Tuesday, March 09, 2004
More Conversations with Eric Raymond, Part 4
We picked up our discussion after I copied him on some responses to Richard Stallman's comments.
---------------------------------------------------------
> Bruce thinks he's in Richard's camp, so the argument
> you think you have appers to be with me.
Eric, you sweet-talker you.
> There are reasons to doubt this; see my analysis in
> "The Magic
> Cauldron". But let's suppose it is true. Predicting
> the consequences
> you wanrt to draw from this would still require
> several steps you
> haven't justified. Here's one: you haven't shown
> that the incomes of
> most programmers are dependent on software sale
> value -- and, in fact,
> this turns out not to be the case, either
> empirically or in theory.
I'm not saying that it is. Yes, MOST programmer
revenue is generated from "custom software
development," which would include consulting. MOST,
however, does not equal all, and I think that the
"open source" community is rather cavalier about
destroying a model that provides a strong source of
revenue for software as such. As I explained in past
emails, there are a number of advantages to strong
revenue generation, such as providing incentives for
developers to concentrate on the needs of
non-technical users.
In fact, if the open source community was more willing
to consider proprietary software, if nothing else for
the revenue, they could operate under a hybrid model
that provides MOST of the benefits of open source
software (access to source code) while availing
themselves of a decent source of revenue to allow them
to continue concentrating on JUST their product.
> No, instead you get the much worse kind of
> fragmentation that
> results from closed-source vendors "differentiating"
> for
> competitive advantage. Crap like this nearly
> destroyed the
> Unix market in the 1980s and explains why moving
> your documents
> from one word processor to another is either
> difficult or
> impossible.
An Unix was open source (in the sense that the source
code was readily available). That isn't the case with
Windows or the Mac OS (except for the BSD core). The
fact that differentiating happens, though, shows why
the market consistently organizes around ONE solution.
It's the reason most of the market organized around
Windows, and the reason most of the Unix market is
organizing around Linux (and even specific distros of
Linux). The software market lacks inherent standards,
and so the market rationally chooses to
disproportionately favor one provider in order to
generate the cost-saving economics of scale the market
needs and wants.
We picked up our discussion after I copied him on some responses to Richard Stallman's comments.
---------------------------------------------------------
> Bruce thinks he's in Richard's camp, so the argument
> you think you have appers to be with me.
Eric, you sweet-talker you.
> There are reasons to doubt this; see my analysis in
> "The Magic
> Cauldron". But let's suppose it is true. Predicting
> the consequences
> you wanrt to draw from this would still require
> several steps you
> haven't justified. Here's one: you haven't shown
> that the incomes of
> most programmers are dependent on software sale
> value -- and, in fact,
> this turns out not to be the case, either
> empirically or in theory.
I'm not saying that it is. Yes, MOST programmer
revenue is generated from "custom software
development," which would include consulting. MOST,
however, does not equal all, and I think that the
"open source" community is rather cavalier about
destroying a model that provides a strong source of
revenue for software as such. As I explained in past
emails, there are a number of advantages to strong
revenue generation, such as providing incentives for
developers to concentrate on the needs of
non-technical users.
In fact, if the open source community was more willing
to consider proprietary software, if nothing else for
the revenue, they could operate under a hybrid model
that provides MOST of the benefits of open source
software (access to source code) while availing
themselves of a decent source of revenue to allow them
to continue concentrating on JUST their product.
> No, instead you get the much worse kind of
> fragmentation that
> results from closed-source vendors "differentiating"
> for
> competitive advantage. Crap like this nearly
> destroyed the
> Unix market in the 1980s and explains why moving
> your documents
> from one word processor to another is either
> difficult or
> impossible.
An Unix was open source (in the sense that the source
code was readily available). That isn't the case with
Windows or the Mac OS (except for the BSD core). The
fact that differentiating happens, though, shows why
the market consistently organizes around ONE solution.
It's the reason most of the market organized around
Windows, and the reason most of the Unix market is
organizing around Linux (and even specific distros of
Linux). The software market lacks inherent standards,
and so the market rationally chooses to
disproportionately favor one provider in order to
generate the cost-saving economics of scale the market
needs and wants.
