Re: os discuss

[Adam Tauno Williams <awilliam@xxxxxxxxxxxxx>]

>>>>...In a more level playing field, where smaller companies
>>>>are competing for smaller pieces of the market, it's not 
>>>>as clear that this laid-back approach to deadlines would work.
>>>>And, by extension, that the Open Source model would work.
>This is an interesting supposition. I am reminded of what
>Watson (of Watson and Crick, of DNA fame) said when asked
>about what other scientists thought of their work:
>   "Only 8 other people in the world undstsnd it, and since
>    they were all working on it too, the probably hate our
>    guts."
>There is a lot of motivation to produce good software here, and
>since there may not be enough of a market there to make selling
>it worthwhile, it will either be GPL'ed or sold at VERY high 
>prices.
>>>I'm not sure I follow this.  What are some "smaller"
>>>markets that you speak of?  Niche applications (IMHO)
>>>is Open Source's home turf, and where it really
>>>shines.  Someone with ONE purpose can create a truely
>>>awesome tool for that ONE purpose,  more out of
>>>personal passion to fill the need, than necessarily
>>>for financial gain (which may be a "side" benefit for
>>>solving a problem).
>Agreed. This is the initial motivation for most programming.
>>>For example,  note the absence
>>>of a pre-eminate word processor that is GPL-ed.  I
>>>think the OS model falls aport more often when
>>>attacking broad and general problems.
>Oh? Operating systems are not examples of attacking "broad and general
>problems"? I think there are other reasons for this:
>  - In recent years, Word really has dominated things. There
>    are NOT a lot of companies duking it out for market share,
>    unless you ignore Word.

Yes, but Word doesn't run on any "other" operating systems,  and that issue has been haunting all those other "OS"'s for some time.

>  - Good WP programs are HARD to write. One person can spend
>    YEARS of spare time on this sort of project, with very little 
>    hope of gaining attention or respect until it is really ready
>    for prime time. If/when you suceed, there's a lot of attention
>    to be gained, but gratification is very much delayed.

Granted.  Especially when that developer has to use his spare development time to address ALL the issues relating to the word processing experience. See below...

>  - The lead time is so long that technology underneath the project will
>    change, which delays delivery even further.
>A good example of the above is Abi-Word, which is just about ready for
>prime time after 4-5 years of development, mostly by one person.  I don't
>think it was XML-based from day 1, either.

It most certainly was not,  I've watched both GNOME and Abiword progress for some time now, and I think it illustrates one of the strong points of Open Source.  Code re-use.  While GNOME was mucking around with "boring" things like ORBs, XML DTDs, and "object models" the progress of Abiword seemed slow and steerless,  hampered by a need to do everything.  When GNOME finally started to form into something that more resembled the consistence of Peanut Butter vs. warm grape jelly,  Abiword "adopted" alot of the GNOME architecture (and corresponding libraries) and began to make noticable progress.  "Trivial" issues like how to cut-n-paste, file formats, we're addreses for them by something else.  This is one reason I'm optimistic about Open Source.  While I don't beleive the "no applications" myth is currently true,  I also percieve that the pot that is the Open Source community is just now reaching a slow boil. Discussion of issues like code re-use in something other than kernels!
 is quite recent.  Another axample of this is Palm Pilot support.  Because the people who wanted to support the Palm Pilot under Linux used existing "components" (asks gnomecc for configuration information, runs as a panel applet) it more or less reached maturity in a matter of weeks, and works VERY well.

>>I think we agree here, Adam.  I was referring to an
>>environment where there are many companies vying for
>>a small share of a large market.  A word processor is
>>a good example.  Ideally, however, this would also
>>occur in the Operating System arena.  Right now, we
>>have a few major players doing Op Systems.  If there
>>were, say 25 with no one having any more than say, 10%
>>market share, then would the O-Source model fall apart?
>>My thinking is that it probably would.
>I don't see why; can you explain?

Actually I think OS is more prone to produce and maintain the aforementioned enviroment than is commercial/closed software.  If I'm on 10% of the market, and I'm in it for the cash,  that's a pretty tough row to hoe.  I  think the current M$ totalitarianism is the natural result of commercial/close software developement.  All the investor cash is going to flock to the biggest player,  and he'll use that to advance his, say 5%, OS dominance over other players until they're driven off the field.  Wall Street doesn't care one iotta about product quality, corporate ethics, or anything else (until it gets the CEO dragged into court, and then only momentarily).

>>>>This article cited below is either naive or biased in
>>>>it's allusion to the Open Source model being flawless.
>>>I'd say optimistic, not naive.  I think business will
>>>"take over" alot of Open Source project in time
>>>(witness GNOME),  but I don't think that will make it
>>>necessarily less open.
>Perhaps so, I see no need to make them anything other than open.
>>This sort of says that the Open Source model is separate from 
>>the business model....
>Yes, this is true, and I think the term "business model" is very much
>abused. I think of Open Source as a DEVELOPMENT model, not as a way of
>doing "business". What Open Source/GPL methods does is obviate the need
>for some methods of conducting businesss around software, such as shrink
>wrap licenses, binary-only packaging, and proprietary controls over pro-
>gress.
>>Maybe this is where my thinking is flawed.  Can a business run an open
>>source project?  
>Why not?

I think the answer is very much yes.  They make a good match.  First, developing software is expensive, if I can share those costs all the better.  Second, does anyone make money selling software?  Not that I ever heard.  You make money by supporting, integrating, training, and selling services pertaining to that software.

>>Maybe the answer is yes, but the product would be something other than 
>>what the project is producing.  
>Most software is not produced as a "product"; most software development is
>in the area of applications, and anyone who Open Sources/GPL's an
>application is probably getting a lot of value from it from elsewhere --
>USING it!

I'd assume in this case that we are talking about companies that are in the software business,  not just companies who happen to develope software to solve there own needs (a very small % of companies).