gnu-arch-users
[Top][All Lists]
Advanced

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

Re: [Gnu-arch-users] planning 2.0?


From: John Arbash Meinel
Subject: Re: [Gnu-arch-users] planning 2.0?
Date: Fri, 21 Apr 2006 08:44:35 -0500
User-agent: Mail/News 1.5 (X11/20060309)

Karel Gardas wrote:
> On Fri, 21 Apr 2006, Ludovic Court�s wrote:
> 
>> Peter Conrad <address@hidden> writes:
>>
>>> You have to balance cost against usefulness. If 8.3 filenames
>>> were the limit on today's windows, the usefulness of sticking
>>> to the limit would probably outweigh the cost.
>>
>> I'd like to point out that this is clearly a political argument --- I
>> can't believe that you'd honestly defend such a technical flaw.
>>
>> Basically, you seem to imply that Windows support is important nowadays
>> because Microsoft has (almost) a monopoly; you then deduce that free
>> software projects should thus abide by the technical rules set by
>> Microsoft, no matter how bit-rotten they are, no matter what
>> alternatives exist, etc.
>>
>> I personally do not support such arguments.
> 
> OK, and what about if some free software developers do need to support
> Windows to fulfill their user needs and they decide to use your RCS for
> development?
> 
> If you are taking political view into account, then please also think
> about more broad picture than just your narrow DRCS view.
> 
> Cheers,
> Karel
> -- 
> Karel Gardas                  address@hidden
> ObjectSecurity Ltd.           http://www.objectsecurity.com
> 

I think the biggest reason to support windows is simply because without
it, you cut out a huge portion of your market share. And by that, I mean
even the people who only develop on POSIX platforms, but don't want
*their* stuff to be accessible only on POSIX.

Sure you cut out windows developers (which you don't really care because
they don't fit your philosophy anyway), but then you also cut out people
who try to write cross platform stuff.

I realize windows support is a pain. The path length is actually a POSIX
non-compliance on the part of tla, because Windows tells you their
MAX_PATH is 256 characters, and tla just ignores that fact.
The bigger support issue is path locking. (You can't delete an open
file, or rename over an existing one, etc)

The MAX_PATH bug is an issue because tla won't even work on cygwin,
which implements much more of POSIX for windows. But if tla just didn't
repeat itself so much, it would work just fine. (See my posts in the
archive, but basically the longest path is a temporary file while
building a revision library, and amounts to repeating the same
information about 3 times, (9 if you count repeats of the category name)).

John
=:->

Attachment: signature.asc
Description: OpenPGP digital signature


reply via email to

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