Distribution Service
David Battle
I am toying with the idea of providing a distribution service for Linux.
My motivation is to make enough money so that I can work full time on Linux,
hopefully improving and extending it in the process.
I would appreciate feedback from people who think this is a *Bad* idea as
well as those who think it is a *Good* idea. Please see the QUESTIONS below.
Here is my thinking:
For a small fee I would provide the following:
* A copy of the Linux Boot and Root disks with the Latest Release
* A Disk with Patches which have been Posted to the Net
* Optional Software disks, with Dependencies clearly marked
* Printed Introduction and Installation Instructions
* 2nd Day Air or Overnight shipping (Overnight costing extra) (US only)
* A telephone support line which you can call for help with Linux
* A BBS and/or FTP site where you can pick up Patches and New Releases
You would of course be free to copy and re-distribute any and all the material
provided if you so chose.
In the event that this distribution service was an overwhelming success,
excess profits would be turned back into Linux Research and Development,
buying new machines, hiring employees, supporting a connection to the
Internet, advertising to attract more potential Linux hackers, making the
BBS and Support Line Toll Free, etc.
QUESTIONS
I would like to ask you the following questions, please send answers to
bat...@cs.utk.edu:
1. Do you think this service would be useful?
2. Can you think of anything additional which could be provided which would
make you more likely to be interested in such a service?
3. Is there some media other than 3.5" diskettes which would be preferable
or more universally acceptable?
4. What do you think would be a reasonable price for this service, considering
media, packaging, and shipment costs and the time required to prepare a
distribution?
5. Do you think people would be willing to pay this price?
6. Do you think I would have any time left to hack on Linux? Do you think
I would make enough to support myself? Will I get tired of doing this?
7. Do you have any other comments which you would like to make about this idea?
Thank you all for any feedback you can provide,
-David L. Battle
bat...@cs.utk.edu
Jim Winstead Jr.
morning (before noon, even!), and I wasn't altogether clear, so I'll
have another go at it:
In article <ksn005...@utkcs2.cs.utk.edu> bat...@cs.utk.edu (David Battle) writes:
>I am toying with the idea of providing a distribution service for Linux.
>My motivation is to make enough money so that I can work full time on Linux,
>hopefully improving and extending it in the process.
What's enough money? Subsistence, comfortable, filthy rich? Really,
that's a quite vague goal. It's also interesting the number of people
that have been able to make very significant contributions to Linux on
their free time, with extremely little to (more likely) no monetary
reward. I cannot see Linux being a full-time thing for anyone at this
point, really.
>I would appreciate feedback from people who think this is a *Bad* idea as
>well as those who think it is a *Good* idea. Please see the QUESTIONS below.
If you haven't guessed, I fall into the first camp. :)
>For a small fee I would provide the following:
>
> * A copy of the Linux Boot and Root disks with the Latest Release
>
> * A Disk with Patches which have been Posted to the Net
>
> * Optional Software disks, with Dependencies clearly marked
>
> * Printed Introduction and Installation Instructions
>
> * 2nd Day Air or Overnight shipping (Overnight costing extra) (US only)
>
> * A telephone support line which you can call for help with Linux
>
> * A BBS and/or FTP site where you can pick up Patches and New Releases
You're really only offering two things that haven't been done for you
already: mailing disks, and telephone support. The installation
instructions are at least partially written (see INSTALL-0.95a :), and
there are a number of people working on better installation
instructions and intro-type material.
Other people have already volunteered (although I don't know how
seriously) to distribute Linux on floppy. If someone started trying
to sell Linux on disk for what I thought was an unreasonable price, I
would seriously consider offering to make the same sort of thing
available, for my cost only. (Disks, packaging, shipping.)
As for phone support, well, I don't see the demand, really.
>You would of course be free to copy and re-distribute any and all the material
>provided if you so chose.
Of course, the GNU Copyleft requires it. :)
>In the event that this distribution service was an overwhelming success,
>excess profits would be turned back into Linux Research and Development,
>buying new machines, hiring employees, supporting a connection to the
>Internet, advertising to attract more potential Linux hackers, making the
>BBS and Support Line Toll Free, etc.
You're dreaming! Have you tried adding up the cost of all these
wonderful things your enterprise is going to buy? All of the things
you mentioned would cost in the thousands, each.
>I would like to ask you the following questions, please send answers to
>bat...@cs.utk.edu:
Well, I've posted them, since I think this issue concerns the Linux
community as a whole, really - you won't be the last with this idea. :)
>1. Do you think this service would be useful?
