The Open Source Dilemma for Governments 163
Sam Hiser writes "Tom Adelstein, open source consultant and Member of the Open Government Interoperability Project ("OGIP") working group, offers another incisive article in which he discusses the costs in the terms of lives and dollars when local governments do not deploy open standards-based software for data sharing. Asks Adelstein, 'Can local governments afford to create redundant applications to meet new Federal standards for first responder alerts, emergency services, law enforcement, broadcasters?' He posits that Open Source collaborative initiatives may provide the only solution for the US if the people want to create a safer environment."
Well i would have thought this is obvious (Score:4, Interesting)
Simon.
Government not supposed to work that way (Score:5, Interesting)
No! With or without open source, we can't afford such nonsense.
This is another clear example of the overgrowth of the role of the federal government. They're going to run our local governments deeper into debt with these ridiculous unfunded mandates that may be wildly inappropriate for a given locality. The constitution clearly states the roles of the federal government and leaves the rest to the states and localities. This along with over-regulation of personal lifestyles that's going to come with public healthcare, are the biggest disasters on the horizon.
Re:it's already been admitted (Score:4, Interesting)
Microsoft's is doing it's best to keep the bleeding to a minimum, but more companies and governments are realizing that moving away from their dependency on MS is a Good Thing(tm)
Re:what's the dilemma? (Score:3, Interesting)
Tactical considerations (Score:3, Interesting)
Open Source collaborative initiatives may provide the only solution for the US if the people want to create a safer environment."
Here's another related thought. (And, this is not intended as a slam on Microsoft)
Open Source systems (bazaar) are often much more stable than commercial systems (cathedral) just because of the number of bug hunters, and when it comes to military apps, stability is absolutely crucial. Would you really want your military systems to blue screen or dump core right in the middle of a firefight?
gov't lacking in expertise and money for software (Score:5, Interesting)
This article is really talking about standardization and consistency across government organizations -- a huge job.
Imaging thousands of individual offices who have operated in a certain way for a hundred years. Imagine all of the paperwork, homemade spreadsheets, interoffice memos that spawn secondary spreadsheets, etc. This unfortunately is how the US government works.
Now imagine someone coming in and promoting replacing whatever random assortment of tools is in use with opensource tools. This means retraining. This means new hardware. This means *A CHANGE*. Uh oh.
Is this the right long-term thing to do? Yes!!
Is this going to be easy? NO!
In order for this to be successful, it will have to have very important people behind it pushing it from the top down and funding the proper resources (hardware and people) where necessary to bring the government into the 21st century.
I for one, certainly hope it can be done, and it would be great for the US and the rest of the world (except Microsoft) if it can be done with opensource software.
Re:The reverse would seem to be true (Score:2, Interesting)
Re:Text for the soon to be slashdotted.. (Score:2, Interesting)
Some might consider this off-topic, but I would be willing to bet the site doesn't get /.ed. Why? this page is mostly text.
Time will tell.
LK
Re:The reverse would seem to be true (Score:2, Interesting)
2. Small commercial company B reuses A's source, provides service to their own town's police. The cost is minimal. Rinse, repeat.
3. The small commercial companies collaborate to improve the software. The cost is absorbed by service contracts and is split among all involved towns.
Much better than reinventing the wheel N times.
Re:The reverse would seem to be true (Score:3, Interesting)
On the contrary. Niche apps are custom programmed, either in house or contracted and cannot usually be sold again. This would be the perfect place for FOSS -and- companies working on FOSS.
This is a big world. Other communities usually have the need similar niche programs. Modifications are necessary, but most companies aren't so pervasive, that they know who requires this niche product, or known to provide it, and/or cannot provide the modifications.
But, when those niche products are FOSS, those communities can hire a local company to provide the modifications for themselves (and others).
Huh? (Score:5, Interesting)
What the hell is he talking about? In the previous paragraph he writes:
If the Internet failed to follow accepted standards, it simply would not work
So the Internet works because it "follows standards", and we know MSIE (price: free) has the largest share of the browser market. So MS hasn't broken the Internet. Can someone give an example of what he's talking about? And don't tell me Kerberos because it's not the example you're looking for (MS did not co-opt it - MS extended Kerberos in accordance with the spec).
He started out reasonable and then got shrill. He throws out statements like, "Seventy-five percent of the municipalities and schools in the United States cannot afford proprietary software" So...that means 75% of the municipalities are either a) running OSS, b) using pen and paper, or c) pirating all their software. A source reference would have been nice.
Oh no...he has recommendations too:
the states should require the use of Open Standards and Open Source Software when applicable
When applicable? So, who decides when the software "applies"? Availability? Cost? (cost of development for a custom solution vs cost of COTS software) Everyone knows offshore development is cheaper - since he beats the fiscal drum so loudly does he also advocate sending any custom programming jobs overseas? He did have one good idea:
If we can pay for software one time and share it with all government entities, we empower Americans to participate in the security of the homeland.
Solution: site licenses for America!
Re:gov't lacking in expertise and money for softwa (Score:4, Interesting)
The crux is standardization, or, for you DBAs out there, normalization across applications instead of databases.
