Open Source and Our Government
A couple days ago, I saw a tweet by Henry Blodget and replied:
@hblodget @TheStalwart yup. here is the seminal book on that http://t.co/nNGStKDiCy this is not war. surges don't work. open source it
— Fred Wilson (@fredwilson) October 21, 2013
I am really upset by the problems with healthcare.gov. Leaving aside all the issues with Obamacare, and I hope and pray this discussion does not downgrade into a debate about that, I am very excited about the potential of marketplaces and marketplace economics on the price, availability, and transparency of healthcare insurance. It is way too complicated to buy healthcare insurance today and it costs way too much. The Internet and the power of marketplace economics has the potential to change that.
But our government has badly botched the construction of healthcare.gov and is now proposing a tech surge to fix it. More people, more money, and more promises thrown at a badly broken process. This will end about as well as Afghanistan and Iraq.
I'd like to suggest another way. Open source the healthcare.gov project, or at least all the components that easily lend themselves to open source. I think that some of it may already be open sourced. But instead of hiring an army of contract developers who will cost us so much money, harness an army of volunteers, who are likely better engineers, who will do the work for free.
That's what is increasingly done by technology companies and so much of the software that runs the web these days is open source. Why can't the software that runs our government be open sourced too? If you think this is a good idea, you can sign this petition. I signed it yesterday.
There is a lot going on in this area. My colleague Nick posted this link on usv.com today. GitHub now has a "subgit" on government projects. That's awesome and I hope we see the healthcare.gov codebase show up there soon.
We get this far and fumble go into the end zone!This hurts so much.
Mark Sanchez must be running this thing
Ooooh, so much hate.
more like Steve Mills. : )
Sounds like my Northwestern Wildcats.
Except unlike NU or the Jets, this was guaranteed to fail from the start. Was a poorly written bill that became law.
And NU Football isn’t guaranteed to fail from the start? You’re too kind!
They won a bowl game.
I really don’t understand what’s so complicated about things now.Last year I went to ehealthinsurance…had 50 quotes in 7 minutes and a policy the next day. That’s a fairly vibrant marketplace.It would only improve by allowing providers to bid across state lines.Why the hell is govt setting up marketplaces to begin with?On the plus side…at least we are proving to a whole new generation how utterly incompetent and inefficient the “rule by force” crowd truly is.
i should get and post the matrix of choices for an employee my foundation is hiring i was shown the other dayi could figure out a CDO easier than simple coverage for a single healthy employeeits a fucking disaster
Too many choices?
too many options to choose between the various features of the plans and i don’t know what half of them even meani will get the matrix and post it here so we can all see what i am talking about
Sounds like a UI and filtering layer problem…not a marketplace problem
It’s a (solvable) stochastic optimization problem.
Re: Too many options…I am reading this book that a friend recommend – Insanely Simple http://www.amazon.com/Insan…The author talks about how Jobs paired down the number of choices and models…something like when he returned to apple – he relaunched w/ 4 options – Business vs personal and Desktop vs laptop.The author also talks about Dell ( and other OEMs) about how many different SKUs and options they have – they make it too hard.
I truely don’t understand how anyone (such as Dell or HP) can look at Apple’s success and not see how critical simplifying the product lines was to it.
… and, the choices don’t always indicate the terms under which you can be dropped, how much you’ll actually pay under certain conditions and so on. Anyone who’s dealt with this — as I have for both my parents — knows what a jungle it is, and how trying to give and get healthcare often degenerates into managing insurance and providers each trying to milk the other and/or the patient.
Yep, the truth is that health insurance is a very complex “non-negotiable” contract between the payer and the insurance company. Like most contracts, most of the complexities are hidden and obscure. The odds of a healthy person reading and understanding the legalese/medicalese language is slim to none, and that was slim who just left the building.
Years ago I used to get those from Blue Cross for health plans in printed form. You had to sit down and take the time to understand it. Not something that will jump out at you if you don’t have time on your hands and want to make a quick decision.It’s actually quite similar to trying to get setup with Amazon Web Services (AWS) or even taking advantage of their free offering tier.
that’s not the end of the f–ing disaster, though. remember, the poor employee then has to navigate the coverage ongoing.there is no year in the last 13 when I have not argued processing around a group of treatments submitted for myself, child or wasband to insurance. he actually covered the UHC case where they were accused of foot-dragging which equalled criminality. I always remind them of that on the pone and get a manager, speeds things up.
Your foundation? Where have I been sleeping, under a rock?
you have a foundation?
They are not actually setting up marketplaces, it will all devolve to single payer anyway.Kinda funny how the military software to kill people is cutting edge and gives us things like the internet. Software to help heal people? Lol, not so good……perhaps a reflection of priorities and competences.
marketplaces don’t work with lots of rules and constraints. That’s what they set up. Conversely, look at the rules on trading at exchanges. Rule 514 is a list of about 8 things that need to happen for a trade to occur at CME.Let’s look at Fredland marketplaces. Kickstarter, CircleUp etc don’t have a ton of rules or constraints. The community manages the marketplace.Government could never do this-no matter the party. Ceding control is not in their DNA.
Military software is written by hackers; healthcare.gov by empty suits.
You misspelled “evolve” !
This isn’t “rule by force” – this is for-profit businesses lobbying and winning their own interests.Your so-called “rule by force” could have decided to open source it, and then your argument supporting your hate falls apart.
But they didn’t, did they?
*facepalm*You missed the point.
No, I got it. The government *could* have done things an efficient way that worked. But they didn’t. Why not?