Useful, sure. Feasible, not a chance. Like I said, the demand is
simply not there. Most of what you're offering duplicates freely
available services. Support is even freely available through mail and
this newsgroup.
>2. Can you think of anything additional which could be provided which would
>make you more likely to be interested in such a service?
Not really.
>3. Is there some media other than 3.5" diskettes which would be preferable
>or more universally acceptable?
You're going to have to support 5.25" HD and 3.5" HD, at least. There
are many people without 3.5" disks, even on 386's, and many people
have their 5.25" HD drives set up as their boot disk.
>4. What do you think would be a reasonable price for this service, considering
>media, packaging, and shipment costs and the time required to prepare a
>distribution?
Disks, assuming a modest collection of utilities with the boot/root
floppies: $5 (for 10)
Shipping: $4 (US Postal Service, with box)
Additionals: $2 (printed instructions, etc. This is being *very*
generous.)
Time: $2 (this obviously does not reflect 'start-up' time,
which would be minimal)
That's about $13, and I'm assuming you're not going to do anything
silly like buy a computer for this venture and work it into the
price. I'm also assuming very little advertising, which may not be
fair. Even then, it shouldn't increase the price beyond $15.)
>5. Do you think people would be willing to pay this price?
Sure, but I don't think that's the price you had in mind, since it's a
bare-minimum, no-profit cost.
>6. Do you think I would have any time left to hack on Linux? Do you think
>I would make enough to support myself? Will I get tired of doing this?
You would not make enough to support yourself, unless you can live
outdoors and feed off the land. You will have plenty of time to
'hack' on Linux, because I can't see the distribution game being a
real happenin' business, and I think you will soon see it's futility.
(Then again, I'm not being real optimistic. :)
>7. Do you have any other comments which you would like to make about this
>idea?
I think this is not a suitable direction for Linux' future. Linux has
been written entirely by volunteers who have been working on their own
time, and I don't think that should change. I also don't think it's
fair that someone take what has been written for free by people and
try and sell it to turn a buck (i.e. make a living doing so). How
fair is that to those of us who contribute our time freely? I know I
spent in excess of 20 hours last week getting the root floppy together
and distributed, and I know I'll be spending a fair amount of time
each week preparing for the next release.
>Thank you all for any feedback you can provide,
Well, it may not be the most encouraging feedback, but it's the best I
can give. This sort of operation is not without precedent, however,
as I mentioned in mail. Cygnus Support is reportedly doing quite well
by supporting GNU software - however, the market for GNU software is
quite large, and there's actually people to support. I don't see that
sort of market for Linux in the near future. (Speaking of near in
terms of years.)
--
Jim Winstead Jr. (CSci '95) | "Catch a fish!"
Harvey Mudd College | -Geddy Lee,
jwin...@jarthur.Claremont.EDU | San Diego Sports Arena
Disclaimer: Mine, not theirs! | January 20, 1992
I Reid
>In article <ksn005...@utkcs2.cs.utk.edu>, bat...@cs.utk.edu (David Battle) writes:
[stuff deleted]
>I agree with Jim Winstead; almost all of what you propose to provide is
[more stuff deleted]
I was going to mail David Battle directly but since there is discussion going
on here I'll just add to it. I have to say that I firmly agree with Jim
Winstead and the second person (with the amazing mail address.... do you have
to type all of that to login?) about the large amount of free work that is
going into Linux. In my own case, with my beginner's guide I'm trying to test
everything I write which involves (for the HD partitioning bit which will be
out as soon as my HD stops misbehaving and I stop corrupting my only floppy
copy of the rootimage :-)) repartitioning my HD to a full DOS only setup,
repartitioning to DOS and Linux, noting down EVERYTHING that happens,
reinstalling 2 filesystems, writing it all up, removing ambiguities and trying
to make it all sound easy and straightforward etc etc etc. Why should someone
else freeload on the things everyone else has contributed for free.
[Rant mode off]
'Nuff said?
Iain
c t wilson
> My motivation is to make enough money so that I can work full time on Linux,
> hopefully improving and extending it in the process.
>
> I would appreciate feedback from people who think this is a *Bad* idea as
> well as those who think it is a *Good* idea. Please see the QUESTIONS below.
