Request for Literature: How come it’s so hard to explain Open Source to institutions?

My guess is that accountability is one of the big issues – not so much whether support is good or not in practice as to who is responsible in case of failure. They need “one neck to throttle” and with a lot of OS software it’s not clear whose neck that is.

But then you mention this in HCE re: perl, so I guess you already know that.

As an aside, my own experience with OS software is that “simple to set up, and easy to use” is not universally true – I’d say it’s not simpler than closed source as a rule.

…and with a lot of OS software it’s not clear whose neck that is.

It’s always the CIO’s neck. If the CIO picks whatever is recommended by the forrester research, his neck is more or less safe whatever the failure; and if he picks a cheap but good alternative – he’s sitting on a barrel of gunpowder.

abb1’s refinement of Tom’s argument is right on the money here. It’s simply “no-one got fired for buying IBM” syndrome; the main change in the industry over the last 30 years is that Accenture, EDS and CapGemini have joined IBM on the list.

It’s simply “no-one got fired for buying IBM” syndrome

But they should because Lotus Notes sucks.

Oh, come on. This post might well have been “How come those stupid old stuck-in-the-muds don’t GET IT that there’s a wonderful new era Gift Economy where all is roses and unicorns. Can anyone point me to literature explaining why they’re soooo dumb?”
(“Free, easy, and good confuses them”)

Bleh. I’m a big advocate for open source software. But if you think it’s somehow sneer-worthy that “‘Just try it and see how it goes’ is actively upsetting”, well, that’s Kool-Aid pushing, not academics.

Open source software has been produced by a lot of people who haven’t been concentrating on their work. They have either been doing it on their employer’s time, or at weekends when they should ideally have been thinking up new ideas for their employer. Choosing open source signals employees that this behaviour is acceptable, even desirable. It’s the loss of psychological control, the Panopticon complex. Clever employers like Google and Pixar see that it’s in their interests for employees to “get a life”, but they are very much a minority as yet.

Seth, you’re a good guy, but jeez, take your meds. What clay is pointing out is that there are alternatives to the standard pay-mucho-dinero-to-get-shafted system, and that corporations don’t seem to be that intersted [whether he’s right or wrong on the second point, I don’t know].

As for experimentation, it can be done, and is done every day by corporations. And as others have pointed out, hiring BigNewSoftCorp, Inc. to install very large corporate systems is a serious gamble. How many corporations have pissed away how many billions by having very large makeovers of IT systms in the past two decades?

Simple. People naturally assume value is correlated with cost.

As a consultant who implements systems based on open source technology, allow me to suggest that you hire a consultant who implements systems based on open source technology.

Seriously, though: downloading a tarball and handing it to IT is not the same thing as finding a vendor. The people who are wary are, to some extent, right to be wary. Hiring a professional to handle the project means you may not save as much money as you’d been hoping, but provides a level of support and institutional attention that frequently can’t be attained in-house. There are still good reasons to go with an open source technology when hiring an outside firm, though — not the least of which is that you can find other vendors/contractors to support it down the line should you want to stop using the original one.

If your application really is as simple and unimportant as a WordPress site, though, then it may be appropriate to undertake configuration in-house. My suggestion in that case is to pretend that getting the site live is an urgent priority. Stand up a simple FOSS installation — then present a plan for evaluating proprietary systems and, if necessary, transitioning to them. There’s no easier way to convince people that FOSS is sufficient than showing them.

Slocum at #6: Risk aversion and path dependence are not the only reasons for the prevalence of: “Nobody gets fired for buying IBM”. There are two other good reasons, neither of which involves the belittlement of the intelligence, capabilities or judgment of CIOs.

1. Major purchase decisions are often (perhaps almost always) made under time pressures and in conditions of uncertainty. In such circumstances it can be cost-effective, efficient and rational to simply choose whatever option most of your peers are currently choosing.

2. For many (again, perhaps most) products there are network-good effects, since fashion comes into the buying and use of nearly all products (even raw materials commodites). When these effects exist, the utility of one purchaser’s choices depends on the choices made by other purchasers. (Note that this is something different to path dependence.)
Once again, it can be rational for a decision-maker to choose what other people are currently choosing.

@6: …trade journal stories about other companies doing the same things for much less…

But it’s a huge hurdle to overcome without a capital, lobbying infrastructure, etc. Opensource can’t bribe them, so, I’m afraid, the only solution is to discredit them, to intimidate them, to get rid of them.

It may be worth noting that the same effect can also be found between expensive and cheap comercial software as well. In a large organsation managing to convince everyone that a $200 SW is as good/better than a $5000 (or $50 000) can be very difficult. There are two reasons for this (that I know of, there are probably lots more)
Firstly the $200 SW does not come with a sales organisation that does all the stuff that is expected/wanted: Answering RFIs and RFQs in detail, coming to several on-sites visits, doing test installations etc.
Secondly, and perhaps even more important, the feeling that a $5000 SW must be better is very human and difficult to overcome, especially in large organsations where the actual decision is far removed from the users.