Let’s define ‘politics’ as deliberate manipulationvia deliberate obfuscation for a partisan end.> Why the hell is govt setting up marketplaces tobegin with?It had nothing much to do with “market places” — anyclaims of ‘market places’ are just ‘politics’ as inthe definition above.For now, the ACA ‘market places’ are really mostlyjust ways of allocating medical insurance subsidiespaid out of Federal general revenues. Indeed,the big point about the failure of the Web sitewas just politics to hide the magnitude of thesubsidies; that is, a user had to give a lot ofinformation to permit a lot of queries of a lot ofdatabases, mostly not related to healthcare,so that a ‘quote’ could be generated withoutletting the user know about the real cost orthe subsidies. The Web site failed due mostlyto the need to gather all that information fromthe user and then the remote databases. Thereason for not letting a user know about thereal costs was politics as in the definition above.So, a user could believe that the great pricethey were getting was just due to the greatresults of ‘competition in the free market’and cut out the insurance company middleman obscene ‘profits’. Now, the greatprice was just due to subsidies. And thegreat coverages were due to the ACA lawand associated regulations. So, suddenlyhealth care was “affordable”.But from all I’ve read, any concept of a ‘marketplace’ has next to nothing to do with the realdirection and goal of the ACA. Instead, the realgoal is just ‘single payer’.Then to keep down costs, (1) have some statisticalwork in DC take data from a physician and thenspecify exactly what procedures are ‘authorized’ –so, we will have ‘statistical’ medicine. People willdie. (2) Have “The Commissioner” go to companiesin the health care industry and set prices and clawback profits.One of the standard issues in ‘free’ health care isthat there is no limit on how much a person can use.E.g., I know a woman who had two babygirls and just across the street a pediatrician.Right: That mother and her babies crossed thatstreet so often the town should have put in astop light just for them. Right: The babiesdidn’t need a pediatrician and, instead, the motherneeded a psychiatrist for her OCD. No joke. Or, “Doctor, I haven’t felt better in years, and Ithink it’s time I did.”.There’s more, but Fred wants us to keep downthe flames.
This interview with Kathleen Sebelius put me somewhere between sad and irate: http://www.cnn.com/2013/10/…”I think we talked about testing going forward. If we had an ideal situation and could have built a product in five years, we would have taken five years. But we didn’t have five years.” – Kathleen Sebelius????!!!!!!!!!They didn’t need five years. They needed to find and let folks who know what they’re doing, do it.And you only “thought” about testing?!!!!!!!!!I simply have no words for this.
This is what happens when people who rule by force are tasked with trying to build something that people actually want.You can be quite sure that the mechanisms for punishing those that don’t participate will work as planned on day one.
The sad thing is I bet this sort of thing is repeated in government and enterprise all over the country every single day.
I could think of two appropriate ones (to quote this era’s Barnum parody): ‘you’re fired.’
Last I heard about Ms. Kathy is that shewanted to shut down all the coal firedelectric generating plants in Kansas.Yes, Kathy, humans are evil, coal isfilthy, and evil humans are using filthycoal to destroy the pure, pristine, precious,fragile, delicate, 100% all-natural environmentand the only redemption will be return towalking, bicycles, and horses as in aCurrier and Ives painting of the goodold days before electric lights, electricmotors, telephones, radio, TV, computers,the Internet, etc. Save us Ms. Kathy,save us from filthy humans and evilcoal or is it evil humans and filthy coal?But, as usual, the redemption will involvesacrifice. Yes, character buildingsacrifice!!! Go for it Ms. Kathy!
Isn’t open source the way that 9 women can deliver a baby in a month? I love the open source idea. I’ll sign.Forgive my side snigger – @hblodgett is our most infamous Exeter classmate.
You are not the only Exeter grad in this community as well.Not me of course!
Oh, I know I am not the only Exie, but the only one from that year. And a lot of them got thrown out of school.
my roommate got expelled senior fall for something stupid. it really impacted him long term. i had coffee with him a couple nights ago here in nyc.
We had quite a group of hell raisers and pranksters, so it was always surprising who got caught. That was the laugh with Henry, he got “expelled” later, from some where else.Getting the boot works for some people and not for others, like your roommate, sadly.
totally. its a powerful (and productive) wake-up call for many, but in this case it really impacted his family life. his younger sister was not accepted as a prep (it was believed his expulsion was related to the decision) and then she applied again as a lower, got in and graduated. our most notorious classmate was a “panty bandit” that over 4 years had stolen and aggregated $6k of underwear that was eventually found floating in the exeter river in individually marked baggies! Some weird stuff I tell you. He fled NH and has never been heard of by anyone in our class since! I am not joking.
got ‘expelled’ from Wall Street too.
I tried to sign the petition. Logging in I’m told that I don’t have an account, and when trying to sign up I’m told I already have one. Dead end.I’m done with the government tying to build websites. Everyone else should be, too.
Maybe this is close to the end for big outsourced IT contracts like that. If anything, IT is getting less complicated and easier to manage, given all the advances in software technology.Using a Taleb euphemism, contractors have no skin in the game.The Romans required that engineers that built bridges needed to spend some time under the bridge they built. The English went further and required the engineers and their families to do that.
thats a super cool history tidbit!
The Egyptians simply buried them in the pyramids.
seriously?I thought the mummies buried in pyramids are important people of that time of history … like kings,queens,warrior etc.,
And the builders as well.Or so I was schooled to believe.
nope. It was part of the local taxes to work on the pyramids. Dedicated artisans for the Mid and Late dynasties in a dedicated town. They got paid in beer!http://en.wikipedia.org/wik…
nothing you read in the mainstream is true about the pyramids. they were most likely created by extraterrestial beings and are basically hyperdimensional energy devices (extract energy from other dimensions of the universe).
energy devices … now that one made me LOL.Good one kid …keep reading similar stuff :-).
there it is! … thats the kid comment I look for in the comment section 🙂
lol you know it!!! 😀
talking of contract employees … it is believed that the people who built Taj Mahal (one of the old-wonders of the world) have to give away their thumb after it was built.Take the thumb off who built it so that it cannot be reproduced.
i heard about that. it is radical!
TWO THUMBS DOWN !
Darwinism at its best!
it works well!
I see where you’re going here 🙂 Perhaps making Congress have to use the exchange in the same ways us “less important people” have to would be the quickest way to get the issue actually fixedAs opposed just keelhauling a bunch of people for show on TV and then really doing nothing to fix the underlying problem …. !!
Wow. That history bit sent me on wikipedia for 30 mins.
As the less than tech-but getting there-Ear to the rails appreciator of the collective intelligence at this digital water cooler- I heard this broadcast clip yesterday on NPR traveling between appointments. Seemed highly relevant to today’s discussion. http://www.npr.org/blogs/al…
is ‘Henry Blodget’ a real person or a nom de plume/ guerre? such an unlikely sounding name, straight out of a lampooning pamphlet read in London coffee shops in the eighteenth century.govonomics is a serious business, with serious people making serious sums of money out of it.open source is their black death.
