Linux-Misc Digest #590, Volume #24 Wed, 24 May 00 21:13:02 EDT
Contents:
Re: how to enter a bug report against linux? ([EMAIL PROTECTED])
Re: how to enter a bug report against linux? (John Hasler)
Re: Gnome graphical ppp dialer ? (Brian Moore)
Re: ssl certificate server (Paul Rubin)
Re: Need ideas for university funded project for linux (JEDIDIAH)
Re: resume download program ("Lonni J. Friedman")
Printing man pages ("Brian")
Re: removing LILO from mbr (Juergen Heinzl)
Re: Xterm with transparent background ? (Max Heijndijk)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.advocacy
Subject: Re: how to enter a bug report against linux?
Reply-To: [EMAIL PROTECTED]
Date: Wed, 24 May 2000 23:50:26 GMT
ggq15$[EMAIL PROTECTED] writes:
>bug tacking and testing crap system aint never gona be pushed down
>our throats, get over it, I agree with Peter, this is how it always
>was and always will. This is the Linux way.
Maybe it won't always be that way. But if (or when) it changes, it won't
be because someone feels that doing the bughunting in a certain way is
the done thing.
If it changes, it will be because things start to stop working. When
the workload gets too much for Alan and Linus, or when maybe they are
just plain sick of it all, a different solution will be found. Last
time Linus was growing a bit tired of dealing with this whole Linux
thing, Alan Cox became a sort of buffer between him and the teeming
wilderness of the world. These days, getting a patch into a Linux kernel
usually gies through two steps --- first getting into the *AC kernels,
taking a bit of a beating there (and a lot of inspection), and then
passing from Alan to Linus, and thus into "the real thing".
Actual system development used to be discussed in alt.os.linux, then
comp.os.linux, then comp.os.linux.development.system (was there a time
where c.o.l.dev didn't have subgroups?), and these days, is almost
exclusively done on open mailing lists.
It's possible that within a few month, things will change again. The
linux kernel is getting bigger, the goal for the post-2.4 kernels are
quite ambitious and radical, and, weird as it feels saying this about
linux, the need for backward compatibility with earlier, sometimes flawed
implementations is getting more important. Maybe the current system is
unable to cope with these issues. It's possible.
But until the *need* for change arises, the old adage of "if it ain't broke,
don't fix it" is as true as ever. Linux has come very very far since I first
started using it in 1992. Undoubtedly a lot of its success had to do with
pure luck, with doing the right thing at the right time by sheer chance.
But luck is also what you make of it --- and Linux has avoided the splits
that have so harmed the *BSDs. For whatever reason, Linux seems to have
harboured a culture of non-escalation in conflicts between developers....
or at least of not washing dirty laundry in public. The worst I can think
of was the whole GGI thing, and even that seems to have blown over nicely.
Sometimes people realize that Linux isn't going the way they want it to,
or that they are not interested anymore --- especially the network code
went through several transitions of maintainers, and has had a couple of
pretty extensive (i.e. almost complete) rewrites --- but that sort of thing
tends to happen peacefully.
Then again, at the same time, Linux has managed to develop leaps and bounds.
A certain amount of "what do I care what I said last year" attitude helps.
Linus once stated that Linux would always be an x86 only OS. Oh well,
he was wrong. I remember the first, horrible SLIP hacks, and I seem to
recall them being preceded by comments that such couldn't be done. Well,
it was. And it was done again, a lot cleaner, but still a bit ugly. And
again.
Linus once had the brilliant idea of having two devices for each serial
port, with different locking behaviour. Solved a lot of problems people
had. Turns out that down the road, it also introduced other problems,
and that the problems solved in the first place could be solved differently.
So it was changed again.
What also helps is someone at the helm who doesn't take himself too seriously,
and at the same time *does* take important things seriously. Having the
guts to call a x.even.1 release "The brown paper bag release" carries
all the right messages --- this is something that shouldn't have happened,
this is something to be a bit ashamed of, but hey, shit happens, so let's
go and fix it, right? And with a big grin, too. [ For everybody's reading
pleasure, at the end of this post I'll add an email Linus sent to the
AXP mailing list 3.5 years ago :-]
And much as it may rub some people the wrong way, at its core linux is
by geeks, for geeks. The whole motivation behind work on linux never had
much to do with money, and to this day, it seems that is still true.
