Strange error

[Alexandre Bergel]

Hi!

Does anyone has already seen this error?

[Johan Brichau]

Yes, this happens all the time in Pharo2.0

Did not yet get to investigating where it comes from...

On 17 Apr 2013, at 20:00, Alexandre Bergel <alexandr...@me.com> wrote:

> Hi!
>
> Does anyone has already seen this error?
>
>

> It seems to happen randomly. Evaluating the expression "ConfigurationOfRoassal project" times to time raise this error.
>
> Cheers,
> Alexandre
> --
> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
> Alexandre Bergel http://www.bergel.eu
> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
>
>
>
> <Screen Shot 2013-04-17 at 1.58.02 PM.png><PharoDebug.log>--
> You received this message because you are subscribed to the Google Groups "Metacello" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to metacello+...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[Esteban Lorenzano]

all the time?
I have report of that... it is happening because something weird happens when something that I don't know changes the stack (in vm)
yes... it is not very clear :)

I'm planning to work on that next week.

Esteban

[Johan Brichau]

I only get it during Metacello loads, so it was not really obvious for me that it could be a VM problem.
But then again, I have been doing only metacello loads in Pharo2.0 for now :-)

Good to know that it's a known problem!

[Norbert Hartl]

I experienced that as well. If my memory is not fully broken it depends on the fact that there are dirty packages and pharo is raising a dialog. I had this with 1.4 where zinc was dirty and the configuration pulled in a newer version. I fixed it by changing my builder scripts to

[ ((Smalltalk at: #ConfigurationOfEmcee) project lastVersion ) load ]
on: MCMergeOrLoadWarning do: [ :warning |
Transcript crShow: 'Ignoring with Load anyway ', warning printString.
warning resume: true ].

Norbert

[Dale Henrichs]

Esteban,

If it helps, I've recorded[1] some of the information about the bad behavior that has been reported by Metacello users (Christophe and Alexandre) and the bad behavior will go away if you use a different vm. It can show up in Pharo1.4 if using the right (er wrong) vm ...

Dale

[1] https://github.com/dalehenrich/metacello-work/issues/143

[Norbert Hartl]

Ah, yes, I remember that changing the vm solved it. Hmmm, not sure, I should take notes

Norbert

[Tudor Girba]

Hi,

I also encounter this problem often in the Moose image, and I have really no idea where it comes from. The observation that it has to do with dirty Monticello packages could be something given that in Moose we right now do have such dirty packages due to Grease.

Dale: I looked at your comments, and I could actually reproduce it also in Pharo 2.0.

Here is what I did:

- download the latest moose image:
https://ci.inria.fr/moose/job/moose-latest-dev-4.8/lastSuccessfulBuild/artifact/moose-latest-dev-4.8.zip
- download the Pharo 2.0 VM you mentioned
http://gforge.inria.fr/frs/download.php/32303/Pharo2.0-mac.zip
- execute:
ConfigurationOfGlamour loadDevelopment

Doru

www.tudorgirba.com

"Quality cannot be an afterthought."

[ChristopheD]

Yes, I got exactly the same errror.

It is reproducible both on Pharo 1.4 and Pharo 2 and really looks like a VM problem.

I "solved" this problem by using another VM.

Regards,

Christophe.