|
From: Szakacsits S. <sz...@si...> - 2004-02-26 16:15:23
|
On Thu, 26 Feb 2004, Anton Altaparmakov wrote:
>
> > Secondly and most importantly, you _need_ Windows to do this. And that's
> > silly :-)
>
> True. But then maybe we need a utilitiy that can upgrade ntfs 1.2 to
> 3.x rather than having mkntfs create 3.x.
I don't think so. For me it seems to be
- more difficult
- more error-prone
- "everybody" uses or will use XP/W2K/Longhorn (version 3.1), why
the pain for the conversion, e.g. when in [qt]parted, diskdrake,
etc one wants to create an NTFS.
- it would mean more work for development, e.g. for tests we should
always do the two steps 'mkntfs + ntfsconvert' instead of just
'mkntfs' and this is the most important at present since we
don't have much free time. Of course we could wrap them by a
script, just sometimes forget about it, etc, etc.
In short it would be more useless and resource demanding for us, who has
very limited free time anyway.
> After all you might already have data on the partition and don't want
> to backup, mkntfs, restore...
Very, very, low priority, at least for me. To be honest I'm not interested
at all.
However what I'm interested in is to have the basic tools to foster the
development. ntfsresize soon ready, then write is coming, ntfstouch,
finish/revise ntfstruncate, etc.
We need the _tools_ to make regression testing and debugging as easy and
fast as possible and at present we have neither mkntfs 3.x nor ntfsdebug,
the two most important tools for any serious filesystem development (and
ntfs is being one of the most complex in the world).
> I have always found these [IN|OUT] things confusing and obfuscating.
> It's quite clear from the code what they are...
Actually, it's not always clear from your code, at least for me ;)
If you were writing short, easy to read and revise functions then it would
be close to redundant but often you write code being very-very long and
hard to see _immediately_ its purpose and correctness.
> Ok, ok. I give in, put them in wherever you like but at least preserve
> some sane formatting like:
>
> @paramname:<tab>[IN]<space>description
>
> Is that a deal? (-:
That's ok with me also :-)
Szaka
|
