E. Charters wrote:

> It takes a lot of money and time to stay ahead of the curve on a distro.

No. You need one person to put out the distro. PV does that already. Instead
of phone support, as you suggest, you need a forum with somebody keeping an
eye on trolls and picking up the bits and pieces that are worth adding to the
site.

That's two salaries, no rent. Add a few contributors giving a hand once in a
while and able to take the relay if anybody gets sick. Let's say $200,000 /
year for salaries. At only 25$ / year per user, you need 8,000 to keep the
distro going. If you have a good distro, getting this kind of money should be
no problem. If you can't get this kind of money, quitting is an option.

Slackware has a nice business model. Countrary to Debian's, its little
packages get out the door fast and swaret has showned they we aren't lame in
any way despite their simplicity. Slackware is easy to maintain and it uses
the best that is available from the Linux community. There's no mumbo-jumbo
nonsense tools to maintain in order to keep the customer hooked and this is
fine to me and to most Slackware users. There's no /community/ to nourrish
endless discussion, and only one platform available directly from Slackware.

But maintaining a /store/ to provide pretty much the same as is avalaible on
the net or from any Linux CD reseller, is stupid. What's the use of shipping
plastic all over the world, mainly when, as has often been told here, people
receive the said plastic, weeks, if not months, after the distro is out and
they have already downloaded the distro? This business model is dead. It died
years ago.

But it seems The Little Man does everything he can to piss off people so they
don't send money.

When I began using Slackware 3 years ago, I didn't have a cent in my pockets
and wasn't sure I'd stay with Slackware. But, say, compared to the Mandrake
and Bed Rat, I liked the no nonsense approach. Then there was this problem
with the missing link to the /dev/cdrom and I was given shit when I reported
it. I finally found out the booby-trap was intentionnaly set to /encourage/
beginners to take a subscription.

But beginners are not those you should expect money from. Often, they come
from Windows and they want to find out if 1) Linux works for them 2) if
Slackware is really the distro that's for them. Setting booby-traps is a very
bad welcome.

The die-hard trollers here say they love Slackware so much they will gladly
accompany it to its last rest. Should you suggest to include a firewall
configuration utility, or swaret to check dependancies, which they are in no
way obliged to use, they get offended because it makes Slackware easier to use
for beginners. To them, beginners must either learn the hard way or opt for
another distro, then get to Slackware.

When I came from Windows, 3 years ago, there was no Guarddog to configure
iptables. I had to learn how to write a basic firewall for a desktop. Thanks
to Daniel Robbins' text on IBM's site, and in no way to The Little Man and His
Trolls, I finally was able to.

But just finding proper instructions, getting to Robbins' text, was a pain.
When you know nothing on a subject, finding good instructions is a tough job.
And whatever I've learned then, I pretty much lost because I'm not in the
business of maintaining firewalls, servers, networks, whatever. I'm just
maintaining my desktop.

I don't believe beginners are in general as patient, stubbornly oriented
towards a goal, as I have been. They might very well begin with Mandrake, Suse
or Bed Rat, get used to their fancy tools and System V booting, and never get
to Slackware. No doubt, that's the bet those companies are taking.

Then, there are all those new Debian-based distros moving ahead: Mepix,
Ubuntu, Knoppix (if it may qualify as a distro), Kanotix, etc.

Somehow, people who put out those distros manage to keep in touch whith their
users and that's why the users appreciate the product. They don't get
slack-pkg shoved down their throat without any explanation. They have a little
on-site community they can rely on. They appreciate.

After 3 years of using Slackware, I know it suits my needs. What's more, I now
have enough money to send my 25$ a year, nothing asked in return, no problem,
even more if need be. But shit, what do I see?

The Little Man sings in chorus with the aols trolls, having a good bragging
time on irc, and the result is, when he gets sick, the distro is no longer
maintained and the signature on his Trust-Gus-Brazil messages don't even check.

Whatever you plead him to provide, he won't provide or he'll provide some
piece of crap without any explanation. Obviously, this fuck-head wants to ruin
his distro to become a legend of the good ol' times.

