[Noisebridge-discuss] City of Berkeley explanation of why not using open source.. Open Source vs. "proprietary" software

[travis+ml-noisebridge at subspacefield.org]

On Tue, May 31, 2011 at 08:18:53AM -0700, travis+ml-noisebridge at subspacefield.org wrote:
> > > City programmers are competent at creating small,
> > > specialized applications for those needs that can't be filled by commercial
> > > software -- customizing databases and operating systems exceeds their skill
> > > set.
> > 
> > Again, the idea that the Rent Board needs to write a custom database
> > engine to use an open source database is a total red herring. Someone
> > is feeding them misinformation.
> 
> I think they're talking about webapps, but confusing it with databases
> and operating systems.

Sorry to ramble so much, I know I'm all over the place.

There is a certain interesting point here.  If you have the ability to
create a custom application for some city government at a job that
pays $75k/yr, you probably have the ability to work on commercial
software for a company in the valley and earn $100k/yr+, so they might
not be able to retain that talent.

To hire such talent, they'd have to be a judge of technical competency
without relying on credentials such as experience and certifications;
you know, grab them right out of college.  They may not have the
ability to tell a good programmer from a bad one.

As a result, they end up buying solutions and not requiring much from
their employees, which means the job is boring and doesn't attract
quality talent, forming a sort of self-fulfilling prophecy.

Such shops are probably more likely to be comfortable buying a
solution from some company who takes care of all of this, and simply
hiring sysadmins, troubleshooters, and system integrators.

I know I've seen that even at the federal government level, so it must
be more true of local and state governments.  There's a sort of gap
for talented people between "just graduated, no experience" and "bay
area programmer making six figures" that is a sort of no-man's land.

Note that one could get around this by using interesting technology
that isn't quite mainstream enough yet.  For example, I've found that
apps written in python are about ten times easier to modify than those
written in C/C++, if for no other reason than they're ten times
shorter.  Probably this is more true of, say, Ocaml.  You could
possibly retain your Ocaml programmers since there aren't many
companies hiring them :-) Since the problem domain isn't interesting,
perhaps by using interesting technology, you could manage to still
attract and retain quality talent.

However, the more I think about this, the more I kinda think it won't
work, despite my best hopes.  Governments tend to be more conservative
and risk-averse (since they don't often turn a profit, so why should
they care about possible gains - they just want to avoid mistakes),
and that doesn't attract the kind of people you'd need to make this
work.

However, you could possibly swing it by somehow involving the
university.  Maybe a programming class could write the relevant
software as a class project, and they would have the chance at hiring
some of them to come and continue development.  That'd be a nifty kind
of exercise that wouldn't require city funds or buy-in to get started.

BTW, Garter had an interesting study on the ROI of F/LOSS recently.
(I don't have a link but one of the sections was titled something like
"there's no ROI in spite").  That has some authority that an individual
wouldn't have.

Okay, I'm gonna stop rambling on this thread :-)
-- 
http://www.subspacefield.org/~travis/
He who lives by the computer, dies by the computer.
If you are a spammer, please email john at subspacefield.org to get blacklisted.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 834 bytes
Desc: not available
URL: <http://lists.noisebridge.net/pipermail/noisebridge-discuss/attachments/20110531/5dca59da/attachment-0003.sig>