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

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

Re: [Gnu-arch-users] arch 2.0 survey followup


From: Mark Flacy
Subject: Re: [Gnu-arch-users] arch 2.0 survey followup
Date: Mon, 01 May 2006 14:37:07 -0500

On 2006.05.01 02:27, Andy Tai wrote:

I don't fully agree. Steve Job's using gcc for NextStep to Mac OS X did not/does not matter to the GNU Project's goal of a free OS, but the gcc contribution by Apple (and many chip makers whose product may be more used in car engines than in user hackable personal computers) has been valuable to the development of gcc. Good Windows support in GNU Arch may attract industrial support which may not matter to most GNU software but will help the GNU Project (including GNU Arch) nonetheless.

You are comparing apples to aardvarks. GCC doesn't require the same level of file system integration as Arch 2.0 would.

Windows file systems have three basic differences from the unix ones:

1) They are case insensitive but case preserving. This has a major impact in determining if two files really have the same name or not. It also has major impacts on file name manipulation.

2) In addition to the two file names implied by the previous bullet, there is a third file name: the 8.3 name. More fun in determining if two files are the same or not.

3) They do not allow the deletion of a file that is opened by some other process. (FWIW, I have been told that behavior can be overridden but is the default.)


In addition, a call to Python's os.popen3("exec_path_with_a_space_in_it") method will fail on a Windows platform but not on a Unix/Linux one. Perhaps this is a Python specific bug.


I happen to maintain a Python tool which runs on Windows and *nix systems to automatically order and compile over 1500 Java packages. The count of "God, I hate Windows" comments in the code are non-zero.


There are possible *economic* reasons to support the Windows platform, assuming there is any money to be made by Arch 2.0.

As far as *technical* reasons go, I think that the coding effort to design around Windows file systems assumptions could be better spent doing almost anything else. It's not my project and it's not my design, so I think that I've said all that I need to say on the subject.




reply via email to

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