Lol that was what I first thought years ago when I heard his name. He is real, though not to be taken seriously. I considrr him a useful contrarian indicator.
he’s real. I dislike Business Insider. Great headlines, stupid stories with no meat, and click throughs that build page views. Almost never click on one of their links anymore.
Last year U.S. Govt IT budget was $82 Billion. Most of that money went to contractors and vendors. The IBM’s and Oracle’s along with PriceWaterHouse or KPMG etc don’t have a single bigger revenue stream from anywhere else. To me what is happening with this rollout of healthcare.gov is nothing shocking, clusterfucks like this happen time and time again and whole bunch of folks have managed to make a living out of it. Technologies are never the issue, the people with their own fiefdoms and agendas make sure things go a certain way.IBM and Oracle earnings depend on clusterfucks like this and if you are a shareholder in them through your 401k’s then you wish,hope and pray for them to continue to sustain fuckups like this.This is big business and a few heads will roll but money will be made, there shall be a McKinnsey study done and a few million dollars to find-out what happened and a binder of Best Practices (bs) will be provided in glossy power-point presentations. Oh blimey this is the world we live in
off-topicI hear French Healthcare System is the best on earth though it is fully government funded….wonder how a government funded program can run so efficiently.We (India) have one of the crappiest healthcare systems on this planet….again gov-funded …people who can afford private are treated on par with western.
The French system has easy access to preventative care. And pharmacists can practice medicine and write prescriptions, so you can go to your local pharmacy for basic things, instead of going to the ER like in the US — and ER care is one of the chief expenses of insurance in the US.The French healthcare system is better because it has the same quality at a cheaper price. And its just better put together.
Brandon for pres.
it’s better because that’s why people have revolutions.
I find it hard to figure out how so much money ($394m) could have spent on the project in such a short amount of time – how many people are working on it?http://money.cnn.com/2013/1…
Well costs tend to go up when it’s required that you have 4 slobs smoking cigs standing around watching for each programmer that is working.
Awesome, so you are against smoking!
Fat Start Up. New Business model.
Brad Feld also said Forget the tech surge; do a “Tech retreat and reset” instead.http://www.feld.com/wp/arch…
This is what the GOP has been calling for since Scott Brown won Teddy Kennedy’s seat.
Fred, You’re not the only one who’s called for this. The problem is, there needs to be clear explaination why “open source” doesn’t mean “lack of security” (and can in fact mean more security, done right) — I have heard a number of comments expressing this concern. (We could then talk about how our education system is broken; you could provide links to points you’ve made on that.) PS Your “take our survey link” in footer of email newsletter links to a no longer live page.
Agree that open source is the best way. One problem with coding something like this though is all the rules and regulation, all the if/then categories. The law is written poorly.One of the great hallmarks of open source coding and good businesses is they get the economic incentives correct. There are so many exceptions and loopholes in this law it has to be very very difficult to work through. As soon as you get one segment coded, you discover that you have to recode it due to an different interpretation of another section.I have seen a lot of figures as to the cost to build this albatross. Somehow, being government it doesn’t surprise me. If the military can buy million dollar screws, why shouldn’t it cost hundreds of millions to build a working website?Since I don’t want this to denigrate into a right /left thing, I will keep my opinion to myself. (But I’d scrap the whole thing and start over)
ehealthinsurance, esurance, kayak, priceline, hotels.com, elance….and a multitude of others have figured this out already.It’s astounding the difference in competence that “self-interest” can have on a project.
Not only is the scale much larger but the companies that you mentioned above didn’t open on day one with a ton of traffic and interest. Just like ebay didn’t or craigslist. And they certainly didn’t have the eyes of the world on them. And they didn’t have the same issues with authentication of individuals and working with companies that weren’t perhaps 100% onboard with what was going on. The integration part.
Start by sending everyone involved a copy of The Lean Startup.
They wouldn’t get through the first chapter. Lean and government (run by any political party) don’t go together.
Yeah, then maybe a thigh master, from the same Guthy Renker bin of nonsense… Lean is great, the popularized form is asinine….
It is worth looking at what has happened in Oregon over the past three weeks. They hare already reduced the number of uninsured people in Oregon by 10 percent. (in three weeks!) And yes their website works. Surely they are doing something right.
.Kind of the low hurdles, no?Your website works, the scheme works.In the end, is it all not about just simple competence?JLM.
Actually it is a bit more. It is the fact that Oregon wholeheartedly embraced the Affordable Care Act and spent the time and money to get it going for all of its citizens. Worked closely with the health insurance industry and with social service departments within the federal government and state government. They did not fight it, the way that Rick Perry and some other republican governors have fought it. The end result–more citizens with health care–just like the law intended. Imagine, if every state did what Oregon has done. If you push politics aside and just implement the law to its fullest as it was intended–all citizens benefit. That is true competence.
.You’re letting your politics show.In Texas, we are quite content not to have a state sponsored exchange. That was our right under the law and that is the option that Texas elected. It works just fine for us. Thank you.As a businessman, I provided insurance — health, dental, vision, life — and a wellness program in several different companies I founded or lead for over 33 years.I did not need any government to get me to take care of my employees. That’s the way Texas likes to solve its problems.In Texas, we do not think it is a good law.Texas is thriving — absent Texas jobs, there is NO net job creation in the balance of the country. We must be doing something right.Perfect? Not yet, but we’re trying.Whether Oregon embraced it or not is immaterial, they were competent in implementing it. Software programmers do not program in Democrat or Republican code, no?JLM.
Nothing to do with my politics. The mere fact that Oregon embraced it early, increased the chances of its success. I would expect to see similar results in time in Arizona, where a Republican governor has fully embraced the law as well–they just did it later than Oregon, so will need more time to execute well. Great that you provided insurance for your employees. But you as one employer can not cover all the uninsured people in Texas–which has one of the highest rates of uninsured citizens of any state in the US. So programmers can code under the direction of Republican or Democratic governors–it is just those Dems or Reps that fully embrace the law will have better success. Success in coding follows political competence over the long term, the controversy over the national site will simply die over time as it is fixed and the states successes are seen as well.
