Aw: Re: Update on Trinityrnaseq packaging

["Steffen Möller" <steffen_moeller@gmx.de>]

Hi Michael,
 

> On Mon Feb 16 2015 at 5:46:42 PM Andreas Tille <andreas@an3as.eu> wrote:Hi Michael,

On Mon, Feb 16, 2015 at 07:09:18AM +0000, Michael Crusoe wrote:

>> I had some quick look into this and did some changed (please pull).
>> Thanks for the changes.  I realised that some large files are installed into
>>
>>   /usr/lib/transdecoder/sample_data
>>
>>The name of this dir might support the suspicion that these data are not
>>necessary for using transdecoder but rather for testing.  If this is
>>true I'd recommend separating these files into a package
>>
>>   transdecoder-data
>>
>> (which has architecture all).
>> 
>Since this is a Perl program I guess the entirety should be Architecture: all? Since there is less than 10 MB of sample data I prefer to keep it with the main program. Sample >data with example usage helps users understand how the program works, with what file formats, and how to use it.

I propose having the sample data together with a -doc package. The separation
of the executable parts into a package of its own is meant to help the I/O
for the dynamic construction of some runtime environment (chroot, docker,
vm, cloud instance ...) when some workflow specification refers to Trinity
as a dependency.

Many thanks for your tremendous effort on all those packages!

Best,

Steffen