[Scons-dev] Cross-language support
William Blevins
wblevins001 at gmail.com
Wed Jul 29 20:19:28 EDT 2015
Jason.
FYI. The failing tests you listed match the list from June. The
action-test.py may be new.
https://bitbucket.org/scons/scons/pull-requests/237/issue-2264-cross-language-scanner-support/diff#comment-7323143
V/R,
William
On Wed, Jul 29, 2015 at 8:16 PM, William Blevins <wblevins001 at gmail.com>
wrote:
> Jason,
>
> As previously stated in the pull request comments, I expect some tests to
> fail currently because of the recursive header install issue (topic #1):
> <lang>SUFFIXES, HeaderInstall.py. The other tests may or may not be
> related.
>
> There is more value in looking at those tests first. It may also be
> valuable to go back 3-4 commits to the closest ancestor of the default
> branch and run tests there, so that we can see if some of the tests failed
> previously. Also, if we think its a problem of the current location of
> that divergent head, then we can see if a rebase fixes some of them.
>
> V/R,
> William
>
> On Wed, Jul 29, 2015 at 7:09 PM, Jason Kenny <dragon512 at live.com> wrote:
>
>> ok not any better...
>>
>>
>> Failed the following 16 tests:
>> test\Batch\action-changed.py
>> test\CPPSUFFIXES.py
>> test\DSUFFIXES.py
>> test\Fortran\FORTRANSUFFIXES.py
>> test\HeaderInstall.py
>> test\MSVS\vs-9.0-exec.py
>> test\Win32\mingw.py
>> test\exitfns.py
>> test\implicit\IMPLICIT_COMMAND_DEPENDENCIES.py
>> test\import.py
>> test\long-lines\signature.py
>> test\scons-time\run\config\python.py
>> test\scons-time\run\option\python.py
>> test\sconsign\script\SConsignFile.py
>> test\sconsign\script\Signatures.py
>> test\sconsign\script\no-SConsignFile.py
>>
>> I would need to dig down to see what is failing and why
>> Jason
>>
>>
>> *From:* Jason Kenny <dragon512 at live.com>
>> *Sent:* Wednesday, July 29, 2015 5:38 PM
>> *To:* SCons developer list <scons-dev at scons.org>
>> *Subject:* Re: [Scons-dev] Cross-language support
>>
>> ok I have it fixed... I am rerunning the tests now
>> Jason
>>
>> *From:* Jason Kenny <dragon512 at live.com>
>> *Sent:* Wednesday, July 29, 2015 5:30 PM
>> *To:* SCons developer list <scons-dev at scons.org>
>> *Subject:* Re: [Scons-dev] Cross-language support
>>
>> I must have messed up. I only see debugCount and JniHeaderDir
>> bookmarks at the moment.
>>
>> Jason
>>
>> *From:* William Blevins <wblevins001 at gmail.com>
>> *Sent:* Wednesday, July 29, 2015 4:50 PM
>> *To:* SCons developer list <scons-dev at scons.org>
>> *Subject:* Re: [Scons-dev] Cross-language support
>>
>> That's a SCons fork. Not an HG branch (in case its a terminology
>> problem). Make sure you are using the CrossLanguage bookmark.
>>
>> V/R,
>> William
>>
>> On Wed, Jul 29, 2015 at 5:45 PM, Jason Kenny <dragon512 at live.com> wrote:
>>
>>> branch *SCons_20150323 *
>>>
>>> I can will setup from scratch again and try again.
>>>
>>> Jason
>>>
>>> *From:* William Blevins <wblevins001 at gmail.com>
>>> *Sent:* Wednesday, July 29, 2015 4:38 PM
>>> *To:* SCons developer list <scons-dev at scons.org>
>>> *Subject:* Re: [Scons-dev] Cross-language support
>>>
>>> Jason,
>>>
>>> Are you sure you are using the correct commit revision?
>>>
>>> None of those tests are failing on my Linux distro, and I haven't
>>> modified any of those tests either. Plus, tests that I am expecting to
>>> fail are apparently passing.
>>>
>>> V/R,
>>> William
>>>
>>> On Wed, Jul 29, 2015 at 5:06 PM, Jason Kenny <dragon512 at live.com> wrote:
>>>
>>>> So a quick pass of the tests has these failures
>>>>
>>>> Failed the following 15 tests:
>>>> test\Batch\action-changed.py
>>>> test\MSVS\vs-9.0-exec.py
>>>> test\Win32\mingw.py
>>>> test\exitfns.py
>>>> test\implicit\IMPLICIT_COMMAND_DEPENDENCIES.py
>>>> test\import.py
>>>> test\long-lines\signature.py
>>>> test\option\debug-count.py
>>>> test\option\debug-multiple.py
>>>> test\option\debug-objects.py
>>>> test\scons-time\run\config\python.py
>>>> test\scons-time\run\option\python.py
>>>> test\sconsign\script\SConsignFile.py
>>>> test\sconsign\script\Signatures.py
>>>> test\sconsign\script\no-SConsignFile.py
>>>>
>>>> *From:* Jason Kenny <dragon512 at live.com>
>>>> *Sent:* Wednesday, July 29, 2015 2:32 PM
>>>> *To:* SCons developer list <scons-dev at scons.org>
>>>> *Subject:* Re: [Scons-dev] Cross-language support
>>>>
>>>> Ok let me update the branch and re-run everything.
