Monday OCT 3, 2006 Tuscany regular IRC chat log.

[Rick <cr...@gmail.com>]

[11:35] <jboynes> morning all
[11:35] <rfeng> morning
[11:35] * simonnash has joined #Tuscany
[11:35] <cr22rc> hello
[11:35] <jboynes> I'd like to chat about the M2 release but that's 
likely to soak up a lot of time
[11:35] <jboynes> before doing that, is there anything else others would 
like to bring up?
[11:36] <simonnash> M2 is my main interest at this point
[11:38] <jboynes> ok, I'll kick off on where I think we are and I'd like 
feedback from everyone about what's left to do
[11:38] <jboynes> we need to track this on the list but we can share 
info here easily
[11:38] <jboynes> I just posted tags of the parent pom and buildtools 
that everything depends on
[11:39] <jboynes> these are things that the other distros need in place 
to build
[11:39] <simonnash> your post to the list mentioned SDO
[11:39] <jboynes> I'd ask everyone to review them and respond to the 
vote thread
[11:39] <simonnash> did you mean SCAas well or just SDO?
[11:40] <jboynes> I meant SDO - they will be needed for DAS and SCA as 
well but I believe SDO is just about ready to go
[11:40] <simonnash> ok
[11:41] <jboynes> and SDO is a pre-req for DAS and SCA
[11:43] <jboynes> kgoodson: is SDO ready to go?
[11:44] <kgoodson> not quite, sorry,  otp,  can i get back to you in c. 
45 mins
[11:44] <jboynes> ok
[11:45] <jboynes> I take back the Q anyway  - I think we still need to 
finish legal review
[11:45] <jboynes> and we had the issue yesterday of how we were going to 
package the samples in the distro
[11:46] <jboynes> one thing we did yesterday for SDO is move the samples 
into that tree
[11:47] <jboynes> so instead of samples/sdo we now have sdo/sample
[11:47] <jboynes> same with the distribution goal
[11:48] <jboynes> that means all the sdo stuff is in one source tree 
(and hence in one source distro)
[11:48] <jboynes> which makes tagging it much much easier
[11:48] <simonnash> why are we not changing spec/sdo to sdo/spec?
[11:48] <jboynes> we thought the spec would be a separate distro
[11:48] <jboynes> as we would want to rev our impl separately
[11:50] * lresende has joined #tuscany
[11:50] <simonnash> is the SDO spec at some standard level such as 2.1 
or 2.01?
[11:50] <jboynes> 2.0.1 I belive
[11:50] <simonnash> or is it between these levels?
[11:50] <simonnash> i know Frank had put some things in from 2.1
[11:50] <simonnash> not sure if they impacted spec APIs
[11:50] <jboynes> ah, they may have
[11:51] <jboynes> I've been wondering about that but its what the sdo 
folk seem to want to do
[11:51] <simonnash> we should check.  if it is not a standard level then 
it is really tied to our impl
[11:52] <jboynes> I'd be dubious about releasing a non-standard level
[11:52] <jboynes> but then I have JCP scars ;)
[11:52] <simonnash> if our impl needs it then we need to release it
[11:53] <jboynes> i don't think we should release anything in the 
commonj namespace that does not conform
[11:53] <simonnash> let's find out first if there is an issue
[11:53] <jboynes> yeah
[11:53] <simonnash> anyone on the chat who can confirm?
[11:54] <simonnash> are you thinking we would package the sdo spec 
separate from the impl?
[11:54] <jboynes> separate and as part of
[11:54] <jboynes> a user can get it separately
[11:55] <jboynes> but they also get it as part of the impl's distro
[11:55] <simonnash> ok
[11:57] <rfeng> for milestone releases, I think it's fine to only have 
the impl's distro. would it simplify anything?
[11:57] <bhdaniel> Yes, Frank added some 2.1 methods to TypeHelper in 
spec/sdo
[11:57] <jboynes> is it all of 2.1 or just parts of it?
[11:58] <jboynes> and is 2.1 released?
[11:58] <simonnash> not yet
[11:58] <bhdaniel> Just parts
[11:58] <jboynes> sounds like we have an issue then
[12:00] <jboynes> let's take that to the list and move on
[12:00] <jboynes> ok?
[12:01] <jboynes> lresende: where are we with DAS?