Well, so be it -- as everybody says, it's his distro -- but I'll have to be
excused for not contributing. After all, at this point, the sooner the distro
dies, the longer the legend will live. As Neil Young puts it, it's better to
burn out than to fade away.

So, after 3 years of getting used to Slackware ways, I'm looking for a new
distro to send my money to. It might be straight Debian updated on testing(1),
Knoppix, Kanotix or even Vector. After 3 years of using Slackware I still have
my nose up in the air, trying to sniff out which way the wind is turning.
Shit, could I have ever done without this!

(1) BTW, did you learn that apt-get /might/ be checking signatures on packages
for Sarge, something swaret already did one year ago?

> It appears to me that Linux did not get the support of the business
> community because the revenue model is hard to sell. You cannot make
> it with free downloads as attractive as they are, and 40 dollar 4 disk
> sets. Support contracts as nice as they were worked out in practice but
> also scare the bejeesus out of people at 25 dollars for 15 minutes.
> Given that people anticipate about 300 questions a year and 15 all
> niters to fix problems, the cost begins to mount to infinity in their
> minds.

Either those people are right in expecting that much need for support and they
should definitely stay with Windows, or they should be proven wrong by /both/
on-site support and good support contracts. I see no effort whatsoever on
Slackware's part to achieve the latter.

> I admit one thing, when Slack had email support, it was slick and their
> people on the phone were great. Even complex upgrading problems became
> almost understandable.

Maybe this should be provided, but really, as a last resort. When distros get
to relying on support for raising money, on-site instructions tend to get not
so clear. Then, with good reason, companies begin to fear the 25$ for 15 min
calls. The organisation must be kept lean and mean.

> I always thought there should have been docs
> like this.

Why provide the doc, if you get paid for providing the information to one
custoner at a time? And, of course, who will want to loose their time, and
eventually, their life, providing this kind of /information/, save for a breed
of jackasses who get clients infuriated?

> Offhand I would say a distro would need 4 help people of extremely good
> skills with people and distros, and about 6 programmers all full time.

Why would you need 6 programmers? Doesn't dustribution mean distributing the
work of programmers? Isn't it exactly what Slackware is doing? Anything wrong
with that? Don't Swaret and Guarddog provide standard tools that can be used
by any distro? What's the use of distro specific tools? What do you figure has
to be invented for Slackware that's not already available? A better security
check to Webmin? Well, maybe

> PV will be on the mend I would think for a few more months.

Well, you're certainly better informed than I am. To tell you the truth,
though I wish all the best to the Little Man, I couldn't and certainly
shouldn't care less. You know, Ford, GM, GE, etc, have survived the /death/!!!
of their founders. A company's scope should go beyond that.

The trollies on this ground pretend that if ever the Little Man was to
disappear, a whole pack of developper would raise to hold the flag. But the
Little Man has been sick for months, he's apparently still staggering, and
nothing happens.

Despite the Little Man promises while he was sick, Slackware still is and will
continue to be a one Little Stubborn Man organisation. It's had advantages in
the past and now has huge drawbacks.

Slackware won't survive the way things are going now. It won't survive for me,
for beginners and companies alike. It might survive for some time for aols
trolls. The time it takes to drive it to the graveyard.

GP

Well I have used Slackware for ten years. I do not use Windows and I
don't think I have used much it for perhaps 3 or 4 years. It's just that
the frustrations with Windows are very vile, whereas the frustrations
with Linux are no less time consuming, but there always seems that there
is hope for some kind of solution.

On the other hand that has to change as the need for certain commercial
programs grows. I will be forced to use Windows. I dread using XP.
I have used it and it is as slow_as_molasses_in_novaya_zemla. Win
98 is 5 times faster.

There are problems with Slackware that started around 7.0. The uprgrade
past glibc6 started breaking things and this has continued until this
day. The distros seem smooth until you try to get some things working
that others take for granted. Like IDE CD-Writing or running GUI's in
some cases. One time he left the crypto libs of a distro, saying there
was some sort of problem with licenses, and exploting etc.. of course
if you don't have them you cannot run a site where remote logins use
passwords.