.You make my point perfectly.Oregon got started early and was competent in its execution.The Feds got started late — a measure of management expertise — and were incompetent.The results speak for themselves.”Success in coding follows political competence…” — extract “political” and we are in perfect agreement. It is really not a political issue, it is an issue of competence.I applaud the site being fixed and want it to be fixed. Probably unrealistic to expect it to be fixed while simultaneously trying to operate it.The notion of a one year delay for implementation seems a management consideration no longer a political consideration.Statistics in Texas are funny because of the high concentration of illegals and small businesses. Tx has more small businesses per capita than any other state.Texas has an excellent system of clinics which provide free health care to folks. I served on the People’s Community Clinic board for a number of years and saw it up close and personal.JLM.
Ohio embraced it (Republican governor). Having big trouble
Yes, but I believe Ohio started late in the process, so had very little prep time. Unlike Oregon, that embraced it from the start and had much more time to plan.
Saw on CBS this morning how they thought it would be great to give the “shop prices” option. You give zip and choose either 49> or 50<. The younger is based on a 27 yr old, the older on 50. Just as much of the problem is stupidity in many corners.The big danger is those who need pre existing coverage will go back 1,000 times to sign up, the younger healthy will say fuck it, pay penalty and stay out.
German laws on now in Github!In true open source spirit, a developer post Germany’s federal laws. It allows the community to use Git and some custom tools to convert it into a simple format and figure out what’s changed.http://www.wired.com/wirede…
It’s brilliant for immediate real-time notifications and peer review – from layperson to the lifelong experts – in any field of interest who could then state their positive or negative issues relating to it or its wording, etc.. Allowing a platform for discussion.
put the… government… in the hands… of the people?what a novel idea!SIGNED.
.The entire issue comes down to one of simple competence. The issue of competence is the common thread in most things this government has screwed up in the last 5 years.I actually do not fault Pres Obama — though he is the Incompetent in Chief, no doubt — but rather the IT folks who should have simply reached out to the best Amazon, eBay CTO and hired him for 3 years to undertake the project.Websites of this complexity are like building a high rise office building. The developer hires the best design, construction, management, property management talent and manages the process like an orchestra conductor.The failure is clearly in the process of wire diagram, alpha, beta etc.A year before going live, they should have had 100 geeks and 4 orangutans testing the crap out of this website.This has nothing to do with politics and everything to do with competence.Who will be held accountable? Nobody. Hell, I suspect Sec Sebelius will get the damn Bronze Star with V device before this is over. She has a degree in poli-sci and an MBA. She should have been able to handle this.Of course, we did get a speech out of it all. Something goes wrong in the Obama administration — speech, Rose Garden, Greek chorus. Real problem solvers.JLM.
Test with orangutans – love that. Or like @msuster says, build for your mother, or Feynman explaining quantum mechanics to a housewife – KISS.
JLM, One (of the many) problems with gov’t work is the extremely complex bidding specs and process – those who win aren’t necessarily experts at the work itself, but are all too often just experts at the bidding process itself.
yes, I couldn’t believe the amount of money consultants could make advising on bidding.
but rather the IT folks who should have simply reached out to the best Amazon, eBay CTO and hired him for 3 years to undertake the project.I think you are giving way to much credit to those individuals. It’s the “magic bullet” and halo. Ignoring synergy of assembled groups.What they have done is in combination with who they have been able to hire and work with at those companies. Not just what their abilities are. And of course luck.This is one of the reasons someone can literally have no government experience and actually lead the government. They are able to surround themselves with people that you or I couldn’t hire. Experts at their beck and call.Personally (to bring up a theme I mentioned in another comment) I don’t think that building a government healthcare website rises to a “manhattan project” level of attractiveness.
.The real issue is this — can it be fixed while still attempting to operate?This strikes me like trying to overhaul an engine in a moving truck.I think they need to take it to the shop and work under ideal conditions.JLM.
Great MIT Tech Review article “The Costly Paradox of Health-Care Technology” (http://bit.ly/1erdrmg) that asks:”In every industry but one, technology makes things better and cheaper. Why is it that innovation increases the cost of health care?”Any thoughts on how this might play out going forward?
Healthcare isn’t broken, Politics is broken. The same mess in healthcare is going on in a dozen other areas of Government – Immigration, Education, Healthcare, Finance. Big business and the Unions have the Gov over a barrel. Fed & State
Healthcare is absolutely broken.Compensation is tied to procedures done not outcomes. The system is optimized to do stuff whether it is necessary or not.
+100 for “tied to procedures done not outcomes”Just like billing by hours on contracted governmental IT project.
Came across this post yesterday, says healthcare.gov’s front end was on Github repository but they took it down for some reason: http://blog.dobt.co/post/64…
Being a former defense contractor and understanding how government contracts are awarded I am not surprised at the Healthcare.gov website debacle.The government contracting process does not attract the most innovative and nimble companies.It’s disappointing that the President could not leverage some of his most ardent backers from the Tech scene to write a tighter RFP.or oversee this important project.Back in July at TechStars Patriot Boot Camp US CTO Todd Parks alluded to his frustration with the cumbersome bidding process. I hope this might facilitate changes.
It’s disappointing that the President could not leverage some of his most ardent backers from the Tech scene to write a tighter RFP.or oversee this important project.So what you are suggesting is that the President enlist his contacts to get people from within their companies that are “the best or nearly the best” to take time off from what they are doing (earning money for shareholders) to get a government project to work. More or less a Manhattan project approach.I don’t think there is anything that would have indicated that doing this should or would involve that kind of effort. And even if it did if I was a shareholder I would not support doing that.
I was advocating specific people like Shervin Pishevar, Chris Cox, Chris Sacca, etc to do so.This is an approach that is taken in the Saudi Arabia where well off people do their normal job and then do their government job for free.I am a big cheerleader of getting people from Tech to check in to the political process full time not after the fact or just on boutique issues like Startup Visa etc.
For the people you mentioned what specifically makes them the magic bullet or even close to it? I’m not seeing it.
