yade-dev team mailing list archive
-
yade-dev team
-
Mailing list archive
-
Message #05891
Re: qt4 looks nice!
-
To:
yade-dev@xxxxxxxxxxxxxxxxxxx
-
From:
Janek Kozicki <janek_listy@xxxxx>
-
Date:
Wed, 13 Oct 2010 14:01:49 +0200
-
Face:
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
-
In-reply-to:
<1286914101.32553.18.camel@gamma.arcig>
Václav Šmilauer said: (by the date of Tue, 12 Oct 2010 22:08:21 +0200)
>
> > How about adding another tab to those "Simulation, Display, Generate,
> > Python", called: "Omega" or "root body" or "Scene" and use my qt4-tree
> > serialization interface? Or your python based qt4 serialization?
> > We could even add tabs for each scene savedTmp do memory.
>
> You are welcome to do any of that. For scene, what else is there besides
> what is already in other places? Please no "Omega" or "rootBody", those
> are programmer's name which have nothing to do in the interface. The
> idea with saveTmp'd is nice, I had it some time ago, but never had
> resources to do it.
>
> The qt4-tree serialization interface is IMHO not useful, since it has to
> serialize everything every time, will not honor Attr::triggerPostLoad
> attribute, has not access to python docstrings and so on. In short, c++
> is too heavy for UI.
OK. Too bad that I am confident with my qt4 tree. I will look into
your implementation. Unfortunately useful results are less likely
this way.
> > besides I see some minor issue when trying to 'generate' something,
> > e.g. a TriaxialTest via GUI. The error from python interface is:
> >
> > --> 125 self.generator.generate(filename)
> > 126 if self.generatorAutoCheck:
> > 127 O.load(filename)
> >
> > RuntimeError: unregistered class
>
> That is (likely) a bug in boost::serialization. If you were following
> the list, we had this issue w/ ubuntu 10.10, then it went away without
> any obvious reason (so I assume it was a bug in some libs which is now
> fixed). You will have the same error for every save, not just for
> generators.
I see, so debian squeeze has currently an affected version of boost.
--
Janek Kozicki http://janek.kozicki.pl/ |
Follow ups
References