Defenestrated

Google! - grants and open source

I spotted this post on Slashdot about some old Google pages that are available on WayBackMachine.com, which is an archive of Internet sites and web pages. One of the pages in the archive is the old About Google! page, and it highlights some very interesting facts.

Funding for the research that made Google possible was provided by NSF, NASA, and DARPA (through grants, I presume), and an outfit called Interval Research. So Federal grants helped create possibly the most widely-used web application in the world and, in my mind, one of the cornerstones of the Internet's usefulness. I don't believe many people know this, and they should. Taxpayers should recognize that their investment in scientific research, via government grant-making, creates massive returns for them and for the benefit of the world.

Secondly, the developers made significant use of open source technologies, including GNU, Linux, Python, and Gimp. They didn't use fancy development tools, nor did they use a proprietary operating system. To revolutionize information indexing and searching, they employed stuff they could get and use at no cost. I'm sure this is less true nowadays at Google (although I'd be surprised if the majority of their development is done using anything other than fairly rudimentary tools) but the fact they started with tools available to anyone shows that open source is democratizing and a powerful enabler for innovators everywhere.

NASA seeks OS input through SourceForge

For the first time ever, NASA is hosting an open source project somewhere other than NASA! Java PathFinder has just been upped to SourceForge. This is a major step towards truly open, collaborative development of key government systems by utilizing the massive brainpower of society-at-large. The project is also very interesting: JPF (as it's known) is an "explicit state software model checker, systematically exploring all potential execution paths of a [Java bytecode] program" to find problems in the code.

The government already has its own repository for government-owned software components that can be shared across the government, called the Component Organization and Registration Environment (CORE). It's always been a pet peeve of mine (and most of us at TCG) that the government doesn't reuse more of its own systems in a more strategic way. There are a number of reasons why this doesn't happen, and I'll post more about it in a few days. But putting this to one side for a moment, I'm much more enthused by the government's use of SourceForge than CORE because the former is truly open, truly democratic, and gives the government's investments in its own systems -- in the systems built for the benefit of the American people -- back to the people it serves.

PDF SpeedUp / Microsoft Metro

I've always been frustrated by how long Adobe Acrobat Reader takes to load. PDF SpeedUp automatically turns off all the extraneous plug-ins, advertisements, and other nonsense that slows Reader down. I picked up this this tip from a Slashdot discussion about Microsoft's plans for a PDF-like technology called Metro that will be in the next version of Windows.

League of Gentlemen movie trailer

If you enjoy very very black humor, you'll enjoy The League of Gentlemen. The League consists of three writer-actors and one additional writer who produce the most horrific (literally) comedy that I've ever seen. After three highly successful BBC series, they've now produced a movie, for which the trailer is available. In all seriousness: if you're not partial to comedy that cuts extremely close to the bone (and political correctness), stay well away from this. Otherwise, embrace the unmatched talents of The League and laugh yourself silly.

Grants.gov and Macs

I just read through a series of comments from readers of MacInTouch regarding Grants.gov's "marginalization" of Macintosh users. The crux of the complaint is that the PureEdge Viewer, which users employ to complete grant application forms on their desktops, is a PC native application. There is no Mac version right now, so Mac users are being told to use VirtualPC as a workaround. That's not an acceptable long-term solution, as the comments at MacInTouch attest, but I think the furore needs to be put into context.

One poster on the site, David Frackelton, has it most-of-the-way right. Grants.gov is an enormously challenging undertaking, with 26+ agencies standardizing on a small number of forms and data sets, and migrating electronic and manual application receipt processes to use Grants.gov. In effect, the Grants.gov team are changing the way the government publicizes grant opportunities and receives grant applications. At the same time, the initiative has to prove to the Hill that it's worthy of funding and attention, and the only way to do that is to demonstrate adoption among the target audience

While the research community is awash in Mac users, they represent only a fraction of grantee clients. Grants.gov chose to invest most of their resources into the Windows client to fulfill its immediate objectives. A decision was made very early on to use a different business model for Grants.gov than is used by NSF's FastLane. The former is a "post office" for grant applications and aims to be nothing more. FastLane, on the other hand, is a complete web application, replete with the business logic and complexities that go with it. Grants.gov needed to launch and be production quality quickly; developing a complex web application was never a viable option, although it might be in time.

