[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: gEDA-user: Users of verilog and VHDL gnetlist backends -=(BREAKAGE IN 1.4.0)=-



Hi Peter C., and All,

Was there a clearly posted RFC to move the 3 top level scheme variables
"packages, ..." to the gnetlist-post.scm to be included in the 1.4.0
release?  Normally that kind of things should not happen if we
follow certain guildlines that if you are trying to fix 
gnet-spice-sdb.scm,
it could be fixed without globally affecting others.

IMHO, if you intend to change globally something, then there should be
clearly posted RFCs for them, so that people do not have to check
every changes you made which should normally not affect them.

Although to fix this is simple, but I am a bit concern what else
could be changed in a similar uninformed fashion in the future,
especially those features which might dictates gEDA's future
architectural direction.

I would like to suggest a forum post "geda-RFC" similar to the
"geda-announce", so that the RFCs which have global significance
are clearly visible to everyone.  This is also a good way to
avoid unintended mistakes.


Best Regards,
Paul Tan




-----Original Message-----
From: Peter Clifton <pcjc2@xxxxxxxxx>
To: geda-user <geda-user@xxxxxxxxxxxxx>
Sent: Sat, 9 Feb 2008 1:49 pm
Subject: gEDA-user: Users of verilog and VHDL gnetlist backends 
-=(BREAKAGE IN 1.4.0)=-



WARNING, USERS OF VHDL and VERILOG NETLIST BACKENDS, DON'T UPGRADE TO 
1.4.0

Hi all,

I was doing some work looking at adding attribute output to the verilog
backend, and discovered some breakage. (It is dead in 1.4.0).

I've done some tests, and it seems that the VHDL backend is also
affected.

This was an unfortunate side-effect of code changes introduced for
slotting with spice-sdb between 1.3.0 and 1.4.0.

I'm working on fixes to the appropriate .scm files, but until those are
ready (and will go in 1.4.1), I'd suggest that anyone needing those
back-ends does not upgrade past 1.3.0.

Anyone interested, I have a gnet-verilog.scm backend fixed, and also
handling attribute output for modules. (Useful for simulation, esp. when
gnucap's Verilog-A support is out!)


Sorry for the breakage, it slipped past the gnetlist test suite when I
made the changes for spice-sdb. I hope to increase the gnetlist
test-suite coverage as a result of this problem. It turns out that only
a handful of the backends were being tested by it at all.


Regards,

--
Peter Clifton

Electrical Engineering Division,
Engineering Department,
University of Cambridge,
9, JJ Thomson Avenue,
Cambridge
CB3 0FA

Tel: +44 (0)7729 980173 - (No signal in the lab!)



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


________________________________________________________________________
More new features than ever.  Check out the new AIM(R) Mail ! - 
http://webmail.aim.com


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