PetitParser 1.6 instead of 1.5. Tests only work when no Nautilus browser is open. Some notification problem when creating classes. OSProcess ci has lots of failing tests, so we don't know if it actually works⦠Diego & Stephan Verstuurd vanaf mijn iPhone
On Wed, May 21, 2014 at 04:35:33PM +0200, Stephan Eggermont wrote:
PetitParser 1.6 instead of 1.5. Tests only work when no Nautilus browser is open. Some notification problem when creating classes.
OSProcess ci has lots of failing tests, so we don't know if it actually works???
Diego & Stephan
The tests for OSProcess require VM support. Until recently, an interpreter VM was required to run the full OSProcess test suite because it relies heavily on #forkSqueak to generate cooperating VM processes for the tests. Eliot has recently added an enhancement to Cog that provides the required support. When that change is added to the Pharo VM, the tests will run. The changes are in SVN in branches/Cog/platforms/unix/vm/ and the commit notice is at: http://lists.squeakfoundation.org/pipermail/vm-dev/2014-March/014964.html Dave
On Fri, May 23, 2014 at 5:12 AM, David T. Lewis <lewis@mail.msen.com> wrote:
On Wed, May 21, 2014 at 04:35:33PM +0200, Stephan Eggermont wrote:
PetitParser 1.6 instead of 1.5. Tests only work when no Nautilus browser is open. Some notification problem when creating classes.
OSProcess ci has lots of failing tests, so we don't know if it actually works???
Diego & Stephan
The tests for OSProcess require VM support. Until recently, an interpreter VM was required to run the full OSProcess test suite because it relies heavily on #forkSqueak to generate cooperating VM processes for the tests.
Eliot has recently added an enhancement to Cog that provides the required support. When that change is added to the Pharo VM, the tests will run.
The changes are in SVN in branches/Cog/platforms/unix/vm/ and the commit notice is at:
http://lists.squeakfoundation.org/pipermail/vm-dev/2014-March/014964.html
I experienced that PipeableOSProcess broke after calling in for a given amount of invocations. Having a solid OSProcess thing in Pharo is really a must have... Are those changes only in a plugin? That I can integrate. Phil
Dave
On Fri, May 23, 2014 at 11:20:18AM +0200, phil@highoctane.be wrote:
On Fri, May 23, 2014 at 5:12 AM, David T. Lewis <lewis@mail.msen.com> wrote:
On Wed, May 21, 2014 at 04:35:33PM +0200, Stephan Eggermont wrote:
PetitParser 1.6 instead of 1.5. Tests only work when no Nautilus browser is open. Some notification problem when creating classes.
OSProcess ci has lots of failing tests, so we don't know if it actually works???
Diego & Stephan
The tests for OSProcess require VM support. Until recently, an interpreter VM was required to run the full OSProcess test suite because it relies heavily on #forkSqueak to generate cooperating VM processes for the tests.
Eliot has recently added an enhancement to Cog that provides the required support. When that change is added to the Pharo VM, the tests will run.
The changes are in SVN in branches/Cog/platforms/unix/vm/ and the commit notice is at:
http://lists.squeakfoundation.org/pipermail/vm-dev/2014-March/014964.html
I experienced that PipeableOSProcess broke after calling in for a given amount of invocations.
Having a solid OSProcess thing in Pharo is really a must have...
Are those changes only in a plugin? That I can integrate.
Hi Phil, No, the change I am referring to is part of the main VM, not OSProcess or OSProcessPlugin. Can you describe the problem that you had with PipeableOSProcess? I am guessing that you may have been doing something that resulted in unclosed file handles, which eventually will cause a failure when the OS resource limits are exceeded. Dave
David T. Lewis wrote
I am guessing that you may have been doing something that resulted in unclosed file handles, which eventually will cause a failure when the OS resource limits are exceeded.
Yes, I've run into that quite a bit. IIRC remembering to "proc close; closePipes" does the trick... ----- Cheers, Sean -- View this message in context: http://forum.world.st/Coral-is-green-tp4759893p4760543.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.
On Tue, May 27, 2014 at 6:45 PM, Sean P. DeNigris <sean@clipperadams.com>wrote:
David T. Lewis wrote
I am guessing that you may have been doing something that resulted in unclosed file handles, which eventually will cause a failure when the OS resource limits are exceeded.
Yes, I've run into that quite a bit. IIRC remembering to "proc close; closePipes" does the trick...
Duly noted!
----- Cheers, Sean -- View this message in context: http://forum.world.st/Coral-is-green-tp4759893p4760543.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.
participants (4)
-
David T. Lewis -
phil＠highoctane.be -
Sean P. DeNigris -
Stephan Eggermont