[ruby-talk:6060] Re: Cygwin conflicts

Hi,

Lewis Perin wrote:

> "Conrad Schneiker/Austin/Contr/IBM" <[e-mail]> writes:
>
> > [...CygWin Ruby installation woes...]
>
> Maybe it's worth while taking a step back from the details of getting
> Ruby (installation) to work in the Cygnus Win32 environment and
> consider CygWin from the standpoint of, well, Ruby world domination.
>
> Would Perl have the Win32 penetration it has without the ActiveState
> native port?
>
> Would Emacs have the Win32 penetration it has without the NTEmacs
> port?
>
> If the answers to the above are no and no, then should Ruby's future
> on Win32 OSes (and possibly its future, period) be yoked to Cygnus?

Well, I think that even if you answer both the above questions "probably
almost as much", you still raise an important question.

I think the main problem is how much additional/continuing work would it
take to develop/maintain a native Win32 port, and who will volunteer to do
it? IIRC, Microsoft for a practical purposes funded the original Win32 port
of Perl, and AFAIK, Microsoft still continues to subsidize some significant
fraction of ActiveState's Win32 support/development work on Perl (more of
which now is also applicable to the Unix ports, e.g. Unicode support).

There is a related issue that .Net looks like it will soon raise the bar for
what could be considered as full-fledged native Win32 port. I don't know how
Cygnus is going to play in this arena, if at all.

Now from the standpoint of Ruby world domination, I think the best
short-term strategy for Win32 with respect to presently available resources
is to stick with CygWin. (I think the % of people that run into the problems
I've seen will be a very small fraction of all users, and unless they are
one of the even more rare cases of people doing development work frozen to a
back level of CygWin, they can certainly deal with the obnoxious and
inconvenient work-around of upgrading CygWin.)

And I think the best medium-term strategy for Win32 would be to aim for
ActiveState's support (and/or someone looking to compete with ActiveState) 1
or 2 years from now, when there will be 3 or 4 English Ruby books and when
Ruby has a much larger user base on both Win32 and Unix-related platforms,
and when Ruby has much greater visibility. (Alternatively, someone could get
Dave and Andy a big ASP or whatever contract that included deploying native
.Net versions of Ruby.)

Conrad