That's not an option for everyone. It's nice that someone came up with a
'standard' software interface for Linux.. I think it's a good idea.
> : o Drivers being able to throw away their init. code once complete.
> Yes, to save several 100s of whole bytes with average 10 drivers; could
> almost be 2-5 kbytes per system. Could be really important one or two
> days before someone wants to get 0x86 running without any main memory.
Maybe it's a hold-over from DOS, but I merely stated it for the benefit of
Linux and general streamlining..
> : o Very trivial, but a global min/max function [I notice in a lot of spots,
> : are implemented per .c file with a note usually saying "Why isn't there a
> : global function for this?"]
> Don't forget the general argument parser for INLINE macros.
I'm merely mentioning something I have seen in a few places in the source and
it might save only a few bytes.. but I saw it and thought it might be useful,
so I mentioned it.. I'm not out attacking anyone or anything, as I kind of take
this message as.
> : o Ability to nuke source code that is not specific to your configuration.
> : Not necessarily a default option, but an option, none-the-less.
> Yes, a one-byte kernel really would fit into the master boot record.
I'm not talking about the actual kernel binary, I'm talking about the source.
To compile many apps, you still need to keep [at least a portion] of the Linux
source around for them to compile correctly. Allowing people to hack off 'the
unnecessary' parts saves disk space. Not everyone can go hardware crazy and
get 2GB hard-drives. I'm not saying that the source is huge, I'm just merely
stating, what I thought was a nice 'idea'.. it doesn't have to be implemented.
I meant my original message to be helpful. If I'm not wanted, I'll leave I
guess.
> : o Uploading various patches to ftp.lmh.ox.ac.uk for a centralized deposit.
> You mean directly at everybody's door? Everything?
Someone offered their services specifically for what I was mentioning and
publicizing that address can't hurt, can it?
> : o Only re-make modules if necessary
> Supposed to work already.
"Supposed" doesn't always mean "does".
At least as of the last few 1.99's, it didn't. If 2.0 already does, good.
Ignore that section.
> : o Fix the boot-up display so that "kswap started" doesn't appear in the
> : partition check.
> Lately I heard an interesting definition of "multi user ability" from
> someone who got drunken by DOS/Windows: "totally different users may
> utilize the same machine, right one after the other".
This is merely cosmetics.. take a "Joe Blow" user evaluating Linux.. "Gee, Why
does it say that in the middle of the partition check? That looks
unprofessional.. "
> : o Breaking sbpcd.c apart cleanly, per family.
> Crazy. This can make me really unwilling, boy.
> If you really want to do a "job" with sbpcd, take it apart and put it
> together for supporting those damned Funai E24xx drives (in an "isolated"
> version). They are mis-using the interface lines in a way which does not
> allow "multiple families" on one cable.
> You can get the full information you need. Even if you don't manage it
> there is the hope that it would take you away from your stupid task.
Maybe you call me boy for my enthusiam, but I'm certainly not. I'm doing this
in my free time since it is something I feel I can tackle and that other people
might find useful. I didn't release my original 'hack' of sbpcd until about
20 people asked me for copies.. I did so specifically stating not to bother you
with questions about what I did since you'd probably just get angry. It
obviously has. This is the first negative response I've gotten from my original
post. I find it rather rude.
I'm not trying to put you down or anything, but I think that sbpcd.c tries to
be too 'all-encompassing'..
> sbpcd does not only support totally different drive families - it also is
> supporting them alltogether on one cable. Currently, you can mix Matsushita
> CR-521 plus Longshine LCS-7260 plus Panasonic CR-562 plus CreativeLabs CR-563
> plus CreativeLabs CD200F plus Funai E2550 or MK4015 or 2800F plus Teac CD-55A
> plus ECS-AT Vertos 100 up to a total of four one one cable, and you can
> have up to four cables (more possible).
Realistically, how many users out there have that many different types of
CD-ROM's in their machines? How many even have 2? Very little.
How many people enable enough of the debugging information to require a pause
between each syslog entry? Very little.
How many systems are locked up by over-ambious address probes?
I'm not out to get you, I'm only here to help.
> I did not assemble it to let it tear apart by playing kids, really.
I'll be a college senior this fall. I'm turning 21 soon. That was uncalled for.
If you're trying to scare people away from helping the Linux development
process, you're doing a good job.