[Pan-devel] Re: Re: Various problems when launching pan CVS (bis repetita)

[Duncan]

GISQUET Christophe posted <address@hidden>, excerpted
below,  on Sat, 09 Jul 2005 15:20:28 +0200:

> There is indeed a philosophical/ethical debate on whether to use and offer
> open source/free software on a proprietary platform like the one you
> quote. This is a matter of personal choice, so let's keep technical, and
> have the focus on the development side.

Wow.  I hadn't intended to get into that and made the comments in a more
general sense, but with the issue now on the table, I can't resist at
least the following.

I do think it's "neat" to be able to do ports to closed source platforms,
and support, for instance, the KDE and GPL-QT4 on MSWormOS projects, as
well as GTK+ and PAN, and Firefox and OOo. However, that's not for me. 
Save for open standards reference implementations, where I /could/ be
motivated to work on a BSD licenced project, my motivation comes from the
community benefit, both from my code and from others that might find it
useful to add to or change.  Without that benefit enforced by GPL (or at
least LGPL) style licenses, while I certainly acknowledge the value of all
the BSD licensed code I may use, there's simply no motivation for me to
contribute anything more than trivial bugs and/or fixes.  I've no less
respect for those that choose a BSD license, or those that port to closed
platforms, respecting the differences in personal motivation folks may
have, but it's just not something I'm personally interested in
contributing to, myself.

> It also depends on how Pan developpers want to get my help. I thought my
> previous reports were borderline betweend -devel mailing list and
> bugzilla? Therefore, how about opening bug reports on bugzilla? I'm not
> sure how developpers prefer their bug+patch submissions.

Charles definitely prefers them /eventually/ in Bugzilla, because it makes
tracking them in an orderly way much easier, so even bugs and code that
isn't implemented immediately remains recorded for possible use later. 

That said, I'm a fairly strong believer in the lists (or newsgroups as I
access them thru gmane.org), for discussion of the merits, before it's
filed away in bugzilla, at least if it's something that would change
PAN's intended behavior and not just a bug fix. There have been a few
features that went into PAN that were later removed because folks either
HATED them, or some liked them but there simply wasn't strong enough
support for them in the PAN community to justify the code and menu bloat
in PAN.  (If there's one thing Charles has demonstrated a number of times,
it's that he /likes/ a lean and mean PAN, functions that don't get used
much get dropped.)  It is my belief that were these vetted on the
groups/lists first, at least some of them would have been found to not be
worth the effort in the first place, and the code would have never been
made part of PAN, only to be removed later.  Thus, any behavior changes or
feature additions should IMO be posted for discussion to the group, here
if its an API or similar "dev only interest" item, to the user group if it
will involve UI or behavior changes visible to the end user.

After any discussion on the groups/lists, with any suggested changes taken
into account, /then/ it can (and should) be posted to bugzilla for
tracking and possible eventual inclusion.  I'd go ahead and post it there
even if there's little interest on the groups, as long as the response is
not overwhelmingly negative (as would likely be the case for any suggested
changes negating the GNKSA approval, for instance -- Charles is a strong
supporter of that, so PAN is as well, and many regulars in the groups
including me are as a result -- if we weren't, we'd very possibly be using
other readers).  First, even if there's little group interest, that
doesn't mean Charles won't find it personally interesting.  Second, again,
there's the archiving effect.  Even if the code or idea behind it isn't
merged now, it's entirely possible it might become useful later.  A
discussion on the list will be lost with time, or anyway even with the
archives difficult to find again. A filing in bugzilla will be easily
found at some point in the future, if necessary.

> None is free software, and there is no alternative to graphical
> closed-source clients. Anyway, my focus is on Pan, which I really like.
>   I also think it is a worthwile newsreader compared to those clients.
> Besides, there is a technical thrill to use a software you can check what
> it is made of.

Very valid points!  I do actually tend to agree, but at the same time,
honestly find it difficult to understand anybody who'd go to all the
additional lengths required to get and keep PAN running on MSWormOS, when
there are so many other options available, without ultimately finding that
moving to an entirely open source platform, either the BSDs or Linux,
better fits their goals.  Of course, as with the above motivations for
development thing, that's just me.  I do realize and in fact appreciate
the variety in folks that would cause them to want to go to the lengths
necessary to get and keep PAN working on MSWormOS, even if I don't fully
understand the reasons why they might be so interested.  Without such
personal variations, the world would be a VERY boring place!

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman in
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html