You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@tuscany.apache.org by kelvin goodson <ke...@thegoodsons.org.uk> on 2008/02/04 18:45:16 UTC
[SDO Java] IRC chat on release contents
This is a follow-up invitation for you to influence the current SDO Java
release content that Amita is managing. We tried having an IRC slot in
Amita's "working" day (actually starting at 9:30PM her time; summary posted
at [1]), but didn't get participation, which may have been due to the
early start from a US West Coast perspective, so I will host another session
on Friday afternoon UK time, precise time to be confirmed, but I would
suggest 5PM GMT / 9AM PST. Please post if you plan to attend, or would
attend if the time were different.
I'm cross posting to tuscany-user and tuscany-dev here, but please respond
to tuscany-user only, as I assume all who are subscribed to tuscany-dev are
also subscribed to tuscany-user.
We will use the usual Tuscany IRC channel on server irc.freenode.net,
channel #Tuscany
Kelvin.
[1]
http://mail-archives.apache.org/mod_mbox/ws-tuscany-user/200802.mbox/browser
Re: [SDO Java] IRC chat on release contents
Posted by kelvin goodson <ke...@thegoodsons.org.uk>.
Thanks for the summary Amita, comments below
On 07/02/2008, Amita Vadhavkar <am...@gmail.com> wrote:
>
> Please see below complete summary of the IRC chat we had on 4,5,6th Feb.
> The
> next one is as specified in previous response.
>
> There was an IRC chat where me and Kelvin participated on Feb 4th, 5th and
> 6th , 2008
> for SDO Java 1.1-incubating release. Below is the summary.
>
> 1) license files under distribution and sdo-api need to be
> reviewed/corrected
> as well as copyright text fro sources/resources under sdo-api need to be
> reviewed/corrected based on latest from OSOA and based on when the
> source/resource
> files are changed.
>
> 2) SDO Java will follow the same strategy of providing LICENSE files under
> src\main\resources\META-INF as before
>
> 3) http://cwiki.apache.org/confluence/display/TUSCANYWIKI/SDO+Java+Project
> is the place to find all unresolved JIRAs list marked for Java-SDO-next,
> Please see columns"Category" and "Comment" for summary of what was
> discussed
> in IRC.
>
> 4) Went through the list one by one and covered all JIRAs as below- ## are
> the ones which are important, need to discuss in detail on IRC and ML
> and see if can go in this release.
> **********************************************
> must have:-
> 1]1293 - important, patch is available, need to review and apply the
> patch,
> if somebody with OSGI expertise can pitch in, will be very helpful
I don't think this is what I meant to convey. I'm happy we've had excellent
OSGi skills assisting in the production of this patch, so I think what's
required is for me to carefully understand all that's been done here and to
apply the patch.
2]1063 - useful - Kelvin
> **********************************************
> nice to have:-
> 3]1862-workaround available in 1842 and no user issues on it so far
I don't think this is a nice to have, I think it is a defer. Apologies if I
misled, but there's a workaround in place, and this sort of issue is an
unresolved topic for discussion in the spec.
4]1838-Amita will check the existing solution against the last comment from
> Ron in ASF JIRA comments - *Note* Please check Amita's comment in JIRA and
> respond.
I took a look at your investigations, but won't have the time to analyse
them in depth before the IRC chat today.
*Kelvin's comment* If this facility for java serialization requires use of
> the SDOObjectOutputStream, document it in javadoc/website.
> ##5]1021 - + point is it's an omission from the spec, but it needs
> considerable work, important one - check in Friday IRC
> 6]257 - Paul have suggestions, partial fix. Amita need to check it and try
> to get it in the release.
> Resolved------->7]1514 - Amita - ML[das] and reassign comp to sdo-impl ->
> Kelvin will do it - give test case and track which revision(564600) did
> the
> change
> 8]1688 - Kelvin checking with Ron, check EMF fixes and see subclassing
> 9]1689 - similar to 1688 , Ron's comments on 1688, 1689 will be helpful
> 10]1835 - ML to Fuhwei - ask more details,what constraints apply, what
> investigation led to conclusion about "appinfo" - record these details in
> JIRA
> 11]1868 - patch available, test case available - Kelvin will look at it
>
> 12]1659 - linux surefire logs available, can check from that, in Kelvin's
> debian env could not reproduce the problem. if issue is not revealed by
> logs
> alone,
> then time required to recreate will be large.
