Web lists-archives.com

Re: GitForWindows vs. Cygwin

It doesn't use cygwin, therefore it's not related. Since this is a
cygwin mailing list, that means the first two replies were correct —
this isn't the place to discuss it.

When the vast majority of the code in the MSYS2 runtime is exactly
the same as the code in the Cygwin runtime, and Git for Windows uses
msys-2.0.dll in a number of places (which is essentially a renamed
copy of cygwin1.dll) the question was fair. Particularly given Cygwin
has its own git package, it's only natural to wonder how that differs
from Git for Windows, and how it differs from the probably lesser-
known msys2-git package (which is built in very nearly the same
configuration as the Cygwin git package, posix style).

Only GfW and Msys2 Git projects can answer how Cygwin components are used by
their projects. Most (all?) of Cygwin winsup and (all of?) newlib are available
under permissive BSD(-style) licences that allow all uses of the code.

It's possible that the developers of those GfW and Msys2 Git packages have moved
on from those projects and the current maintainers don't have the basis to answer.

One might expect some of those maintainers to be subscribed to this list and
able to answer, but they may not, and just check web archives, or they may
choose not to get into this discussion here, although they could perhaps suggest
a more appropriate forum, or take the discussion off-list.

I've been watching this thread closely, but I had decided not to post since it
seemed like this discussion is unwanted by a part of the community.

On the surface, this discussion is about Git, so the most on-topic place to discuss
is the Git public mailing list.  I personally don't think the Cygwin mailing list
is off-topic.  Because most(all?) Windows Git distros use Cygwin code, it might be
even more useful to ask here (though I assume Johannes subscribes to Git ML and
he would be able to answer all the intricacies pretty accurately).

So far, Tony Kelman's descriptions were pretty spot on (that's why I didn't respond
to his call for corrections -- I couldn't think of any).

David Macek

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature