Mail Archives: geda-user/2015/10/07/23:36:45
On Thu, 8 Oct 2015, Kai-Martin Knaak wrote:
> gedau AT igor2 DOT repo DOT hu wrote:
>
>>> Well, seen from a distance, this is just what Markus wrote. You did
>>> not feel like joining the "official geda team". You sure have your
>>> reasons. And you surely are not alone. There is no need to feel
>>> insulted.
>>
>> Nope, he literally wrote: "don't want to join anybody"
>
> The context was the geda main-line-developer-community (for lack of a
> better word). Quotes taken out of context have a tendency to sound
> differently. I can understand how you took the sentence as an attack.
I think it is you who get the context wrong. The funny thing is, that he
brought up my name out of the blue in that context only for a personal
attack. Read that whole mail again. Also read his answer to my answer
which again assures it was indeed a personal attack.
> But maybe you can likewise see that Markus may have referred to main
> line development. While he certainly did not express gratitude,
His answer would have been "sorry, I used the wrong words, I meant you
didn't want to join [my team] / [official geda team]". Please read the
thread again, his answer made it clear it was a personal attack and my
interpretation was right.
Anyway, I'll stop this thing here, as I'm pretty sure we won't agree. At
the end of the day, it doesn't really matter...
>
>>> While you technically forked the code. You still did not go all the
>>> way. That would have included complete separation of website,
>>> communication
>>
>> I do have a completly separate website. I did publish the URL.
>
> The page at
> http://repo.hu/projects/pcb-rnd/
> is a collection of diffs to main line pcb. A website for a completely
> forked project would focus on the project itself, not purely on the
> diffs to its parent. No offense, just a diagnosis.
The project is a collection of diffs to an old version of the mainline
pcb. The page itself does not contain patches, but documentation on the
patches.
I do not know what exactly you expect, should I copy the user
documentation of the parts that are the same in both mainline and my fork?
If so, why is it bad if they can be simply found at the mainine's web
page? Or a fork is a fork only if it changes at least 2/3 of the code?
>
>
>> I do not mess with anyone. Since pcb-rnd is compatible with pcb, I
>> regard it as a relative of the geda effort and I think the geda-user
>> list is appropriate place to discuss it.
>
> Whatever the reason, this is not complete separation as a complete
> fork would have it. Again, no offense, just a diagnosis. A semi-, or
> four-fifths-fork is not inherently better or worse than a full one.
So a complete fork has to copy the page of the original project and never
link or refer to it? Or should it try to change all code, just to make the
volume of differences bigger? Or should it totally rewrite all
documentation but can keep most of the code?
Sorry, I really don't get what a proper fork means for you.
Is gedahead a proper fork in your opinion? If so, what exactly makes it
proper and how does pcb-rnd lack that thing?
>
>>> This sends ambiguous messages. But it is totally ok, too. Just
>>> accept that
>>
>> Ambiguous message like what? I tried to make my intentions clear, if
>> you still have doubts, feel free to ask.
>
> Ambiguous as in:
> "I do not want to cooperate but I do want to cooperate."
I've written this down many times, but let's do it again: I generally want
to cooperate when it does work. When it does not work between me and a
specific team/project, I either try to change the things that make the
cooperation non-working for me or I can not join the given effort. In case
of pcb this has happened. I disagree with many little details about how
the project is run (management-wise, on some technical details like VCS,
goals and directions of the projects, etc).
I first shared my concerns on the list, and I saw change to the
directions I find good are impossible. It's fine, it's the decision of
the project. From then on, I had three ways to deal with it:
1. Lean back and be an user in the consumer sense. When series of bad
(for me) improvements were added, I tried to complain on the list but when
the consensus was sort of "wontfix", I just stuck with an older version
that actually worked.
2. Try to join the project anyway - I actually did submit some real minor
patches at some point, but because of the differences mentioned above I
got demotivated very fast and unintentionally falled back to 1. Before
people start to bring up names: it's nobody's fault, the project takes one
direction and I prefer another.
3. Fork the project and run the fork in the way I think is best.
Given the actual choices, after trying all three, I think what I am doing
is cooperation. There are more ways than one to cooperate. Joining the
mainline is only one; joining gedahead is another one. Implementing
something from scratch (e.g. Peted) is another one. And forking the
existing stuff publicly is one too. It can have benefits for the mainline
and the community. I implement features people can look at, evaluate and
try. If similar features are needed in mainline PCB you had a prototype
for free whether you chose to take the code 1:1 or implement it in a
totally different way or use pcb-rnd to learn the given feature is a bad
idea and should not be implemented in mainline.
I hope this clears any ambiguousity about my intentions. If not, just ask.
Regards,
Igor2
- Raw text -