invalid cache hostname 
>> "memcached://unix:/tmp/memcached.sock"
>> Feb 13 02:25:00 host rb-site[7497]: 
>> /usr/local/lib/python2.7/dist-packages/Django-1.4.10-py2.7.egg/django/db/models/fields/__init__.py:761:
>>  RuntimeWarning: DateTimeField received a naive datetime (2014-01-15 
>> 00:...upport is active.
>> Feb 13 02:25:00 host rb-site[7497]: RuntimeWarning)
>> Feb 13 02:25:00 host rb-site[7497]: 
>> /usr/local/lib/python2.7/dist-packages/Django-1.4.10-py2.7.egg/django/db/models/fields/__init__.py:761:
>>  RuntimeWarning: DateTimeField received a naive datetime (2014-02-14 
>> 00:...upport is active.
>> Feb 13 02:25:00 host rb-site[7497]: RuntimeWarning)
>> Feb 13 02:26:04 host rb-site[7497]: ERROR:root:Uploading SSH key failed: PID 
>> check failed. RNG must be re-initialized after fork(). Hint: Try 
>> Random.atfork()
>> 
>> FYI, again, this thing is being run like so via systemd:
>> 
>> rb-site manage /path/to/site.com runfcgi daemonize=false host=localhost 
>> port=8025
>> 
>> --
>> Please do not email me anything that you are not comfortable also sharing 
>> with the NSA.
>> 
>> On Feb 13, 2014, at 1:24 AM, Christian Hammond <chip...@chipx86.com> wrote:
>> 
>>> What's the reason we use rbssh instead of ssh? We need more control over 
>>> the environment and SSH keys than ssh allows for.
>>> 
>>> So.. wait, let's take a step back. Are you using a GitLab repository here, 
>>> or are you specifying a standard Git repository and pointing to the local 
>>> path, or are you specifying an SSH or git protocol path?
>>> 
>>> rbssh should only ever be involved if you're using an SSH/git-based path.
>>> 
>>> It'd help if I could see more of your repository configuration. Can you 
>>> also check your reviewboard.log file (turn it on if you haven't already) 
>>> and see if there are any errors or debug info related to this?
>>> 
>>> Christian
>>> 
>>> -- 
>>> Christian Hammond - chip...@chipx86.com
>>> Review Board - http://www.reviewboard.org
>>> Beanbag, Inc. - http://www.beanbaginc.com
>>> 
>>> 
>>> On Wed, Feb 12, 2014 at 11:18 PM, Greg <g...@kinostudios.com> wrote:
>>> Err... what's the reason for rbssh? I think I have password authentication 
>>> turned off with ssh (requires key).
>>> 
>>> $ rbssh u...@host.com
>>> /usr/local/lib/python2.7/dist-packages/paramiko-1.12.1-py2.7.egg/paramiko/client.py:95:
>>>  UserWarning: Unknown ssh-rsa host key for host.com: 
>>> f315ab3debbdf29ae833df8ba0cab48c
>>> Error connecting to server: No authentication methods available
>>> 
>>> FYI the git repo is located on the same server, and should be readable by 
>>> the same user.
>>> 
>>> --
>>> Please do not email me anything that you are not comfortable also sharing 
>>> with the NSA.
>>> 
>>> On Feb 13, 2014, at 1:07 AM, Christian Hammond <chip...@chipx86.com> wrote:
>>> 
>>>> The generally recommended setup is to use Apache with mod_wsgi, but 
>>>> there's also wsgi support for nginx, plus services like gunicorn or uwsgi 
>>>> (though we don't provide any official support for these).
>>>> 
>>>> I don't know for a fact that fastcgi is the culprit. It's just the one 
>>>> thing that stood out from what I know about your configuration that 
>>>> differs from mine.
>>>> 
>>>> From Googling around, I can see that this was a known error with older 
>>>> versions of Paramiko (1.7.x from the looks of it). I'd imagine 1.12.1 
>>>> should be new enough. It doesn't look to be a bug with Review Board itself.
>>>> 
>>>> What happens if you attempt to SSH into that server (or another) using 
>>>> 'rbssh' instead of 'ssh'? Do you see the error?
>>>> 
>>>> Christian
>>>> 
>>>> -- 
>>>> Christian Hammond - chip...@chipx86.com
>>>> Review Board - http://www.reviewboard.org
>>>> Beanbag, Inc. - http://www.beanbaginc.com
>>>> 
>>>> 
>>>> On Wed, Feb 12, 2014 at 11:00 PM, Greg <g...@kinostudios.com> wrote:
>>>> So what does that mean for nginx users?
