According to the open source movement [this article is concerned
with open source, not free software (although free software
certainly shares some of the funding problems), to clear up any
ambiguity], considering a piece of software someone has written,
one should not use it unless one has the source code.
That's not strictly true. If that is a view of some open source
advocates, it is not the dominant view. The dominant view is that you
are better off, overall, using an open source application, because
someone can fix the application. Of course that may be you,
or the original developer, but it might also be the clever kid next
door who you let use your playstation 2 on your giant TV since he can
always sort out your PC for you.
The reasoning is that if one has a problem with it, one cannot resolve
the issue without outside help. As far as I can see, although this
is certainly a distinct advantage for say Google, who with a staff
of highly trained engineers could easily tweak the Linux or BSD
kernel to suit their requirements, its advantages in ensuring
quality and reliability are far from assured.
True enough. Although I disagree with your views, I know that
success for Open Source is not asured.
For example, in propounding the open source solution in John Goerzen's
paper on the ethics of free [open source] software he says that
the famous case of the USS Yorktown, that the 'problem behind all
this is proprietary software'.
This claim is one that Mr. Goerzen fails to adequately
establish. His arguments can be summarized as follows:
lack of peer review means that closed source software is intrinsically untrustworthy
the 'fact' that closed source means knowledge is not shared, something he says is unethical
By contrast he argues that from utilitarian grounds open source is
better insofar as it tends to maximize the sum total of happiness,
and, most specifically that 'free software is the most beneficial
for the greatest number of people.'
To consider his first argument, namely that the absence of peer
review makes closed source software untrustworthy, I would argue
that in fact peer review is *more* rather than less common with
closed source software. To take an example, Microsoft operating
systems typically spend upwards of a year in external testing,
whereas open source software tends to follow Eric Raymonds's
famous Bazaar principle, where software is released little and
As others have already pointed out, you missed the point here. If I
was the top GUI expert at Microsoft, per review of my code would need
to involve my peers reviewing my code. Testing is
not peer review.
The difference between the two can easily be seen. Anyone who
used an open source OS and GUI environment, simply by clicking through
each option. In my experience there would also be a considerable
number of software crashes.
This depends on which applications you're talking about. I do have
one application that is unstable on my main machine - freeamp - and I
know why (I'm using slightly out-of-date versions of some libraries it
depends on). Also of course we know that Mozilla isn't mature yet and
Netscape has had some problems. Other than that though my Slackware
machines are quite stable (the gateway has an uptime of 99 days as it
happens, which is a mail, shell and webserver).
By contrast my girlfriends Windows98 machine has had a lot of
problems (I have had to use the restore CD twice to reset the hard
disk), and my WindowsNT machine has crashed out completely several
times this year. My experience is not at all uncommon.
Here you don't prove any kind of point, merely giving us one data
point which is unconvincing when generalised into a sweeping
conclusion, to say the least. For us to accept this kind of trash talk
about open source software being generally crashy just with a little
clicking, you would have to find some evidence. In contrast, I would
probably want to cite some
of these links.
There are a number of reasons for this as I see it:
lack of money
Since the open source movement is associated with software
that is without price, there is little money to fund fulltime
programmers, marketing to attract new people to the project,
or commercial testing.
An opposing view :
Since the open source movement has a low barrier to entry and
immediate feedback it pulls in increasing numbers of highly talented
developers who start doing it as a hobby and end up finding jobs where
they can do it full-time - Larry Wall, Brian Behlendorf, Alan Cox,
Steven Tweedie, Donald Becker, Ted T'So, dare I say Rusty himself
pretty soon, these are just the ones you and everyone else here will
probably have heard of. I wouldn't be surprised if the free software
community has more programming talent working in it, right now, than
Microsoft has ever had (and I say this as someone who is fully aware
of the might that Microsoft brings to bear on problems when it wants
lack of direction (i.e. the ability to be able to say: 'Right,
you get that bit done or you're fired')
No! That's one of the best bits! I'm sure if the programmers at
Microsoft fixed the stuff they thought was important Windows GDI
resource leaks would no longer kill their mainstream consumer
For example, let us consider one of the top open source games, Freeciv, and its nearest commercial competitor, which is probably Alpha Centauri. In the
making of Alpha Centauri, the software house would work something like this:
'We need x programmers, x video guys, and x voiceover artists.'
