Mail Archives: geda-user/2016/07/24/01:36:40
On Sat, Jul 23, 2016 at 02:30:12PM +0200, Roland Lutz wrote:
> On Sat, 23 Jul 2016, Vladimir Zhbanov (vzhbanov AT gmail DOT com) [via
> geda-user AT delorie DOT com] wrote:
> >at least before the point where [Xorn] would be comparable with geda-gaf
>
> Xorn isn't a competitor to gEDA/gaf. It's a set of libraries which provide
> some functionality whose purpose is to be useful to gEDA.
Let's be honest, it is. It uses other languages than ones are used
now in geda-gaf. You pretend it to be better while it does only a
part of work you can do in geda-gaf. Probably, I would support
your efforts, if our project were small. However, looking
pragmatically, and seeing a scale of work which is needed to move
the project forward (including all compatibilities issues), seeing
the number of contributors, and already having what we have, I
think it would be better to just work on it as it is, fix the C
side, extend the Scheme side, write documentation, and not to
break it all. Anyways, you know, Scheme is an interpreter, and it
can use the `system' calls the same way any shell does, so
everybody would be able to invoke his/her scripts in it as I've
already showed to the community. Therefore your work won't be lost
as is, it just can be done othewise.
>
> Xorn doesn't make much sense all of itself. It's meant to be used in
> conjunction with and as part of gEDA/gaf. The core libraries are not
> specifically tailored to gEDA, so it may, in the future, support other
> applications, too, but currently it's specifically made as a component for
> gEDA/gaf.
>
> >I see all the people who mention their preferred languages
>
> I didn't choose Python because it was my favorite language, as I think I've
> made clear in the past. I chose it because it was the least PITA and gets
> stuff done.
I believe it is because you've been familiar with it. When I
started to work with gEDA I didn't know neither Scheme nor Python
and have chosen to learn Scheme because it was used in the
project. You've chosen the way more convenient to you.
>
> >I see Roland tries to rewrite anything in Python.
>
> What I'm doing is not at all about "rewriting it all". I can only guess you
> are referring to the netlister. When I refactored it, translating it to
> Python was about 5% of the work (and I automated most of that). The other
> 95% of the work was untangling the ways the netlister works and putting it
> into a nice, understandable and, most of all, maintainable form. I actually
> considered translating it back to C once the refactoring was done, but
> decided against it because I didn't see much added value in that and would
> lose some of the flexibility.
While I admit our Scheme base is not perfect (because most of our
contributors don't try to have PhD here, and do what they can,
included me), I see it would be better to work on it rather than
re-write it all in other language in order just to
refactor. Because, again, people cannot learn many languages at
once to contribute everywhere, they tend to wait, which language
is the preferred one in the project.
I could repeat your last sentence in this paragraph, substituting
Scheme and Python for Python and C. I'm trying to be as polite as
I can, and to be as little discouraging as I can. I value your
work and sometimes I see great ideas in it, though I don't see
much added value in changing the language we use, but see more and
more frustrated users (including me). So my desire is to stop
this, to choose some direction, and move there.
> >Xorn still fails in various places and on different machines for me (at
> >home or at work), and I have no time to work on it, so I wouldn't release
> >it as is now.
>
> If you provided me with details, we could work on that. On every machine on
> which I've tested Xorn, it compiles and tests without problems, so there's
> only so much I can do without your cooperation.
>
I'm very busy man (somebody would laugh at this statement here,
but I can defend myself with the fact that I have five children, a
wife and two cats ;-)). Having some time at work I'm trying to
work on gEDA. I cannot work even on all reported bugs that I know,
that are related to my work and I know how to reproduce them and
am suspecting where to search for to fix them. I know, writing all
this is longer than writing a report, but I cannot. It usually
happens when I forget to use --disable-xorn if I do `make all
distcheck' when working on some bug/feature. So I was waiting for
other users to contribute in xorn before deciding if there is a
sense to include it in the project. Yet, I'm not inspired.
--
Vladimir
- Raw text -