>>>> 
>>>> I already spent a day getting this thing working with nginx. :-\
>>>> 
>>>> --
>>>> Please do not email me anything that you are not comfortable also sharing 
>>>> with the NSA.
>>>> 
>>>> On Feb 13, 2014, at 12:55 AM, Christian Hammond <chip...@chipx86.com> 
>>>> wrote:
>>>> 
>>>>> Hi Greg,
>>>>> 
>>>>> We're running the same versions of PyCrypto and Paramiko, but I haven't 
>>>>> managed to reproduce the problem.
>>>>> 
>>>>> It's possible it has something to do with fastcgi. These days, wsgi is 
>>>>> the better supported option (really across the board on all Python-based 
>>>>> webapps). We've seen various issues with fastcgi over the years, and now 
>>>>> that wsgi is more standard and maintained, we may end up dropping fastcgi 
>>>>> in the future.
>>>>> 
>>>>> Christian
>>>>> 
>>>>> -- 
>>>>> Christian Hammond - chip...@chipx86.com
>>>>> Review Board - http://www.reviewboard.org
>>>>> Beanbag, Inc. - http://www.beanbaginc.com
>>>>> 
>>>>> 
>>>>> On Wed, Feb 12, 2014 at 10:51 PM, Greg <g...@kinostudios.com> wrote:
>>>>> Thanks for the quick reply Christian.
>>>>> 
>>>>> paramiko 1.12.1.
>>>>> 
>>>>> Site is being started like so:
>>>>> 
>>>>>   rb-site manage /path/to/site.com runfcgi daemonize=false host=localhost 
>>>>> port=8025
>>>>> 
>>>>> Also, having problems with memcached, but for that I created another 
>>>>> thread, hope you can help there too.
>>>>> 
>>>>> Thanks,
>>>>> Greg
>>>>> 
>>>>> --
>>>>> Please do not email me anything that you are not comfortable also sharing 
>>>>> with the NSA.
>>>>> 
>>>>> On Feb 13, 2014, at 12:39 AM, Christian Hammond <chip...@chipx86.com> 
>>>>> wrote:
>>>>> 
>>>>>> We're working on a better repository form that will be a lot easier to 
>>>>>> use, and automatically determine many of the fields, but in the 
>>>>>> meantime, I'll make a note to improve the help text.
>>>>>> 
>>>>>> Service URL is the path to the root of your GitLab server. Same with the 
>>>>>> URL field for the bug tracker. (I know it's inconsistent.)
>>>>>> 
>>>>>> What version of Paramiko do you have?
>>>>>> 
>>>>>> Christian
>>>>>> 
>>>>>> -- 
>>>>>> Christian Hammond - chip...@chipx86.com
>>>>>> Review Board - http://www.reviewboard.org
>>>>>> Beanbag, Inc. - http://www.beanbaginc.com
>>>>>> 
>>>>>> 
>>>>>> On Wed, Feb 12, 2014 at 10:30 PM, Greg <g...@kinostudios.com> wrote:
>>>>>> I'm having this same problem with GitLab. :-(
>>>>>> 
>>>>>> Running Debian Jessie and PyCrypto 2.6.1.
>>>>>> 
>>>>>> Help very much appreciated.
>>>>>> 
>>>>>> Also: it's great that this supposedly supports gitlab, but I have no 
>>>>>> idea what to put into the fields, and the documentation isn't helpful at 
>>>>>> all.
>>>>>> 
>>>>>> What goes into "Service URL"? Where is this documented?
>>>>>> 
>>>>>> What do I put for the bug tracker URL?
>>>>>> 
>>>>>> Not a single example is given. :(
>>>>>> 
>>>>>> Thanks,
>>>>>> Greg
>>>>>> 
>>>>>> 
>>>>>> On Wednesday, January 15, 2014 4:33:04 AM UTC-6, Christian Hammond wrote:
>>>>>> Hi Juia,
>>>>>> 
>>>>>> This is likely a PyCrypto-related issue. It's fortunately not 
>>>>>> GitLab-specific, and can likely be fixed with a dependency upgrade.
>>>>>> 
>>>>>> Can you tell me which OS/Linux distro you installed this on?
>>>>>> 
>>>>>> Also, can you run:
>>>>>> 
>>>>>>     $ python
>>>>>>     >>> import Crypto
>>>>>>     >>> Crypto.__file__
>>>>>> 
>>>>>> This should give you a path to the version of PyCrypto you have 
>>>>>> installed. Can you show me that path?