>
> 13]1831 - Amita will check, looks useful, can have caveats in that once
> you've made something pluggable they might diverge
> and therefore a 3rd party version might rely on implementation details
> that
> are subject to change
> or they might miss out on fixes that we apply to the Tuscany version. If
> we
> made that clear that it was the users repsonsibility to
> take vcaar of these issues, then it sounds like a good commmunity type
> thing
> to do
>
> ##14]761 - check during next IRC-Kevin - put comment in jira if it is
> still
> a requirement-Amita
> what is the memory cost of having to init number of HelperContexts?
> Each one has at constrcution an instance of all HElpers that require to
> manage their own state -- i.e. not CopyHelper or
> EqualityHelper, but all the others . The state of each of these helpers
> is
> not large until you start registering types
> so i think there would only be significant cost if there were duplication
> in storage of metadata across typehelpers
> so the problem with multiple HelperContexts would be when loading XML
>
> 15]1182 - get tech findings from kelvin , see if anybody pops up, license
> side is clean, there are other examples of use of this in apache
>
> ##16]1361 - check on next IRC - *Kelvin* , check in IRC, easy, significant
> user impact
>
> 17]1360 -
> http://www.mail-archive.com/tuscany-dev@ws.apache.org/msg26306.html make
> comment in JIRA about both approached and apply 1360-Amita.patch.
> This was all the business where EMF has an issue that it cant handle the
> empty string as a valid enumeration. Make that comment in JIRA and then
> apply the 1360-Amita.patch. Wait for one day for Kelvin's response on ML ,
> i.e. till 7th Feb.
sorry, didn't get to it. I'm pretty sure that i was leaning towards your
approach, at least in the short term, because of the EMF deficiency that
surfaces in the more elegant approach.
**********************************************
> next release:-
> 18]1725-the scope is big and better justice can be given in next release,
> test schema is provided but no test code, javajet experience is useful
> here
> 19]1479 - Frank has suggested a workaround. And final solution can be
> supporting "large model" pattern similar to EMF
> 20]1180 - nothing done on spec, the workaround is not to use r/o
> properties
> 21]203 - minor JIRA with workaround and no watchers
> ##22]1002 - kelvin want to check on it in next irc, ideally should be
> resolved. so if time permits try to get it in this release.
this gives mixed messages since it is in the "next release" section. I
think that's because I'd like to see it done, but can't imagine getting to
it in the short term.
23]1129
> 24]477- defering to see first the spec discussion, also there will be
> performance hit
25]1433 - there will be performance hit
> 26]182 - check with Sebasitien and Frank on ML and mark their responses in
> JIRA comment - Kelvin
I added the SCA Assembly Model component to the JIRA to bring this back on
the SCA radar
27]1817 - too big
> ##28]1038***- talk in next IRC - is imp and major, axis2 databinding
> 29]152 - good new feature
>
> 30]1192 - depends on 2.1.1 spec - more detailed comments -
> what it requires is a specific list implementation ,so all demand
> created
> open content properties are
> created is-many = true, if you do a getList() on a property using string
> "foo" then a Property is created,
> and it must be referenced by the list, that measn that if the list
> contents
> is one, then getInstanceProperties
> on the DataObject retursn the Property with name "foo". If the list is
> emptied, then getInstanceProperties does
> not retuen the prop with name foo, and then if the same list has a new
> value
> added to it, then we still have hold
> of the original Property, by virtue of the reference held in the list. so
> we woulld get == as true for the Property
> instance returned before and after the emptying of the list
>
> Is this already this way in Tuscany ? - not sure, need to check
>
> This is a clarification that will be in the 2.1.1 spec, but it wouldn't
> hurt if it worked that way in our impl
> Kelvin will take a look
>
> What is - alter behaviour such that the demand created property is
> available
> in the applications metadata after the load operation. ?
> What will change after load operation?