[12:01] <simonnash> yes ok to taking that issue to the list
[12:03] <jboynes> lresende: where are we with DAS?
[12:03] <jboynes> guess lresende  is lurking
[12:03] <jboynes> ok, where are we with SCA?
[12:04] <lresende> jboynes: i have just sent the mail for RC1
[12:04] <lresende> i was waitting for a place to post it, and kgoodson 
helpedme over the weekend and is now available at :
[12:04] <lresende> http://people.apache.org/~kelvingoodson/das_java/RC1
[12:04] <bhdaniel> Will we need to move samples/das to das/samples?
[12:05] <jboynes> "need", no
[12:05] * Venkat has joined #tuscany
[12:05] <lresende> not sure if we will need that, for das we have a 
binary and sample distribution
[12:06] <jboynes> you're including the samples in the source distro though
[12:07] <lresende> y
[12:07] <jboynes> how do we tag that?
[12:07] * isilval_ has joined #tuscany
[12:07] <lresende> not sure :) if taging is the problem, we can move 
from samples/das to das/samples
[12:09] <jboynes> I think it would make it easier to move
[12:09] <jboynes> I think it would make it easier if it was moved
[12:10] <lresende> ok, i can work on this
[12:10] <kgoodson> hi, i'm back now and watching this -- i haven't had a 
chance to come up with a proposal on samples inclusion yet
[12:11] <jboynes> kgoodson: as you have recent experience, did moving 
the samples under sdo make cutting the branch easier?
[12:11] <kgoodson> i would have thought there would have been a maven 
pattern for this
[12:11] <kgoodson> as to ease,  yes,  a bit easier
[12:11] <kgoodson> and moving distro/sdo to sdo/distro helps too
[12:12] <jboynes> lresende: should I do the same for das?
[12:12] <kgoodson> so the only thing now is having to treat spec and 
implementation separately,  but we have gone through that and decided 
that the split is something that is worthwhile
[12:13] <jboynes> (you need a committer to move things around :) )
[12:13] <lresende> sure :)
[12:13] <jboynes> ok
[12:13] <jboynes> kgoodson: if you missed it we have an issue with sdo 
spec and are taking it to the list
[12:14] <kgoodson> ok,  i will take a look,  is it htere yet?
[12:14] <jboynes> no
[12:14] <jboynes> been too busy chatting :)
[12:14] <jboynes> back to SCA?
[12:15] * Venkat has quit IRC
[12:16] <jboynes> let me ask the big questions - which extensions should 
be in?
[12:16] <jboynes> axis?
[12:16] <rfeng> +1 for axis
[12:17] * Venkat has joined #tuscany
[12:17] <jboynes> rfeng: is that +1 to have it in, +1 it's ready, or 
something else?
[12:17] <rfeng> +1 meaning it should be in
[12:18] <jboynes> "should" or "must" ?
[12:18] <rfeng> for M2 release, right?
[12:18] <jboynes> yes
[12:19] <rfeng> I feel it's "must"
[12:19] <simonnash> must
[12:19] <jboynes> anyone else?
[12:20] <cr22rc> didn't think it would up for debate but +1
[12:21] <jboynes> ok
[12:21] <jboynes> is it ready?
[12:21] <rfeng> I give it +1 except for the async ws support
[12:22] * jsdelfino has joined #tuscany
[12:22] <rfeng> cr22rc? what's your view?
[12:23] <cr22rc> I think basically it is for synch. but not a whole lot 
of testing done
[12:23] <jboynes> cr22rc: has there been enough testing to release it?
[12:24] <jboynes> and rfeng do we release with async as it is now?
[12:24] <cr22rc> no not in a real world testing IMO ... I'm just  
starting to see some interaction on the wire. But have not found a show 
stopper yet.
[12:25] <rfeng> we got async half-way, there are some discussions on the 
dev-list
[12:25] * YangZHONG has joined #tuscany
[12:26] <rfeng> if M2 is to be released this week, I'm not sure if we 
can finish the async part, needs ignacio's input
[12:26] <isilval_> wrt async, we're working on finalizing the 
integration w/databinding
[12:27] <jboynes> I don't think we should release something that doesn't 
work
[12:27] <isilval_> hopefully in the next couple of days we should be 
able to get helloworldws sample (client and server) working
[12:28] <jboynes> unless we don't think that the stuff that doesn't work 
doesn't matter
[12:28] <simonnash> i think we need to wait until it works well enough 
to release
[12:28] <rfeng> agree
[12:28] <jboynes> right - we can have shortcomings but we need to 
document them
[12:28] <cr22rc> IMO things seem rather rough
[12:28] <jboynes> but what we say works better work :)
[12:28] <simonnash> agree to a point.. as long as the shortcomings 
aren't too major
[12:29] <jboynes> so rfeng isilval_ , when do you think it will be "good 
enough" to release?
