[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
RE: [cdt-dev] Spawner treats environment differentlyon WindowsandLinux
|
>In this case I tend to think that the best way to go for now given
where
>we are in our release cycle is to do a lightweight temporary solution
by
>implementing an executable search in the internal builder and passing
>the full path to an executable to the exec. This will not require any
>native code changes and recompilations.
I agree with Mikhail because of where we are in the release cycle. That
is, limiting the fix limits the potential regressions that the fix could
cause. I think it would be appropriate to change all spawners in the
3.1.1 release.
Regards,
Leo
>-----Original Message-----
>From: cdt-dev-bounces@xxxxxxxxxxx [mailto:cdt-dev-bounces@xxxxxxxxxxx]
On
>Behalf Of Chris Recoskie
>Sent: Thursday, June 08, 2006 8:18 AM
>To: CDT General developers list.
>Subject: RE: [cdt-dev] Spawner treats environment differentlyon
>WindowsandLinux
>
>I don't see this as being solely an MBS issue. Users are going to want
the
>same behaviour when invoking an external debugger, for example. I
think
>it's just a matter of time until someone complains about it in that
context
>as well.
>
>===========================
>
>Chris Recoskie
>Team Lead, IBM CDT Team
>IBM Toronto
>http://www.eclipse.org/cdt
>
>
>
>
> "Sennikovsky,
> Mikhail"
> <mikhail.sennikov
To
> sky@xxxxxxxxx> "CDT General developers list."
> Sent by: <cdt-dev@xxxxxxxxxxx>
> cdt-dev-bounces@e
cc
> clipse.org
>
Subject
> RE: [cdt-dev] Spawner treats
> 08/06/2006 08:08 environment differentlyon
> AM WindowsandLinux
>
>
> Please respond to
> "CDT General
> developers list."
> <cdt-dev@eclipse.
> org>
>
>
>
>
>
>
>Hi all,
>
>Since I've got only a few responses to the problem, and most of them
are
>from the MBS guys, I would assume that the issue is currently affects
>the internal builder and its clients only, i.e. no one is
>implementing/using SIProviders and other functionality that intends to
>adjust the PATH environment before running an external tool.
>
>In this case I tend to think that the best way to go for now given
where
>we are in our release cycle is to do a lightweight temporary solution
by
>implementing an executable search in the internal builder and passing
>the full path to an executable to the exec. This will not require any
>native code changes and recompilations.
>
>What do others think?
>
>NOTE: still this behavior should be moved to the Spawner in the future
>since with the new project model we are going to allow other parts of
>the CDT e.g. the CIProviders use the same environment as used by the
>build process to invoke external tools. This means that the issue will
>be relevant for other CDT parts in the future as well.
>
>Mikhail
>
>-----Original Message-----
>From: cdt-dev-bounces@xxxxxxxxxxx [mailto:cdt-dev-bounces@xxxxxxxxxxx]
>On Behalf Of Chris Recoskie
>Sent: Thursday, June 08, 2006 3:27 PM
>To: CDT General developers list.
>Subject: RE: [cdt-dev] Spawner treats environment differentlyon
>WindowsandLinux
>
>My opinion is that it's better to have it working on some platforms
than
>none. As I said before I don't think any ISVs are relying on this
>behaviour, but it does affect users, and I think we should help the
ones
>we
>can.
>
>I should be able to arrange for an AIX build once I figure out who to
>ask
>around here. Not sure I would be able to get this in place for Friday
>though.
>
>Doug, doesn't the Foundation have a suite of build machines for our
use?
>How do we get to those?
>
>===========================
>
>Chris Recoskie
>Team Lead, IBM CDT Team
>IBM Toronto
>http://www.eclipse.org/cdt
>
>
>
>
>
> "Sennikovsky,
>
> Mikhail"
>
> <mikhail.sennikov
>To
> sky@xxxxxxxxx> "CDT General developers list."
>
> Sent by: <cdt-dev@xxxxxxxxxxx>
>
> cdt-dev-bounces@e
>cc
> clipse.org
>
>
>Subject
> RE: [cdt-dev] Spawner treats
>
> 07/06/2006 04:55 environment differently on
>
> PM WindowsandLinux
>
>
>
>
>
> Please respond to
>
> "CDT General
>
> developers list."
