Here's the chat log from the Tuscany weekly chat on Monday 5th of June.
The topic discussed was finishing off the M1 release, and that was mainly
about which of the website JIRAs needed to be completed before announcing
the release. Most of the JIRAs for architecture and design documentation got
pushed out to be done in a later release.
Session Start: Mon Jun 05 16:01:44 2006
Session Ident: #tuscany
* Now talking in #tuscany
* jsdelfino has joined #tuscany
* dkulp has quit IRC (Remote closed the connection)
* dkulp has joined #tuscany
* cr22rc_away has joined #tuscany
* simonnash has joined #Tuscany
* lresende_ has quit IRC ("Trillian (http://www.ceruleanstudios.com";)
<ant_> Hi, 4:30 now, shall we start?
<jboynes> yep
* kgoodson has quit IRC (Read error: 104 (Connection reset by peer))
<jboynes> can I ask about the release?
<jboynes> I updated the files for the change to NPL (in the root and in
rhino)
<jboynes> are those the only ones that need to change?
* rfeng has joined #tuscany
<jsdelfino> I propagated these changes to the java-M1 branch
<simonnash> hi. Simon Nash is here.
<jboynes> k, thanks
<ant_> so can we announce an M1 or do we need more voting?
<jsdelfino> just the top level license and the license file under
container.rhino are changing
<jboynes> I got the impression we didn't need to revote
<ant_> great
<cr22rc_away> given no change in code I don't see then need to revote
<jsdelfino> I have the distro on my machine now, will upload to my home page
later today
<jboynes> thx
<jsdelfino> I'll need some help from somebody familiar with the web site
updates to update our download page
<cr22rc_away> sure
* cr22rc_away is now known as cr22rc_
<ant_> so we should have an M1 today?
* slaws has joined #Tuscany
<jsdelfino> but, before we switch to the new download page pointing to the
M1 distro and we announce it, I think we need to finish the work tracked by
the JIRAs in the M1-website category
<jsdelfino> all the doc that people in the team wanted to have for this
release, and volunteered to work on (see our Wiki page) is part of the
release...
<cr22rc_> do we more approval from apache to make it public ?
<jsdelfino> so I don't think it's a good idea to publish the distro without
these documents, which are really the entry point into the release
<jsdelfino> as far as I know, we have approval from the incubator to make
the release public
<ant_> jsdelfino, you mean JIRAs like TUSCANY-293, 294,295,296 etc?
<jsdelfino> but again, the distro without the correct doc is not going to
give users a good impression
<simonnash> i agree with Sebastien. we would expect publishing the release
to create a lot of interest. We want to give people a good experience with
getting started on using it.
<jsdelfino> here's the link to the JIRA issues in the M1-website category:
<jsdelfino>
http://issues.apache.org/jira/secure/IssueNavigator.jspa?reset=true&&pid=12310210&resolution=-1&fixfor=12311108&sorter/field=priority&sorter/order=DESC
<cr22rc_> getting started on using it (not extending ) is in the binary. I
don't know if I agree whole heartly that we need all that resolved
<jsdelfino> not sure if the link works for all, or if it's user-specific,
but if it doesn't work for you, just browse from the Tuscany project to the
issues under M1-website
<ant_> link doesn't work for me
* cr22rc has quit IRC (Read error: 110 (Connection timed out))
<cr22rc_> do we really need to document how to extend. right off the bat...
there is a lot of benfit for users to use just sca aspects IMO
<jboynes> it worked for me
<jsdelfino> rick, check this link for example:
http://incubator.apache.org/tuscany/SCA_Java_Tools_documentation.html
<jboynes> yeah - I don't see these as blockers
<ant_> under M1-website there things like "document the architecture of
xxx". are these likely to get done given whats going on?
<jboynes> none seem to apply to your typical end-user
<simonnash> top of the JIRA list is one about tools discpreancies
<jsdelfino> the link I just gave, the doc is completely out of sync with the
release
<cr22rc_> joboynes ... thats the angle I take
<simonnash> i thought one of our goals was to enable extension
<jsdelfino> the other question is: I thought one of the big motivations for
this release for people to be able to extend Tuscany, not just for SCA users
<cr22rc_> maybe we should go through each and determine if it is really a
must have ?
