Internal ZIP bug, "more" command, dynamic memory

[Volker Augustin]

Am Tue, 31 Oct 2000 schrieb OKUJI Yoshinori:
> About asm.diff: Your code doesn't work correctly, because prot_to_real
> and real_to_prot destroy the value of %ebp (see the comments!). So
> your code just trys to read data from a bogus drive.
> 
> About cat.diff: I don't like such an ad hoc way. Why don't you add a
> new command, say, "more" instead of hacking "cat"?
> 
> Okuji

Internal ZIP bug:
  Yes, I figured that deficiency in the code in asm.diff myself.
  Do you have any other ideas how to get an internal zip drive to work?

"more" command:
  You are right, hacking cat was not the right way. I will change that.

  I will also have to reimplement the grub_strsplit function, even if I don't
  need it in "more" anymore. The current implementation of grub_strsplit
  is buggy, since a second call to grub_strsplit kills the results of the
  first. However, to implement it correctly I NEED dynamic memory. So,
  the next question is:
  Should I include my dynamic memory library (free/malloc) or should I
  dump the implementation of grub_strsplit?
  For grub_strsplit, it would suffice, to allocate a fixed memory block in
  lower memory and handle that dynamically (tested, works).

Volker