Mail Archives: geda-user/2019/01/30/12:24:57
On Wed, 30 Jan 2019, John Doty wrote:
>So, we have a three way fork, with pcb-rnd centered on repairing the
>architecture of pcb, but with its own schematic capture, and gEDA becoming a
>pcb-centric tool, no longer much of a kit. Neither of these forks seems
>focused on maintaining the flexibility of gEDA as a primary goal: Lepton is
>keeping that dream alive.
Most of what you wrote about pcb-rnd is inaccurate.
Disclaimer: the only reason I write this mail is to straight out some
facts about pcb-rnd. Because of the context I have to refer to some other
projects - but I am not making any requests or suggestions to those
projects.
"pcb-rnd centered on repairing the architecture of pcb" - this is only
partly true: we did rewrite a lot of infrastructure. While having a strong
infrastructure is an important goal, it is not the only goal and we are
not centered around it. We did a lot of other things (it's are easy to
figure with 2 minutes of research): we introduced a very long list of new
features and bugfixes of existing/old code (that we didn't replace and are
not infrastructural), we wrote a lot of docs too.
If I had to name something pcb-rnd is centered around it would be:
providing a real good, flexible pcb editor for users with the UNIX/hacker
mentality, to be used as a tool in a toolkit. Rewriting infra is only one
of the tools perfecting that. (It's funny because pcb-rnd does a lot of
things that you are constantly talking about as most desired toolkit
approach stuff.)
Your sentence suggests pcb-rnd is not flexible, without backing up that
with facts or references. Which is no surprise: pcb-rnd _is_ flexible and
is not tied together with any schematics editor. In fact, pcb-rnd supports
schematics import from 8 different (specific) schematics tools and
supports 4 generic netlist/forward-annotation methods.
See: http://www.repo.hu/projects/pcb-rnd/user/09_appendix/bridges.svg
"but with its own schematic capture" is simply false. Please check your
facts before posting:
pcb-rnd does _not_ have its own schematics capture and there was and is no
plan to have one. There are plans to have a _separate_ schematics editor
(cschem), and lately my preferred one is xschem. But there were exactly
zero plan ever to tie pcb-rnd together with _any_ specific schematics
capture tool in any way. Both xschem and cschem are for supporting
multiple flows and software, not narrowed down to pcb-rnd either. You
often suggest the situation is "lepton-eda is toolkit and everything else
is monolith integrated rigid tailored-for-one-workflow hack", but the fact
is that pcb-rnd is really a good tool that combines very well with a large
numer of other tools to form different tookits, supporting a real huge
number of applications/workflows.
The "Neither of these forks seems focused on maintaining the flexibility
of gEDA as a primary goal" is not true at all. pcb-rnd does maintain at
least the same flexibility, or even higher flexibility than lepton-eda or
gEDA. Fact: besides interfacing up and down, we are also interfacing
sideways. We can load and often save in the file format of "competing"
layout tools. This way pcb-rnd can be used as a converter tool or
preprocessor or postprocessor in a non-pcb-rnd workflow where the user
doesn't GUI-edit anything in pcb-rnd (but in kicad, eagle, protel, etc.).
Does lepton-eda do any sideway interfacing, e.g. loading eagle or
kicad/eeschema or tinycad or ltspice schematics or symbols?
Flexibility _is_ among the primary goals of pcb-rnd: we constatly hook up
with new tools and build more bridges. We are interested in building up a
good, flexible, toolkit-approach ecosystem - we have a dedicated
subproject for that (coralEDA).
The only (very much bent) interpretation of your sentence that is true is
that we are not working on making gEDA more flexible: we don't think the
goals (on toolkit aspects and flexibility) could be achieved within the
gEDA framework. There are projects that are open for cooperation and see
the big picture and agree on the benefit of combining tools - and we tend
to team up with them. Anybody is welcome, we don't care where cooperating
project are coming from, but we don't push it when we see some projects
don't want to join.
So please don't confuse the level of cooperation pcb-rnd gets from your
favorite tool with pcb-rnd's flexibility. Please don't try to define what
is not a goal of pcb-rnd: you have no role in pcb-rnd, you can't set or
change our goals. Please don't make up non-existing parts ("own schematic
capture [of pcb-rnd]") and don't spread that as if it was a fact.
Regards,
Igor2
- Raw text -