One of the examples he gives talks about differing field names (last_name versus surname, for example). Well, sorry, but that has nothing to do with whether you're using SQL Server or MySQL and everything to do with standardizing architecture.
But how does one do that across an entity as large as a government? How do you tell programmers they must use only these field names? And how much will it cost to rename fields in existing applications, and ensure all the links, dependencies, etc., are rectified as well? It's not really anything to do with the platform; at the least, it doesn't have anything like the impact the author suggests.
An important issue, as the author says, is that for many applications (such as SAP and JD Edwards), no open source equivalents exist. This is a big problem for purchasers, because it makes them wonder how long open source will take to give them the applications they need (or if they'll ever come). They may have to pay big bucks for that other software, but it integrates with their existing applications and it's a known quantity. Never underestimate the power of familiarity.
And, although I hate to be a grammar nazi, the author might just find himself being taken more seriously if he learns how to use words properly.
Pool all Government software. (Score:4, Interesting)
Some existing body, like the GAO, could administer the pool and send CDs to any community, state or federal department that would require the software.
Re:Well i would have thought this is obvious (Score:1, Interesting)
1. Most of the folks that are in the position to make development decisions - do NOT have IT backgrounds. After all, salaries for state/local civil servents are not up to industry levels, promotion is usually more politically motivated than merit oriented, and it takes so long to move up the ladder to get into an IT management position. These factors conspire to mean that decision makers only go with recognized companies (Red Hat/Suse/Mandrake etc are not known entities) - or they go with the companies who have big / well supported sales staff Micro$oft, Ci$co, etc).
2. Technically, state/local governments are VERY conservative in their decision making process. They typically want to hear how many other similar organizations have ALREADY acquired the COTS product - that's how they feel they can be successful. They do not have a good history with large scale developments in-house.
3. Technical Staff in most state/local IT shops do not receive the pay/training - nor have the skills for the latest and greatest... many are still using AS400's and Cobol extensively. JAVA/PERL/C on INTEL/AMD servers more risky than they are comfortable with.
4. Funding. The bottom line. Most state/local IT shops do not have a real budget... they get in line with other public services and hold their hand out... it's amazing how much gets done by way of Federal Grants. The Feds could go a LONG way in demanding OSS and standards when they aprove grants... that will change the course of events.
5. A big "gotcha" that happens when state/local AND FEDERAL goverment pays for development - the contractor shows where they have contributed from internal funds - to the extent - that they get to sell the same systems again and again... the public is NOT getting their $$$'s worth.
My project uses LINUX, XML (the same JXDD referenced in the posting), JAVA, PostgresSQL, Apache/Tomcat. And we are making sure the contractor has no strings attached so we can freely release the code to all Criminal Justice activities. We also intend to let it be well known that OSS does work with Criminal Justice projects...
I'm one Public Servent who spends like it IS MY OWN MONEY... every cent is counted - just as I do myself. I too am a taxpayer... and I am under public review... I do NOT want my name on the news/newspaper for poorly spending the publics $$.
In general a very good writeup... with lots of valid points - - despite the dilletantes and naysayers comments here...
Re:The Open Source Software Institute... (Score:4, Interesting)
I have a lot of experience with software for local law enforcement agencies. This particular area is a morass of smaller and larger companies, each with their own software packages that may or may not interoperate with their competitors. I've seen a number of small police agencies that have been trapped by trying to support a package from a vendor that either (a) is out of business, or (b) is no longer supporting the package/version in question. In my judgement, a good open source package supporting local law enforcement could make major improvements to the situation.
Having said that, I don't think the existence of such a package would, in fact, put many of these vendors out of business. Most local police departments don't have the expertise to manage the installation (including data conversion!), tailoring, etc. that is required for any such package. There would still be plenty of opportunity for companies to provide this as a service.
However, the fact that the underlying package was standard, and known and understood by more than the employess of a single company would help insulate the local police from the problems that arise when their support organization moves on, for whatever reason.
In addition, there is a big drive these days for national, state, and regional, and local interoperability between law enforcement agencies. Anything that helps to standardize data models, etc. etc. would be a big help in this area.
Re:gov't lacking in expertise and money for softwa (Score:3, Interesting)
Publish a namespace reference as a RFC, dictate that all governmental entities that are having custom apps developed adhere to those guidelines, and that they submit addendum to the maintainer of those guidelines when they are adding named feilds to the list.
The programmers have access to the spec before they bid on the job, and the spec is included with the customers criteria.
Document formats, not software (Score:3, Interesting)
If someone wants the government to use their software then their software must be capable of saving to the government standard GPL format.
Government documents will always be accessible.
Goverments will be free to switch software and not worry about format incompatibility.
They can choose to use the best software for their formats...free(dom) software or proprietary.
The playing field will be leveled. No document lock. A software package will compete on its pricing and merits.
Chances are all of these benefits will transfer to the private sector as the sheer volume of government documentation will force the inclusion of government standard gpl formats into software made for the private sector.
As a bonus the GPL will get a shot in the arm as far as legitimacy go.
The government formats will also spread and be improved being GPL as anyone will be free to use or change it.
If the government sees a nice modification they can make it the standard.
Steve