I feel the need to shed some light on this subject from the viewpoint of
one MUCH less capable than the others here.
Recently I purchased a PPC 7200. It is a beat up old mac that I could pick
up cheap. I have used intels since I was a boy (SysV3.3 on 386!) and saw
some characteristics of the mac that were lacking in the intel boxes.
Finally, I tried to use the box to do some video grabbing only to find
that the 2.2, 2.3, and 2.4 kernel all failed. I Attempted to read the
code, and use kdb, only it doesn't appear to work on the mac very
well. And I don't have a mac serial to 9/25-pin adapter. So I have to go
to Orlando to get one. This is a two hour drive for me, and not worth my
time, thus gdb is out of the question. I have sent some questions out to
that community and they have been unresponsive on this one (I guess no one
uses the happauge wintv on those boxes).
The result is a powerpc that will be used for little more than a file
server. (I am greatful that it can be used for this much.)
Some more options for this little machine would have been useful at the
time. And while I realize that the push for intel far outweighs the push
for Mac or Sparc, there is a following that could use some (more) kernel
level help. This really is an issue that spans far beyond just which, if
any, kernel debugger should be available in the kernel sources. This gets
back to the organization on the whole.
While I might be calling on a giant with a roaches voice, where does Linus
stand on this?
Mechanics fought computers and the analysis tools that went with them for
years. Now, a good mechanic can turn around more cars than ever possible
with accurate use of the excellent tools available to him.
Please don't nix this tool before it gets a chance to prove its worth.
I sometimes feel that the linux community, possibly from the top, has
forgotten that this community was founded on making 'it' work, and work
well. Part of this is incorperating good ideas where they are seen. And
discarding good (and of course bad) ideas where they are not valuable.
Some of the other OS's (notably the BSD's) have some good ideas. Possibly
the manos debugger is one of them. We should not discard those ideas
because 'It has been done this way since 1993!' Please don't forget
this! This type of mindset could be very costly in the long-run.
Thanks for your time, and I hope this has been useful.
-- Matt Berglund One guy for a unified OPERATING SYSTEM (not just a kernel) on all platforms, from Palmtop to Mainframe. And linux has the best shot.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to email@example.com Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu Sep 07 2000 - 21:00:25 EST