PureEdge has been promising a native Mac PureEdge Viewer for some time. It's still on it's way, and that's a disappointing fact. Ironically, the people most annoyed by the absence of the Mac client are research scientists -- the very group that lobbied so hard for the government to accept electronic applications in the first place!

However, many of their needs can now be met by other means. MIT has been piloting the "Grants.gov Applicant System-to-System Interface", which enables organizations to submit grant applications using a Web Services interface. (A similar interface has been available to grant-making agencies on the back-end of Grants.gov for quite a while. We developed software to utilize it for USDA.) So grantees at MIT and other institutions with their own grants management infrastructure will soon ignore the PureEdge client altogether and use their own systems to submit grant applications.

It's also important to note that (to my knowledge) there is no e-forms client software for Macs that can meet Grants.gov's other functional requirements...although I'd be happy to learn I'm wrong.

To establish a presence and to serve as many of the American people as possible, Grants.gov chose to use PureEdge Viewer as its e-form solution. (Incidentally, the Army and a bunch of other folks recently standardized on PureEdge too.) People are absolutely right to be fuming that a Mac client still isn't available for Grants.gov. But it's not rooted in ignorance, or stupidity, or malice. Rather, it's a consequence of priorities, money, and availability.

AmeeTee - Bastion of sanity

One of my work colleagues has started an experimental blog and has bravely posted an unpopular sentiment among tech-types: Linux Griping. Amy's right on target with this. My comment on the page is:

After being mightily surprised by the ease of installing Red Hat on an old Dell laptop, I quickly became frustrated by missing files and dependent RPM’s and whatnot. I think that’s the beauty of OS X: it’s taken Debian Unix and put a consumer-friendly interface on the entire OS. That’s a very tough job to do well, and Apple’s pulled it off with panache. That said, Microsoft and the Linux desktop advocates have a different set of problems to address, legacy architecture and non-proprietary hardware being just the tip of the iceberg. Until Red Hat, SuSe, Lindows, and others make Linux far more consumer-friendly, Windows will continue to rule the roost…. At least, until Apple sells more iPods and the world creates a new monopoly for us all to hate in Apple!

Nina Paley's animation

Check out Nina Paley's animations and, in particular, her wonderful take on The Ramayana. Each episode has its own bluesy-soundtrack. The animation is wonderful, and the effort needed to condense such a complicated and lengthy work into these 3-minute episodes is almost unimaginable to me. (Thanks to BoingBoing for the heads-up.)

I'm getting stoopider every day

According to this BBC report, researchers have discovered that excessive use of technology lowers intelligence.  More specifically, people who are distracted by e-mail and text messages suffer from a 10-point drop in their IQ.  The report isn't clear whether the drop is temporary (e.g. you're less smart if you're constantly distracted) or a permanent effect of e-mail and text message devotion.  Assuming the former, I try to block off time for e-mail every day, and ignore it the rest of the time.  If I didn't, I'd never get any substantive work done. However, if the latter outcome is true, I may need to have my driving license revoked. How do you manage the relentless information flow?

A more controversial take on the AFFIRM event

I see that Jason Miller at GCN has focused on John Sindelar's pledge to resolve the governance issues with the Grants Management Line of Business.

AFFIRM cross-agency service centers discussion

On Tuesday evening I attended an AFFIRM panel session titled "Best Practices in Developing Cross-Agency Service Centers: Success Factors and Pitfalls" and was intrigued by the discussion around the Lines of Business (LoB) initiatives.  OMB is very serious about this, and has some very smart people leading it.  Their insights and statements were very interesting.

