OT: SCO Forum

[Brian K. White]

----- Original Message ----- 
From: "Fairlight" <fairlite at fairlite.com>
To: <filepro-list at lists.celestial.com>
Sent: Thursday, June 22, 2006 6:41 PM
Subject: Re: OT: SCO Forum


> >From inside the gravity well of a singularity, Bill Campbell shouted:
>>
>> There's nothing cheap about my losses on the Caldera/SCO IPO :-).
>
> Shoulda gone for the RH IPO.  As I recall though, you hate RH.  :)
> Invitation was based on whether you'd submitted a bug to their bugzilla.
>
>> Obviously you think that spending the time and resources to attend it is
>> worth it for Microlite's business.  Certainly you see a much wider 
>> picture
>> of existing SCO users than I do.
>
> "I'm shrinnnnnkinnnnnnggggggg!"
>
>> I have a hard time understanding why developers would develop new 
>> products
>> for SCO systems, given SCO's past history, destroying their network of
>> Advanced Education Centers and Authorized Resellers, suing their own
>> customers, and seemingly adopting a litigatation vs innovation strategy.
>
> "But it's easier to use than Linux."  I actually had someone tell me that
> the other day.  Sure it was.  Until they wanted the latest (ie., latest
> -developer- version, not SCO-released version) apache, PHP, etc., ad
> nauseum.  Then it costs them.  And actually it's not easier, as they still
> "need" (their word, not mine) Webmin to administer sendmail--what they can
> admin of it.
>
>> I find I can provide systems running on Linux, FreeBSD, and OS X far more
>> easily than I can on SCO.  If SCO systems provided significant advantages
>> in reliability or performance, it might be worth the extra effort to use
>> them, but I certainly haven't seen any reasons to do so for the last 8
>> years or so.
>
> If SCO wasn't a complete nightmare to port to, it might have advantages.
> They must have -really- had to bastardize perl to get it working on OSR5.
> I tried that compile (which I've done many times on 3 other platforms,
> including Solaris, with zero hitches) and the first time it wouldn't
> compile.  The second time I finally convinced it to compile but it failed
> 84%+ of its own tests.  I gave up.  And you can't use most XS modules with
> SCO's version.  They fail to load the blib files.  I -think- it's because
> they compile with their own compiler, while the modules are compiled using
> gcc (a vastly outdated gcc, I might add--and one that's buggy).  Their
> devkit isn't even worth getting, as it's not ANSI, and so many things 
> these
> days require ANSI C that it's just...well, pointless.
>
> If -all- you're doing is running a canned app or five and a few shell
> scripts, it might be okay.  I wouldn't use it for anything I actually
> needed to get something non-canned done on though.  It's horribly
> time/cost-inefficient when trying to use OSS software.
>
> mark->

Finally, even I agree.
(Well, not with everything. SCO's native devkit is perfectly ansi. Inability 
to get something "ansi" to build, even one that builds on sun, means 
nothing. Even the odd missing or mutated function or call that everyone else 
seems to have means nothing.)

It's a long story, and I stand by my, uh, stance, while I was was still 
behind SCO, but finally things just don't add up anymore even to my 
BS/popularity/fad/speculation rejecting sensabilities.
Not always for the same reasons depending on the product, but always it's 
less practical than using Linux even though most of my arguments against 
Linux are still just as true as ever!

I wrote and re-wrote this mail 3 times trying not to write a book backing up 
this statement but no matter how I try to summarize it comes out huge, so 
"it's a long story" and my promise that there is lots and lots of thinking 
and first hand experience and observations behind it will have to suffice. 
This list is just not the place for that dissertation.

Brian K. White  --  brian at aljex.com  --  http://www.aljex.com/bkw/
+++++[>+++[>+++++>+++++++<<-]<-]>>+.>.+++++.+++++++.-.[>+<---]>++.
filePro  BBx    Linux  SCO  FreeBSD    #callahans  Satriani  Filk!