> There's a different emphasis there, there may be a mixture of two issues
> here
> We should wait until the 2.1.1 discussions have completed before we
> address
> this
>
> 31]1493 - ongoing discussion with Agfa to discuss how we might absorb some
> of the ways they address SDO issues
>
> 32]1685 - useful new addition
> 33]551 - useful new addition
>
> "mostly" won't fix - need a confirmation on ML
> 34]67 - there has some work done in this area and JIRA is not clear on the
> particular issue. HelperContext came up after this issue. Some
> work done on DefaultHelperContext to work with thread local context. So,
> discuss and make appropriate comment in JIRA so the 2 watchers will
> get notification if the JIRA is closed as won't fix.
>
> 35]69 - first announce as summary on ML planning to close 67., 69 and then
> if noone objects, do it
> this is the only remaining subtask of 67, that we decided we would close
> IIRC, and see if anyone reopens it
> Please responde if there is some objection to close 67 and 69
>
> 36]832->dont close, defer, outside the SDO spec, at serialization the
> graph
> should be closed,and refs outside the closure of graph are broken,
> EMF doesnt have this restriction, and maybe in the future SDO won't,
> Fuhwei
> as acknowleged the fact that this is outside the spec, and is
> asking for SPIs
>
> 37]1327 - Ask Fuhwei if this is for xsd2java or wsdl2java and mark change
> in
> comp/release, at present is marked Fix Version - SCA release
> 38]1918 - make comment, dont change anything else, has dependency on spec
>
> 39]1932 - mark as sdo next and is nice to have - its not too difficult to
> do, and has some synergy
> with the issue of enumeration facets we discussed earlier , lets see if
> anyone wants it on Friday
>
> 40]2014 - possible nice to have - kelvin will put a comment, David is
> looking at it.
> **********************************************
> Appreciate all help in getting all important JIRAs in this release
>
> 5) After Friday's IRC - based on the final discussion - will mark in ASF
> JIRA, the "must have" JIRAs for Java-SDO-1.1, till
> then all discussed are marked for Fix Version "Java-SDO-Next". Also
> appropriate comments will be appended to JIRAs
>
> 6) Did not discuss any CTS JIRAs are these won't be part of the release.
>
> Kelvin, will you please see if by mistake I have missed any details?
>
> Regards,
> Amita
>
> On Feb 6, 2008 5:07 PM, kelvin goodson <ke...@thegoodsons.org.uk> wrote:
>
> > Lets confirm this at the suggested time.
> >
> > Kelvin.
> >
> > On 05/02/2008, Jeff Davis <jd...@hireright.com> wrote:
> > >
> > > The 9AM slot works for me, and I will try to attend.
> > >
> > > jeff
> > >
> > > On Mon, 2008-02-04 at 17:45 +0000, kelvin goodson wrote:
> > > > This is a follow-up invitation for you to influence the current SDO
> > Java
> > > > release content that Amita is managing. We tried having an IRC slot
> > in
> > > > Amita's "working" day (actually starting at 9:30PM her time; summary
> > > posted
> > > > at [1]), but didn't get participation, which may have been due to
> > the
> > > > early start from a US West Coast perspective, so I will host another
> > > session
> > > > on Friday afternoon UK time, precise time to be confirmed, but I
> > would
> > > > suggest 5PM GMT / 9AM PST. Please post if you plan to attend, or
> > would
> > > > attend if the time were different.
> > > >
> > > > I'm cross posting to tuscany-user and tuscany-dev here, but please
> > > respond
> > > > to tuscany-user only, as I assume all who are subscribed to
> > tuscany-dev
> > > are
> > > > also subscribed to tuscany-user.
> > > >
> > > > We will use the usual Tuscany IRC channel on server
> irc.freenode.net,
> > > > channel #Tuscany
> > > >
> > > > Kelvin.
> > > >
> > > > [1]
> > > >
> > >
> >
> http://mail-archives.apache.org/mod_mbox/ws-tuscany-user/200802.mbox/browser
> > >
> > > ---------------------------------------------------------------------
> > > To unsubscribe, e-mail: tuscany-user-unsubscribe@ws.apache.org
> > > For additional commands, e-mail: tuscany-user-help@ws.apache.org
> > >
> > >
> >
>
Re: [SDO Java] IRC chat on release contents
Posted by Amita Vadhavkar <am...@gmail.com>.
