gEDA for Windows?

[Joerg]

Since quite a while I see a hint here or there that someone wants to
compile a Windows version containing some of the gEDA modules. But
somehow that seems to always fizzle.

Is there a realistic chance we'll ever see a somewhat "official" Windows
version? Maybe along the lines of "Here, use it but please don't call us
for support."

I know that it's possible to compile it for Windows. However, lots of us
are more hardware guys, don't have compilers, don't have much experience
doing that, and would need something that installs itself just like
other CAD software does. After that we could be productive for the group
by contributing library models, tricks and so on. For many reasons
hardware guys can't migrate away from Windows, mainly because things
like the mechanical CAD, beamfield simulators and the like would then quit.

 

[Cesar Strauss]

Since quite a while I see a hint here or there that someone wants to
compile a Windows version containing some of the gEDA modules. But
somehow that seems to always fizzle.

It's one of my goals to compile the Windows version of some of the
gEDA tools, starting with the schematics editor and the pcb layout
tool. I'm not part of the core team of developers, but I do contribute
some source code to the project, with that goal.

Is there a realistic chance we'll ever see a somewhat "official" Windows
version? Maybe along the lines of "Here, use it but please don't call us
for support."

I think there is a good chance. Not only that, I'll be glad to answer
Windows-related questions posted here or at the gEDA user mailing
list. I expect others would help as well.

I know that it's possible to compile it for Windows.

Almost possible, I'll say. Much work has already been done, but a few
minor user-visible issues remain. I would not wish to make available a
Windows version until most of these are solved, it could impact
negatively on the reputation of the project as a whole. But it's only
a matter of Time.

However, lots of us
are more hardware guys, don't have compilers, don't have much experience
doing that, and would need something that installs itself just like
other CAD software does.

Very understandable.

After that we could be productive for the group
by contributing library models, tricks and so on.

That would be great.

For many reasons
hardware guys can't migrate away from Windows, mainly because things
like the mechanical CAD, beamfield simulators and the like would then quit.

I know that feeling...

Regards,

Cesar

 

[Joerg]

It's one of my goals to compile the Windows version of some of the
gEDA tools, starting with the schematics editor and the pcb layout
tool. I'm not part of the core team of developers, but I do contribute
some source code to the project, with that goal.


I think there is a good chance. Not only that, I'll be glad to answer
Windows-related questions posted here or at the gEDA user mailing
list. I expect others would help as well.


Almost possible, I'll say. Much work has already been done, but a few
minor user-visible issues remain. I would not wish to make available a
Windows version until most of these are solved, it could impact
negatively on the reputation of the project as a whole. But it's only
a matter of Time.

So there is hope. That is great to hear.

It may be ok to release something that's not perfect if it clearly says
so on the download page. Maybe with some hints as to what might not
quite work yet.

Very understandable.


That would be great.


I know that feeling...

Almost every week someone tells me to use Linux and all the crashes and
stuff would go away. Then I ask them whether my .NET driven scope would
still work. "Umm, probably not."

Incompatibility actually goes as far as Vista. Several of the SW
packages I use are almost guaranteed not to work there. At least that's
what the mfgs said. Luckily one can still buy XP (done that again about
a month ago).

 

[Joel Koltner]

Almost every week someone tells me to use Linux and all the crashes and
stuff would go away.

That's optimistic anyway -- buggy software has a lot more to do with the guy
who programmed it than the particular operating system it's running on.
Windows "blue screens" (OS crashes) are about as rare as Linux kernel panics
(crashes) these days. (Although when it comes to Vista they made so many
changes that you can't really accept Microsoft's claim that it's a
"compatible" operating system with XP... :-( )

Have you looked at virtualization software such as VMware? Let's you run
Windows and Linux (and other OSes) all side by side on the same box.

 

[Joerg]

That's optimistic anyway -- buggy software has a lot more to do with the guy
who programmed it than the particular operating system it's running on.
Windows "blue screens" (OS crashes) are about as rare as Linux kernel panics
(crashes) these days. (Although when it comes to Vista they made so many
changes that you can't really accept Microsoft's claim that it's a
"compatible" operating system with XP... :-( )

According to compatibility notices from several mfgs of software I use
it definitely isn't 100% compatible. Which is why I don't use Vista.

Have you looked at virtualization software such as VMware? Let's you run
Windows and Linux (and other OSes) all side by side on the same box.

I've only heard that it can take a long time to make everything work in
the box, virual RS232 over USB and such vital things. But no, I haven't
tried yet. Basically Windows does everything I need.

 

[Joel Koltner]

I've only heard that it can take a long time to make everything work in the
box, virual RS232 over USB and such vital things.

Yes... if you need the virtual computer to control Real Hardware (not just
hard drives, your Ethernet card, keyboard/mouse, etc.), you are better off
just getting a second machine. For something like gEDA though, a virtual
Linux box could work quite well.

BTW, you might find this site useful if you haven't come across it before:
http://www.fritzing.org/development/market-overview . The KiCAD guys made a
smart move in using wxWidgets for their GUI -- it makes it pretty trivial to
support releases on multiple platforms. Hmm... I see that Magic is still
around for VLSI design... that's cool -- as an undergraduate I remember some
people using it. I used John Beetem's Galaxy program in a couple of digital
design classes way back when, but it seems to have died when he left the
University over a decade ago now.

 

[Joerg]

Yes... if you need the virtual computer to control Real Hardware (not just
hard drives, your Ethernet card, keyboard/mouse, etc.), you are better off
just getting a second machine. For something like gEDA though, a virtual
Linux box could work quite well.

2nd machine is out, there are already three PCs here and they must run
Windows. To be honest I don't know how to roach a virtual Linux box onto
a Windows machine.