>>>>
>>>> Jason
>>>>
>>>> *From:* William Blevins <wblevins001 at gmail.com>
>>>> *Sent:* Wednesday, July 29, 2015 10:30 AM
>>>> *To:* SCons developer list <scons-dev at scons.org>
>>>> *Subject:* Re: [Scons-dev] Cross-language support
>>>>
>>>> Jason,
>>>>
>>>> I'm not sure. I remember that you were helping me look into the
>>>> recursive Install behavior. Plus, possible parts incompatibility?
>>>>
>>>> I haven't ran the tests on a windows box at all. I don't expect
>>>> anything new, but it should be done before it ships :)
>>>>
>>>> V/R,
>>>> William
>>>>
>>>> On Wed, Jul 29, 2015 at 8:44 AM, Jason Kenny <dragon512 at live.com>
>>>> wrote:
>>>>
>>>>> Which MS toolchain on windows do we need tested? I can do a test
>>>>> run this afternoon.
>>>>>
>>>>> Jason
>>>>>
>>>>> *From:* William Blevins <wblevins001 at gmail.com>
>>>>> *Sent:* Wednesday, July 29, 2015 1:11 AM
>>>>> *To:* SCons developer list <scons-dev at scons.org>
>>>>> *Subject:* Re: [Scons-dev] Cross-language support
>>>>>
>>>>> Happy with it*
>>>>>
>>>>> Until the review is complete, and I can perhaps get a few more guinea
>>>>> pigs to try it out, it's hard to make a concrete projection. I have 3 items
>>>>> listed out that need to be given a thumbs up/down at a minimum:
>>>>>
>>>>> * Item #1 is still mostly outstanding. I'm not sure how to address,
>>>>> please see the discussion Jason and I started a few weeks ago under the
>>>>> pull request comments.
>>>>>
>>>>> * Item #2 requires no changes (to my knowledge), but someone with more
>>>>> QT knowledge may say otherwise.
>>>>>
>>>>> * I may already have most of the patch for Item #3 (IE. no scanner for
>>>>> key) based on today's feedback.
>>>>>
>>>>> I also still need to request test runs on Windows, and for toolchains
>>>>> that I may not have installed or instructions to install them (EG. D).
>>>>>
>>>>> V/R,
>>>>> William
>>>>>
>>>>> On Wed, Jul 29, 2015 at 1:58 AM, William Blevins <
>>>>> wblevins001 at gmail.com> wrote:
>>>>>
>>>>>> As soon as we are happen with it :)
>>>>>>
>>>>>> On Wed, Jul 29, 2015 at 12:30 AM, Bill Deegan <
>>>>>> bill at baddogconsulting.com> wrote:
>>>>>>
>>>>>>> O.k. let me push 2.3.5 with the visual studio 2015 stuff.
>>>>>>> Then we'll changed to 2.4 merge slots. stabilize. release.
>>>>>>> Then this code? (2.5?)
>>>>>>>
>>>>>>> -Bill
>>>>>>>
>>>>>>> On Tue, Jul 28, 2015 at 6:33 PM, Jason Kenny <dragon512 at live.com>
>>>>>>> wrote:
>>>>>>>
>>>>>>>> I have been using the slots drop for a while with Parts. I think
>>>>>>>> it is ready. It does have a notable improvement in speed and memory size. I
>>>>>>>> would before getting this out as officially earlier than later.
>>>>>>>>
>>>>>>>> Jason
>>>>>>>>
>>>>>>>> *From:* William Blevins <wblevins001 at gmail.com>
>>>>>>>> *Sent:* Tuesday, July 28, 2015 8:31 PM
>>>>>>>> *To:* Dirk Bächle <dl9obn at darc.de> ; SCons developer list
>>>>>>>> <scons-dev at scons.org>
>>>>>>>> *Subject:* Re: [Scons-dev] Cross-language support
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Tue, Jul 28, 2015 at 7:16 PM, Dirk Bächle <tshortik at gmx.de>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> Hi guys,
>>>>>>>>>
>>>>>>>>> sorry for chiming in so late.
>>>>>>>>>
>>>>>>>>> On 28.07.2015 23:44, Gary Oberbrunner wrote:
>>>>>>>>>
>>>>>>>>>> Yes, that's how we've done it in the past. Sounds like doing it
>>>>>>>>>> at the same time as slots would be perfect.
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>> Doing this in parallel with the "slots" change sounds good to me
>>>>>>>>> too. +1
>>>>>>>>>
>>>>>>>>
>>>>>>>> I'm not opposed to releasing in the same update as slots, but since
>>>>>>>> the cross language code reviews haven't been finished, I don't want to
>>>>>>>> delay slots since it is ready now.