[12:29] <simonnash> and agree completely that what we say works must work
[12:30] <isilval_> couple of days for the first milestone
[12:30] <cr22rc> how concerned are we with the end user experience that 
when things don't work they get something meaningful on how to fix it?
[12:31] <rfeng> but we can leave the code as-is (assuming it's not 
breaking other features and we're not announcing the not-working-yet 
feature), right?
[12:33] <jboynes> rfeng: I don't see us ripping code out
[12:33] <jboynes> but what we say is there better work (at least to the 
extent that we say it works)
[12:34] <rfeng> agree, just want to make sure
[12:34] <jboynes> IMO, no surprises for the user
[12:34] <jboynes> fits with cr22rc' s comment - meaningful error 
messages help
[12:34] <jboynes> could be classified as part of "works"
[12:35] <jboynes> so, axis - couple of days?
[12:35] <rfeng> yes from me
[12:35] <isilval_> that's what I'm thinking
[12:35] <jboynes> ok
[12:36] <jboynes> next, celtix?
[12:36] <jboynes> not seen any activity on it recently and given all the 
changes that would make we question whether it should be included
[12:36] <simonnash> there are degrees of brokenness with error msgs.  
having any kind of error msg is better than the "die horribly" scenario
[12:36] <jboynes> also dkulp said a long time ago that we should really 
migrate to cxf
[12:37] <cr22rc> is anyone supporting it ?
[12:37] <jboynes> so IMO we should cut celtix for now
[12:37] <dkulp> Right, but cxf isn't "ready" yet.
[12:37] <jboynes> dkulp: is that ok with you?
[12:37] <dkulp> Yep. I'm fine with it.
[12:37] <jboynes> anyone else?
[12:38] <dkulp> All the code is easily recoverable later when we start 
the migration.
[12:38] <jboynes> and we can release it as an extension
[12:38] * rajith_home has joined #tuscany
[12:38] <dkulp> Actually, can it be "moved" to a "contrib" section instead?
[12:39] <jboynes> "contrib" being something between sandbox and trunk?
[12:39] <dkulp> Well, kind of a "semi-supported" part of trunk.
[12:40] <jboynes> yeah - in-build but not in-distro?
[12:40] <dkulp> Build's and supposedly works, but you're on you're own 
if you use it.
[12:40] <jboynes> +1 from me but that something we should push to the list
[12:40] <dkulp> Right.   In the source distro, but not the binary distro.
[12:41] <jboynes> I can think of quite a few things in that category
[12:41] <cr22rc> does it have samples ?
[12:41] <dkulp> Actually, all of this needs to be pushed to the list.   
Decisions are not allowed to be done on IRC.
[12:41] <dkulp> That aside...
[12:41] <rajith_home> jboynes, I talked to the WG about your comment on 
a component might expect a raw JMS message instead of a object[]. so 
they are going to add that to the spec :-)
[12:41] <dkulp> There is a celtix sample someplace.
[12:41] <jboynes> rajith_home: om
[12:41] <jboynes> ok
[12:41] <cr22rc> dkulp : does it work ?
[12:41] <simonnash> is the source distro the whole of trunk?
[12:42] <jboynes> can we come back to that - trying to discuss the 
release atm
[12:42] <dkulp> According to Jervis, yes.
[12:42] <dkulp> However, the Celtix stuff is very limitted.
[12:42] <jboynes> simonnash: it's easiest if it is
[12:42] <dkulp> Only SDO (does not use databinding stuff), no tomcat 
support (only standalone), etc...
[12:42] <jboynes> well, the "sca" bit anyway (sdo and das being separate)
[12:42] * YangZHONG has quit IRC (Read error: 104 (Connection reset by 
peer))
[12:43] <jboynes> in the interest of time, celtix -> "contrib" for this 
release?