>
> <cdt-dev@eclipse.
>
> org>
>
>
>
>
>
>
>
>
>
>After looking into this problem for the other platforms we support it
>looks
>like the Spawner for Aix, Macosx, Qnx and Solaris platforms seem to
>treat
>the environment in the same way as the Linux Spawner. (NOTE: I did not
>test
>that, this is only my assumption since it seems that only on windows
the
>launch is performed via a separate launcher process).
>
>So in case we decide to go ahead with the Windows Spawner behavior, the
>Aix, Macosx, Qnx and Solaris Spawners should be updated as well and I
>won?t
>be able to do that since I don?t have these OSes.
>
>I?m not sure what approach is generally used with regard to Spawner:
>should
>the Spawner change/update process be consistent across all supported
>platforms?
>In this case I don?t think we?ll be able to do this for the 3.1 since
we
>should do all the necessary changes and testing by the end of this
>Friday.
>In this case I propose to postpone the fix for the 3.1.1
>
>In case it is OK to do it only for Linux for now, I?m OK with
committing
>it. But before that I need a volunteer to build and test the update for
>Linux.ppc.
>
>Mikhail
>
>From: cdt-dev-bounces@xxxxxxxxxxx [mailto:cdt-dev-bounces@xxxxxxxxxxx]
>On
>Behalf Of Sennikovsky, Mikhail
>Sent: Wednesday, June 07, 2006 8:48 PM
>To: CDT General developers list.
>Subject: RE: [cdt-dev] Spawner treats environment differently on
>WindowsandLinux
>
>Hi All,
>
>I've implemented a quick fix for the linux Spawner.
>The script is attached to the corresponding bugzilla (
>https://bugs.eclipse.org/bugs/show_bug.cgi?id=145737).
>See https://bugs.eclipse.org/bugs/attachment.cgi?id=43732&action=view
>
>I've tested it on Linux x86 and it works fine, though I've made only a
>little testing so far.
>
>In case we all agree to make the Spawner treat the environment on Linux
>as
>it does it in Windows, we should decide whether or not to include the
>fix
>in the 3.1
>
>The major impact of this bug currently is the MBS Internal Builder and
>all
>its clients: Internal Build functionality, Building a single resource
>functionality.
>
>Mikhail
>
>
>From: cdt-dev-bounces@xxxxxxxxxxx [mailto:cdt-dev-bounces@xxxxxxxxxxx]
>On
>Behalf Of Sennikovsky, Mikhail
>Sent: Wednesday, June 07, 2006 3:07 PM
>To: CDT General developers list.
>Subject: [cdt-dev] Spawner treats environment differently on Windows
>andLinux
>
>Hi All,
>
>While working on the fix for the MBS Internal Builder-related issue (
>https://bugs.eclipse.org/bugs/show_bug.cgi?id=145099), I?ve noticed
that
>the Spawner treats environment differently on Windows and Linux (see
the
>bug https://bugs.eclipse.org/bugs/show_bug.cgi?id=145737)
>Although it might not seem to be a really critical bug especially when
>we
>are close to our release, I?d like to here your opinion on what should
>be
>the right way of handling environment: the way used by the Windows
>Spaner
>or the way used by the Linux one, since the decision might affect other
>components using Spawner, e.g. the SIProvider mechanism, MBS Internal
>Builder i.e. the way the Path environment variable is treated by those
>components.
>See the bug https://bugs.eclipse.org/bugs/show_bug.cgi?id=145737 for
>more
>detail and please comment to that bug if you have any ideas regarding
>this.
>
>Thanks,
>Mikhail
> _______________________________________________
>cdt-dev mailing list
>cdt-dev@xxxxxxxxxxx
>https://dev.eclipse.org/mailman/listinfo/cdt-dev
>
>
>_______________________________________________
>cdt-dev mailing list
>cdt-dev@xxxxxxxxxxx
>https://dev.eclipse.org/mailman/listinfo/cdt-dev
>_______________________________________________
>cdt-dev mailing list
>cdt-dev@xxxxxxxxxxx
>https://dev.eclipse.org/mailman/listinfo/cdt-dev
>
>
>_______________________________________________
>cdt-dev mailing list
>cdt-dev@xxxxxxxxxxx
>https://dev.eclipse.org/mailman/listinfo/cdt-dev
