tech-pkg archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: wip/iverilog

Hash: SHA256

On 13.12.2015 16:22, Greg Troxel wrote:
> Kamil Rytarowski <> writes:
>> My proposition is to upgrade pkgsrc version with Icarus Verilog
>> from wip/iverilog and optionally pull in modifications from 
>> wip/verilog-current.
> You straightening out all the patches, filing them with upstream, 
> etc. is of course most welcome.
> Presumably you have already checked in with the MAINTAINER before 
> embarking on this?

I checked that Makoto was the person in recent years who contributed
changes to it. He's addressee of this mail too. So called
verilog-current from pkgsrc is 6 years old.

And since I don't understand local modifications (there are no
comments why we are doing what we do), I leave the rest to the maintaine

>> Do we need 4 packages with the same tool, but different name? Can
>> we obsolete them and stick to stable and optionally -git
>> version?
> Since this seems to be a single upstream, I don't see any reason
> for more than one package with a release (in pkgsrc proper) and a
> snapshot one (perhaps in wip, since we tend to avoid other than
> releases in pkgsrc).
>> Can we stick to new iverilog name? For me 'verilog' is like name 
>> 'editor' for an editor.
> As for the name, I see your point about 'verilog' properly being
> the name of a language, not a program, but the upstream tarball is
> called verilog- and unpacks to that.  Plus it's been that way
> forever and has not generated complaints.

Since I took tarball from GitHub the tarball is iverilog for me. I was
surprised that 'verilog' might be the same thing.

> I don't follow "stick with"; the precedent is "verilog", so any
> proposal for a change is just that.  Generally I view changes as
> having a cost and tend to avoid them (cvs history, updating, etc.).
> All in all I would leave the package as verilog because that's the
> upstream tarball name.  You could file a bug asking them to change
> it to icarus-verilog or iverilog and we could of course follow
> suit.
> But if you're going to take care of this, I don't object - that's
> just my graybeard rant.

My proposition comes from the fact that pkgsrc one is outdated and
unlikely actively maintained (and used?). Otherwise I wouldn't come
with this thought.
Version: GnuPG v2


Home | Main Index | Thread Index | Old Index