They will then hire those staff and the product will be produced. By contrast, the free equivalent works on a haphazard basis whereby that which is produced
is determined by those people who happen to volunteer for the
Or we could consider Daikatana. A-ha, a-ha, ha-ha-ha-ha-ha-ha.
Thus Freeciv is without any sound effects, video, etc., and also
has inferior graphics, all of which detract from one's enjoyment
of the game (not to mention that it exhibits one of the major
problems with open source, namely lack of innovation).
By contrast we have a game using out-of-date game engine technology,
badly, with inept plot devices, which is years late, has burnt
millions of dollars in investor money for no apparent reason, and
exhibits only tiny innovations over games like Doom and Quake.
Even so, I have to concede that games is one of the last frontiers for
open source, I don't know if it will ever overtake the best commercial
efforts. Nor do I feel my freedom threatened by market dominance in
pure entertainment like I do by market dominance of key strategic
enabling technologies like operating systems and network protocols.
is my contention that open source is a fundamentally incorrect
model for software aimed at the consumer.
Characteristics of the consumer:
- little or no programming knowledge (and therefore the
so-called advantage of having the source code is no such
- low tolerance of technology for its own sake
- little understanding of computers
For them, open source software holds no benefits compared to the
leading commercial equivalents from Microsft and Apple.
This is obviously wrong : consumers can get precompiled linux software
at low prices which is often more than adequate (web, email,
word-processing, simple photo editing etc) so just on price alone the
option of using open source software gives them more options. Now, the
fact that the source to these things they use is open gives them even
more options. Options are good. Apple obviously agrees. IBM, Sun,
Compaq, SGI all obviously agree. It seems that Microsoft is the odd
man out here.
As such, the consumer Linux
distributions I believe are doomed.
Open Source != consumer distributions. This is a bit of a straw man
Even if the idea of pre-packaged open source operating systems down
at Fry's or Wal-Mart fails, that really doesn't tell us much about the
whole movement. Slackware for example used to be just one guy working
for Walnut Creek CD-ROM, and even now is only about four I think
(Patrick, David, Chris and Logan). It's pretty hard for Microsoft to
think down to that level. Very little promotion or advertising, mainly
mail order products, yet here you have a Linux that works for a medium
size server, a desktop, all kinds of things that the commercial,
proprietary competition would like to charge you hundreds of dollars
for. Even if Red Hat and Mandrake and TurboLinux and SuSE all went
bankrupt simultaneously the community would pick up the major pieces
and I'd continue to get a good OS for my needs very cheaply. Pretty
The problems are:
insufficient funding. Open source businesses typically depend on
business models that stand no chance of ever making any money.
Not only is this wrong, it's semingly ignorant, so I'll just try to
Because of the
mistaken believe that making money out of software is somehow immoral (a bizarre belief, considering that everyone must make money to survive), they
rely on 'donations',
You must be talking about the FSF here, but not any of the other Open
Source businesses I've ever heard of. It certainly would be a bizarre
belief for businesses, but only the FSF (a charity) is
founded on it. Other companies, say, Loki for example give away some
stuff to inrease demand for their products. So does Microsoft.
on selling services, and on limited and voluntary sales of products they could download for free. Although to a certain extent the
market has wised up to this, as seen by the fact that Corel's Linux division was sold for a miserly £5million, I still believe that businesses like
thekompany.com, and Nautilus, which rely on selling vague
services or on giving the core product and charging for
addons, it still persists.
No, most open source companies are effectively consulting
companies. They make money by doing things that customers ask
them to. A few also try to make tangible products. None of
these believe it's immoral to charge money or make profits, so
I can't possibly agree with this either.
It is unfortunate for open source that this socialist tendency
persists so much -
I don't mean to offend, but your argument is so far into the
weeds here it's difficult to want to address any more points,
but I'm going to try and just put across some well-known
rebuttals. Not even the FSF is socialist. Richard Stallman
foresaw a world perhaps there are less jobs overall for
programmers, because of a reduction in totally wasteful
duplication of effort, but for those that did work they would
be prized consultants working with a growing body of shared
knowledge and competing by skill against their colleagues.
Once you move away from there and look at, say, RedHat or
TurboLinux, I don't discern anything vaguely resembling
Microsoft would not be able to afford produce the world's best word
processor if they had given Word away and just charged for the thesaurus.
