On W= ed, Apr 1, 2015 at 11:00 PM, <= gedau AT igor2 DOT repo DOT hu> wrote:

On Wed, 1 Apr 2015, Russell Nelson wrote:

IMHO, the gedasymbols library is fine. If it were easier to use, people
would use it more. Here is what I suggest:=C2=A0 o Take the current set of<= br> symbols and footprints and incorporate them into gedasymbols.
=C2=A0 o Nuke the current set of symbols and footprints.
=C2=A0 o Replace it by a cached copy of gedasymbols.
=C2=A0 o Make it easy to refresh the cached copy using, say, rsync or somes= uch.
=C2=A0 o ONLY, and I mean ONLY pull footprints out of the cached copy.

In other words, when you create a footprint or symbol, the only way to use<= br> it is to pull it out of gedasymbols. So when you create a symbol or
footprint, you can simultaneously creating it on gedasymbols. If you happen=
to be offline, well, you can pull it out of your cached copy. But when you<= br> refresh, the refresh works bidirectionally. Everything you've created becomes available to everyone else.

Would be the moment I had to fork gschem and rest of geda as well. I simply= don't believe in such centralized/cloud "we tell you where and ho= w to store your own stuff" methods. Some implications:

1. licenses: what if someone wants to have a symbol/fp lib that is not dist= ributed or distributed under different conditions that's not applicable= on gedasymbols? E.g. you are a small company trying to make proprietary de= sign (contract work?) using open tools and for whatever legal reasons you a= re not allowed to share any parts of the design (including new symbols/fps)= .

You use GPL software. Why do you obje= ct to a GPL parts library?

=C2=A0

2. sandbox: what if I want to have 200 variants of a symbol/fp for whatever= experiment? Do I have to pollute the central library and everyone will alw= ays pull all them?

Yes. Some contributi= ons would be worthless. Isn't that always going to be the case?

=C2=A0

3. in case you still allow local symbols of some sort: what if I have multi= ple systems I want to use my own local symbols on? I will use version contr= ol on them (just like now). What if a central symbol added later starts to = collide with my own?

My proposal makes = no allowance for local symbols. Isn't that clear enough?

=C2= =A0

I think this kind of cloud approach would decrease usability/flexibility of= the tools big times. This kind of thing is more appropriate to web based d= esign tools.

So you don't use gedas= ymbols? You only use the symbols/footprints that comes with gschem/pcb? Or = else you create your own proprietary symbols and never share them? Well the= n you would never sync with gedasymbols, so what about this proposal is dis= pleasing you?

=C2=A0

Regards,

Igor2