On Thu, 23 Nov 2000, tc lewis wrote:
> whatever happened to the linuxvirtualserver.com idea? a sort of coalition
> of consultants or what not. i forget whose idea that was. joe mack's?
> yours? =)
I did a bit of work on this earlier this year. There were 2 parts to this,
neither of which I could get a reasonable plan for
1. using the name linux as in LinuxVirtualServer in a commercial context.
I talked to Maddog, who works for Linux International and who licenses the
name "linux" for commercial use, at the Ottawa Linux Symposium. He did not
want the name Linux used commercially in Linux Virtual Server as it would
preclude anyone else from making a Linux Virtual Server whose function was
different to ours and which used different principles/coding. The problem
as he saw it was that we were taking up too much of the Linux name space.
To use LVS in a commercial context, we'd have to call it something like
the "Marvellous Linux Virtual Server".
This precludes commercial use of linuxvirtualserver.com as a webname.
(We have the dommainname but can't use it commercially.)
If we were to do anything commerically with LVS it would have to start
with a different name and we would initially suffer from lack of name
2. Setting up a group of consultants under any name to do LVS work.
The original idea was to have a list of people who were prepared
to design, install, maintain or give talks about LVS. People looking
for commerical help on LVS would look on this page and pick likely
people according to criteria which became difficult to define.
It would be impossible to credential any of these people. No-one in a
position of authority exists in LVS. The nearest thing is Wensong who
everyone defers to by virtue of him having started the project single
handed and who writes most of the code. Still Wensong isn't in a position
to be able to say whether any particular person is suitable for doing
commercial LVS work and he has no interest in doing so. Wensong was quite
happy for me to try to set something up like this, but I was in no better
position than to be able to attest to someone's ability to do commercial
installs. I could only say whether someone had been helpful on the LVS
list. However there are people on the list who know a lot about LVS who
don't say much. What the customer wants to know, is if the LVS expert is
reliable, can solve the problems they have, is competent and has a
personality compatible with theirs.
The alternative would be to not make any declarations as to the competence
of the LVS experts and let potential customers of the LVS service try to
figure out who was best for them. Presumably they could go to the list
archives and see whether the expert made any useful suggestions or not.
However since the customer didn't understand LVS, they wouldn't be able to
tell the good suggestions from the bad ones. So this approach while it
would appear to be offering commercial support for LVS, would not be able
to consistently offer high standard service with a minimum of digging
around by a potential customer. It would just be show for us and
frustation for customers.
Another alternative was for people to be self selecting and ask people to
pay 100U$/yr. Money from this fund could be used to pay for plane tickets
for Wensong, while he was a student (this phase is just about over) run a
LVS conference in Hawaii or Switzerland :) or whatever. However someone
pointed out that while 100U$/yr was not a lot of money for a person
working in a developed country, it would be beyond the means of many
people (working or students) in other countries. If I were to keep track
of this money, it would require bookwork and being accountable to various
state and federal agencies here, as well as being accountable to the LVS
experts. I could easily see much of this money and all of my spare time
going just to maintain accountability of this rather small amount of
money. I wouldn't have time to work on LVS anymore.
So the first two problems were finding a new name for LVS and not
having any way to deliver good commercial grade support.
Other problems included maintaining the list of people. We wouldn't want
a person on the list to have a webpage and valid e-mail, but who never
answered their e-mail. Do we drop them or are they sailing their yacht
around the world and fully intend to answer their e-mail when they get
Related to this was not being able to make any money from such a group.
Maddog said that people had tried such groups before, but it wasn't a good
business plan and he didn't know of any that had made any money. The only
plan know to work was setting up a business yourself, hiring people
accountable to yourself and finding LVS work for them.
As for having a list of commercial companies doing LVS work on the webpage
- I thought at one stage of having a roadmap of the software pieces used
in LVS, who the contact people were, and what they were working on. This
would be 3 or 4 lines. Examples of this were the people who at one stage
were interested in going off writing L7 code etc and Thomas Proell who has
written a scheduler for web cache real-servers. As part of this I asked
commercial people who had written software used in LVS for entries in the
road map. One of them practically demanded a half page advertisement in
marketeese-speak, based apparently on the pivotal role he perceived his
company playing in the development of LVS.
I presume it is possible to have a listing of commercial companies on the
LVS webpage, but it would require control of the webpage by someone who is
perceived by everyone who comes to the webpage as having no conflict of
interest in the matter. Who on this list is internationally recognised as
having no interest in making money from their LVS skills (or being
impartial about those that do) and is prepared to sit on commercial people
who would like to abuse the system?
At the moment it's just simpler to let commercial people do their own
advertising. Potential customers have access to search engines and can
find commerical support for LVS in a few seconds. The
linuxvirtualserver.org website is currently a users and developers site.
It's not clear that we would gain anything from having commercial listings
on this site. Problems in doing so have already surfaced.
Joseph Mack mack@xxxxxxxxxxx