Please see below complete summary of the IRC chat we had on 4,5,6th Feb. The
next one is as specified in previous response.
There was an IRC chat where me and Kelvin participated on Feb 4th, 5th and
6th , 2008
for SDO Java 1.1-incubating release. Below is the summary.
1) license files under distribution and sdo-api need to be
reviewed/corrected
as well as copyright text fro sources/resources under sdo-api need to be
reviewed/corrected based on latest from OSOA and based on when the
source/resource
files are changed.
2) SDO Java will follow the same strategy of providing LICENSE files under
src\main\resources\META-INF as before
3) http://cwiki.apache.org/confluence/display/TUSCANYWIKI/SDO+Java+Project
is the place to find all unresolved JIRAs list marked for Java-SDO-next,
Please see columns"Category" and "Comment" for summary of what was discussed
in IRC.
4) Went through the list one by one and covered all JIRAs as below- ## are
the ones which are important, need to discuss in detail on IRC and ML
and see if can go in this release.
**********************************************
must have:-
1]1293 - important, patch is available, need to review and apply the patch,
if somebody with OSGI expertise can pitch in, will be very helpful
2]1063 - useful - Kelvin
**********************************************
nice to have:-
3]1862-workaround available in 1842 and no user issues on it so far
4]1838-Amita will check the existing solution against the last comment from
Ron in ASF JIRA comments - *Note* Please check Amita's comment in JIRA and
respond.
*Kelvin's comment* If this facility for java serialization requires use of
the SDOObjectOutputStream, document it in javadoc/website.
##5]1021 - + point is it's an omission from the spec, but it needs
considerable work, important one - check in Friday IRC
6]257 - Paul have suggestions, partial fix. Amita need to check it and try
to get it in the release.
Resolved------->7]1514 - Amita - ML[das] and reassign comp to sdo-impl ->
Kelvin will do it - give test case and track which revision(564600) did the
change
8]1688 - Kelvin checking with Ron, check EMF fixes and see subclassing
9]1689 - similar to 1688 , Ron's comments on 1688, 1689 will be helpful
10]1835 - ML to Fuhwei - ask more details,what constraints apply, what
investigation led to conclusion about "appinfo" - record these details in
JIRA
11]1868 - patch available, test case available - Kelvin will look at it
12]1659 - linux surefire logs available, can check from that, in Kelvin's
debian env could not reproduce the problem. if issue is not revealed by logs
alone,
then time required to recreate will be large.
13]1831 - Amita will check, looks useful, can have caveats in that once
you've made something pluggable they might diverge
and therefore a 3rd party version might rely on implementation details that
are subject to change
or they might miss out on fixes that we apply to the Tuscany version. If we
made that clear that it was the users repsonsibility to
take vcaar of these issues, then it sounds like a good commmunity type thing
to do
##14]761 - check during next IRC-Kevin - put comment in jira if it is still
a requirement-Amita
what is the memory cost of having to init number of HelperContexts?
Each one has at constrcution an instance of all HElpers that require to
manage their own state -- i.e. not CopyHelper or
EqualityHelper, but all the others . The state of each of these helpers is
not large until you start registering types
so i think there would only be significant cost if there were duplication
in storage of metadata across typehelpers
so the problem with multiple HelperContexts would be when loading XML
15]1182 - get tech findings from kelvin , see if anybody pops up, license
side is clean, there are other examples of use of this in apache
##16]1361 - check on next IRC - *Kelvin* , check in IRC, easy, significant
user impact
17]1360 -
http://www.mail-archive.com/tuscany-dev@ws.apache.org/msg26306.html make
comment in JIRA about both approached and apply 1360-Amita.patch.
This was all the business where EMF has an issue that it cant handle the
empty string as a valid enumeration. Make that comment in JIRA and then
apply the 1360-Amita.patch. Wait for one day for Kelvin's response on ML ,
i.e. till 7th Feb.