>>>>>> 
>>>>>> You can then try to upgrade, either through a new package from your 
>>>>>> distro, or by doing:
>>>>>> 
>>>>>>     $ sudo easy_install -U pycrypto
>>>>>> 
>>>>>> Christian
>>>>>> 
>>>>>> -- 
>>>>>> Christian Hammond - chi...@chipx86.com
>>>>>> 
>>>>>> Review Board - http://www.reviewboard.org
>>>>>> Beanbag, Inc. - http://www.beanbaginc.com
>>>>>> 
>>>>>> 
>>>>>> On Wed, Jan 15, 2014 at 2:21 AM, Julia S.S. <hit...@gmail.com> wrote:
>>>>>> Hi, I just updated review board to 1.7.21 and I'm trying to add a new 
>>>>>> repo from gitlab with the new gitlab support, as I have no gitlab 
>>>>>> account configured I'm leaving <Link new account selected> and filling 
>>>>>> user and pass, I'm also adding redmine.
>>>>>> When I click on save I get Unknown error when linking the account: PID 
>>>>>> check failed. RNG must be re-initialized after fork(). Hint: Try 
>>>>>> Random.atfork()
>>>>>> 
>>>>>> What should I do to fix it?
>>>>>> 
>>>>>> Thanks,
>>>>>> Julia
>>>>>> 
>>>>>> -- 
>>>>>> Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/
>>>>>> ---
>>>>>> Sign up for Review Board hosting at RBCommons: https://rbcommons.com/
>>>>>> ---
>>>>>> Happy user? Let us know at http://www.reviewboard.org/users/
>>>>>> --- 
>>>>>> You received this message because you are subscribed to the Google 
>>>>>> Groups "reviewboard" group.
>>>>>> To unsubscribe from this group and stop receiving emails from it, send 
>>>>>> an email to reviewboard...@googlegroups.com.
>>>>>> 
>>>>>> For more options, visit https://groups.google.com/groups/opt_out.
>>>>>> 
>>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> -- 
>>>>> Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/
>>>>> ---
>>>>> Sign up for Review Board hosting at RBCommons: https://rbcommons.com/
>>>>> ---
>>>>> Happy user? Let us know at http://www.reviewboard.org/users/
>>>>> --- 
>>>>> You received this message because you are subscribed to the Google Groups 
>>>>> "reviewboard" group.
>>>>> To unsubscribe from this group and stop receiving emails from it, send an 
>>>>> email to reviewboard+unsubscr...@googlegroups.com.
>>>>> 
>>>>> For more options, visit https://groups.google.com/groups/opt_out.
>>>> 
>>>> 
>>>> 
>>>> -- 
>>>> Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/
>>>> ---
>>>> Sign up for Review Board hosting at RBCommons: https://rbcommons.com/
>>>> ---
>>>> Happy user? Let us know at http://www.reviewboard.org/users/
>>>> --- 
>>>> You received this message because you are subscribed to the Google Groups 
>>>> "reviewboard" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send an 
>>>> email to reviewboard+unsubscr...@googlegroups.com.
>>>> For more options, visit https://groups.google.com/groups/opt_out.
>>> 
>>> 
>>> 
>>> -- 
>>> Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/
>>> ---
>>> Sign up for Review Board hosting at RBCommons: https://rbcommons.com/
>>> ---
>>> Happy user? Let us know at http://www.reviewboard.org/users/
>>> --- 
>>> You received this message because you are subscribed to the Google Groups 
>>> "reviewboard" group.
>>> To unsubscribe from this group and stop receiving emails from it, send an 
>>> email to reviewboard+unsubscr...@googlegroups.com.
>>> For more options, visit https://groups.google.com/groups/opt_out.
>> 
>> 
>> 
>> -- 
>> Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/
>> ---
>> Sign up for Review Board hosting at RBCommons: https://rbcommons.com/
>> ---
>> Happy user? Let us know at http://www.reviewboard.org/users/
>> --- 
>> You received this message because you are subscribed to the Google Groups 
>> "reviewboard" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to reviewboard+unsubscr...@googlegroups.com.
>> For more options, visit https://groups.google.com/groups/opt_out.
> 
> 
> 
> -- 
> Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/
> ---
> Sign up for Review Board hosting at RBCommons: https://rbcommons.com/
> ---
> Happy user? Let us know at http://www.reviewboard.org/users/
> --- 
> You received this message because you are subscribed to the Google Groups 
> "reviewboard" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to reviewboard+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

Reply via email to