Kiki Novak wrote:
> Hi everybody,
>
> No, this is NOT a troll. Let me explain. I live in a small south French
> village (Montpezat, 800 people), and recently the mayor fixed a budget to
> get the public library equipped with what they call a "salle multimedia".
> There will be like ten PCs, one server and nine clients, so the people in
> the village can surf on the internet and check their mail, and so on.
>
> I'm currently trying to persuade the mayor to hire me as a (paid) part-time
> sysadmin for that thing, and my arguments go in that direction: take the
> money that you save on ten windows licences, ten office xp licences plus
> all the program licences (multiplied by ten also), _plus_ all the money you
> will save in the future (by not having to buy forced updates and so on),
> and with this money, pay a half-time sysadmin like me that will also
> migrate the town hall's PCs under Linux, _plus_ you get some courses.
>
> Question. I'd like to evaluate the approximate costs of _maintenance_ for a
> ten-PC-network running Windows XP. I know this is an easy joke, here, and I
> also hate Microsoft products with a passion. But what would you estimate,
> roughly? Try to be serious, even if you find it hard)
>
> Or formulated otherwise: how many hours / week or per month to repair the
> system under Windows?
>
> (Down here, we call the MS admins "sneaker admins" because they're always
> running from one machine to the next))
>
> Cheers,
>
> Niki
>
> PS: I forgot to mention. In case it works, my whole south french village
> (public library, town hall, school) would work under my favorite distro)

One vast area that you have not seemed to notice and one that will cost
the village a lot of money, is the time and moneys to train their users.
This is a major area of a.) cost for the village in time (to train users
and time lost being productive until training is completed) and in
money, and b.) a very high cost to abandon what they are using now.

I would suggest that you do a 'pilot program' for them, by taking on
just the server portion, first, and show how much they can save on
server licensing. Once that is done, then look at converting one or two
of the workstations to Linux and do some casual training. Soon, you will
be queried about how much you would charge to finish converting their
entire system over, once they see how much they are *not* spending on
license fees.


My opinion, yours may differ.

--

humjohn AT aerosurf DOT net