**********************************************
next release:-
18]1725-the scope is big and better justice can be given in next release,
test schema is provided but no test code, javajet experience is useful here
19]1479 - Frank has suggested a workaround. And final solution can be
supporting "large model" pattern similar to EMF
20]1180 - nothing done on spec, the workaround is not to use r/o properties
21]203 - minor JIRA with workaround and no watchers
##22]1002 - kelvin want to check on it in next irc, ideally should be
resolved. so if time permits try to get it in this release.
23]1129
24]477- defering to see first the spec discussion, also there will be
performance hit
25]1433 - there will be performance hit
26]182 - check with Sebasitien and Frank on ML and mark their responses in
JIRA comment - Kelvin
27]1817 - too big
##28]1038***- talk in next IRC - is imp and major, axis2 databinding
29]152 - good new feature
30]1192 - depends on 2.1.1 spec - more detailed comments -
what it requires is a specific list implementation ,so all demand created
open content properties are
created is-many = true, if you do a getList() on a property using string
"foo" then a Property is created,
and it must be referenced by the list, that measn that if the list contents
is one, then getInstanceProperties
on the DataObject retursn the Property with name "foo". If the list is
emptied, then getInstanceProperties does
not retuen the prop with name foo, and then if the same list has a new value
added to it, then we still have hold
of the original Property, by virtue of the reference held in the list. so
we woulld get == as true for the Property
instance returned before and after the emptying of the list
Is this already this way in Tuscany ? - not sure, need to check
This is a clarification that will be in the 2.1.1 spec, but it wouldn't
hurt if it worked that way in our impl
Kelvin will take a look
What is - alter behaviour such that the demand created property is available
in the applications metadata after the load operation. ?
What will change after load operation?
There's a different emphasis there, there may be a mixture of two issues
here
We should wait until the 2.1.1 discussions have completed before we address
this
31]1493 - ongoing discussion with Agfa to discuss how we might absorb some
of the ways they address SDO issues
32]1685 - useful new addition
33]551 - useful new addition
"mostly" won't fix - need a confirmation on ML
34]67 - there has some work done in this area and JIRA is not clear on the
particular issue. HelperContext came up after this issue. Some
work done on DefaultHelperContext to work with thread local context. So,
discuss and make appropriate comment in JIRA so the 2 watchers will
get notification if the JIRA is closed as won't fix.
35]69 - first announce as summary on ML planning to close 67., 69 and then
if noone objects, do it
this is the only remaining subtask of 67, that we decided we would close
IIRC, and see if anyone reopens it
Please responde if there is some objection to close 67 and 69
36]832->dont close, defer, outside the SDO spec, at serialization the graph
should be closed,and refs outside the closure of graph are broken,
EMF doesnt have this restriction, and maybe in the future SDO won't, Fuhwei
as acknowleged the fact that this is outside the spec, and is
asking for SPIs
37]1327 - Ask Fuhwei if this is for xsd2java or wsdl2java and mark change in
comp/release, at present is marked Fix Version - SCA release
38]1918 - make comment, dont change anything else, has dependency on spec
39]1932 - mark as sdo next and is nice to have - its not too difficult to
do, and has some synergy
with the issue of enumeration facets we discussed earlier , lets see if
anyone wants it on Friday
40]2014 - possible nice to have - kelvin will put a comment, David is
looking at it.
**********************************************
Appreciate all help in getting all important JIRAs in this release
5) After Friday's IRC - based on the final discussion - will mark in ASF
JIRA, the "must have" JIRAs for Java-SDO-1.1, till
then all discussed are marked for Fix Version "Java-SDO-Next". Also
appropriate comments will be appended to JIRAs
6) Did not discuss any CTS JIRAs are these won't be part of the release.
Kelvin, will you please see if by mistake I have missed any details?
Regards,
Amita
On Feb 6, 2008 5:07 PM, kelvin goodson <ke...@thegoodsons.org.uk> wrote:
> Lets confirm this at the suggested time.
>
> Kelvin.
>
> On 05/02/2008, Jeff Davis <jd...@hireright.com> wrote:
> >
> > The 9AM slot works for me, and I will try to attend.