Re: #23:
(from http://www.joelonsoftware.com/articles/CamelsandRubberDuckies.html, discussing how to price software, if you’re selling it)

“Software is priced three ways: free, cheap, and dear:

Free. Open source, etc. Not relevant to the current discussion. Nothing to see here. Move along.

Cheap. $10 – $1000, sold to a very large number of people at a low price without a salesforce. Most shrinkwrapped consumer and small business software falls into this category.

Dear. $75,000 – $1,000,000, sold to a handful of rich big companies using a team of slick salespeople that do six months of intense PowerPoint just to get one goddamn sale. The Oracle model.”

Clay –

I think you are conflating lots of different types of software in your most recent comment. What you say applies to stand-alone, single-user applications without any consequences for the organization’s mission or processes, and without major interactions with legacy systems and applications. But these are not properties of most major commercial systems. Exactly how does Boeing “try” a new enterprise-wide logistics system before deciding to use it? Exactly how does AT&T “try” new network switching management software without actually deploying it? Even the mere process of installing a complex, multi-user, multi-purpose, multi-site, mission-critical application in order to adequately evalute it may take a team of a score or more people to plan, let alone try.

This behaviour is not risk-aversion or stupidity on the part of CIOs, it is rational commonsense. Corporations have good reason to be cautious about messing with their mission-critical systems and operations.

Clay, let me semi-apologize because this kind of stuff sets my teeth on edge, and that’s particularly relevant here because I’ve been a hardcore Linux advocate as a programmer. So I’ve run into this situation personally. The sort of la-la land implicit above is the easiest way to lose. You didn’t say “all is roses and unicorns” either. The point was your post implicitly assumes Open Source systems are so obvious better in a broad sense that any organization not rushing to immediately convert at all costs must be staffed by fearful timid corrupt dullards.

I know that sells on the conference-club. But it’s really irritating the next time I have to deal with a manager and explain no, it’s not about some sort of anti-corporate ivory-tower ideology.

To amplify what peter just said at #27, running a large project in corporate management is very different from what’s valued by a perpetual-beta early-adopter mentality.

The class of dilemma I am interested in is when your application is as simple and important as a WordPress site.

The question is one of leverage—software is important for its effects, so if the goal is social effects (more talking, more shared awareness, more co-editing, whatever) then the simpler apps may work better. If we were just in a pair of boxes labled ‘Complex and Important’ and ‘Simple and Unimportant’, there wouldn’t be any issue.

I may have spoken imprecisely. When I say simple and unimportant I’m referring to the choice of software specifically — of course its utilization may serve a very important role for a given organization. But how hard would it be to replace if it broke, or was found lacking, or you discovered you couldn’t support it within the organization? If it’s a blog the answer is “not very”. There isn’t a big downside to rolling the dice and seeing how it works.

Contrast this with, say, the selection of a VoIP system for your office. The phone system may be much less important to an enterprise than its website, but switching phone technologies is likely to be a much bigger pain than switching from WordPress to Movable Type.

I say all this from experience, incidentally. We could be forced to migrate a dozen WordPress sites to Drupal, MT or whatever before it’d be half the hassle that dealing with our FOSS VoIP system (Trixbox) has been, although on a day to day basis the phones are far less important.

Hmm. I am not sure I can add to the thread, but I am grateful for the post. My university has been moving from D2L to Blackboard for the past two years. Its painful. Both software packages under perform and are difficult to use. Needless to say my school is part of ‘large state colleges and university system’ with a central bureaucracy in the state capital that micromanges IT for all the branch campuses. (When ever power gets cut to the central servers all the schools go down…)

My buddy who teaches at a small liberal arts college down the road is using moodle and loves it. Its amazing, flexible and cheap. We can’t have moodle, or even our own wiki server, because the central office has forbidden it. We are all using D2L and thats that.

One last thing, the central office’s IT budget has grown steadily while tuition has gone up and salaries have staid flat, system wide for the past decade. But we have D2L… which just got sued by Blackboard.

Phooey.

“Having worked with enterprise-class systems and software for almost 20 years now, I find that (1) that you are correct about that perception (2) that the reality is that no entity smaller than Boeing has any leverage over any systems/software supplier, and that actual instances of “choking necks” and filing lawsuits do not exist.”

The perfect example of this is the disastrous NHS IT overhaul. Admittedly, it wasn’t helped by the most ill thought out procurement process in IT history, but it was made even worse by the fact that the IT providers have control over the code, so they can’t be turfed off for underperformance without having to start all over again.

As someone who previously worked in government procurement, I can tell you that the answer is really quite simple. There is no identifiable supplier who can be held liable if there is a serious problem. The U.S. Government is actually prohibited from accepting volunteer services, and it is the same sort of mentality that keeps it from using open source.

Not exactly scholarly literature, but here are some notes from a talk on the subject by Ira Fuchs, whose work at the Mellon Foundation helps fund some higher-education-centric open source projects.