>
I agree with Jim Winstead; almost all of what you propose to provide is
*already* being provided for *free*. Linux has, through the generousity
and hard work of Linus, and the hard work of many, many others, become
the *nix of those who wish to avoid commercial channels. I think Jim's
suggestions for service prices are about right. I don't think, however,
that $15 dollar distributions could sustain your service unless it was
not much more than (or just) a hobby. I plan to do some ports as soon
as I am able to expand my diskspace and give gcc 2.0 a try (the libs for
1.40 are not quite complete, though very stable); I will do this for the
joy of it and as my contribution to Linux. If you do set up a distribtuion/
support service, I hope it's in the same spirit.
Tom Wilson
Paul H. Rubin
In article <ksn005...@utkcs2.cs.utk.edu> bat...@cs.utk.edu (David Battle) writes:
>I am toying with the idea of providing a distribution service for Linux.
>My motivation is to make enough money so that I can work full time on Linux,
>hopefully improving and extending it in the process.
What's enough money? Subsistence, comfortable, filthy rich? Really,
that's a quite vague goal. It's also interesting the number of people
that have been able to make very significant contributions to Linux on
their free time, with extremely little to (more likely) no monetary
reward. I cannot see Linux being a full-time thing for anyone at this
point, really.
>In the event that this distribution service was an overwhelming success,
>excess profits would be turned back into Linux Research and Development,
>buying new machines, hiring employees, supporting a connection to the
>Internet, advertising to attract more potential Linux hackers, making the
>BBS and Support Line Toll Free, etc.
You're dreaming! Have you tried adding up the cost of all these
wonderful things your enterprise is going to buy? All of the things
you mentioned would cost in the thousands, each.
I don't know, maybe Linux isn't ready for this, but in principle the
idea doesn't seem too bad and there is precedent for it. The GNU
project is being worked on by a mixture of volunteers and full-time
paid hackers, and the Free Software Foundation gets the money to pay
the hackers by distributing tapes and manuals (but not phone support).
The FSF salaries are a lot lower than industry would pay comparable
programmers: they are intended to provide people willing to work on
GNU full time a way to keep a roof over their heads while doing so.
In this sense, anyone working for the FSF needs to have a streak of
volunteerism. I was the first paid FSF programmer (in 1986-7) and found
it to be a very rewarding experience. Working on free software
around first-classs programmers was infinitely more satisfying
than standard sleazy software company jobs, even though it paid a
lot less.
Morgan Schweers
I'd like to point out that what Mr. Battle was suggesting is extremely
similar to the situation which occurred with BSDI and the Net/2 distribution
of BSD Unix. The situation COULD be the same, if Mr. Battle decided to invest
a great deal of energy, completeing the Linux transition to a commercial class
OS, but making the decision to *NOT* release his code. Thus he would be
offerin an encumbered version of Linux, but he'd be able to name his price
because the current version of Linux wouldn't have all the features (like X)
that he'd add in.
However, this is an extremely larger effort than what I gleaned that
Mr. Battle was interested in. Substantially uninteresting to the quick buck
mindset, I think...
-- Morgan Schweers
--
Hacker, Furry, SF reader, gamer, art collector, writer. 24 hours isn't enough.
m...@netcom.com | I'm a practicing furry! Some day I hope all the practice
Freela @ Furry | will pay off, and I'll grow fur! -- me
K_Balore @ Furry |___________________ CLEAN C:\USR\SPOOL\*.* [SigVir] /SUB
Hi! I'm a .signature virus! Add me to your .signature and join in the fun!
Drew Eckhardt
>Greetings,
> I'd like to point out that what Mr. Battle was suggesting is extremely
>similar to the situation which occurred with BSDI and the Net/2 distribution
>of BSD Unix. The situation COULD be the same, if Mr. Battle decided to invest
>a great deal of energy, completeing the Linux transition to a commercial class
>OS, but making the decision to *NOT* release his code. Thus he would be
>offerin an encumbered version of Linux, but he'd be able to name his price
>because the current version of Linux wouldn't have all the features (like X)
>that he'd add in.
>
> However, this is an extremely larger effort than what I gleaned that
>Mr. Battle was interested in. Substantially uninteresting to the quick buck
>mindset, I think...
>
> -- Morgan Schweers
Are you familiar with the GPL? Like much other good software these days,
Linux falls under it. Basically, it says you can make any changes you
want to something, but if you distribute it publically in any
way (comercial, or otherwise) any derivative work is also covered by
the GPL, which means it must be freely available with source,
standard system libraries and headers excluded.
Mr. Battle could sell a commercial class Linux, but it would also be
freely available.