You must know that they will never open source it. And as I always wonder about my tech friends and persons I admire: why would you think government would get this right? You say: “The Internet and the power of marketplace economics has the potential to change that” but those are private, voluntary exchanges. Shouldn’t a healthcare marketplace be the same?
I’m not sure why anyone is surprised by this. It has nothing to do with Obama or healthcare. Big hourly SW contractors don’t optimize for robust, simple, easy to implement systems. Once they have the contract the more problems, the more they get paidLook into the debacles that have been the FBIs computerization efforts. Look up Sentinel and Virtual Case File, the latter they eventually abandoned.
As usual the correct answer by you.The only question I every have is that if its well planned or just happens through incompetence.I’m really not sure about that. Part of me would actually like to be well planned from the top because I suppose I could respect that.But in my heart I think you just have a bunch of people that don’t give a shit about what they are doing.
I don’t think it’s “planned” in a sneaky way. But you optimize what you measure, and consultancies measure billable hours.
Again you are correct, succinctly put. Goes from the top to the bottom.
We don’t have the details on exactly what the cost covers (equipment needed, ongoing support, how many years – at least I haven’t seen it).All we know is that things aren’t working. Like with anything else people forget the price if they are happy. Here they are unhappy do it is an issue.I’m not one that particularly likes to see “normals” passing judgement on what the correct cost or pay for something should be. More or less same reason it’s bad to try someone in the media.Now of course knowing the full details anything is possible.All this has been happening for years. It’s only an issue now because (like with everything else) it resulted in pain and embarrassment and the shit hit the fan.Like in boating (as you know) there will always be a need to have a bilge pump to keep some water out. You just hope that it’s not so much as to sink the boat.
Part of the complaints going around is that there are only a limited amount of people who want to bid on government contracts. Reasons given include difficulty and red tape and all the normal reasons. (Which is a bs reason.)So this is just an example of people not wanting to go for the low hanging fruit of the easy buck, or the big brass ring payout. Or perhaps the “best and the brightest” simply find it boring to do government work because there is no halo around it and no articles in Techcrunch, HN, or (now) USV.com.This really (stay with me here) goes down to the whole thing with popular culture and the media. People are driven by what the group finds sexy and important. On HN people are apologetic about “lifestyle” businesses. As if not taking a gamble and making money is something to apologize about.My brother in law (the opera singer) is paid $1000 for 3 hours work if he sings at old age homes. But he doesn’t do that often because it’s not viewed as cool as singing in an opera house or an overseas gig. (Even though there are rich old people at both places oddly enough.)Interestingly the example of Walmart and Sam Walton comes to mind. Made his money by going to the small towns that all the big guys ignored.Perhaps the “better engineers” should stop laughing at the government contractors for doing a job that they don’t want to lift a finger to do with their super skills.
You have never gotten a big government contract have you?I have.I can tell you three things:1. If you save money by doing something less expensive than you wrote in the grant you are in trouble (had that happen).2. The way you make money is through the 50% overhead rate and 17% profit rate, the less you spend the less you make. (learned that one quick the hard way)3. The government does not care how much it costs, they just want to make sure you don’t make too much (as a percent of the contract) Every rule is configured that way.
A government contract yes. A big one no. Gaming things? Of course was obvious to me from the start.I don’t think you are going to eliminate waste in this type of system. As you say “they just want to make sure you don’t make to much”. Ntim the wasted money flows back into the economy as well.Same with getting your car repaired or a home repair. You just want to make sure you don’t get ripped off to much. If the dealer doesn’t make any money he won’t be there to fix your car when you need him. If you chisel the handyman to much he won’t show up when you need him.I realize of course that the scale of the government contracts is huge. But I wonder if there is really a way to put this genie back in the bottle without some other consequence that we haven’t considered. (I don’t think there is).
The point here though is (to use the mechanic analogy)The government doesn’t necessarily in some cases seem to care if you charge $1M for a $10K job as long as you only make 10% “profit” …. there are a lot of ways to bid up a $10K project to cost $1M w/out making official profits of more than X%With your car repair, hopefully one would be a little ticked off if a mechanic charged you $10K for a $1K repair even if they only “made” $100 profit on each deal ….However, I have seen the government now pushing toward “Lowest bidder” but even that is flawed (hence I kind of agree with your “cant put the Genie back” comment because then everyone just has to lowball to get he contract and then blow through the budget once they win it ….)Simple requiring a huge overhead-ridden company to make a bid that is competitive with what a much smaller leaner company charges does not make them able to actually deliver than that any more than increasing someone salary to “rock star” levels will make that person said rock star….I think we need a lot *more* but much smaller companies doing this work and in general breaking projects down into smaller pieces that are easier to predict …Either that, or have a ginormous tech company like Amazon or Google do it, who actually have experience getting this stuff right …..
Simple requiring a huge overhead-ridden company to make a bid that is competitive with what a much smaller leaner company chargesGood points but with the above in theory a big company should have more scale efficiency and resilience than a small company who can go “poof” much easier if a few key people leave. Not the same with Boeing or even a thousand person firm (or pick your number).By the way I don’t think google is an example of a company that makes easy to use consumer front facing software although they obviously have strengths in some of the areas that are necessary (but not all). And they tend to lose interest in things pretty easily and fold them when they lose focus.Not to mention the fact that that isn’t in the hiring description for google to do this type of work. Ditto for Amazon. And both companies most importantly don’t have experience in having to deal with “normals” and have to say, get an insurance company or state on board. (I could be wrong about this though.)In other words if the people who you hired work there because you are building world class vacation resorts they are not going to be happy if you say “for the next 3 years you will be doing Section 8 housing.