> >
> > jeff
> >
> > On Mon, 2008-02-04 at 17:45 +0000, kelvin goodson wrote:
> > > This is a follow-up invitation for you to influence the current SDO
> Java
> > > release content that Amita is managing. We tried having an IRC slot
> in
> > > Amita's "working" day (actually starting at 9:30PM her time; summary
> > posted
> > > at [1]), but didn't get participation, which may have been due to
> the
> > > early start from a US West Coast perspective, so I will host another
> > session
> > > on Friday afternoon UK time, precise time to be confirmed, but I
> would
> > > suggest 5PM GMT / 9AM PST. Please post if you plan to attend, or
> would
> > > attend if the time were different.
> > >
> > > I'm cross posting to tuscany-user and tuscany-dev here, but please
> > respond
> > > to tuscany-user only, as I assume all who are subscribed to
> tuscany-dev
> > are
> > > also subscribed to tuscany-user.
> > >
> > > We will use the usual Tuscany IRC channel on server irc.freenode.net,
> > > channel #Tuscany
> > >
> > > Kelvin.
> > >
> > > [1]
> > >
> >
> http://mail-archives.apache.org/mod_mbox/ws-tuscany-user/200802.mbox/browser
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: tuscany-user-unsubscribe@ws.apache.org
> > For additional commands, e-mail: tuscany-user-help@ws.apache.org
> >
> >
>
Re: [SDO Java] IRC chat on release contents
Posted by kelvin goodson <ke...@thegoodsons.org.uk>.
Lets confirm this at the suggested time.
Kelvin.
On 05/02/2008, Jeff Davis <jd...@hireright.com> wrote:
>
> The 9AM slot works for me, and I will try to attend.
>
> jeff
>
> On Mon, 2008-02-04 at 17:45 +0000, kelvin goodson wrote:
> > This is a follow-up invitation for you to influence the current SDO Java
> > release content that Amita is managing. We tried having an IRC slot in
> > Amita's "working" day (actually starting at 9:30PM her time; summary
> posted
> > at [1]), but didn't get participation, which may have been due to the
> > early start from a US West Coast perspective, so I will host another
> session
> > on Friday afternoon UK time, precise time to be confirmed, but I would
> > suggest 5PM GMT / 9AM PST. Please post if you plan to attend, or would
> > attend if the time were different.
> >
> > I'm cross posting to tuscany-user and tuscany-dev here, but please
> respond
> > to tuscany-user only, as I assume all who are subscribed to tuscany-dev
> are
> > also subscribed to tuscany-user.
> >
> > We will use the usual Tuscany IRC channel on server irc.freenode.net,
> > channel #Tuscany
> >
> > Kelvin.
> >
> > [1]
> >
> http://mail-archives.apache.org/mod_mbox/ws-tuscany-user/200802.mbox/browser
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: tuscany-user-unsubscribe@ws.apache.org
> For additional commands, e-mail: tuscany-user-help@ws.apache.org
>
>
Re: [SDO Java] IRC chat on release contents
Posted by Jeff Davis <jd...@hireright.com>.
The 9AM slot works for me, and I will try to attend.
jeff
On Mon, 2008-02-04 at 17:45 +0000, kelvin goodson wrote:
> This is a follow-up invitation for you to influence the current SDO Java
> release content that Amita is managing. We tried having an IRC slot in
> Amita's "working" day (actually starting at 9:30PM her time; summary posted
> at [1]), but didn't get participation, which may have been due to the
> early start from a US West Coast perspective, so I will host another session
> on Friday afternoon UK time, precise time to be confirmed, but I would
> suggest 5PM GMT / 9AM PST. Please post if you plan to attend, or would
> attend if the time were different.
>
> I'm cross posting to tuscany-user and tuscany-dev here, but please respond
> to tuscany-user only, as I assume all who are subscribed to tuscany-dev are
> also subscribed to tuscany-user.
>
> We will use the usual Tuscany IRC channel on server irc.freenode.net,
> channel #Tuscany
>
> Kelvin.
>
> [1]
> http://mail-archives.apache.org/mod_mbox/ws-tuscany-user/200802.mbox/browser
---------------------------------------------------------------------
To unsubscribe, e-mail: tuscany-user-unsubscribe@ws.apache.org
For additional commands, e-mail: tuscany-user-help@ws.apache.org