<jsdelfino> yes simon, you were quicker than me :) agreed
<jsdelfino> rick, good idea
<jsdelfino> let's start with 184
<jsdelfino> that's the link I posted
<simonnash> go it
<simonnash> got it
<jsdelfino> IMO this one is pretty bad, needs to be fixed
<simonnash> this one seems important
<rfeng> agree
<jboynes> I'd say 184, 160, 138 and 344
<cr22rc_> ok since this is something the user of sca will use ... ok
<jboynes> the rest are nice but not blockers
<jsdelfino> can we go one by one? in the order displayed by JIRA?
<jsdelfino> and we just vote on these, we can make it quick
<simonnash> they are not blockers for someone just using the release. but
thye will inhibit [eople getting drawn in at a deeper level.
<jsdelfino> +1 to fix 184 from me
<jboynes> trying to speed things up :-)
<cr22rc_> 184 +1
<rfeng> +1
<ant_> ok 184 +1
<jsdelfino> ok, if nobody objects let's move to the next one
<simonnash> +1 if I am allowed to vote:-)
<jsdelfino> 291
<cr22rc_> not a blocker for me
<jsdelfino> 0 from me, not for users, not related to the release
<rfeng> +0
<simonnash> this raises a larger issue of how important do we think testing
is
<simonnash> it's a bigger issue than the release
<jsdelfino> yes, it is very important, but the question is: do we block the
release until this one is fixed, IMO the answer is no
<ant_> 291 out
<cr22rc_> ok but lets put that as an item to tackle later
<simonnash> ok with me. guess tha is 0
<jsdelfino> ok, next: 289
<simonnash> i would say this is important
<cr22rc_> is this still changing ?
<rfeng> +1
<jsdelfino> my vote, +1 to fix it for this release
<rfeng> jboynes already has the draft on wiki, right?
<cr22rc_> I thought I saw messages that this is really different in the
sandbox ?
<cr22rc_> aka Tuscany II :-)
<jboynes> rfeng, I did some but had no feedback from anyone
<jsdelfino> rick, maybe this is changing, but what we're documenting is the
M1 release
<rfeng> I read it, looks good to me
<jboynes> thanks :-)
<jsdelfino> looked good to me too, jboynes, can you just wrap it up then?
complete the few sections that are missing
<cr22rc_> hmmm.... wow pushing something on users that we know already is
changing ... no one else see a concern with that?
<jsdelfino> rick, isn't that the point of having milestones? or are you
questioning the value of the M1 release, given that we know things may be
changing in the future :)
<rfeng> doc will be evolved to reflect the changing code
<cr22rc_> I see value it getting people an SCA implementation regardless of
extending or not.
* slaws has quit IRC ("Bye")
<jsdelfino> so for now we have two +1 to get 289 fixed, from rfeng and
myself, can others vote as well?
<simonnash> i think we document what's there, and deal with changes later
<simonnash> +1
<haleh> agree
<cr22rc_> -0
<jsdelfino> ant_, jboynes? any opinion?
<ant_> other than the doc already in wiki, do we have someone volunteering
to do 289?
<jsdelfino> dan?
<jboynes> as ant_ said, any volunteers?
<rfeng> I guess I can help
<cr22rc_> I thought it was done ? Only move from wiki to website ?
<jsdelfino> ok then, +1 with rfeng volunteering, if no further objection
let's move to the next one: 293
<jsdelfino> 293, 294, 295, 296 are all in the same category I guess
<jsdelfino> this is about documenting the architecture of the runtime, java
container, js container and model
<ant_> i agree with cr22rc's concern about pushing something we know is
changing, also i don't think the all the extension work was ever really
completed
<cr22rc_> I don't see these a blockers for the binary release...especially
since we again know it's changing
<simonnash> 294 talks about removing some obsolete material
<jsdelfino> I already did that part of 294, the rest of the work is to
document the M1 model
* kevin has joined #tuscany
<jsdelfino> now I hear what sounds like objections on 289 from rick and ant,
so let's reopen 289...
