Re: why not TortoiseGit

From: Nigel Magnay <nigel.magnay@gmail.com>
Date: Nov 3, 2008, 13:39 UTC
Message-ID: <320075ff0811030539q476ab4f4kd332a37acdc1ca87@mail.gmail.com>
In-Reply-To: <alpine.DEB.1.00.0811031425080.22125@pacific.mpi-cbg.de.mpi-cbg.de>

Show 10 quoted lines

>> But it doesn't look like a hard project to me, just requires stripping
>> out a lot of junk and re-patching callouts to a git executable (which
>> could be the standard git tools) and a minimal git library that knows if
>> files are dirty.
>
> I only wish that people would put their code where there mouth is.
>
> At least with GitCheetah, we have working code, _and_ an opportunity to go
> cross-platform.
>

Well, hey, I don't care there's no TortoiseGit. I looked at these things back when I had colleagues stuck on Windows, and at the time wanted to try and wean them off SVN.

The shell-icon overlay limit on Windows looked a significant problem to me, and a good reason for at least re-using that bit of code (which is common to even tortoiseCVS). It looked like it had been through a significant number of iterations to get platform shell subtleties right.

I even looked at wacky things, like using IKVM.Net and JGit to hack it quickly, but that's a non-starter because of MS' stupid one-clr-per-process.

That's what I found. Maybe it'll be useful for anyone else that wants to continue. Since it's not an itch for me any more, and it won't feed my children, until someone that cares enough does something there won't be one.