Samba and other sorts of software have similar frustrations. The trouble
is, if one thought of throwing the cash at them to solve the problem is
there any guarantee that the problem would go away for X dollars? That
sort of solution is not apparent. What a customer needs is not only
a solution but one that he can see is one before he applies it. Nacht
und Nebel is not an environment you want to put a customer in. And with
Linux there is lots of nacht und nebel.

There is lots more to putting out a distro and supporting it than
one man can handle. There is hordes of testing of packages, and it
could use people just writing docs too. You will never do it with
only a couple of guys. Red Hat did a so so job and they spent 18
million dollars getting to 6.2. Most of that went into the GUI
but there were lots of other areas. If you looked at their
changelogs, you cannot tell me that one person or even three people
could make 1000 changes in one program in six months, AND do all the
other stuff to check out packages. And do phone support. No way.
There is a lot to it. Start up scripts alone are quite an excercise.
Their mass is not that great, but they have a lot of exigencies
to take care of.

Even an old lowly distro like Yggdrasil that had many useful features
that were "stolen" by other distros like Slackware and RedHat (graphical
install for one) had several programmers and business people working
there. To give you an example of penetration, Yggie sold 75,000 copies
of Plug and Play linux in one year. Just for a laugh I installed it a
year or so ago on a modern system and it worked not bad. One thing it
had was a completely rebuildable source tree. Every program on the disk
could be rebuilt in one make that checked for dependencies and
conflicts. Believe me on 90 Mhz pentium that took a while. He still had
not solved the color map problem in X, but at least he was trying
something nobody else had. He tried for a few years
to make a distro that probed and recognized all hardware and set it up,
automagically but it never worked. He needed about a million bucks and 5
or 6 programmers. He could have got it too, but that was not to be.

Most of the distros worked with Indian and Chinese offshore programmers
as the predominant group. The Hat did for one. But they want real cash
for their work.

Trying to shoehorn a lot of people into the unix paradigm simply is too
tough. It could be done educationally as in Mexico, China and India but
our people have turned their backs on it. Teachers in Canada, university
or high school are not deploying Linux, (well, not widely or
pervasively). Linux should be a natural for education. Gates has horned
his way in there, thrown some cash and computers around and everybody is
bought. Sickening really. I can see why engineering departments do not
use Linux often as few specialty programs are written for Linux. An
exception is Ryerson here and there I have seen.

I would not criticize PV that much. I don't always agree with some
things that happen, but by and large, it is a terrific job. It
is fast, small and complete compared to RH. I ran the hat in a
commercial environment for a while. It was not that bad, but adding a
program was more complex and adding a new kernel or recompiling the
kernel was no better than Slack. Worse I would say. Writing scripts
or changing scripts to do network and server stuff was as bad as any
OS. In fact it was a lot harder than a Slack system. I admit one
thing, that if you stayed with a simple hardware setup, that the
Hat stayed crash free. I could not always say that if one were running
a large GUI in Slack. I have seen a few problems there.

Your problem with dependency checking may not be that well figured
out. You have to ask yourself, is dependency checking at installation
time what you want to do? What if you run into a problem? What
do you do? And if you install a package, and it has a problem with
dependencies you get the message the instant you try to run it.
I have seen Debian systems that had no way to remove a package
and then install a new one of a later version. No way. you were
stuck completeley. No way to upgrade a package. You had to be
a complete genius to figure out how to fix it. The thing to do
was to completely reinstall. Apt get, like rpm's are wonderful
when they work. When they don't, you might as well be trying
to fix the space shuttle with a monkey wrench, during take off.

