[long] Re: Legal traps in open source

[Simon Fowler]

On Wed, Oct 30, 2002 at 03:49:10PM +1100, David Gibson wrote:
> The trouble is that software engineering at this point is simply not a
> comparable discipline to civil engineering.
> 
> Take most books on the subject and you'll find that they're a mix of
> obvious platitudes and waffle so turgid and vague as to be completely
> meaningless.  With a bit of stuff thats Just Plain Bollocks for
> flavour.  If you're very lucky and it's one of the better books it
> might have a handful of good ideas or techniques mixed in with that.
> Of course, it might also take one pretty good idea, pursue if beyond
> all reason ("hey, you don't need intelligence, common sense or
> contingency plans if you just Embrace The Process") to the point that
> it is counter-productive.
> 
> Maybe one day software engineering will resemble real engineering, but
> personally I doubt it:  abstract things are very different from
> physical things.  It shouldn't be surprising that building them
> requires a different approach.
> 
I think this is the most important point of all here. Software
engineering just plain /isn't/ engineering, no matter what all the
people who teach it or who've done their time^Wdegree and are making
money out of it say. The best that can be said for it right now is 
that it's a craft - in the same sense as the old guilds and the like. 

If software development is going to turn into a real engineering
discipline it's going to have to grow up. Dropping the stupid
faddishness, Language of the Month crap, One True Design Technique
of the Month, and so on and so forth, would be a very good start
down that road. Growing some balls and taking some real
responsibility for your work would be an even better place to start. 

One other thing to note is that if you use the same kind of skills
heirarchy in software as you do in a traditional engineering
project, you're going to end up a steaming pile of shit: a 'code
monkey' who can write /good/ code will have the same level of skill
as the 'software engineer' who can come up with the overall design.
Trying to treat one as a highly skilled profession, and the other as
a lowly "you can type, you've got the job" type thing is just plain
stupid. If anything, that's going to be what stops software
development from becoming a real engineering discipline: engineers
are really expensive, and if you have to have them doing /all/ the
work, the total cost will get insane. So I can see things going
along as they are now until liability costs get so high people
/have/ to get their heads out of their arses and work out how to
make software engineering /real/ engineering. But probably not in my
working lifetime . . .

Simon

-- 
PGP public key Id 0x144A991C, or http://himi.org/stuff/himi.asc
(crappy) Homepage: http://himi.org
doe #237 (see http://www.lemuria.org/DeCSS) 
My DeCSS mirror: ftp://himi.org/pub/mirrors/css/ 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 232 bytes
Desc: not available
Url : http://lists.samba.org/archive/linux/attachments/20021030/8ca55377/attachment.bin