Moving from SCO to Suse Linux

[Fairlight]

At Thu, Jul 29, 2004 at 02:34:05PM -0700 or thereabouts, 
suspect Enrique Arredondo was observed uttering:
> >
[John said]
> > Good try... not going to happen. If you want another copy of filePro on
> > another platform, it is going to cost you the price of that new platform.
> > Anything else is sort of like saying, I'm done with my Ford now I want a
> > Pontiac, won'tyou give me one for $1?  :-)  Like I said good try.

I politely disagree with that.  I mean, if someone bought something I
had that they were using on SCO, and all they're doing is -switching-
platforms, not running two separate servers, I don't see the problem.
If they're upgrading versions, they should only have to pay the upgrade
fee that would occur with an upgrade on the same platform.  If its the
same version, just a different platform, maybe a $50 fee for handling the
request and the attendant paperwork.  If I'm reasonably certain they're
simply migrating platforms, I fail to see where they should be gouged.
They could be migrating from 'doze to *nix after wising up.  Should they be
penalised for gaining IQ points?  I think not. :)

To borrow your analogy, John, I see it really as no different than taking
the custom 150w 10-disc stereo out of your Ford and putting it in your
Pontiac.  (Likewise with hubcaps, radar/laser detectors, whatever...it's an
add-on accessory.)  Yeah, you'll have some labour charge to reinstall it,
but you -don't- have to go buy a whole new stereo itself, just because you
changed what car you installed it in.

I suppose it could be said that I'm looking at it differently because I
currently deal with a pretty portable language.  But even for programs
that I have compiled for specific platforms--if someone wanted to just
migrate, how can I justify more than a token fee for handling the request
and giving them access to the new files?  Assuming I trust my customer to
not be ripping me off, I don't see the problem in supporting a migration
path if I have reasonable assurance that's what's going on.  I see it as
providing superior service.

> If I move to 5.0.7 it'll cost me about $4,000 for licesing 256 users (or
> more!) because all my current licences are not carry on style. As for Linux
> it's about $1200 upto 32 CPU's with Suse and about $800 with REDHAT.

How many CPU's do you actually have, Enrique?  I mean, SuSE's costs for
enterprise are basically for support.  If you're putting together a huge
system with 32 CPU's, sure, go for it and hedge your bets on the chance
that something pops up.  If you're putting together a quad CPU machine
or close to it, heck, get the $79 Professional 9.0 and have done, unless
you'll need their support in any event.

I mean, just what are we talking here, firepower-wise?

Regardless, I'd go with SuSE over RH.  The product is better engineered,
and at least SuSE gives a damn.  I spoke with RH about their Enterprise 3
product's totally buggy version of perl that actually -crashes- (which they
claim to have known about partially already--they think...they couldn't
swear it's the same issue but thought it was).  They said they were
escalating it, and would let us know.  That was...oh, about a month ago,
probably a little more.  Haven't heard a thing, and they haven't released
an updated perl.  If RH can screw up something as integral to their system
as perl (look at the dependancy list sometime) and let it ride this long
without caring, I don't -want- them.  As much as they're reaming people,
it's inexcusable.  The client -has- paid support for the ES3, and it
hasn't done them a whit of good in that respect.  Yes, it -can- be worked
around, but it shouldn't -have- to be.  And it's for that reason they'd
have to pay -me- to -take- RH's current ES3 offering.  I don't trust it.

I haven't fully trusted anything they've done after RH 7.3.  After having
delved into more of the internals, ripping apart their SRPMS to use their
configurations to use to seamlessly integrate upgraded packages for
security and features after EOL (duct tape and baling wire time), and
seeing what a mess they made of things, I'm surprised their stuff didn't
break far more than it actually did, even back then.

> Oh man, If I could get a key generator for OSR507 and make it 512 licenses
> that's the only way I could definately wanna use SCO more over Linux.

Probably not the wisest thing to say publicly, in light of the current
climate and goings-on in the industry.   Or at least not the wisest way to
phrase your point.

mark->
-- 
Bring the web-enabling power of OneGate to -your- filePro applications today!

Try the live filePro-based, OneGate-enabled demo at the following URL:
               http://www2.onnik.com/~fairlite/flfssindex.html