Also, amazingly, it never seemed to have much to do with prestige.
The motivation, for most developers, for most developers always has been
to create a better tool to fix real problems. Well, either that, or to
do something that scores high on the cool-o-meter. While this is one of
Linux' weaknesses (the "real problems" of geeks are not often the same
as the real problems of secretaries, and what geeks find cool often
completely eludes "normal" people), it is also a strength. It means that
things get done *right*, rather than just done. And it also means that
the whole thing has all the aspects of a crazy, geeked-up late night
collaboration. If you write code, and I can find a way to make it break,
then I did you a favour --- better find out now than later. You might have
to completely redesign, but if you are a geek, you'll understand that
that isn't my fault --- thus no grudges held. You might still work especially
hard to find a way to break my next piece of code, but hey, that's why
you get to see it. And then we'll all head down to the pizza joint and
plan some new cool project on the placemats. Well, that last part is hard
when we are on different continents, but that's the sort of atmosphere.
When talking about the development model for Linux, you have to always keep
in mind that that's the culture it grew from, and the people who actually
do the development are living in this culture. These are the same kind of
people who consider trying to write as-unreadable-as-possible C programs
for the IOCC a fun way to unwind at 4am after spending 17 hours straight
hunting an obscure bug.
Now, one thing that kind of people resents is being told what to do ---
especially when they don't think it is the right thing to do. Again: For
a geek, doing things *right* is a huge part of the motivation for doing
things at all. Take away the belief that he (or she) is doing it *right*,
and that motivation disappears. Geeks need to work on things they are
interested in, fascinated by --- that's when they will work best (and
hardest, and longest). Setting them to work on things they don't care about
breeds resentment, disassociation from the work, and thus poor results.
But give them a sexy problem.....
Committees, bug tracking systems and so on are viewed in much the same
way as 9-to-5 workdays, ties and suits or the whole concept of "billable
hours" --- as some management nonsense that gets in the way of doing the
coding, doing it *now*, doing it *right*, and having fun while doing it.
They might be the solution to a set of problems --- but those are
certainly not the problems the geeks have any interest in solving,
and as linux is by geeks, for geeks, convincing the people who drive
linux to get excited about solving those problems is going to be very
very hard.
Of course, what all this condenses to is that Linux' development process,
in its current form, is a barely tamed chaos, with no guarantees whatsoever.
Most developers only work on linux because they want to, and if they can't
get excited about the issue that you care about, well, tough, nobody is
going to work on it. If Donald Becker decides today that he has had enough,
and that he wants to spend the next five years writing Visual Basic
viruses instead of Linux network drivers, well, that's his choice, and
nothing anoybody can do about it. Everybody does what they are interested
in, when they want to, however they think it's right. Take it or leave it.
While that might be a frightening idea to contemplate for someone who
likes org charts, written policies and well established rules (and believe
me, I can sympathize with that attitude --- I am German by birth ;-),
it seems to have worked well. The results are there, they are impressive,
and there are no signs of everything falling apart anytime soon. So it
might be a good idea to just sit back, keep quiet, and let the bunch
of crazy guys and gals for whom "that's impossible" is a challenge rather
than a reason to shy away do what they do best. If nothing else, it will
certainly be interesting to see what "impossible" thing they'll dream up
next, and then somehow make work --- an awful lot of the world of computing
as we know it today started out in that way, as someone deciding to attempt
something out of pure curiosity and for the feeling of the challenge.
Ok, that's it! Rant's over! Back to coding ;-)
Bernie
P.S.: Here is the promised email:
Date: Wed, 16 Oct 1996 19:10:43 +0300 (EET DST)
From: Linus Torvalds <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Subject: Re: Househunting?
On Wed, 16 Oct 1996, Jim Nance wrote:
>
[...]
> Are you moving to the US? You didn't go and get a job did you?
Ok, folks, DON'T PANIC.
[ inline: Linus running around in circles and tearing his hair out ]
Now, it had to be told some day. I have a life. There. I did it. I came out
of the closet. It's sad, it's shocking, but it's true. I even have a
girlfriend [ collective GASP ], and it all adds up to the fact that it had to
happen sooner or later. I'm getting a RealJob(tm).