Yes, also good points … there may not be a super good answer ultimately.I agree re: googles usability being …. lumpy at best I guess (to say nothing of the total mess that ensues when you try to sign on to multiple gmail accounts in firefox … whattamess)Maybe it should be the case that big companies can only have so much % of their business from the government ? Or that they have to rotate people out into “real jobs” periodically ?On a project I was on, in order to support around 6-8 engineers, OUR project office “Simply couldnt operate” with less than 22-24 people and since it was a joint project, then the main contractor actually had over 30 people just for a few engineers !Supposedly most of that was filling in paperwork and “proving” where the time and money was spent (dont know for sure — way above my pay grade) but just the sheer number of people involved and the huge attendance lists for even simple meetings still makes me shiver in dread
We agree 1000%. And I am glad because this is what the topic was supposed to be, not political.Erik points out you get paid by billable hours you bill hours.You point out you get overhead reimbursed with a profit rate, you build some mother loving big overhead.The problem with overhead is that once you put it in, its like being overrun with weeds. You almost have to scorch the earth and start over.In business, somebody does that for you, it moved to the press, and is going to move to the Universities.It happened in Detroit, but I don’t know what the solution is for the Federal Government. I would start by moving the seat of the Federal Government every four years.
Wow, another voice for open source. Fred, read Mythical Man Month. Your use of open source here clearly falls under Brooks’ magic bullet umbrella. Let’s consider the TONS of open source projects that have failed. And as a software developer, I have to laugh when someone on wall street wants all the software to run the nation’s health care written for free. The highest paid developer in history makes a rounding error on what a slub at a hedge fund takes home, in a losing year. (You were right though, Blodget’s comment does go back to MMM…)
Hate to sound like an R, but this is Solyndra, Part Deux and it’s time for not another ridiculous speech, but the details (facts), and ALL of them. I read a bit last night that the first part of the development effort was done by a Canadian contractor (hilarious in itself) who had a checkered past. But the #s were completely whack, something like $85M!! Um, this exchange is ‘here are a set of plans, compare and then purchase.’ I seriously would have to strain to see how that would take 10.
Does anyone know what stack/framework/tools they used?And presumably they have to interface with a lot of legacy systems and have a high privacy/security requirement.
That’s my understanding too — lots of legacy systems. The pricing and plan offers are based on a complex algorithm of data from a number of sources. What I don’t get is it appears a number of these sources contain relatively static data. But instead of pulling the data into a local datastore, periodically updated, instead it appears they are making live calls to legacy systems, many of which are surely old AS400 type systems without modern niceties like RESTful APIs, etc. The reliability off all these other systems becomes an issue. Heck, just the pipes to these old systems could be an issue. It’s also like a Six Sigma rolling yield problem. If you need 30 systems to provide the right data and they each do their job right 99% of the time, your yield will only be something like 29-30%. Just too many points of potential error from what I see.Having been in and around the consulting industry for a number of years, I can easily seen how lots of hours got billed, and many meetings held, but precious little new/good code written. Not an industry that values makers a whole lot. And I think this problem requires a makers mindset
I think that it is very easy to carp at the result. But it’s hard to see how a Minimum Viable Product would be helpful. And if you aspire to be at all comprehensive its going to be a complex clusterfuck.
Part of what makes open source successful isn’t just cost: it’s that the projects can start with a small scope. Big, useful systems never start big; they evolve from small, successful ones.Comparisons with Facebook and other big sites fail at understanding that crucial point. None of these had to support 100 million people on day 1.So while open-sourcing the components is a useful idea, the way that will work is politically unpalatable. Take a random state, a selection of a few cities or counties, and roll out the site there first.
The Open Source / GitHub model is also a great model for education, especially in light of all the controversy around Common Core.Recognizing that Common Core is, at its more fundamental, just a standard, there is no pedagogical reason per se that schools must use (buy) corporate developed common core aligned curriculum and lesson plans. In my capacity as a CEC (13) council member, PTO officer, and Ed Grad School CTO, there are two big complaints, among a number, I hear about Common Core aligned curricula: 1) it’s too expensive for schools, and 2) teachers themselves don’t feel they have input on, and control over, the curricula. Simple availability is still another issue – at my sons’ school in Brooklyn, it’s nearly the end of October and we’re still waiting for much of our new Common Core curriculum to arrive from the corporate publisher, to whom our school and district paid a ton of money.Instead of relying on a top-down, centralized, and expensive model, how much better might it be to open source the development of curriculum, lesson plans, etc.? A teacher in Austin has a great common core aligned 4th grade science class? Great – post it and let others modify, branch, fork, etc. Just like in the open source software world, companies could resell packages of the open source curriculum and create services (e.g. programs to teach how to use them) on top. But the fundamental building blocks would remain no-charge and open for modifications by teachers, schools, parents, etc. Imagine just having the ability as a parent to lend your expertise in, say, math to your child’s teacher as s/he prepares a lesson plan?Platforms like BetterLesson and Edmodo are doing some of this already, and it’s a great start. As we think about the application of open source models in government, I am excited about what is possible in education.(And as to ACA, I agree that those parts that can be opened to an open source solution should be. Better to spend the money on awards to the best solutions to key ACA site problems developed at nationwide Obamacare weekend hackathons than to spend that on $400/hr consultants and their PowerPoints. Maybe time for NYTechResponds to healthcare.gov?)
Signed! I am a huge proponent that as much software as possible that is funded with out tax dollars (at all levels of government) needs to be open-sourced.Highly doubt they will do it as it’s going to expose the utter crap that the $2000+ per day contractors from Booz Allen and company produce while reaping in huge profits.Probably be some bogus reasoning about “Security and trade secrets of the companies involved”IMO if you develop software for the government, then unless it’s super-secret military stuff, you need to be prepared to release the source so the people can see what they are actually getting for their $500M or whatever it cost to make ….
Signed! I think one point that kind of gets lost in the “it’s broken” tirade is that the procurement system of the government is fundamentally flawed. It rewards major contracts based on seniority in most cases and there is no due diligence to make sure that government is not paying above market prices. There have been some improvements in which the government states what it will pay for a service, but there is still a long way to go.
.Given the government’s ability to manage and manipulate technology as demonstrated by the NSA’s use of technology, it is frankly a bit surprising they are not a bit better at software related endeavors.JLM.
I think the NSA has a better recruiting pitch and lure and sheen.Seems much more intriguing to work for the NSA doing computer spying than to worth for a contractor building a healthcare website.Same reason if you are NBC in NYC at Rockefeller plaza you are going to be able to get a better grade of applicants that someone on main street in Paramus NJ.You yourself have significant military experience. Wouldn’t be the same if the stories you told had to do with time served in the (pick a third world country) army.
Look up Trailblazer.http://en.wikipedia.org/wik…The NSA has the same issues.