Donald Knuth produced (arguably) the world's best typsetting engine
and didn't even charge for the fonts. Just one counter-example.
Still further, the belief that making money out of software is
somehow damaging is fundamentally misconceived.
Yes, but almost nobodyhas that view.
While closed source software has
grown up, so to has the economy - high software spending is a
*good* thing, not bad.
I don't think this is true. How does it improve the
competitiveness of, say the USA vs. China?
The massive growth in the economy has been fueled by commercial companies making money, whereas open source ultimately aims at making all
software 'free', which would undoubtedly be harmful.
"Undoubtedly"? More like completely open to debate at this point.
inadequate product - whereas commercial companies such as Microsoft have armies of people employed in usability testing, the fact, as explained
above, that open source can *never* match the resources of
means that the product will never be as advanced or as easy to use as the
paid-for alternative [note that there are certain circumstances where open source can compete].
The common reply to this is that absence of resources is not an impediment, since open source depends on volunteers, but this makes the fundamental
assumption that there are enough people who would rather make software for free than make money making commercial software.
Plenty of projects show the power of all volunteer unpaid labor,
however a lot of open source is now being written by people as
their day job (e.g. at places like IBM, Sun and SGI) which is
not something you seem to consider at all.
the pool of volunteers or the quantity of their free time will
never be large enough to build a 'complete' open source
Not true, see above
as explained above, commercial companies producing open source are not typically viable, and so do not have anything like the resources of the
commercial sector with which to compete.
Sun, IBM, Compaq, SGI, RedHat, TurboLinux ...
the lack of money and commercial incentive means that open
source produces very little innovation, and so is always
You would not have ex-Apple engineers forming Linux startups to do
better desktop if they didn't think innovation wasn't viable
in open source.
Having, I believe, debunked the myth that open source can ever
produce a sustainable and complete consumer software ensemble,
well, no, I don't think so, but even so consumer software is
not the be-all and end-all of open source nor closed -source
I return to one of the first
arguments made, namely that closed source impairs does not allow people to learn.
This is a very flimsy argument, and I would in fact argue the reverse - at present colleges and learning schemes are heavily funded by profit-making
businesses, but if open source succeeded these businesses would be redundant, which would in fact cause even greater damage to learning since this
funding would stop.
I don't know what to make of this - I don't even understand
what you are getting at
Furthermore, the net result of this would be that people would be discourage from software production as a career, since it would no longer represent a
profitable career path, and so they would probably pursue a career as a doctor or a lawyer. This would be a great loss to the nation, since the quality of
software would decline, as highly intelligent students would go
It is true that open source might create many less overnight
millionaires a la Microsoft. If such people really became doctors
instead that would not be a great loss to the nation however
In conclusion, I'm not arguing necessarily that open source is
always necessarily inappropriate, but rather that for consumer
software it certainly is.
I disagree with this, and whilst my rebuttal is pretty rushed, I
think I've made it clear how completely you have failed to prove your
point, at least to me.
In more specific cases, it might present a useful solution - for
example, for high-end military applications or servers maintained by
experts there are certainly advantages to an open system; however,
these cases are relatively restricted - since I see little commercial
potential in free software, these have to return to the roots of open
source - to the limited number of highly dedicated hackers producing a
small range of software (such as Unix kernels). It is here that there
can be union between the two opposites - a movement that believes in
free software, and those who make money out of it. Thus OSX represents
a good example of the sort of project open source is ideally suited
for - a defined Unix kernel is the ideal project for open source, in
that it requires relatively few resources other than programmer time.
I have to ask, do you actually have any open source yourself?
I wanted to do a good job demolishing this piece, but without being
rude or unfriendly. In the end I found the piece itself so fractured
and difficult to follow (I must admit I hadn't finished reading it
when I started replying) I fear my rebuttal adds "more heat than
light". Also I know I've come over as hostile in places. Well I
apologise for that.
Clearly when you scratch me, off comes the placid live-and-let-live
covering and out comes a bit of a free software zealot, so I learnt
something here. No doubt by now someone has written half as much and
said twice as much. Oh well...
p.s. sorry if I mangled any quotes or formatting
Chris Morgan <see em at mihalis dot net>