[Debichem-devel] Bug#776662: Bug#776662: Going forward with jmol in debian

[Tobias Hansen]

Am 15.03.2015 um 12:40 schrieb Michael Banck:
> Hi,
> 
> On Sun, Mar 15, 2015 at 09:27:27AM +0100, Julien Puydt wrote:
>> we need the new jmol for the sagemath packaging effort ;
>> unfortunately, none of us has any clue how to package java code.
>>
>> From looking at the sources, upstream is using the classical "ant"
>> builder, so I guess on the other hand the debian java team would
>> find it pretty easy to help on the matter.
>>
>> So which way do we go ? Can someone take care of the matter within
>> the debichem team, or should we ask the debian java team to take
>> over ?
> 
> The problem is not building jmol - the problem is packaging the
> additional requirements since the last version that is in Debian.
> 
> I believe the main issue was jmolspec or something (sorry I am on crappy
> internet right now and cannot check), which was a mess to package by
> itself.
> 
> One alternative (if sagemath does not require it) would be to come up
> with a way to hack out that additional feature, if that is possible.
> 
> I am not to keen on moving jmol packaging over to the java team, as e.g.
> the cdk package is lingering along there as well.  I am more than happy
> to add people who are interested in helping with jmol packaging to the
> debichem team, though.
> 

Although sagemath installs JSpecView itself, I don't think it needs a
"viewer for spectral data in the JCAMP-DX format" since it uses jmol
only as a viewer for its 3d plots. If the package doesn't secretly
provide other functionality as well it's quite possible that sagemath
can work with a jmol version without that.

Sagemath does however use jsmol.

Best,
Tobias