[Debichem-devel] Bug#560412: Found a fix

[Michael Banck]

Hi Vincent,

On Mon, Dec 14, 2009 at 12:02:06PM +0100, Vincent Fourmond wrote:
>   I've found a very simple fix for this problem. The cmd.png function
> now takes a prior argument that makes it output previously traced
> things. It defaults to prior = 0 (ie do not use what was done before).
> I think the old behavior was equivalent to prior = 1 (though it did
> not exist). In any case, here is a patch that works for me, and does
> not seem to break anything else.

> diff -Naur pymol-1.2r1/modules/pymol/exporting.py pymol-1.2r1-fixed/modules/pymol/exporting.py
> --- pymol-1.2r1/modules/pymol/exporting.py	2009-07-12 14:00:40.000000000 +0200
> +++ pymol-1.2r1-fixed/modules/pymol/exporting.py	2009-12-14 11:59:20.016059145 +0100
> @@ -283,7 +283,7 @@
>          
>  
>      def png(filename, width=0, height=0, dpi=-1.0, ray=0,
> -            quiet=1, prior=0, format=0, _self=cmd):
> +            quiet=1, prior=1, format=0, _self=cmd):
>          '''

Unfortunately, this patch seems to have changed expected behaviour:

I believe nowadays the preferred way to get a rendered png is "png
file.png ray=1 [...]", and with our patch setting prior=1, people only
get the non-raytraced screenshot.  So basically your original issue in
reverse.

I have tried backing out your patch and I cannot reproduce your original
problem anymore, i.e. PyMOL does not render a second time on png even
with prior=0.

So I'll remove your patch for the next upload, let me know if that
causes trouble for you, in case you are still doing batch PyMOL.


Michael