A few points and questions
markh at linuxfromscratch.org
Sat Sep 29 03:39:14 PDT 2001
On Sat, 29, Sep, 2001 at 11:31:19AM +0100, Paul Campbell wrote:
> I have been waiting patiently for progress, checking in to the CVS book here
> and there, to see how things are going.
Slowly unfortunately but they are moving believe it or not.
> I have noticed a lot of the bugzilla stuff being worked on, that appears to
> be superflious to the system, like hdparam and the like. Things that will
> not be on the mind of someone just finishing the LFS book.
Depends - I have to use hdparm on one of my hard drives to disable DMA
otherwise the damn machine crashes within about 10 minutes. Also,
hdparm isn't difficult to do - it's just that I haven't got around to
> Possible the goal of the book is set too high. I think you could axe a load
> of stuff for the momment and think carefully about what are the essential
> things that peeps will want in their LFS system as soon as the base is
> finished. Rather than make the book about *everything* that you may want
> after LFS, I feel the BLFS book can follow the style of the LFS one, if it
> isn't essential, it don't get in.
Hmm - that's difficult. Who decides what is essential? For a server,
sound isn't essential - for a desktop it is. BLFS isn't supposed to be
linear - people are meant to use it to guide themselves.
> Make the BLFS book about an extended base
> system with X etc. Rather than a bulky disrto from the hints. The later
> making the book much harder to support and maintain or even finish.
I'm looking for a release towards the end of the year. I know it seems
a long way away but we ARE writing this book pretty much from scratch (pun
intended) and it can't be done overnight.
> Choose what apps make an extended base system, and set a dealine for
> completion of version 1.0, don't try and convert every single hint though,
> they will still be there to use.
We aren't intending to just convert existing hints.
> The other thing is, can you avoid the non-defacto-standard apps. I hate the
> hints that say ... ".. everybody uses crond, but we going to tell you how to
> use fcron, or bcron or someother wacy item....) With fcron I forgive them,
> but NOT djbdns, that apps really winds me up, that daemon tools foo breaks
> with too many traditions and ways of getting things done.. You have to
> create an entirely new way of running daemons, and the compile/install
> proceedure was written by someone who clearly has no regard for linux system
> standards. Some of the hints also hold too much personal preference rather
> than standard setup advice.
Djbdns is included as an *alternative* to bind. Bind will be included
just as soon as someone gets around to writing the instructions for it.
Are you volunteering? ;-)
Some people like "non-standard" apps. Personally I use bind but that's
my preference. I happen to know that Jeff who wrote those instructions,
prefers djbdns - this way, people can choose.
We are trying to just stick to standard setup advice. We are trying to
be FHS compliant in all of the instructions (which is one of the many things
which the QA team helps check). This is a reason why just putting
./configure && make && make install for every package isn't what we do -
the book would be rather pointless if we did and you'd end up with files
all over the damn place ;-)
> I would like to start helping out, just my Open University term is reaching a
> end and I have an science Exam in two weeks, so it will be after that. Has
> anyone done samba & nfs? Or a colour bash prompt?
Best thing to do is to:
1) subscribe to blfs-book and blfs-dev
2) get a bugzilla account at blfs-bugs.linuxfromscratch.org
3) Install CVS on your system and pull and read the xml to get a feel
One of the things we DO need is people to TEST instructions. Jeff
Bauman is in charge of the QA team. He listens in on this list (and
hopefully blfs-dev) and so if you want to help with that then speak up.
> I hope you dont think I'm being harse, I just feel that there may be too much
> being taken on for a version one effort. The book needs an set deadline to
> encourage the effort too.
> If version one only contains X, KDE Gnome and one or two apps, it will still
> be of good use and valuable to new comers. Version two can include more and
> so on.....
I'm not incredibally keen on specific artificial deadlines. Look at LFS-3.0 -
we released it WHEN it was ready. I agree that we need to release 1.0
ASAP thought but I *don't* want it to be not worth releasing. I'm not
sure about releasing 1.0 with just X, KDE and Gnome. I'd rather wait
another 2 months and get a really good book out there. Other people may
(and probably will) disagree though.
Phew - another long email...
Hope this clears stuff up though
Mark Hymers BLFS Project Leader
markh at linuxfromscratch.org
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe blfs-book' in the subject header of the message
More information about the blfs-book