begin quoting Lan Barnes as of Fri, Oct 20, 2006 at 12:19:57PM -0700: > On Fri, Oct 20, 2006 at 11:34:01AM -0700, Stewart Stremler wrote: > > one of these days I should look at the CVS code "in anger", > > just to see how hard it would be to add some features... > > Why? No, _really_? Because that's exactly what the SVN development crowd > did, keeping wherever possible scrupulous compatibility with the theory > and command structure of CVS. So what you propose has been done and done > well already.
Dependency hell is a huge argument against "done well". I don't think the SVN folks were wrong to go off and roll their own; I am pretty convinced that what they did once they made that decision *was* wrong. SVN has one feature going for it -- it's mostly compatible with the CVS commandset/worldview. (I got tla running enough to pull down GST source, but the "worldview" didn't make a lot of sense, and I didn't find any good documentation in my brief search. I intend to revisit the issue, as I need to pull down GST 2.2b...) > Better to make use of your time scanning SVN for possible improvements > ... like adding database labels ;-) There are some problems I'm willing to tackle. Others, I'm not. I'm not willing to put up with _compiling_ SVN, much less trying to improve it. It fails my first quality-assessment test: they didn't KISS. > Hey, if you get serious about it, I might throw in with you[0] ... > provided it's not in <gag> C++. I'm using C++ "in anger" now, and I am amazed that *anyone* would seriously contemplate using it for non-toy problems. I'm tempted to spin up GNAT and to look for a PLI compiler, just to see if the other "does everything" languages are as bad. > [0] I read a military historian who contended that, during WWII, the > side that got Italy as an ally was put at a disadvantage over the side > that didn't. Keep that in mind as you consider recruiting me. Heh. First, we need to find a problem to work on that doesn't eliminate either one of us. -- _ |\_ \| -- [email protected] http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
