[Pkg-running-devel] Bug#789875: subsurface: FTBFS in experimental

Christoph Anton Mitterer calestyo at scientia.net
Wed Aug 26 14:40:43 UTC 2015


On Wed, 2015-08-26 at 13:40 +0300, Lubomir I. Ivanov wrote:
> your approach for convincing is offensive and unwise.
Well if upstreams are effectively hostile against core paradigms of the
FLOSS community, it must expect that people won't be happy with it.


> as a peace of software it now no longer obeys the Linux distribution
> rules; it's setting a bad example at the expense of not being
> available everywhere at anytime in the ecosystem.
Being not available is not the problem by itself. There's many software
which is not packaged for Debian.
The problem is when you have projects which intentionally work against
some of the base principles of open source software, as here: the
apparently expressed wish of upstream to be under fully control of the
program (like being the only place where it's been distributed, people
not adapting it without renaming, etc.)


> otherwise, why would you even care if some odd divelog software is no
> longer distributed by distro X?
It's like when DLRS manufacturers bring out a new mount system... they
promise you everything that this will be *their* system for the future.
People start to invest (money) in it,.. and not rarely it happened that
the manufacturer changed the mount some years afterwards.

The point is:
People "invested" in subsurface, in the sense that they stored all
their data in it,... and part of the decision for that may have been
the believe that subsurface acts like any other typical sane open
source project - in this example: not trying to keep distros from
properly packaging software.

If the subsurface people would have said from the beginning:
"This is our nice fancy new diving log software. Use it, but beware
that we want to keep full control (unless you fork) and in 2 years
we'll try to prevent distros from packaging."
than I guess many people (at least myself) would have never used it in
the first place.




On Wed, 2015-08-26 at 08:54 +0200, Christian PERRIER wrote:
> For sure, I well understand that redisigning Subsurface to use a
> different storage backend is not something one can reasonably 
> considerbut maybe having this as an option could be nice.
Actually, the XML backend already exists (or did so?). I even think
it's the only usable backend in the most recent version of the
official[0] Debian package.


> I do share Christoph's feeling about the "raison
> d'être"
> of distros and, well, I'm never happy when this is not well 
> understood....
Well I guess it's perfectly fine and politically simply the best and
only choice to throw out packages that behave bad,... such hostility
against distributions cannot be accepted as it threatens the FLOSS
model at whole.
And IMHO neither distributions nor the community should accept projects
which call themselves open/free, but basically demand full control and
renaming/forking when one starts to change bits of the "experience"[1].

The FLOSS philosophy isn't just about having some open source license,
but on the other hand making it basically impossible to freely use
software as people wish (in this case: properly packaged in
distributions) with the small side node that people could always simply
fork if they're not happy. That's the Oracle way of handling open
source.
So, right choice from Debian,... it's just a pity that all people who
put their trust into subsurface are now kinda screwed.

Best wishes,
Chris.



[0] Official = the one from Debian
[1] Funny, btw, that this rule apparently only applies to downstreams.
If such upstream does the very same, like subsurface changes the
"experience" of libdivecomputer/etc. a rename is apparently not deemed
necessary by them ;-)



More information about the Pkg-running-devel mailing list