Re: [PATCH 0/8] CMake build system for git

[Junio C Hamano <gitster@pobox.com>]

Sibi Siddharthan <sibisiddharthan.github@gmail.com> writes:

> I have made the following changes as suggested:
> 1) The CMake script works from contrib/cmake now.
> 2) The CMake script parses the Makefile for
> SCRIPT_SH
> SCRIPT_PERL
> TEST_BUILTINS_OBJS
> LIB_OBJS
> BUILTIN_OBJS
> XDIFF_OBJS
> VCSSVN_OBJS
> 3) Philip suggested to change the error message if sh/bash was not
> found on windows.
> 4) CMake now tests for ICONV_OMITS_BOM, NO_ST_BLOCKS_IN_STRUCT_STAT.
>
> Regarding the commits, since 1,2,4 are additions to the script, is it
> acceptable for those changes to be staged in the next commits?

I am not sure what you exactly mean by "to be staged in the next
commits".  But as a new topic (from the point of view of the general
public), please avoid adding a known-broken one in patch 1 and then
following up with a series of "oops, that was wrong and here is to
fix the breakage" patches.

On the other hand, if the final contents added by the resulting
topic is easier to explain and review if built incrementally in
logical steps, please do so.  In other words, a series of follow up
"now we have basics there, teach the machinery to do this, too"
patches is perfectly fine, as opposed to "oops, that was wrong and
here is to fix".

> Regarding the workflow file(main.yml), I modified the vs-build and
> test steps, should I drop the commit or should I keep the changes(a
> modification is further needed if CMake is going to be used for
> vs-build)

It sounds like an integral part of the end result we would want to
get out of this series, so if that is the case, I do not see a
reason to exclude it.

Thanks.