[12:43] <simonnash> ok, until just now I had expected it would have the 
same contents as the bionry... but there is no particular reason why 
that should be the case
[12:43] * YangZHONG has joined #tuscany
[12:44] <jboynes> rmi ?
[12:45] <jboynes> must/nice/no? and if must/nice, is it ready?
[12:45] <dkulp> +1 to celtix -> "contrib"
[12:45] <isilval_> rfeng: are you available for a one-on-one over google 
talk?
[12:45] <rfeng> sure after this chat
[12:45] <jboynes> you could just chat here :)
[12:45] <isilval_> after this chat we can continue on IRC, I meant now
[12:46] <jboynes> :)
[12:46] <Venkat> jboynes... rmi works... just that I must improve 
coverage.. Ant mentioned he will help with that as I was looking into 
the PropertyLoading stuff
[12:47] <jboynes> ok
[12:47] <jboynes> axiom databinding?
[12:47] <jboynes> assume that's a prereq for axis
[12:47] <rfeng> it has to be in for axis2
[12:47] <jboynes> is it ready?
[12:47] <rfeng> I assume so
[12:47] <jboynes> ok
[12:47] <jboynes> sdo databinding?
[12:48] <rfeng> if you use sdo with axis2, yes
[12:48] <jboynes> I assume, "if you use sdo" ?
[12:49] <jboynes> as in, I assume you can use axis2 without sdo :)
[12:49] <rfeng> right
[12:49] <jboynes> is it ready?
[12:49] <rfeng> maybe so far it's the only way to work with complex types
[12:49] <rfeng> I assume it's ready
[12:49] <jboynes> who owns it?
[12:50] <rfeng> I'm, I guess
[12:51] <jboynes> I'll take that as a "yes" then :)
[12:51] <jboynes> jaxb databinding?
[12:51] <rfeng> nobody is driving the test so far
[12:51] <jboynes> I originally had that in there for celtix - should we 
move this to contrib as well?
[12:52] <rfeng> I didn't have a chance to test it with axis2
[12:52] * halehM has joined #tuscany
[12:52] <rfeng> fine with you and we can move it back when it's ready
[12:52] <jboynes> ok
[12:52] <jboynes> javascript container?
[12:53] <jboynes> ant's not here, does anyone else know where we are?
[12:53] <Venkat> I think it has to be enabled for properties... and async...
[12:54] <jboynes> do you think its good enough, or should we move it to 
contrib as well?
[12:54] <Venkat> which should be done once the base the base loaders are 
thro for Properties
[12:55] <Venkat> could we please get to this with Ant... :) in the 
interest of my life
[12:55] <Venkat> my gut feel is with a couple of days work... it must be 
there...
[12:56] <Venkat> When I added e4x and reference to that... I saw it work 
decently
[12:56] <jboynes> ok
[12:56] <jboynes> so - in, needs a couple of days
[12:57] <jboynes> we'll look at it again when ant's around and when 
we're finishing axis?
[12:57] <jboynes> ok?
[12:57] <Venkat> that sounds good :)
[12:58] <jboynes> ok
[12:58] <jboynes> spring?
[12:58] <jboynes> I'm dubious atm but we need input from jmarino and 
probably andy
[12:59] <rfeng> seems so
[12:59] <jboynes> ok, the things we had originally classified as "optional"
[12:59] <jboynes> jsonrpc osgi xmlbenas castor groovy anything else ...
[13:00] <jboynes> I'm figuring "out" or "contrib" unless someone jumps 
up real soon
[13:00] <Venkat> Ruby ?
[13:00] <simonnash> need to leave now.  bye.
[13:00] <jboynes> is it ready?
[13:00] <Venkat> its about the same state as Javascript... just that 
properties need to work
[13:00] <Venkat> I am keeping Async out...
[13:01] <jboynes> so resync in a couple of days?
[13:01] <Venkat> yes...
[13:01] <jboynes> ok
[13:01] <jboynes> samples?
[13:02] <Venkat> jboynes, is there anything up that brings up the 
standalone... say from command prompt..
[13:03] <Venkat> just about need this to complete the rmi samples... so 
that I can have one rmiserver and a rmiclient runtimes..
[13:03] <jboynes> java -jar launcher.jar foo.jar
[13:03] <jboynes> where foo.jar contains the application composite
[13:04] <Venkat> and the extensions ?
