Hi All,
Since it is going to be fifth month from proposing the idea of LTP
intermediate release, wanted to know that whether it is meeting our
expectation from user and maintainer point of view or not.
Since LTP community is putting lot of effort towards making LTP better,
we should also think about the process what we are following to maintain
it better.
Thanks
- Rishi
On Thu, 2008-03-20 at 09:56 +0530, Rishikesh K. Rajak wrote:
> On Wed, 2008-03-19 at 17:37 +0530, Subrata Modak wrote:
>
> > Hi All,
> >
> > Recently we had a discussion amongst some members, regarding the
> > necessity to have at least 1 Intermediate LTP Release. This is necessary
> > to confirm that new additions and fixes has not broken the existing
> > stable LTP tree. As part of this initiative i plan to have the release
> > of the CVS tree around 20th-25th of every month, for all of you to test
> > and validate the treeś stability. We saw that the LTP nightly snapshots
> > don´t work that good. So, we propose to have just 1 Intermediate LTP
> > release before the final release. What do you all say about this ??
>
>
> Thanks subrata for considering my idea, This will be a good start to
> not face any problem with monthly release.
>
> As we see our monthly release is very much important and all community
> totally depend on the monthly release. If this fails community will
> have to go back to previous release and they can't test with new
> testcases which has been newly added.
>
> Thanks
> Rishi
> IBM ISTL, Bangalore
>
> >
> > Regards--
> > Subrata
On Tue, 2008-02-26 at 13:02 +0530, Rishikesh K. Rajak wrote:
> Hi Subrata,
>
> I am using LTP from last year very frequently and i found something to
> follow with using it, so just throwing an open thought and any one can
> modify/remodify this anytime.
>
> Here my idea goes:
>
> First i would like to congratulate you because, LTP is using by
> many of the Linux kernel developers and different kernel community,
> but simultaneously it is very common that you get more and more
> testcase patches by daily basis or weekly basis. When the testcase
> input is having this much frequency the probability of unstability of
> end month released LTP become so high ( may be ).
> So my suggestion goes to have one pre-release of LTP in mid of the
> month or whenever you feel more new/modified testcases has come which
> can effect the stable LTP code, so that when any developer or test
> engineer releases their LTP testcases whether it is new/modified , it
> should be again available as a unstable package to LTP site so that
> everyone can use it and it may happen that the release can also be
> tested on many architecture with many components and older/latest
> kernel. I am not sure if your nightly snapshot provide this ?
>
> Advantage:
>
> 1> Testcase developer prospective:
> When any testcases have been developed by test
> engineer/developer, that is their responsibility to make it work on
> all architecture, but sometime the end results come with some nasty
> problems because, by that time some code changes in the new kernel. If
> you release to community that code can be tested on maximum flavour of
> kernels and we can also certify that will work fine with end month
> release. (experimental).
>
> 2> Getting build failure fixed before stable LTP month end
> release.
> - Suppose in one month 20 new testcases and 10 modified
> testcases you got from community and you merged with CVS after testing
> on one/two different arch, but if you release this to community this
> can be tested on different distros, latest kernel and with different
> device peripherals also. So ltp community can get more failure report
> and then if that bug/failure get solved by community before end of the
> month that can be merged to month end release other wise it will be
> inside unstable only. And possibly we will get less failure/bug so
> that no one can get stuck.
>
> This i came out because of NUMA related failure for dec'07 release. On
> this stable release no one was able to compile LTP, because of build
> failure and i appreciate that you gave quick patch also for not
> compiling NUMA, and no one was using NUMA enabled LTP till exact fixed
> came.
>
> 3> It will be useful more by community because everyone will know
> that the latest version is stable on each arch with the latest kernel
> as well as older kernel, otherwise i have seen such mail saying that
> it work with older version but it does not work with the latest
> version of LTP/kernel, yeah this can be happen because of the new
> development inside kernel, but that should not effect the existing
> testcases.
>
> 4> You will have some stable patch to include to LTP month end
> release and testcase developer can get more time to fix those and
> include in month end release.
>
> As i said this is just my thought there can be more or less by
> community. Suggestions are welcome.
>
>
> Thanks
> Rishi
>
> >
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________ Ltp-list mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/ltp-list
-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list