Re: [cbm4linux-users] Problems with cbm4linux

[Spiro T. <tri...@gm...>]

Hello Dirk,

* On Thu, Jul 14, 2005 at 04:33:36PM +0200 Dirk Jagdmann wrote:

> >One could come to the conclusion that these always changing APIs do
> >no good. It is the Linux' way to force people keeping up.
> 
> You'd better ask this on LKML and be prepared for tons of flames :-)

Yes, I know. I have had this discussion more than once. Nevertheless,
this is something I almost hate about Linux.

 
> You're right at this point. I've now reworked all my changes and came up 
> with 6 patches, which should do only one thing at a time, although some 
> of them patch in the same areas and are therefore slightly dependand on 
> each other.
> I have updated the patches on
> http://sourceforge.net/tracker/?group_id=45917&atid=444478

Ok, it is obvious that they have to depend upon each other.
Nevertheless, with these patches, it is easier to patch by hand if one
wants only one patch and it fails.

I had a look at your patches:

1. The compilation bug for 2.6 (1238239): Does this mean that
   parport_find_number(lp) replaces the loop from before? Doesn't the
   new loop work at all, or is this just a shortcut?
   BTW: Why did you rearrange the for() loop into a while() loop? (just
   out of curiosity, not that it matters)

2. The reset patch: Your description tells us that < 0 is a "smart
   reset", while "== 0" performs no reset at all. Anyway, I cannot find
   a difference between these two. Is this "smart reset" what Joe
   proposed, thus, something to be added in the future, or what is this
   for?
   Anyway, I see you insist on your "no reset" default. Wouldn't it be
   better to use a "safe" setting, thus, perform a reset, and let it be
   disabled by knowledgeable people only?

3. I like the kernel source directory patch (1189489), although I have
   not tested it yet (but what should go wrong with this?).

Please, don't understand me wrong: I do not have anything against people
providing patches, especially if it is not even my own code but
Michael's. ;) Au contraire, I'm very happy other people look into it. I
only want to tell you what my point of view is - and I am hoping to
convince you. It seems it worked, at least partially.


> >As Michael has no means to test the Windows code, this
> >"cross-testing" is mostly my task. As I do own some Linux machines, I
> >want to be able to test the Linux code, too.
> 
> I have a machine which can boot linux 2.6 and XP. And I'm willing to
> help out.

Well, but it is much easier doing development and being able to test it
yourself instead of asking others to do it, isn't it?

 
> I don't insist on dropping support for old kernels if people want to
> keep it. I can live with the code present in the source, as it's
> #ifdef'ed anyway.

Ok, thank you.

Regards,
   Spiro.

-- 
Spiro R. Trikaliotis
http://www.trikaliotis.net/
http://cbm4win.sf.net/