.Good example. If my memory serves me right Trailblazer was cancelled with big public story being it was way over budget and never completed.There was a CYA IG report issued in the mid-2000’s.Inside scoop is the project started in early 2000’s and by 2005 the technology had changed and was no longer useful or attractive.The fact that they cancelled it is OK with me. I don’t expect every effort to succeed.I think the ACA should take a step back, catch their breath and come at it again.JLM.
But instead of hiring an army of contract developers who will cost us so much money, harness an army of volunteers, who are likely better engineers, who will do the work for free.I don’t think it’s right to say that the army of volunteers “are likely better engineers”.What are you basing that on? Why are volunteers better?What the army of volunteers has going for it is:a) They will spend time at their own cost in order to solve the puzzle. These volunteers are likely to be people that have full time jobs paying them for doing the same thing. Somewhere. For somebody. What they lose is their time what they gain is the joy of figuring out something that others have failed at. Bragging rights. I wonder if the government had simply said (not that this would be possible) “hey anybody want to lend a hand at no cost to writing this software” how that would have worked out. I think as you mentioned a short while ago jumping in where there is a framework is much different than doing something from scratch. Which is what this project was. In all fairness.b) Amongst the mass of them, the participants who provide no value won’t matter but there will be someone who will provide value and solve the problems. And as a result the approach will be the hero. And it will make great fodder for the nightly news and the media who love to load on the way government does things.So the approach is likely to yield a better result. Not the people who would either be hired or do the work for free.That’s what is increasingly done by technology companies and so much of the software that runs the web these days is open source. Why can’t the software that runs our government be open sourced too?For the same reason that you aren’t going to open the kimono on everything you do at USV. Because not everything should or needs to be made public.Despite the argument that putting things out there as open makes them more secure there are cases where that is not true. A non software example of this is the system that allowed prisoners to escape in Florida. Known to only a few it was a total outlier that someone could exploit it. Open up to many how the system works and you would have many more exploits and it would be likely that you would spend untold money to keep that system secure.Last example. Gaming the line at Disney using the “handicap” card. Dated a woman with a special needs son in the early 00’s and we went to the head of the line. (It was great…) Because of the internet that info became widely known. Now so many people exploited the system that even the truly handicapped can’t go to the head of the line.
open source does not go with turf protection ..
They did originally plan to release the code on GitHub ( https://healthcare.gov/deve… ), but that release was pulled. I don’t know if/when they’ll put it back. But there is a difference between releasing code under an open source license, and actually engaging the open source community.I suspect they would accept bug reports. I’m not sure they would actually accept code submissions.
why did they pull the release?
I don’t think there was an official announcement. People simply discovered that there was nothing at https://github.com/CMSgov/H… even though that is where https://healthcare.gov/deve… says it will be and where it was on Oct 1.
I don’t know why they pulled the release. In fact, I don’t think they made an official announcement. I remember taking a look at the GitHub page ( https://github.com/CMSgov/H… ) on October 1 and seeing that there was some stuff there (although, apparently, it was incomplete). It’s now a 404.
I don’t think there was a formal announcement that they pulled the release or why. But the GitHub page that they link to (on https://healthcare.gov/deve… ) no longer works. I can’t find any proof that they had a release other than my own word: I personally logged into the healthcare.gov GitHub page and saw a release at one point, and it no longer exists.
Source Code is Data. So it can even be argued that Open Source falls under the Federal Open Data initiative.For a civic tech startup like us, its so frustrating when we come across RFPs with such onerous requirements (e.g. insurance, financial statements, etc), that we’re effectively disqualified from bidding.Granted, these requirements were put in place to protect the public interest, but they effectively bar small companies from participating, limiting participation to big government contractors with specialized teams dedicated to responding to these RFPs.There are experiments like RFP-EZ http://www.whitehouse.gov/i… from the Whitehouse Innovation Fellows program, but in the meantime, hybrid initiatives like checkbooknyc.com can be emulated.The system was developed by one of the biggest government contractors – REI, but the contract stipulates that the resulting source code be made open source https://github.com/NYCCompt…In the healthcare.gov example, civic hackers could take a look at the code without having to have a tech-surge. The added transparency will also incentivize system integrators to build quality code in the first place.
Excuse my ignorance but isn’t the whole problem with the website the connection to government databases and existing systems? If it was a clean project started from scratch there wouldn’t be nearly as many issues.They can’t grant open access to that sort of thing without security and privacy problems – or can that be open sourced to the dev community too?
I’m not a programmer and I know jack-shit about architecting. What I do know a little bit about is common sense. I was on the NYS Healthcare Exchange site yesterday and the UI and UX are appalling. One web screen requires an applicant to input the name of their current healthcare provider and policy number, if applicable. The box to input the provider’s name has a drop down menu with literally hundreds of carrier names listed RANDOMLY WITH NO ALPHABETIZATION. The box where an applicant inputs their current policy number requires a 13 digit alpha-numeric, yet my current policy only has 11 digits which subsequently locks the screen. It’s easy to put blame on lousy programmers, but I presume they’re only as good as the guidance and instruction they receive from gov’t. Moreover, it’s not like our gov’t has a strong track record for implementing large scale initiatives, whether they be military or economic. What’s the expression, garbage in, garbage out. Very disappointing and quite shameful!
