[Gllug] "Open source has its own problems" - article in Computing

Nix nix at esperi.org.uk
Fri Aug 5 21:35:59 UTC 2005


On Thu, 4 Aug 2005, Benedikt Heinen gibbered uncontrollably:
>> Conceptual integrity - good software needs a single designer with a
>> clear vision, and this can't happen with OSS
> 
> I wouldn't go as far as saying it *can't* happen with Open Source, but
> usually I would think it's true (simply because if there are too many
> people just quickly thinking of something they want to patch on - I

Well, ugly hacks generally won't get past the project's upstream
maintainers in ugly form.

The fact that upstream can last for as long as it is interested often
means conceptual integrity *far* exceeding that of proprietary software,
where people may have to stop working because they've moved house,
because the company has relocated, or for any of a myriad of reasons.

> doubt they're going to read through "architectural documents" of the
> software in question before doing the patch.  (That is - if there even
> ARE documents about the architecture used).

You don't need documents for architectural vision: you need the
visionaries to *still be around*, or for the vision to be simple enough
that lots of people grok it (e.g. the philosophy behind Unix).

>> Professionalism - he likens the OSS movement to the games industry of
>> the early 1980s, where apparently bedroom coders produced really bad
>> games which nearly ruined the whole industry
> 
> It can be that way - it depends on the project - and it it's in a
> sense certainly more difficult to achieve in open source. In

Why should we *care* if the authors arne't `professional'? By definition
those who aren't working on it for money will be working on it solely
because they want to (or need something).

Would you have stopped Andrea Arcangeli from working on the Linux MM
system when he started (aged 16 or 17, I think)? After all, he wasn't
a `professional' when he wrote what became the Linux 2.4 virtual
memory manager...

> commercial environments, you only hire people that will work with the
> team (or ditch them again, if they don't work out) and so you can

Er, if someone can't `work with the team' (which I interpret as `piss
people off through abrasiveness') then presumably fewer people will
accept their patches, and eventually they will drop away or learn some
manners.

(Counterexamples exist, generally abrasive maintainers...)

> streamline better.  On the other hand again, in the commercial world,
> projects are often driven by people with little or no knowledge of the
> development process, and hence face "political" problems the kind of
> which are rather rare in open source.

i.e. badly-thought-out or just plain insane designs inflicted on the
code for no good reason? Been there, watched good source trees suffer
accelerated rot, never seen it in the free software world :(

>> Innovation - OSS is mostly just rip offs of proprietary software
> 
> This works both ways - there was open source stuff that made it into
> mainstream, as well as closed source being "copied".

Major examples include DNS (BIND), mail service (sendmail), X11...

-- 
`Tor employs several thousand editors who they keep in dank
 subterranean editing facilities not unlike Moria' -- James Nicoll 
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug




More information about the GLLUG mailing list