Re: gEDA-user: good idea - bad idea ?

["KURT PETERS" <peterskurt@xxxxxxx>]

DJ,

 It seems that you are advocating for "all" newlibs to be in a "personal 
library".  I don't necessarily have a problem with that, but that kind of 
would destroy a naming convention or two (hyperbole), unless gsch2pcb and 
pcb really get the prioritization right as you suggested, possibly using 
some internal attribute.  Is any developer of gsch2pcb ok with all that 
(i.e., is he/she ready to support those changes)?  I suppose it'd be "easy" 
to add the attribute to M4's since you're going to pre-compile them anyways, 
but it may be difficult to add to the already vast amount of newlibs out 
there.
   You have a great point about the necessity to go through each and every 
footprint in everyone's library and "vetting" them for a multiple of 
usability metrics.  That's a tall order.  The main problem is that I think 
right now a lot of people prefer using the newlibs and many of those are 
"better" than the same-named M4.  I think short of some kind of painful 
vetting process, there has to be some way to let the user continue (I use 
that word sparingly since the only way to force non-M4's seems to be to 
move/rename them) to segregate M4's from newlibs.

 Regards,
Kurt
<><><><><><><><><><><><><><><><><><><><><><>
Message: 3
Date: Wed, 3 Jan 2007 21:07:48 -0500
From: DJ Delorie <dj@xxxxxxxxxxx>
Subject: Re: gEDA-user: good idea - bad idea ?
To: geda-user@xxxxxxxxxxxxxx
Message-ID: <200701040207.l0427m9r030585@xxxxxxxxxxxxxxxx>


> What, exactly, does this mean?
> "Bad idea, since we're planning on moving the M4 step to build time in
> the future, making every footprint a newlib style."
>
> Some of the M4's are not-so-great, so if this means that the M4's will be 
> converted to newlib, thus essentially forcing users to use the current 
> M4's, then I'm not in favor.
>   But, since I don't know what that statement implies, I wasn't sure how 
> to respond.

We at least want to pre-compute all the footprints so that the user
doesn't need to run M4 while they're laying out boards.  For example,
this would help Windows installations, which don't have M4 by default.
That would also allow us to use other programs to generate footprints,
like perl or python.

I think keeping them as M4s just because you want to segregate "bad"
footprings from "good" ones is misleading.  EVERY footprint is
potentially good or bad; for example, the 0603-style footprints in M4
are quite good, and maintaining them as macros is very easy.

My comment was simply that in the future, we will only have one type
of footprint.  What you need to do *now* is just make sure that your
personal libraries have preference over the default ones when the
names conflict, using techniques that don't depend on whether the
footprints are M4 or newlib style.

So, using a "newlib_*" prefix is bad, because eventually there will be
no "newlib" just "lib" ;-).  Using a "mine_*" prefix is better, or
making a practice of either fixing existing footprints or providing
missing footprints, while staying with standard names.

Note that PCB now has the ability to store arbitrary attributes in
footprints; mine all have copyright and origin information in them so
I can track them better after they're in the board.




_______________________________________________
geda-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user