<simonnash> sorry i did not relaise tat the JIRA comment meant it was done
already
<ant_> my comment was really also about all these 29x ones
<simonnash> i think 29x have lower priority than 289. not to say that are
unimportant.
<simonnash> ..that they are..
<cr22rc_> My view on that was simply thinking you can turn off people really
quick if they start extending... then wanting to donate and find it obsolete
<haleh> Can we do 292? This will help new developers do the right thing. -
but not a release stopper
<cr22rc_> I don't think it's a qustion on any of these would be great to get
done. Just what stops us from making the binary public.
<jsdelfino> my opinion: +1 to do 289 now, we can do 29x later (even though
they are important, we're not going to delay the release further because of
29x)
<simonnash> true, but it is also a turnoff if people ant to extend and are
unable to do so. maybe not all the extension work is throw away.
<simonnash> +1 for 292, if people are extending then we would like them to
use logging in their code
<cr22rc_> Isn't there an example of that in one of the samples ?
<jsdelfino> there is an example of how to turn logging on, no example of how
to log from your own extension
<cr22rc_> how much do they differ ?
<ant_> i'm with jsdelfino, +1 for 289 as there were volunteers, but leave
the rest of the 29x architecture doc ones to Mx
<cr22rc_> that's about my position
<simonnash> is this including 292? i think it is in a different category
<rfeng> I assume 292 and 291 are in the same bucket
<cr22rc_> will the logging expert please advise how much different is it
from logging in an extension than what we have in the samples ?
<jboynes> I think the samples show how to enable it; when writing code you
need to add the logging statements
* slaws has joined #tuscany
<jsdelfino> 292 is about describing how to add the logging statements
<haleh> Is the logging info somewhere that we can add to the website?
<cr22rc_> How much logging do we have now ?
<jsdelfino> very little
<jsdelfino> afaik we're logging only in 3 or 4 places
<cr22rc_> ok so to users ... do as I say, not as I do :-)
<simonnash> typically this does not get added later. people tend not to go
back to old code and retrofit logging, unless they are directed to do so
<simonnash> maybe this is because the guidelines are not published
<cr22rc_> I think this needs to be done.. but IMO it's not a blocker
<simonnash> without guidelines it is an easy excuse to no do it
<simonnash> not do it
<haleh> It will be impossible to encourage others to fix logging later.
Better document now so it is done correctly
<haleh> I volunteer to finish this if I get the info
* kgoodson has joined #Tuscany
<jsdelfino> thanks haleh, for now I think the question is: is it a release
blocker? do we block the publish of the distro until we have a page that
describes our logging guidelines?
<haleh> Not a blocker IMHO - But must be done immediately
<cr22rc_> Is this just a task of hunting down the few logging stmts and
using those as a guideline ? Or do they not meet what we want either?
<jsdelfino> I think we need a description of how our logging framework works
<ant_> and we need to come up with some sort of logging policy - 'these are
the type of activities where you shold add logging to your code'
<jsdelfino> if all of us and others understand how it's supposed to be done
then we'll have more logging in our code
<ant_> so is everyone now ok with not getting 292 completed for M1?
<simonnash> the framework is the place to start. policy is good too but
without te framework info, no-one can do anything
<cr22rc_> I look at if it's not a release blocker
<jsdelfino> not a release blocker imo
<haleh> Jeremy, you had mentioned once you have something that could be
used. Can we use that as a starting point?
<cr22rc_> if it's a release block we should encourage those that wrote so it
get done immediately if it's not a relase blocker I'll hack it.
<simonnash> ok with this as long as it is done soon. would put this higher
than the othe 29x
<cr22rc_> some time this week ? ok?
<jsdelfino> ok then, 292 is not a release blocker, but we want it get done
soon, can we move on?
