[openbeosnetteam] Re: moving ppp to other location

  • From: "Axel Dörfler" <axeld@xxxxxxxxxxxxxxxx>
  • To: openbeosnetteam@xxxxxxxxxxxxx
  • Date: Thu, 18 Sep 2003 19:13:33 +0200 CEST

"Waldemar Kornewald" <Waldemar.Kornewald@xxxxxx> wrote:
> > > Maybe I will have to sit down on my own...but I already planned 
> > > some
> > > experiments with file/data management after having finished PPP.
> > > Why is it so difficult to add shared library support? Cannot we 
> > > add a
> > > directory /beos/system/kernellibs where all kernel-only libraries 
> > > are
> > > located? Why cannot the userland code be reused for the kernel
> > > modules?
> > Just deal with it - you knew it before, or at least we tried to 
> > tell
> > you. It might come in R2, but certainly not in R1.
> Could you please tell me what would need to be done?

Yes, to think about it deeply before doing anything :-)
Extending the ELF loader to link not only against the kernel should be 
one of the smaller issues, here.

> > Huh? A .a will *never* end up in /system/libs - if it's a public 
> > .a, it
> > would belong to develop/libs/<arch>/
> Yes, in a later mail I wrote that this would be the perfect place for 
> R2/R3
> and you replied to it. ;)

Don't get what you mean. But in any way, it's not public yet, and for 
R2/R3 the .a issue is eventually resolved otherwise.
And I really think we should stop wasting time by discussing that now.

> > > size to around 120-130K, I think (and hope). But this will be 
> > > done at
> > > some
> > How does your list implementation look like when it is that big?
> > Anyway, the large file sizes are something you accepted when going 
> > the
> > "I want a shared library" approach.
> It is Ingo's List template. I do not think the Vector template is 
> smaller.

Probably not much. Are you sure that's the thing which makes it that 
big?
Anyway, about what total size are we talking here?
And which modules would have to be linked against the library?

If it's too much, we could just put all modules under ppp/ together in 
one binary for R1 - not nice but would solve the issue.

Bye,
   Axel.


Other related posts: