Re: [gentoo-dev] Re: "scm" in GLEP 54 (was: Council meeting summary for meeting on May 14, 2009)
> On Sun, 17 May 2009, Ciaran McCreesh wrote:
> About a million years ago, we were going to move all the SCM
> packages into their own category (but it never happened, because
> port001's script didn't work). There was a huge bikeshed debate
> about whether to use vcs, rcs, scm or something else. In the
> interests of getting anything decided, Seemant made an executive
> decision and picked 'scm'.
And please don't mix completely unrelated topics. The discussion at
the time [1] was about moving dev-util/{cvs,git,subversion} etc. to
a new category, and clearly "dev-live" would not be a good choice
for that.
Ulrich
[1]
Re: [gentoo-dev] Re: "scm" in GLEP 54 (was: Council meeting summary for meeting on May 14, 2009)
> On Sun, 17 May 2009, Ciaran McCreesh wrote: > History suggests that if it goes up for debate again, no decision > will ever be reached. If we simply have to decide between alternatives "scm" and "live", then I don't see what should be so complicated about reaching a decision. GLEP 54 doesn't really make clear the connection between the suffix and source code management is. It mentions source code management only shortly in the abstract, and then discusses things like version ordering that are not related to it. And does it really matter if the ebuild obtains its sources via a SCM system, or by some other means? Seems to me that "live" describes the property better. Ulrich
Re: [gentoo-dev] Re: "scm" in GLEP 54 (was: Council meeting summary for meeting on May 14, 2009)
On Sun, 17 May 2009 13:24:27 -0600 Joe Peterson wrote: > Thomas Anderson wrote: > >- Vote on GLEP 54 > >This vote was called for by dertobi123. The vote was on > > whether to approve GLEP 54 conditional on whether GLEP 55 is > > passed. The reason for this is that GLEP 54 is unimplementable > > without the problems mentioned in GLEP 55 being solved. > > I have not seen much discussion lately regarding the choice of the > string, "scm" in this GLEP. I asked the author today on IRC, and he > said he doesn't have a particularly strong reason for "scm" beyond > historical reasons. About a million years ago, we were going to move all the SCM packages into their own category (but it never happened, because port001's script didn't work). There was a huge bikeshed debate about whether to use vcs, rcs, scm or something else. In the interests of getting anything decided, Seemant made an executive decision and picked 'scm'. History suggests that if it goes up for debate again, no decision will ever be reached. Thus, the only sensible thing to do is to let the old decision stand. -- Ciaran McCreesh signature.asc Description: PGP signature
[gentoo-dev] Re: "scm" in GLEP 54 (was: Council meeting summary for meeting on May 14, 2009)
Thomas Anderson wrote: >- Vote on GLEP 54 >This vote was called for by dertobi123. The vote was on whether to >approve GLEP 54 conditional on whether GLEP 55 is passed. The reason >for this is that GLEP 54 is unimplementable without the problems >mentioned in GLEP 55 being solved. I have not seen much discussion lately regarding the choice of the string, "scm" in this GLEP. I asked the author today on IRC, and he said he doesn't have a particularly strong reason for "scm" beyond historical reasons. Since we are stuck with the string once it is adopted, I think we should consider the choice carefully. Personally, I'd prefer "live", since it is what we've been calling these ebuilds for a long time, it's easier to remember (and more "catchy"), and it seems to carry the spirit of what we mean by these kinds of ebuilds. Also, there is a new in-ebuild property with the signifier "live". Comments? -Joe
