I'm crosposting this to all tortoises I know of - the Windows Overlay problem is relevant to all.
(*git-cheetah *is a tortoise in disguise) On 19/04/2008, TK Soh <[EMAIL PROTECTED]> wrote: > > On Fri, Apr 18, 2008 at 12:34 PM, Tom Widmer <[EMAIL PROTECTED]> > wrote: > > > > [copy of > http://bazaar-vcs.org/bzr/bzr.dev/doc/developers/tortoise-strategy.txt] > > > > TK Soh wrote: > > > > I'd say the document details the best strategy for TortoiseHG too, > > simply by replacing BZR with HG and Bazaar with Mercurial. > > > > Except I propose this change: > > > > Share the exact same (eventually C++) shell overlay extension for Bzr > > and HG, so that if you have TortoiseHG and TortoiseBZR installed, they > > actually are sharing a single COM dll, which will presumably pick up > > from the Windows registry the names of the RPC exe programs it needs to > > communicate with (one Bazaar one and one Mercurial one). There's a > minor > > versioning issue here, but this can be solved by versioning the > > capabilities of the RPC server so newer versions of the shell overlay > > extension don't ask unanswerable questions of older RPC servers. > > > > If that's no go, at the very least the C++ code can be shared in its > > entirety, and just built differently for HG and BZR. > > > Stefan Küng of TSVN has proposed a solution to get around the slot > limit of Overlay handlers among the Tortoise clients: > > > http://tortoisesvn.tigris.org/svn/tortoisesvn/TortoiseOverlays/version-1.0.1/Documentation.txt username guest password <empty> Eventually TortoiseHg, and probably all other Tortoise clients too, > will adopt this approach. > > I think that the TortoiseOverlay component could evolve into a separate project, were some of the features mentioned in the bzr tortoise strategy (space-efficient DLL architecture, separate tortoise-processes) would fit nicely, and benefit all tortoises. I would like to contribute to such a project, but I would rather watch a dedicated mailing list. The amount of mailing lists posts I'm trying to follow has reached my processing capabilities, so adding two more (TSVN and TBZR) with a fairly high amount of posts that are not relevant to me is not too attractive. Anybody feel like setting up such a project? Regards, Peer
