On Tue, Nov 3, 2009 at 10:02, Matt Ingenthron <[email protected]> wrote:

> Also, do you have concerns with the MinGW approach?  Since some time/effort
> has been put into that already, if there are concerns/issues I'm sure folks
> would like to hear about them.
>

The biggest problem with MinGW is that for the average Windows developer, it
is a lot of work to understand how to work with a MinGW environment since
everything is different. If you're used to developing on any Unix system you
won't run into common gotcha's, but your average windows developer will
probably just fail or give up instead. This cuts down on possible
contributors, and makes it much less likely that people will download the
latest source code and build it themselves.

All the existing windows ports have had project files for Visual Studio and
been easy to build in that, which is as close to a standard way you get, but
those ports also introduced a lot of changes to the memcached source. Some
necessary that deal with making it a service and using the windows
networking stack, and some unnecessary that change all C99 code back to C89
standard instead, and it's the latter ones that makes it impossible to merge
those back in.

I would argue that it's most important to get a version that can
consistently build on Windows and doesn't break on each little commit you
guys make so that you can grab the latest version and build it, even if
setting up that build environment is a lot of work.


/Henrik

Reply via email to