The session was moderated by Fred Thompson of Unisys (who will soon be at the Council for Excellence in Government) and featured the following panelists:

• Dick Burk, Chief Architect, OMB (former Chief Architect at HUD)
• Teresa Sorrenti, Program Manager for the Integrated Acqusition Environment, GSA
• Melissa Wojciak, Staff Director of the House Committee on Government Reform (Tom Davis' committee)
• John Sindelar, essentially the LoB program manager, although his real title is a great deal longer and less informative

There were several highlights that I noted, with my usual emphasis on grants management:

1. It was said that Karen Evans believes that agencies should keep funds saved by streamlining business processes (e.g. no agency funding penalty for doing good work!).
2. OMB recognizes the overlap between the various grant initiatives and intends to do something about it, soon.
3. Agencies need sustained, top-down management and attention to create the essential catalyst for change.  (This echos the GAO report I referenced in my post on Tuesday, and good management change-agent practice.)
4. Management of IT operations and development of business processes around laws is flawed, and is exacerbated by the short periods for which senior government executives serve.  Rather than serve individual laws and programs, agencies should manage across those programs using consolidated processes and systems.
5. Industry may have opportunities to operate LoB cross-agency service centers, and OMB wants to encourage it.

Dick Burk rightly noted that the LoB initiatives are not excuses for agencies to ignore the many other opportunities they have to streamline operations internally.  Saving money is not the core objective of process improvement and process reengineering.  Rather, he said, those efficiencies are desirable because they help agencies better execute their missions.

Someone asked what would come next, after the current LoB initiatives; what else is on OMB's radar?  Burk's response was that the Federal Enterprise Architecture Business Reference Model  identifies all the potential LoBs...and that the Service Components Reference Model identifies components that cut across agency lines, such as workflow, GIS, and so on.  Later in the session, the following LoB's were identified as potential targets for consolidation at some level of government:

• Telecommunications infrastructure and desktop services
• Budget formulation
• Records management / document management
• Initiatives unique to each agency

John Sindelar provided a startling statistic: there are 498 duplicative systems in the Federal government.  498!  This, it's believed, is caused by too little management of IT at the enterprise level and, to a large extent, the LoB initiative is an attempt to rectify that problem.

One audience member's question was truly excellent: Does the Hill recognize the need to reorganize the Executive Branch as a consequence of the LoB initiative?  The response was that the Hill is waiting to see the results from the LoB, and gauging the position of the Administration, before drafting any concomitant proposals.

Pleasingly, the question of how LoBs will be funded was raised, and the answer seemed to be that appropriators need to be better educated about the value of each LoB initiative so that they can appropriately place monetary resources.  The disconnect between government IT executives, other agency executives, and the Executive Branch is a classic case of common misperceptions of the same problem, and that hasn't yet been addressed fully.  (To be fair, this isn't surprising.  Government is big and complicated.  It takes years to completely comprehend its intricacies, and most senior executives are around for no more than 4 years each.)

Another audience member asked, of the current LoB's, which would be best served by government providers and industry providers?  Human Resources is seen as squarely within the government's control, whereas Financial Management could be provided by industry.  Why?  JFMIP.  Because the standard is well-known and there are many products that comply with it, government can leverage industry investment.  However, John Sindelar cautioned that industry is faced with a big dilemma: without more information from government, how can potential industry providers mitigate the risk associated with investment in development of a cross-agency service center?  The answer is, they can't...which is why he hopes OMB will develop information upon which industry can better base these investment decisions.

I asked one question at the session, based on the GAO report on grants streamlining, and discussed briefly in a previous post: with the missions of Grants.gov, the Grants Management LoB, and PL106-107 all overlapping, how does OMB intend to avoid agencies understandably doing nothing until the situation is clearer?  The answer is that the stakeholders are meeting to address that problem, but that it's a problem unique to grants management -- HR and FM are not affected by it.

All-in-all this was two hours very well spent.  Aside from the really very interesting panel discussion, it's always good to spend time in the company of other AFFIRM members.  As more cross-agency service centers develop, there will be opportunities for government and industry alike to benefit.  More sessions like this will help everyone in more completely understanding those opportunities, rather than their impact being perceived solely as a threat.