Re: [Fwd: python hangs]

Looking at the recent mails about ROOT with Qt, PyRoot, I would like to make some remarks:

 -We will continue to support the Qt interfaces if the people outside the ROOT team who have supported these interfaces so far continue to do so. However, we strongly recommend the direct use of the ROOT GUI instead with no external dependencies. We are convinced that this will pay in the medium/long term.
None of the GUIS made with Qt that we have seen so far are superior to pure ROOT GUIs,
and they have generated more problems/questions than initially anticipated.

 -When reporting problems combining the 3 systems ROOT, Qt and PyRoot, it would be nice if the
problem could be reduced to 2 systems only, eg ROOT GUI and PyRoot or ROOT/Qt with CINT/C++ only.

 -The PyRoot interface has been consolidated across successive versions (thanks Wim).
It would be nice if people reporting problems could make the effort to run the test with more recent versions.

Rene Brun

WLavrijsen_at_lbl.gov wrote:
> Valeri,
>
>
>> Check this:
>> http://root.cern.ch/root/roottalk/roottalk07/0082.html
>>
>
> could be: the Qt3 event dispatcher is not re-entrant (this is fixed in Qt4),
> and in those cases the thread can cause issues.
>
> Herb, do you happen to be running Qt3 (set through .rootrc)? It doesn't (in
> my mind) explain all the issues that you describe, in particular:
>
> "ROOT.gSystem.Exit(0) hangs"
>
> would be rather unexpected, b/c the python GIL prevents mixing of individual
> thread calls during a call through the PyROOT layer (i.e. once the above call
> is initiated, PyROOT no longer sends any gSystem.ProcessEvents() calls).
>
> Best regards,
> Wim
> --
> Wim.Lavrijsen_at_cern.ch -- WLavrijsen_at_lbl.gov -- www.lavrijsen.net
>
> "Stop making excuses for your software." --first step towards quality
> "GIGO is not a valid design pattern." --corollary
>
Received on Wed Jan 17 2007 - 08:52:37 CET