Bug#859130: uploads to mentors, no updates on git repository
Geert Stappers schreef op 2018-06-12 15:34:
At https://mentors.debian.net/package/lina
are several uploads of verion 5.3.0-2 of lina.
Some are from may, most are from 11th of june.
However _no changes_ on the git repository
at https://github.com/albertvanderhorst/ciforth
That is not to be expected, as most complaints by lintian
are about what is in the debian directory, not under control
of upstream.
That means the (ab)used .orig.tar.gz still
gets it's forth.lab and makefile from unknown source.
Since when is an upstream author 'unknown source'?
The author is well known, it is me, see the copyright messages.
Besides, where is the requirement in Debian that an upstream source
should be a dump of a git repository?
So I reject this complaint and maintain that the lina
package is fit for sponsoring.
The ciforth archive is not. I'm not aspiring to add a factory for
generating MS-windows and Apple programs to Debian, as
1. It goes against the spirit of what I want to supply Debian with:
a simple compiler that will spawn a number of e.g.
single board computer related utilities.
A compiler that is by far the most understandable compiler
in the whole of Debian.
2. ciforth probably simply doesn't belong in Debian
3. It would require a tremendous upgrade in quality.
Where I'm 100% behind the quality of lina, not so
much for ciforth. I could not promote adding it to Debian.
Even so, if a Debian developer thinks the ciforth compiler
factory is a valuable addition *to a debian distribution*, (s)he is free
to fork
ciforth and get it accepted.
********************************************************************
* Please note! The famous gcc has no means to generate a native *
* windows compiler. so ciforth would beat gcc to it! *
* Of course chances are better with a compiler designed for *
* simplicity but i'm not foolish enough to gamble that ciforth *
* could pass Debian's quality checks in my life time. *
********************************************************************
See https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=859130#58
where that was previously reported.
I doubt if this ITP is really about
getting a source from a git repository into a debian package.
Of course it is not.
The git repository contains means to get windows
and apple sources that are specifically not going into the lina
package. So it is plain that you don't understand the difference
between ciforth and lina, and that you don't respect the integrity
of an upstream package c.q. upstream maintainer.
That is the reason you're no longer my sponsor.
Remember how this started?
My source package contains
lina.fas
lina.texinfo
forth.lab
There was not even a Makefile . Build with
fasm lina.fas
which is mentioned somewhere in lina.fas
The simplicity is a goal made possible with complexity in ciforth.
Then the whole circus gets started with the remark that lina.fas
is not a "proper source" based on a rule that was supposed to contain
malevolent manufacturers, not hold back free software developpers.
Anyway the "preferred source of modification" rule is obeyed by
supplying everything that could be considered "preferred source
of modification", so that point is covered.
See also
https://github.com/albertvanderhorst/ciforth/wiki/Philosophy-behind-ciforth
An example of usage of the compiler including some parallel
calculations:
https://github.com/albertvanderhorst/primecounting
I welcome a new sponsor!
You will sponsor a very simple package that is
mature and has a very responsive maintainer.
Groeten
Geert Stappers
Groetjes Albert
--
Suffering is the prerogative of the strong, the weak -- perish.
Albert van der Horst
Reply to: