RE: [MiNT] Was: /proc, will be: /sys

["Julian Reschke" <reschke@muenster.de>]

> From: Konrad M. Kokoszkiewicz [mailto:draco@obta.uw.edu.pl]
> Sent: Wednesday, November 17, 1999 4:41 PM
> To: Julian Reschke
> Cc: Guido Flohr; MiNT mailing list
> Subject: RE: [MiNT] Was: /proc, will be: /sys
>
>
>
> > OK, so why not just define a new file attribute combination for the SLB
> > instance, and this is it?
>
> The only problem is that there are some rules implemented (like +a are
> stopped processes etc) and I am afraid of inventing something that breaks
> these. Otherwise, what about $25 for a lib? Gryf?

In fact, I think in addition to the attribute there is also a process state
(wait queue value?) to be defined as well, correct? Regarding the file
attributes -- I doubt that they are used for anything else then -- display,
right?

> I think the implementations are close enough, with the notable except of
> the path search. As I think now about it... the library startup code works

Which could prevent a lot of code from running. It certainly prevents my
examples from running (unless the SLB happens to be in the current
directory, I guess).

> as a separate process (for a while, i.e. before it stops itself), so it is
> both inside and outside the kernel. Anyway, it is user mode... so perhaps
> the library startup stub could do the search (hum, must look at the src).
>
> Otherwise I think the SLBPATH variable value should be just pasted into
> libpath, if the libpath specified by the opening program is NULL.
>
> Anyway, I still miss real, program transparent library scheme.

Could you be a bit more specific?