bug-cvs
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: EMX on DOS


From: Paul Edwards
Subject: Re: EMX on DOS
Date: Tue, 04 Nov 2003 10:59:45 GMT

"Derek Robert Price" <address@hidden> wrote in message news:address@hidden
> You have, in general, been submitting too many changes at once, most
> quite unclearly labeled and sorted and jumbled with other assumptions
> and declarations we are far from agreeing on yet, unless this fix is in
> one of my unread messages.  Please send one per email, with clear,
> concise explanations and include ChangeLog messages and docs and tests
> when necessary.

I have already done that for the "has been added" bug.

> |As a separate issue, yes, I have BASH.  But when I run
> |configure under Bash, it hangs my computer, and I need to
> |cold boot.  I don't know why.  But I never said my computer
>
> Sounds like an issue you should take up with the Cygwin development
> team.

If for some reason I required BASH, I might spend 5 hours
trying to get BASH working.  I have no need for a working
BASH, I already have an arbitrary and unspecified posix
compliant environment, all I need is a posix compliant
version control system and I'll be set.

I've already submitted a minor fix that will create such an
environment.

> Or perhaps, since you are running on Windows, a simple uninstall
> - -> reboot -> reinstall -> reboot cycle will do the trick.  I'd recommend
> the latest Cygwin release if you are using a very old one.  Cygwin has
> come a long way in the last few years.

Or perhaps that will take many hours of work attempting
to recover from a strange set of quirks that that has
introduced.

In actual fact, I am tickled pink that when running
"configure" (not gcc, which I am happy to run, but
"configure"), that when it hung my system requiring a
cold boot, that it didn't take out my FAT allocation
table, which would have cost me all the unbacked up
work, and many days of recovery effort.

> |had the ability to run shell scripts.  What I said is that I have
> |a Posix programming environment, and that I can compile
> |Posix code.  I can and do compile Posix code.  Just as I
> |compile and run C89 code.  THAT works.
> |
> |Other people may have only a partial install of Cygwin
> |that doesn't have bash, or they may have a completely
>
> Then they should install the rest of the tools.

Portable code doesn't require tools.  It merely needs to
folow the standard.

> We are not here to recreate portability tools,

Nope, that's what I'm here for, to convert CVS to be
Posix compliant.

> especially when it is unclear that they will
> remain supported.

This is the beauty of the 1.11.x branch.  They don't need to
be supported.

> These problems have been solved!  Why solve them again?

The problem of having Posix compliant source code has indeed
been solved.  By me.  On the weekend.

> |different Posix system, or may have downloaded a
> |different gcc port that doesn't come with Bash.  Or maybe
> |like me they can't run it for some other reason.  I don't
> |know.
>
> You are still the only person who has ever raised this issue.  Another

For every one person who makes the effort to report a
problem, there's probably 1000 who were too lazy to,
or assume that being a free product, no-one will fix it
anyway.  You get what you pay for etc.

Remember the -j -j to the current version?  Even Larry
had been affected by that in the past.

> reason I'm not inclined to spend much time on it or want to clutter the
> CVS source code with directories of bitrotting source.

Clutter?  If you consider supporting arbitrary and unspecified
Posix systems to be "clutter", can you create a "clutter" directory,
and put the Posix port, plus the EMX and Windows-NT stuff
there.  It may be clutter to you, but the Windows-NT directory
is wonderful to me.

And the Posix one would have been wonderful to me when I
was trying to get the stupid configure script to work on my
unspecified and arbitrary posix environment (Sun Solaris).
HELLO CVS!  I have a posix compiler, just compile the
damn thing!

> |But if they can provide a Posix environment, then Posix
> |compliant code should compile.  No ifs, buts or maybes,
> |if it's Posix compliant, it should work.
>
> CVS is not POSIX compliant, as I've stated before.

Until last weekend, true.  It doesn't need to be that way, it
never needed to be that way.

BFN.  Paul.




reply via email to

[Prev in Thread] Current Thread [Next in Thread]