Re: [vox] Lugod and public schools

[Don Werve <donw@examen.cMAPSom>]

On Wed, May 22, 2002 at 11:06:41PM -0700, Richard Crawford wrote:
> On Wed, 2002-05-22 at 22:28, Don Werve wrote:
> 
> > It's not quite a "subscription" model; e.g., you can run whatever 
> > you want, with as many clients/installations per site as you want -- 
> > you just pay a few hundred bucks a year for us to keep writing patches
> > and adding in new non-bloaty features.  Support costs extra, and I'm
> > eventually going to try to get to the point where we can do on-site 
> > installation and training, where the end-user pays for us to come out 
> > and spend a few weeks training people to set the thing up and use it.
> 
> That's kind of the model I had in mind.  :)

Well, I'm hoping to make it a reality in six to twelve months.

> > Even at $5K a week, it's *still* cheaper than an MS kit; look at the 
> > five-year cost of outfitting workstations with Windows, Office, 
> > Exchange; each workstation has a yearly software cost of about $300 
> > (with discounts), plus server costs (about another $200 per user per 
> > year), endless hardware upgrades to cope with load, and the need to 
> > have a horde of NT admins.  Comes to something like $140K a year 
> > more than a comprable Linux solution.
> 
> $5K/week is roughly $260,000 per year.  Is that really cheaper than an
> MS kit?  I ask honestly, since I've never maintained a full MS shop.

That's for training!  It would cost about $5K a week (per trainer) to fly 
a couple of trainers in to teach a couple of admins how to run the software,
and to run some short classes for the users.  That's for about two or 
three weeks, and then the only ongoing cost should be the maintenance fee.

Sheesh. ;)

So instead of paying $BIGNUM up front for licensing costs, and then 
worrying about finding admins, it'd be $350 for the site license, plus 
around $20-$30K to have faculty trained to use and run the software, 
plus having *real engineers* help the people installing the software,
so it gets done right the first time.

The update system I'm building is modular and automatic (think apt-get), 
and I'm designing things to work around the Unix/X idea -- a powerful 
GUI wrapped around a powerful CLI.  So you can configure it by editing
a text file, or by using a configuration tool.  I'm also designing it 
to use a modular back-end, optional SSL, and plan on doing something 
any business type will call stupid as hell -- while I'm not going to 
GPL the code, I plan on letting users have access to it[1], and I also
plan on *not* charging for private, non-commercial use.  So groups like 
LUGOD can use it for free, as can organizations like the Red Cross and
such.  They're really outside of the target market, and I'm putting this
together to see if a company *not* intent on screwing its customers can
stay in business.

It's an idea; it'll either run, or fall on its face.  Either way, I'll 
have fun, and I'll have an even longer resume, so it's not like it will 
hurt me...

Oh, and I'd say that running an MS shop with a thousand users (the average
school) would cost more than $260K a year.

> > Which is why I'm spending my free time coding. ;)
> 
> Perhaps that's how I can justify spending all my spare time on my Linux
> box to my wife: "It's for the people!"

*grin*

[1] There's already copyright law in place that protects me here; if 
    some enterprising individual decides to steal my code and market it,
    there's enough evidence (javap) to litigate them into the ground.
    I don't understand the logic of needing to hide the code when it
    is already protected by both law and mindshare.

-- 
Don Werve <donw@examen.com> (Unix System Administrator)

Yorn desh born, der ritt de gitt der gue,
Orn desh, dee born desh, de umn bork! bork! bork!
_______________________________________________
vox mailing list
vox@lists.lugod.org
http://lists.lugod.org/mailman/listinfo/vox