[ inline: Linus frantically trying to explain that a university position is
_too_ a real job, while everybody else just stands around and laughs at
him. Is too! ]
I thought I'd be subtle and tell it to the mailing lists first. The idea was
that the mailing list people would tend to be saner than most people who read
the newsgroups and not panic too badly. Well, that was a theory of mine, but
maddog tells me the stampede already started. Bad theory.
Again, DON'T PANIC. It's not the end of the world. It's not the end of me
doing Linux. Unlike my current job at the university (and that _is_ a real
job, you just never noticed. is too!) the new job description will even
explicitly say that I'll be doing Linux part time.
So yes, I'll be in the Bay Area next week looking for a place to live with my
girlfriend and our two cats. We won't actually move there until February next
year or so.
Oh, did I already mention that you shouldn't panic? I did? Good. Don't. It
all just means that I'll be living in the US for a while, and making more
money than I am at the university and doing so in better weather.
Linus
PS. Company called Transmeta. Don't bother looking them up, they aren't
advertising. They are in the business of getting the best minds in the
industry and doing cool things that I won't tell you about. Not selling
Linux, so it's not unfair competition in that field.
--
When I want a peerage, I shall buy it like an honest man
Lord Northcliffe
British newspaper proprietor, 1865-1922
------------------------------
From: John Hasler <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.advocacy
Subject: Re: how to enter a bug report against linux?
Date: Wed, 24 May 2000 21:31:52 GMT
Peter writes:
> Lurking on the kernel list,...
Just to find out if bug that is of little importance to me is known?
> Doug's glad to respond to questions, as you know.
How would I know that? I've never communicated with the man.
> As you know, the kernel list delivers a thousand or so mails a day.
You said email the maintainer, not the list. And why do you write "As you
know"? I don't subscribe to the kernel list.
> Because you are a reasonable person, and Alan is a reasonable person, and
> it is reasonable that the combination of sound with anything would
> interest Alan.
While it might be reasonable for me to presume that Alan (Cox, I assume) is
reasonable, I have no reason to presume that he cares about sound: I
neither care about sound nor follow kernel development closely.
> Because he's the named and registered author and maintainer of the scsi
> driver in question.
Why do you assume that I would recognize the first name of a maintainer
associated with a bug that I have already said that I decided not to
report? I might not recognize his first name even if I had reported the
damn thing to him.
> Is this posting a forgery?
Are you typical of the denizens of the kernel list?
--
John Hasler
[EMAIL PROTECTED]
Dancing Horse Hill
Elmwood, Wisconsin
------------------------------
From: [EMAIL PROTECTED] (Brian Moore)
Subject: Re: Gnome graphical ppp dialer ?
Date: 24 May 2000 20:03:33 -0400
Reply-To: [EMAIL PROTECTED]
In article <[EMAIL PROTECTED]>,
Flemming Bjerke <[EMAIL PROTECTED]> wrote:
>
>The name of the program can be found in Redhat's guide on www.redhat.com.
>But, You can't install it because it requires another program which is
>available form the CD. (Great!) When You installed Redhat, You should have
>selected text installation and custom, then You can get both gnome and kde
>installed, and then You had had the program. It is easier just to use
>netcfg. Make a new interface (PPP, I suppose) and then edit the it.
>
>Flemming
>
Thanks for the reply. I realize what I did now, I installed kde
and not gnome (but it is hard to tell cause some of gnome still
seems to be here).
--
Brian G. Moore, School of Science, Penn State Erie--The Behrend College
[EMAIL PROTECTED] , (814)-898-6334
------------------------------
From: [EMAIL PROTECTED] (Paul Rubin)
Crossposted-To:
comp.infosystems.www.misc,comp.infosystems.www.servers.misc,comp.infosystems.www.servers.unix,comp.os.linux.networking,de.comm.infosystems.www.servers,de.comp.os.unix.bsd,redhat.servers.general
Subject: Re: ssl certificate server
Date: 25 May 2000 00:13:25 GMT
In article <8gggq7$jdm$[EMAIL PROTECTED]>, Diana Block <[EMAIL PROTECTED]> wrote:
>hope, there is somebody, who will help me.
>
>I want to setup a certificate server with linux, which should be an
>autorative server to give certificates to some of my other domains.
>Some of this domains are running under Microsoft IIS, others under Apache.
