[Qgis-developer] What is "is a qgis problem"? (it was Re: [Qgis-user] Scattergram with compiled qwt5.2.0)

[Barry Rowlingson]

On Sat, Nov 7, 2009 at 11:41 AM, Agustin Lobo <alobolistas at gmail.com> wrote:

> will put a significant part of their paid time on debugging, enhancing and
> extending QGIS. This is the case of R, which is, i my opinion, the paramount
> example of success of public domain software, at least in science.

 [legal note] R (and Qgis) is not 'public domain software'. 'Public
domain' is a legal term that generally means out of copyright and with
no usage restrictions. R and the packages in CRAN are under a variety
of open-source licenses including the GPL, and are copyright of
various authors and institutions - it is this copyright that allows
authors to place works under the GPL. [ends]

 My 2 euros:

 Packages are accepted into CRAN only if they pass various QC checks
and if they have an active maintainer. If a maintainer gives up on a
package it disappears from CRAN. This doesn't happen often since an
outgoing maintainer will advertise and a keen user will take it up.
Qgis is similar - Clearly a problem with a plugin in a third-party
repo is not a Qgis issue, and shouldn't be tracked in the qgis trac.
If the developer doesn't fix it then it's open source -- the user can
fix it themselves or pay to get it fixed. Plugins in the qgis repo are
a qgis problem, and if they can't be fixed by maintainers then should
be 'orphaned'.

 Your problem with a student having trouble getting their coursework
done because of a software bug also occurs in proprietary software but
worse - I've seen bugs - serious bugs - in a proprietary stats package
go unfixed for years. Just trying to find a place to report bugs is
often impossible. I had to resort to emailing an old friend who worked
for the company after being unable to find a bug report email address
on their website. I suspected the company thought their program had no
bugs. Try finding a bug tracker or support forum for SPSS even today!

 The situation with R is often that the gap between developer and user
is very small. Packages are often written because that person wants to
use a particular functionality (often from theory they have
developed). With Qgis mostly the theory is well-developed (raster
algebra, geometry calculations) and users just want to use it, and the
motivation for developers is less since they can probably do it all in
Grass anyway.

 Actually I don't know what drove half a dozen or so to gather in
Vienna this week! You're all mad! :)


Barry