Subsurface 2.0 has been tagged

Dirk Hohndel dirk at hohndel.org
Thu Oct 11 01:18:14 PDT 2012


Jef Driesen <jefdriesen at telenet.be> writes:

> On 2012-10-10 16:09, Pierre-Yves Chibon wrote:
>> On Wed, 2012-10-10 at 15:38 +0200, Jef Driesen wrote:
>>> On 2012-10-10 14:27, Pierre-Yves Chibon wrote:
>>> > Libdivecomputer has been accepted into Fedora, builds are running. 
>>> So
>>> > I
>>> > am now preparing Subsurface for review as well.
>>>
>>> Great!
>>>
>>> Just a small comment. I noticed in the bugreport [1] that the 
>>> example
>>> applications are packaged too. I'm not sure that's a good idea. 
>>> These
>>> test applications are intended for diagnostic purposes only and they
>>> aren't really useful for end-users. That's also the reason why they
>>> don't have man pages.
>>
>> The question becomes then, do they need to be installed by make
>> install ?
>
> Good question. I never even thought of that. I usually "install" local 
> software to a directory in my home directory (e.g. 
> --prefix=$HOME/local). That way I don't need root permissions, but as a 
> side-effect, the extra binaries don't cause any problems either. So I 
> never really paid any attention to that.

I think for a distro package of a LIBRARY, the regular package should
only contain the distro. The universal app could be included with a
-devel package instead. The individual apps don't really seem to make
much sense.

>> Should I remove all what goes into /usr/bin/ then?
>
> Maybe keep the universal app until the dctool is ready? That would be 
> consistent with the debian package. They also include only the universal 
> app.

So if an app uses the library and therefore pulls in the library
package, why should that app (even the universal one) be installed?

/D


More information about the subsurface mailing list