EC<:-}


GP wrote:
> E. Charters wrote:
>
>> It takes a lot of money and time to stay ahead of the curve on a distro.
>
>
> No. You need one person to put out the distro. PV does that already.
> Instead of phone support, as you suggest, you need a forum with somebody
> keeping an eye on trolls and picking up the bits and pieces that are
> worth adding to the site.
>
> That's two salaries, no rent. Add a few contributors giving a hand once
> in a while and able to take the relay if anybody gets sick. Let's say
> $200,000 / year for salaries. At only 25$ / year per user, you need
> 8,000 to keep the distro going. If you have a good distro, getting this
> kind of money should be no problem. If you can't get this kind of money,
> quitting is an option.
>
> Slackware has a nice business model. Countrary to Debian's, its little
> packages get out the door fast and swaret has showned they we aren't
> lame in any way despite their simplicity. Slackware is easy to maintain
> and it uses the best that is available from the Linux community. There's
> no mumbo-jumbo nonsense tools to maintain in order to keep the customer
> hooked and this is fine to me and to most Slackware users. There's no
> /community/ to nourrish endless discussion, and only one platform
> available directly from Slackware.
>
> But maintaining a /store/ to provide pretty much the same as is
> avalaible on the net or from any Linux CD reseller, is stupid. What's
> the use of shipping plastic all over the world, mainly when, as has
> often been told here, people receive the said plastic, weeks, if not
> months, after the distro is out and they have already downloaded the
> distro? This business model is dead. It died years ago.
>
> But it seems The Little Man does everything he can to piss off people so
> they don't send money.
>
> When I began using Slackware 3 years ago, I didn't have a cent in my
> pockets and wasn't sure I'd stay with Slackware. But, say, compared to
> the Mandrake and Bed Rat, I liked the no nonsense approach. Then there
> was this problem with the missing link to the /dev/cdrom and I was given
> shit when I reported it. I finally found out the booby-trap was
> intentionnaly set to /encourage/ beginners to take a subscription.
>
> But beginners are not those you should expect money from. Often, they
> come from Windows and they want to find out if 1) Linux works for them
> 2) if Slackware is really the distro that's for them. Setting
> booby-traps is a very bad welcome.
>
> The die-hard trollers here say they love Slackware so much they will
> gladly accompany it to its last rest. Should you suggest to include a
> firewall configuration utility, or swaret to check dependancies, which
> they are in no way obliged to use, they get offended because it makes
> Slackware easier to use for beginners. To them, beginners must either
> learn the hard way or opt for another distro, then get to Slackware.
>
> When I came from Windows, 3 years ago, there was no Guarddog to
> configure iptables. I had to learn how to write a basic firewall for a
> desktop. Thanks to Daniel Robbins' text on IBM's site, and in no way to
> The Little Man and His Trolls, I finally was able to.
>
> But just finding proper instructions, getting to Robbins' text, was a
> pain. When you know nothing on a subject, finding good instructions is a
> tough job. And whatever I've learned then, I pretty much lost because
> I'm not in the business of maintaining firewalls, servers, networks,
> whatever. I'm just maintaining my desktop.
>
> I don't believe beginners are in general as patient, stubbornly oriented
> towards a goal, as I have been. They might very well begin with
> Mandrake, Suse or Bed Rat, get used to their fancy tools and System V
> booting, and never get to Slackware. No doubt, that's the bet those
> companies are taking.
>
> Then, there are all those new Debian-based distros moving ahead: Mepix,
> Ubuntu, Knoppix (if it may qualify as a distro), Kanotix, etc.
>
> Somehow, people who put out those distros manage to keep in touch whith
> their users and that's why the users appreciate the product. They don't
> get slack-pkg shoved down their throat without any explanation. They
> have a little on-site community they can rely on. They appreciate.
>
> After 3 years of using Slackware, I know it suits my needs. What's more,
> I now have enough money to send my 25$ a year, nothing asked in return,
> no problem, even more if need be. But shit, what do I see?