>>>>>>>>
>>>>>>>> The pessimist in me as sees doing two major enhancements in the
>>>>>>>> same release as higher risk; I don't foresee any issues, but its worth the
>>>>>>>> thought if the release overhead isn't too bad.
>>>>>>>>
>>>>>>>>
>>>>>>>>>
>>>>>>>>> -- Gary
>>>>>>>>>>
>>>>>>>>>> On Tue, Jul 28, 2015 at 2:01 PM, Bill Deegan <
>>>>>>>>>> bill at baddogconsulting.com <mailto:bill at baddogconsulting.com>>
>>>>>>>>>> wrote:
>>>>>>>>>>
>>>>>>>>>> Gary,
>>>>>>>>>>
>>>>>>>>>> For such a change we should bump the second digit?
>>>>>>>>>> 2.4?
>>>>>>>>>>
>>>>>>>>>> I agree we should not turn down a change because it will
>>>>>>>>>> cause rebuilds where the past didn't as long as it is now more correct
>>>>>>>>>> (which it should be with this change).
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Yes, "forward" is the way to go. ;)
>>>>>>>>>
>>>>>>>>> Also agree we should be verbose in our notification of the
>>>>>>>>>> impacts of the new change to avoid (as much as we can) "surprises".
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>> I think (better: hope) we did a good enough job for the "slots"
>>>>>>>>> stuff on this. For the scanner changes, I see them more like a fix...so a
>>>>>>>>> single announcement should be sufficient?
>>>>>>>>>
>>>>>>>>> Finally, and just in case I haven't done so already, I'd like to
>>>>>>>>> thank William for all the work he's done on this issue. I couldn't help as
>>>>>>>>> much as I would've liked, but with Gary's support you tackled this down and
>>>>>>>>> brought it to a good end. Kudos to you...bravo!
>>>>>>>>>
>>>>>>>>>
>>>>>>>> Truly appreciated. I have spent a lot of time on this issue
>>>>>>>> despite what one might expect from the number of lines of code.
>>>>>>>>
>>>>>>>> It was honestly my first time working in a "real" python
>>>>>>>> environment outside of scripting, and the SCons code base is rather
>>>>>>>> complex. It instills me greater appreciation for the work that has already
>>>>>>>> done.
>>>>>>>>
>>>>>>>> I also want to thank everyone here for their help past and future,
>>>>>>>> but don't pat me on the back until its done. I might get lazy :)
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>>
>>>>>>>>> Best regards,
>>>>>>>>>
>>>>>>>>> Dirk
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> _______________________________________________
>>>>>>>>> Scons-dev mailing list
>>>>>>>>> Scons-dev at scons.org
>>>>>>>>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>>>>>>>>
>>>>>>>>
>>>>>>>> ------------------------------
>>>>>>>> _______________________________________________
>>>>>>>> Scons-dev mailing list
>>>>>>>> Scons-dev at scons.org
>>>>>>>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> Scons-dev mailing list
>>>>>>>> Scons-dev at scons.org
>>>>>>>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Scons-dev mailing list
>>>>>>> Scons-dev at scons.org
>>>>>>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>>>>>>
>>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>> ------------------------------
>>>>> _______________________________________________
>>>>> Scons-dev mailing list
>>>>> Scons-dev at scons.org
>>>>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Scons-dev mailing list
>>>>> Scons-dev at scons.org
>>>>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>>>>
>>>>>
>>>>
>>>> ------------------------------
>>>> _______________________________________________
>>>> Scons-dev mailing list
>>>> Scons-dev at scons.org
>>>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>>>
>>>> ------------------------------
>>>> _______________________________________________
>>>> Scons-dev mailing list
>>>> Scons-dev at scons.org
>>>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>>>
>>>>
>>>> _______________________________________________
>>>> Scons-dev mailing list
>>>> Scons-dev at scons.org
>>>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>>>
>>>>
>>>
>>> ------------------------------
>>> _______________________________________________
>>> Scons-dev mailing list
>>> Scons-dev at scons.org
>>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>>
>>>
>>> _______________________________________________
>>> Scons-dev mailing list
>>> Scons-dev at scons.org
>>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>>
>>>
>>
>> ------------------------------
>> _______________________________________________
>> Scons-dev mailing list
>> Scons-dev at scons.org
>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>
>> ------------------------------
>> _______________________________________________
>> Scons-dev mailing list
>> Scons-dev at scons.org
>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>
>> ------------------------------
>> _______________________________________________
>> Scons-dev mailing list
>> Scons-dev at scons.org
>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>
>>
>> _______________________________________________
>> Scons-dev mailing list
>> Scons-dev at scons.org
>> https://pairlist2.pair.net/mailman/listinfo/scons-dev
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist2.pair.net/pipermail/scons-dev/attachments/20150729/47a5c502/attachment-0001.html>
More information about the Scons-dev
mailing list