>
>Can somebody give me some statements, what I need and what to do?
>I have no experience in this area.
If you mean web server certificates that standard web browsers can read,
you have to get the certs signed by a CA that the browsers recognize.
Yon can do this (for example) through Verisign OnSite. See www.verisign.com
for more info. Thawte (now owned by Verisign) has something similar.
Equifax (http://www.equifaxsecure.com) is also trying to enter this
market so you might get in contact with them.
You can't just issue your own certificates and have the browsers
recognize them, unless you first install your signing root into the
browsers. If your users are all within your company or something like
that, you can ask them to load the signing root (visit a certain URL
on your server and click "yes" to some security dialogs). There is
some info in the mod_ssl docs (www.modssl.org) about how to set up a
private CA, if that's what you want to do.
If the servers are for public use (e.g. for e-commerce), generally
that method is not workable and you have to pay for public CA-issued
certs.
------------------------------
From: [EMAIL PROTECTED] (JEDIDIAH)
Crossposted-To:
comp.os.linux,comp.os.linux.development,comp.os.linux.development.apps,comp.os.linux.development.system,comp.os.linux.setup,comp.os.linux.advocacy
Subject: Re: Need ideas for university funded project for linux
Date: Thu, 25 May 2000 00:19:11 GMT
On Wed, 24 May 2000 20:51:59 GMT, George Russell <[EMAIL PROTECTED]>
wrote:
>On Wed, 24 May 2000 15:27:51 GMT, JEDIDIAH <[EMAIL PROTECTED]> wrote:
>>>around? Or even two? If an application says it uses GTK+ ver x.y,
>>
>> YES.
>>
>> It allows for other platforms to be supported.
>
>The QPL in no way forbids porting to BeOS / Mac/ Win 32.
Then why isn't anyone motivated to do so?
>
>Feel free to do so. Only TT's port to Win 32 is commercial.
>
>George Russell
--
In what language does 'open' mean 'execute the evil contents of' |||
a document? --Les Mikesell / | \
Need sane PPP docs? Try penguin.lvcm.com.
------------------------------
From: "Lonni J. Friedman" <[EMAIL PROTECTED]>
Subject: Re: resume download program
Date: Wed, 24 May 2000 19:47:45 -0400
[EMAIL PROTECTED] wrote:
>
> Any 'command-line program' which can ressume the file download in case the internet
>link was lost ?
>
> I tried 'wget -c' but it cannot work when the source is a http server ( instead of
>ftp server ).
>
You're doing something wrong, because it always works for me for both
ftp & http protocols. You are placing "http://" before the address,
right?
------------------------------
From: "Brian" <[EMAIL PROTECTED]>
Subject: Printing man pages
Date: Thu, 25 May 2000 00:36:34 GMT
I know I have read this somewhere but I cannot find it now.
How do I print a man topic so it is properly formatted on my printer?
When I send it to the printer, it appears with doubled letters, etc. What
is the correct command; i.e man grep | lpr ..........?
Thanks,
Brian
------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Subject: Re: removing LILO from mbr
Date: Thu, 25 May 2000 00:36:04 GMT
In article <[EMAIL PROTECTED]>, Chris Norris wrote:
>I have read sevral times that I can remove LILO from the mbr by using the
>command "fdisk /mbr" from a DOS prompt. This will replace the mbr with
>DEFAULT values. What I would like to know before doing this is, will these
>default values overwrite my partition information, restoring my disk to
>its original "whole disk" partition?
[...]
Why not lilo -u <device> ? Only the original boot sector will be
restored then, nothing else will be overwritten.
Ta',
Juergen
--
\ Real name : Jürgen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
From: Max Heijndijk <[EMAIL PROTECTED]>
Subject: Re: Xterm with transparent background ?
Date: Thu, 25 May 2000 02:40:31 -0400
Dowe Keller wrote:
> You dont. However, you could run eterm, or aterm with a transparent
> background. I like aterm myself, both work fine, but eterm can be a
> bit of a memory hog.
I downloaded aterm from rpmfind.net. It will not start.
Error message:
aterm: can't open pseudo-tty
aterm: aborting
Am I doing something wrong ? Do I have to give some specific options ?
Can't find it in the man-page.
Bye, M@X
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.misc) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Misc Digest
******************************