>
> The Little Man sings in chorus with the aols trolls, having a good
> bragging time on irc, and the result is, when he gets sick, the distro
> is no longer maintained and the signature on his Trust-Gus-Brazil
> messages don't even check.
>
> Whatever you plead him to provide, he won't provide or he'll provide
> some piece of crap without any explanation. Obviously, this fuck-head
> wants to ruin his distro to become a legend of the good ol' times.
>
> Well, so be it -- as everybody says, it's his distro -- but I'll have to
> be excused for not contributing. After all, at this point, the sooner
> the distro dies, the longer the legend will live. As Neil Young puts it,
> it's better to burn out than to fade away.
>
> So, after 3 years of getting used to Slackware ways, I'm looking for a
> new distro to send my money to. It might be straight Debian updated on
> testing(1), Knoppix, Kanotix or even Vector. After 3 years of using
> Slackware I still have my nose up in the air, trying to sniff out which
> way the wind is turning. Shit, could I have ever done without this!
>
> (1) BTW, did you learn that apt-get /might/ be checking signatures on
> packages for Sarge, something swaret already did one year ago?
>
>> It appears to me that Linux did not get the support of the business
>> community because the revenue model is hard to sell. You cannot make
>> it with free downloads as attractive as they are, and 40 dollar 4 disk
>> sets. Support contracts as nice as they were worked out in practice but
>> also scare the bejeesus out of people at 25 dollars for 15 minutes.
>> Given that people anticipate about 300 questions a year and 15 all
>> niters to fix problems, the cost begins to mount to infinity in their
>> minds.
>
>
> Either those people are right in expecting that much need for support
> and they should definitely stay with Windows, or they should be proven
> wrong by /both/ on-site support and good support contracts. I see no
> effort whatsoever on Slackware's part to achieve the latter.
>
>> I admit one thing, when Slack had email support, it was slick and their
>> people on the phone were great. Even complex upgrading problems became
>> almost understandable.
>
>
> Maybe this should be provided, but really, as a last resort. When
> distros get to relying on support for raising money, on-site
> instructions tend to get not so clear. Then, with good reason, companies
> begin to fear the 25$ for 15 min calls. The organisation must be kept
> lean and mean.
>
>> I always thought there should have been docs
>> like this.
>
>
> Why provide the doc, if you get paid for providing the information to
> one custoner at a time? And, of course, who will want to loose their
> time, and eventually, their life, providing this kind of /information/,
> save for a breed of jackasses who get clients infuriated?
>
>> Offhand I would say a distro would need 4 help people of extremely good
>> skills with people and distros, and about 6 programmers all full time.
>
>
> Why would you need 6 programmers? Doesn't dustribution mean distributing
> the work of programmers? Isn't it exactly what Slackware is doing?
> Anything wrong with that? Don't Swaret and Guarddog provide standard
> tools that can be used by any distro? What's the use of distro specific
> tools? What do you figure has to be invented for Slackware that's not
> already available? A better security check to Webmin? Well, maybe
>
>> PV will be on the mend I would think for a few more months.
>
>
> Well, you're certainly better informed than I am. To tell you the truth,
> though I wish all the best to the Little Man, I couldn't and certainly
> shouldn't care less. You know, Ford, GM, GE, etc, have survived the
> /death/!!! of their founders. A company's scope should go beyond that.
>
> The trollies on this ground pretend that if ever the Little Man was to
> disappear, a whole pack of developper would raise to hold the flag. But
> the Little Man has been sick for months, he's apparently still
> staggering, and nothing happens.
>
> Despite the Little Man promises while he was sick, Slackware still is
> and will continue to be a one Little Stubborn Man organisation. It's had
> advantages in the past and now has huge drawbacks.
>
> Slackware won't survive the way things are going now. It won't survive
> for me, for beginners and companies alike. It might survive for some
> time for aols trolls. The time it takes to drive it to the graveyard.
>
> GP
>

