Re: os discuss (http://www.osopinion.com/perl/story/7263.html)

[Adam Tauno Williams <adam@xxxxxxxxxxxxxxxx>]

>>> The Press tends to treat Linux as if there is a "Linux Corporation", with
>>> industrial-type schedules.
>The press is accustomed to covering the behavior of corporations.  The
>mainstream press will probably not be willing or interested in
>understanding
>open source (behaviors/methodologies) until it has a majority of the
>desktop
>market.  Then and only then will the development standards for user apps
>be at least multi platform.

I don't often hear "understaning" and "the press" used together...  It would be
nice if they'd just report validated facts,  even that would be an improvement.

>>>Roger Smith or GM?  When Linus and his merry kernel men feel there's
>>> nothing to release in that timeframe, the press makes a fuss. Again,
>>>how could GM not release a new model when they said they would?
>You have a lot of frustrated people with MS for that very reason.  It is
>called win9x so we must release by the model year.  (Incidentally 9x evolved
>from the model year of cars)  If they were a little more honest and took a
>release early and often approach, as long as stability was the highest
>priority, they might piss a few less people off.

Microsoft IS moving to a release often schedule:  witness Win98, Win98 SE,
WinME,  all came out within close proximity.  Not only is this "good" for users,
it helps the revenue stream too.

>>Forgetting the schedules for a moment, I'd like to see
>>your explanations as to why Linux is NOT being run like
>>a large monolithic organization.  

1. Communication. Linus, etc..., have nothing to hide, they are very candid
about there objectives, goals, frustrations, etc... Having dealt with large
software companies I can tell you this is very diffrent from the party line: 
yes we will support that, no we won't,  wait for the next release, Oh! it won't
be in that either, wait for the next release.

2. Nothing but lack of excellence is going to stop someone from joining the
inner-circle.  For OS kernels nothing questionable SHOULD be accepted.

3. I can argue my case with Linus directly.  I have had e-mails with Alan Cox, 
they are very open to intelligent questions if you can demonstrate you've
already looked all the usual places.

4.  You painting to narrow.  Linux is not JUST the kernel,  or at least not
pragmatically.  The management of various related projects (glibc, X, etc...)
vary immensely.  This is an asset,  diffrent project may work best under
diffrent kinds of management.

>Linux is not run like a corporation because the development model of
>corporations does not work as well.  (only time will tell for sure) Do
>you want to be able to use the results of developers, bug fixers and testers?Or
>do you want to use the results of the overhead of a corporations:  Janitors,
>interest on capitol investments, marketeers salaries, salaries of managers who
>say release the software before it is stable and a low percentage of
>developers. Software companies have a lot of overhead which I choose to 
>support as little as possible.
>>Eric Raymond's final comments in his Brief History of Hackerdomp point out
>>that the bazzar/Open Source movement flourishes because which has no
>>standards, is released and fed back rapidly, and which involves a great many
>>participants.From my perspective, Linus appears to sit in his great white
>>throne and judge what stays and what goes.

Eric is guilty of a bit of priest-craft IMHO.  Take what he says (which is often
very true in the fundamental perfect-world sense) and mix it with a robust
helping of reality and human psychology.
 
>His throne is built on his own ego instead of a pile of money.  (OK,
>maybe a little bit of money)

Mmmm,  I've read quite a few interviews.  While he certainly seems
self-confident I would never label him as an egotist.

>>In addition, there is a large standards document which
>>must be followed before the code is even considered
>>acceptable.  This again, weeds out contributers.  

Of course, thank goodness.  At the end of the day all code has to compile and be
maintainable.  Standards are the only way to achieve that end.  Every large open
source project I know has such a document.

>>It's beginning to look a lot like Microsoft from here.

Nope,  because I (who work for a less than fortune 500 company) can take my
issues all the way to the top... for free.

>>However, having never attempted to contribute anything,
>>I can only speculate based on my feelings.

Your feelings are not completely wrong,  there is a high barrier to entry.  But
for something like kernel work I feel that's a good thing.  If my spreadsheet
crashes....  If my kernel crashes....

Systems and Network Administrator
Morrison Industries
1825 Monroe Ave NW.
Grand Rapids, MI. 49505