BTW, you might find this site useful if you haven't come across it before:
http://www.fritzing.org/development/market-overview . ...

Quote "To save development time, we will base this version on Eagle and
re-use as much Eagle functionality as possible."

Umm, why are they doing this? Wouldn't that be a waste of grant funds
since a commercial product already exists? Might as well use Eagle which
I do right now. One reason why I want to bail is that Eagle does not
support hierarchical sheet structures and Cadsoft does not seem to grasp
the need for that. V5 doesn't have it either so I won't upgrade. I guess
the Fritzing guys don't understand that either.

... The KiCAD guys made a
smart move in using wxWidgets for their GUI -- it makes it pretty trivial to
support releases on multiple platforms. Hmm... I see that Magic is still
around for VLSI design... that's cool -- as an undergraduate I remember some
people using it. I used John Beetem's Galaxy program in a couple of digital
design classes way back when, but it seems to have died when he left the
University over a decade ago now.

If I switch I'd like that to be to something that has a potential to
become mainstream here in the US, not Europe. One other problem with
Eagle is an utter lack of marketing efforts in the US. In consequence
none of my clients has ever heard of it, let alone uses it. Same for my
layouter :-(

 

[Joel Koltner]

Quote "To save development time, we will base this version on Eagle and
re-use as much Eagle functionality as possible."

Umm, why are they doing this? Wouldn't that be a waste of grant funds since
a commercial product already exists?

They sound a little naive, IMO. See:
http://www.fritzing.org/development/implementation/considerations ... I'm sure
they genuinely think they can address some perceived "need" that no one else,
no date, has... and do so before their grant runs out, or at least make enough
progress to get it renewed.

If I switch I'd like that to be to something that has a potential to become
mainstream here in the US, not Europe.

Hmm... that is a bit of a tough problem, especially if you're trying to keep
the costs down -- it seems U.S.-based EDA tools are almost always noticeably
more expensive than their European or pan-Asian counterparts (there's a
surprisingly large amount of cool stuff that comes out of Oz and Kiwiland
these days... probably China and Japan too if I knew the language...). (This
is especially odd when you have trash like ORCAD Capture which is "maintained"
in India today -- just where are all those maintenance fees going, anyway?)

One other problem with Eagle is an utter lack of marketing efforts in the
US.

I think the first time I saw their ad was in a Nuts & Volts magazine years
ago. I wonder if they ever moved up to at least, e.g., Circuit Cellar Ink?

---Joel

 

[Joerg]

They sound a little naive, IMO. See:
http://www.fritzing.org/development/implementation/considerations ... I'm sure
they genuinely think they can address some perceived "need" that no one else,
no date, has... and do so before their grant runs out, or at least make enough
progress to get it renewed.

Possibly it's about grants, don't know.

Hmm... that is a bit of a tough problem, especially if you're trying to keep
the costs down -- it seems U.S.-based EDA tools are almost always noticeably
more expensive than their European or pan-Asian counterparts (there's a
surprisingly large amount of cool stuff that comes out of Oz and Kiwiland
these days... probably China and Japan too if I knew the language...). (This
is especially odd when you have trash like ORCAD Capture which is "maintained"
in India today -- just where are all those maintenance fees going, anyway?)

Paying for the corporate Lamborghinis maybe

Actually CAD is cheaper here. Always was. When I lived in Europe I
bought OrCad here for half of what it cost over there. Even Eagle is
cheaper here than in Germany (where it is written!). What I do not
understand is why the industry in the US blindly plunks down whatever
OrCad costs, a program that is IMHO not nearly as good anymore as it was
in the DOS days. It's price has tripled since the early 90's.

I think the first time I saw their ad was in a Nuts & Volts magazine years
ago. I wonder if they ever moved up to at least, e.g., Circuit Cellar Ink?

I've never seen an Eagle ad in the US. I stumbled upon it by pure
coincidence when browsing for a new CAD after OrCad blue-screened on me
one too many times.

 

[Marra]

I've never seen an Eagle ad in the US. I stumbled upon it by pure

coincidence when browsing for a new CAD after OrCad blue-screened on me
one too many times.

CAD programming can be highly complex.

When i wrote my program I had to get into things I hadnt done in 10
years of programming professionally.
I had to use factorial arrays for autoplacing.
I had to get into some quite deep algorithms for auotrouting.
Even clearance checking took some jiggery pokery to get the check down
to a sensible time.

What is probably most staggering is the original program was 330,000
lines of Assembler !

Right from the start I had to get involved with a Gerber/Excellon
reader to ensure the output from my program was correct as there
werent many gerber readers in those days.

www.ckp-railways.talktalk.net/pcbcad28.htm

 

[Joel Koltner]

CAD programming can be highly complex.

Marra, as seems to often happen your response here isn't really related to the
discussion and comes off just as an advertisement for your own software.

When i wrote my program I had to get into things I hadnt done in 10
years of programming professionally.

Neither programming nor designing electronics isn't much fun if you aren't
doing new things are a semi-regular basis.

I had to get into some quite deep algorithms for auotrouting.

Too bad you haven't shown us any results of your autorouter vs. anyone else's
(or even your user's manual...). My experience is that most of the "included
with the base package" autorouters produce results that are lousy enough that
most people would probably prefer to route by hand. I have seen spendy
autorouters that are pretty good (i.e., one who takes the time to master them
would definitely save significant time in the long run) -- especially for
large digital designs.

---Joel

 

[k k]

Joerg:

and they must run
Windows.

No of course not. It's just your lasiness.

 

[k k]

Marra:

http://en.wikipedia.org/wiki/Kicad

 

[Joerg]

Joerg:

No of course not. It's just your lasiness.

Then tell me how to hardware-control the ICOM R-1500 or the Instek
GDS-2204 from a non-Windows platform.