[13:04] * kgoodson has quit IRC (Read error: 104 (Connection reset by 
peer))
[13:04] <jboynes> added by the DirScanExtender from the extensions directory
[13:05] <Venkat> ok... will try this one out and get to the ML
[13:05] <jboynes> ok
[13:06] <jboynes> there was talk of samples re-org - is anyone working 
on that?
[13:07] * simonnash has quit IRC (Read error: 104 (Connection reset by 
peer))
[13:07] <Venkat> there is Jojo who owned up to create the Calculator 
Combo sample..
[13:07] <Venkat> I shall be helping him with that... I have raised a 
JIRA for that
[13:07] <jboynes> ok
[13:08] <Venkat> the intention is to show different containers working 
together in an assembly.
[13:08] <jboynes> good intention :) would be cool
[13:08] <jboynes> any idea on when it would be ready?
[13:08] <jboynes> cr22rc: what about bigbank?
[13:08] <Venkat> Jojo is yet to get back on this.. but know for sure 
that he is on it..
[13:09] <Venkat> will ask him to update status on ML tomorrow
[13:09] <cr22rc> well for this moring I got the logon on the wire to 
work.  The client is not consuming it though.  First time seeing client 
/server interacting
[13:10] <cr22rc> I'm being hopeful in a couple of days... but really 
don't know what gottcha I run in to
[13:11] <cr22rc> the deployment story is real messy
[13:12] <jboynes> have you switched to using the plugin yet?
[13:13] <jboynes> I'm going to have to run in a couple
[13:13] <cr22rc> well I do use it . but then i move everything around to 
the old deployment and have my own scdl that loads extensions in order I 
know works
[13:13] <cr22rc> 
http://mail-archives.apache.org/mod_mbox/ws-tuscany-dev/200609.mbox/browser
[13:13] <jboynes> it seems that we need to re-sync in a couple of days 
for axis, js, ruby, possibly bigbank
[13:15] <jboynes> ApacheCon is next week
[13:16] <jboynes> can we get a release tagged by then (all done bar the 
voting)?
[13:16] * lmsurprenant has quit IRC (Read error: 110 (Connection timed 
out))
[13:17] <jboynes> ?
[13:18] <cr22rc> indication of confidence ?
[13:18] <jboynes> can we do it?
[13:19] <cr22rc> how about docs that go along with it ?
[13:20] <jboynes> yep - will we have them?
[13:21] * lresende has quit IRC
[13:21] <Venkat> do we have a list of all that is to be made ?
[13:21] <jboynes> I think your page on the wiki is the best summary
[13:24] <cr22rc> I think we can pull something together.  But there are 
honestly doubts in my mind if the "quality" will be there.  Ok .. now 
everyon throw stone at me.
[13:25] * lresende has joined #tuscany
[13:25] <Venkat> I am with you on this cr22rc
[13:28] * kgoodson_away has joined #Tuscany
[13:28] * kgoodson_away is now known as kgoodson
[13:30] <jboynes> cr22rc: what do you think needs to be done to get the 
right "quality" level?
[13:30] * Venkat has quit IRC
[13:31] * kgoodson is now known as kgoodson_away
[13:31] <cr22rc> more testing as the user does
[13:31] <cr22rc> not just unit test.
[13:32] <cr22rc> fix issues when there are problems that you don't need 
to go into a debugger to isolate the issue.
[13:32] <jboynes> can you be more detailed on "more testing"
[13:33] <jboynes> I mean, more is always good - but how much, and how do 
we get it done?
[13:33] <cr22rc> maybe I'll feel better towards the end of this week 
when I can take samples and just run them.
[13:34] <cr22rc> getting BB up and running is not a cake walk.
[13:37] <cr22rc> so can I ask jeremy your opinion?
[13:40] <jboynes> my opinion?
[13:41] <jboynes> I think we need more testing


---------------------------------------------------------------------
To unsubscribe, e-mail: tuscany-dev-unsubscribe@ws.apache.org
For additional commands, e-mail: tuscany-dev-help@ws.apache.org

Re: Monday OCT 3, 2006 Tuscany regular IRC chat log.

[Rick <cr...@gmail.com>]

Should have been Oct 2 :-)
(can I use not enough caffeine excuse) :-)

---------------------------------------------------------------------
To unsubscribe, e-mail: tuscany-dev-unsubscribe@ws.apache.org
For additional commands, e-mail: tuscany-dev-help@ws.apache.org