E. Charters wrote:

> The distros seem smooth until you try to get some things working
> that others take for granted.

Really?

> Like IDE CD-Writing

Yeah. A dirty joke on 2 releases: 8 and 8.1 . After eight years with
Slackware, don't tell me you didn't manage to fix the missing link!

> or running GUI's in
> some cases.

I never had a problem running KDE.

> Your problem with dependency checking may not be that well figured
> out. You have to ask yourself, is dependency checking at installation
> time what you want to do? What if you run into a problem? What
> do you do? And if you install a package, and it has a problem with
> dependencies you get the message the instant you try to run it.
> I have seen Debian systems that had no way to remove a package

We're not talking Debian here, but Slaclware. I never had a single problen
with swaret.

GP

Well the reason PV does not do swaret is not that he does not
adopt other packages but that swaret can break things badly.
It perhaps has not happened to you yet, but it might. Complex
machines break badly. Also the package is not well maintained.

There is also a philosophical reason that I mentioned. What is
the point of loading a package where it has not been checked
for your system yet? Somebody had to write it and try it out
on some system or other.
If it needs libraries your system has not got, then they
should make a statically compiled version. Or they should
say what library versions it requires in front. As in
documentation. Tbey don't as it is free software and that is
all the time they want to spend on it.

You cannot rewrite someone elses badly behaved C code often.
I have made small changes to drivers, where it was obvious
and translated the odd bit of code that was writen for other
platforms but C code for X is not for the faint of heart. That
incidentally is why there are so few apps for Linux. The code
writing for a GUI is a very steep and slippery mountain range
that keeps on going and going.

Linux is a miracle of compiler, base OS, and GUI tools. Built by
a committee of madmen with a mutual need and reverse engineered
through a wall of cottage cheese. That it works at all should
be viewed as a bloody miracle. Ever try to get two versions of
GCC working on the same machine? Followed Stallman's instructions?
If we ever get attacked by a major power, I think a few people should
save up a few shekels and hire Stallman over to the other side as a
logistics expert. In a few weeks their entire army would be going
round and round backwards in ever smaller circles in myriads of
different places, but the commanders would all be enthusiastically
reporting progress against the enemy. Given a little more time we
could walk in an take over, taking the weapons calmly out of the
hands of weeping soldiers who would be saying things like, "I
pushed the info button and it kept taking me to the top of the
page, saying I was already at the start of the document. Where is
my mother?"

I cry every time I try to get something basic working. but then I
stop and remember Windows 3.11 and installing a new graphics card
with a mouse on com 3. And setting the MTU settings for trumpet.
And Microsoft telling me that my 1000 dollar Basic Professional
Development System would not work across a network, and anyway
being 3 months old, was no longer supported and to buy VB IV Pro
for another 400 dollars. Yeah, I guess Linux is not so bad. They
should solve this upgrade thing though.

EC<:-}

GP wrote:
> E. Charters wrote:
>
>> The distros seem smooth until you try to get some things working
>> that others take for granted.
>
>
> Really?
>
>> Like IDE CD-Writing
>
>
> Yeah. A dirty joke on 2 releases: 8 and 8.1 . After eight years with
> Slackware, don't tell me you didn't manage to fix the missing link!
>
>> or running GUI's in some cases.
>
>
> I never had a problem running KDE.
>
>> Your problem with dependency checking may not be that well figured
>> out. You have to ask yourself, is dependency checking at installation
>> time what you want to do? What if you run into a problem? What
>> do you do? And if you install a package, and it has a problem with
>> dependencies you get the message the instant you try to run it.
>> I have seen Debian systems that had no way to remove a package
>
>
> We're not talking Debian here, but Slaclware. I never had a single
> problen with swaret.
>
> GP
>

E. Charters wrote:

> Well the reason PV does not do swaret is not that he does not
> adopt other packages but that swaret can break things badly.

Really?

> It perhaps has not happened to you yet, but it might.

All sorts of things might happen, but it hasn't happened to me nor to anybody
I know.

> Also the package is not well maintained.

It is not maintained because all Cottyn got for his excellent work was shit.
So he quitted the job.

Thanks again for your opinion! It's getting late.

GP

On 2005-03-02, GP <gilpel@inverse.nretla.org> wrote:
> E. Charters wrote:
>
>> Also the package is not well maintained.
>
> It is not maintained because all Cottyn got for his excellent work was shit.
> So he quitted the job.

Sorry for joining the flame, but I'm just curious -- what did this Cottyn
guy expect? Lots of money and worldwide fame? Well, yes, his work was
rejected by PV. So what? That's life. My own patches to some projects
were rejected several times, and I was rather embarassed, but I didn't
take any offence.

The greatest joy for an open-source developer is seeing that his creations
are used and highly valued by people. Pushing them into a Linux distribution
is not that important.

--
Alexander Ulyanov, maintainer of PosBand roguelike
E-mail: posband_AT_earthsea_DOT_org Web: http://posband.earthsea.org/
Anyone who is capable of getting themselves made President should on no
account be allowed to do the job. -- The Hitchhiker's Guide To The Galaxy