<simonnash> yes, that would be fine with me
<cr22rc_> ok assign it to me
<jsdelfino> I think 346 is important, everybody's running into it all the
time
<jsdelfino> I volunteer to take care of it
<jsdelfino> 407 is not a release blocker IMO
<cr22rc_> hmm didn't think it still had that
<ant_> i just checked. it does, the 2 links at the bottom
<jsdelfino> ok then can u double check and close it if it's already fixed?
thx
<cr22rc_> I'll look at that
<jsdelfino> ok
<haleh> not on the list, but recently someone filed a defect on website that
samples do not build
<haleh> This is following the recent build instruction posting on website
<cr22rc_> fyi in building.txt: Note: Depending on the load of remote Mavan
2.0 repositories you may have to execute mvn several time until required
<cr22rc_> dependencies are all located in your local maven
repository. After you have completed a full successful build you
<cr22rc_> can use "mvn -o" which is much faster.
<haleh> that needs to be fixed.
<haleh> the point is the instruction needs to be corrected on the
javaproject page
<jsdelfino> haleh, which JIRA are you talking about? this looks like 346
<cr22rc_> ok I'll add that verbage to the website ... for 346
<haleh> luciano filed it
<jsdelfino> ok, are we done with 160?
<jsdelfino> I had contributed all the instructions on Wiki, then they got
integrated in the web site
<jsdelfino> is there still work to do on 160?
<haleh> I think it is done. The only piece is that under SCA for
JAVA->Installation.. we need to add info about distribution
<cr22rc_> I think 160 has adequate instruction on the web site. I don't
seen the need for the images
<haleh> so, 160 is done, but we need to update website's link for downloads
<ant_> (5:30. i have to leave in about 5 minutes sorry)
<jsdelfino> ok, good, next is 79, I guess it falls in the 29x category, not
a blocker for the realase?
<cr22rc_> the download binaries will go as resources on the site .. I assume
? once we agree?
<ant_> same as 29x
<jsdelfino> I think we need to publish the distros to the apache
distribution site, right?
<cr22rc_> not sure ... sorry that's probably out of context flow
<jsdelfino> look at the other incubator projects, synapse for example
<cr22rc_> ok ... not sure how we do that
<jsdelfino> 138 is not a release blocker IMO, it's just a link problem on
the Wiki
<jsdelfino> what do you think?
<jsdelfino> same for 215
<jsdelfino> and 344
<cr22rc_> +1 for 138 not a blocker
<ant_> +1 to 138, 215 and 344 not being blockers
<jsdelfino> ok, good, I think we've covered all JIRAs in the list, I'll take
care of adjusting the priority/category/assignee for all the JIRAs we
discussed today
<jsdelfino> the non-blockers will move to the Mx category
<jsdelfino> thanks for your patience, and thanks to those who are
volunteering to help finalize the docs
<rfeng> what's the deadline for 289? I'll limit the scope to only cover how
to contribute new impl/binding types to Tuscany
<rfeng> is it ok?
<ant_> ...have to go, bye guys...
<jsdelfino> yes good idea, the deadline is asap :)
* ant_ is now known as ant_away
<cr22rc_> sounds like its a good start ... we can update the website ... not
like changing the binary
<rfeng> should I just update the wiki and have somebody pull them into the
web site?
<jsdelfino> yeah, the binaries will be up on my web page in about an hour
(it takes so long to upload them for some reason)
<jsdelfino> can one or two of you do a quick smoke test to help verify that
I didn't make any stupid mistake?
<jsdelfino> thanks, ttyl guys
* jsdelfino has left #tuscany
<cr22rc_> I'll download it ... try a select sample or two .. should be
sufficient right ?
* kevin has quit IRC
* jliu has left #tuscany
* kgoodson_ has joined #Tuscany
* cr22rc_ is now known as cr22rc_away
* kgoodson has quit IRC (Read error: 104 (Connection reset by peer))
* kgoodson_ has quit IRC (Read error: 104 (Connection reset by peer))
* dkulp has quit IRC (Remote closed the connection)
