Mail Archives: geda-user/2017/02/14/13:35:34
On 02/14/2017 12:10 PM, John Doty wrote:
>>> gEDA attracted and retained *me*. I started out using it to capture
>>> schematics for ngspice. Then, I was working with a layout contractor
>>> using PH70, an obscure Lincoln Lab layout tool. I was able to cargo
>>> cult a back end for that. Then it was IC design review (back to
>>> ngspice). Then IC design, making netlists for a layout shop using
>>> Cadence. A bunch of work on schematics for various customers with
>>> various layout tool preferences. Abstract symbolic circuit analysis
>>> with a Mathematica back end I wrote. Try navigating that with KiCAD.
>>> gEDA covers a very different space: that’s what it’s for.
>>
>> Very cool. But that's what it's for *for you*. I still think you should fork gEDA for your own use.
>
> OK, what’s your story? Why aren’t you using KiCAD?
My story? I design boards, both small and large ("large" to me means
more than a few hundred components), for both hobby and commercial
designs. I use gschem for schematic capture and PCB for layout.
Sometimes I target ngspice, but for me, everything ends up in solder in
the end. I drive my design flow with Makefiles. I do anywhere from 1-5
designs per month. I do end-to-end designs from conceptualization to
functioning prototypes and small production runs, including firmware if
applicable, and rarely have to interface with other designers/developers
or their tools. (something for which I am thankful)
I'm not using KiCAD because I chose gEDA+PCB back in 2003 and I have
a huge investment in time, symbols, footprints, and finished projects
(some of which occasionally need maintenance updates) all done with
these tools. And, like you, I value flexibility and the toolkit approach.
Because of all the progress-destroying crap and resistance to ANY
change that goes on on this list, I've looked into KiCAD a few times in
the past year or two, but I've never managed to actually get it built.
I could dig in and fix the build problems, but I have other work that I
must do. At least gschem and PCB usually compile out of the box.
But even if I'd gotten KiCAD running, I'd still have the huge of
investment of time and effort I've put into gEDA+PCB, plus the learning
curve of a new software package, plus the un-learning of all the
keystrokes etc that have become automatic for me over a decade and a
half. I'm a busy person; that's a show-stopper for me.
Right now, gschem (v1.8.2), gsch2pcb, and gnetlist do what I need of
them. There are some features I'd like to see, but honestly I could
probably keep using them as-is forever at this point. But I (unlike
you, it seems) would like to see some forward progress with gschem.
Development on PCB all but stopped, years ago it seems, and I'm able
to just get by with what it can do now. Don't get me wrong, I like the
program a lot and I'm very productive with it, but there are problems
and a lot of missing functionality. Pcb-rnd is gaining momentum and
appears to be moving in a very good direction (I test every release, and
some of their development stuff), but until it implements transparency
I'd find it very difficult to use for daily work. I think many people
don't realize how important/valuable transparency is until they actually
use it. The moment it gets transparency, I'll be using pcb-rnd full time.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
- Raw text -