Sorry, Fred:My reading of Brooks was that the main idea was’chief programmer teams’ and “adding people to alate software project makes it later” due toincreased effort for communications and coordinationamong the extra people. Yes, Brooks also said thatnine women can’t make a baby in 1 month.I saw nothing in that book about open source.The first two times I read the book, I read itnon-stop, couldn’t put it down, and from its role as’art’ as in ‘communication, interpretation of humanexperience, emotion’ I’d too often suffered in mycareer writing software until then, was reduced totears.Poor Brooks: Once at Yorktown Heights weshared an elevator, so I told him how much I likedhis book; he was not pleased!As much as I like the book, actually Brooks had aterrible track record: His work on the OS family ofoperating systems for the IBM 360 computers was toomany people, too much money, too many bugs, too muchtime, some really poor design, and really clumsy touse compared with other work at the time onoperating systems, e.g., the DEC PDP 10, work at GE,even the Kemeny-Kurtz Dartmouth time sharing system,and soon IBM’s CP67/CMS and Multics from Project MACat MIT. We are still drawing heavily from Multics.The project CP67/CMS was for the IBM 360/67, with 31bit addressing and virtual memory and virtualmachine, and the software was from a small effort atthe IBM Cambridge Scientific Center in Boston, firstas a tool to help in writing software and in goodshape in 1967. PL/I was also in good shape in about1967 while the work of Brooks was still struggling.The 1960s were “the golden age of programminglanguage design”, but the effort of Brooks, e.g.,JCL, were not among the best. Later the efforts ofIBM’s Mike Cowlishaw on Rexx were much better.Later long Rexx was the key to essentially all ofIBM’s administrative computing on about 3600mainframes running VM/CMS with applications serversin Rexx.There actually was some good work in software fromIBM in those years, e.g., PL/I by George Radin andothers. And soon IBM also did good work on theAS/400, relational database, and more. The storiesI heard much later claimed that IBM had objectoriented software in firmware in 1970!Once I had some ideas on operating system design andwent to talk with Radin. After saying that he likedwhat the team I was on had done for artificialintelligence using PL/I, he response was, “Threetimes in my career I tried to help IBM withoperating systems, and three times I broke my picktrying.”. Apparently he’d concluded that there wasno way to convince the IBM top management ofanything good in operating system design.Indeed, with OS/2, IBM blew it again and, thus,enabled Microsoft: The first versions of OS/2needed PCs that would not be common for several moreyears. Meanwhile Gates brought up a GUI on the PCsthat were actually being sold and, then, withWindows NT, 2000, and XP slowly grew functionalitymore like what OS/2 was trying for. E.g., whencould get a good Windows PC for $5000, my OS/2system had $50,000 in parts. Gates was no dummy.Actually, in a competitive procurement, I was aloneagainst an IBM branch office and IBM’s ‘supersalesman’ Buck Rodgers. Rodgers and IBM offeredsome IBM computing for free, and I blew him and IBMout of the water proposing to pay full price forsomething from Boston. The main problem with whatRodgers was giving away was the operating system –it sucked, and the main reason for the suckage waswhat was left over from the work of Brooks.For the Obamacare Web site, the last I heard wasthat the problems actually were not technical butjust political. The political stuff was from highin the Obama Administration. The political goal wasto give a user a price while hiding the subsidies,but to do this the site had to get enoughinformation on the user to calculate the subsidies.That is, the politics was to hide the amounts of thesubsidies so that Obamacare could still be called’affordable’ as in “The Affordable Care Act”.For the technical side, some of the developmentteams have significant track records.The latest I heard, the development cost was $500+million which seems high. Maybe once the politicsis out there is some good work in there.There’s a really funny parody video of Hitler in therole of Obama going hysterical over all thepolitical aspects and technical failings ofObamacare, how he wants to call it HillaryCare, etc.Since you don’t want to discuss Obamacare itself ashealth care for the US, I won’t, but the reluctanceto discuss it, i.e., the risk of heated discussions,would make my point anyway! Besides to discussObamacare as health care, we would need to dig intothe 4000+ page bill, maybe 100,000 pages ofbureaucratic regulations, the approaches forreducing costs, without tort reform, and, then,whatever “The Commissioner” will decide, which isnearly unknowable anyway.Final remark: Whatever happened to pilot projects?
Late to this post but kudos for the main post and the many thoughtful comments only some of which I had time to scan. My excuse is I’m at the Internet Identity Workshop in Mountain View. Non(open) conference talking about open ID, OAuth, open data and more. Wish I could recommend this as a session. maybe over drinks tonight. Someone commented on failed OSS projects below. On the other hand we have successful examples. The key is starting out with a strong kernel and a clean governance model. Some variation on meritocracy. I don’t know that it can yet be done. Open sourcing the code base is the obvious first step but the construct needs to be more or less in place. Just a few off-the-cuff thoughts. Feels worth it to try if for no other reason than the way in which it could crack open the black box that is government procurement.
That sounds like an interesting conference.I have one question: Has OpenID failed and has OAuth2 been a subset of it or a better implementation? (Meaning is there still a use case for full OAuth though I find the use of http-style URIs for identifiers absurd and counterintuitive.)
It is though it sounds like you would have stoked a more vibrant debate on some of the topics. Lots of believers present! I analyze OpenID and OAuth2 from a business user’s vantage point. I get your question but my interest in OAuth2 is less about re-using credentials and more about streamlining the way I can authorize passage of data from point A to point B. There better be a use case for the whole thing, schism among co-authors relative to 2.0 or not.
I think I might have!I consider permissions one of the big unsolved problems of the app economy.Specifically the lack of granularity is a problem. Every other comment on an Android app in the Play Store seems to be a complaint about how many permissions it needs, it doesn’t help things that Google has entries such as Access system information such as my contacts, or Create or delete files on my SD card or erase my SD card. Cyanogen came up with a great solution, their system pretends to grant permissions to an app but actually sandboxes them. Chrome has a similar problem, most extensions needs to be able to ‘see and modify the contents of all web pages you visit’ when in reality most extensions only want or need access to the DOM of certain pages and could be whitelisted.When it comes to the OAuth use cases, a similar problem exists. Disqus and Twitter grant fairly wide access permissions, such as ‘Access all of my Tweets, timeline and private messages’ or ‘Tweet as me and send private messages’ and for Disqus it’s the same thing but with comments and drafts.What’s lacking I think is a form of Data Permissions Language, an English-like construct (but most likely RDF-based internally) that can make a statement such asEngagio will be notified of every comment posted on Disqus discussions you participate in.Engagio will have access to a list of the Disqus users you communicate with.Or for something like DropboxAppX will have full access to everything in the folder “AppX Data” but will not have access to any other files in your Dropbox.Something like intents could fill in the restYou have chosen to open “Letter to Mom” with Stationary Upgrade, Stationary Upgrade will be able to access this document in your Dropbox from now on, and will be able to edit the contents.You will be able to revert changes made by Stationary Upgrade by restoring a previous version, and Stationary Upgrade will not be able to delete “Letter to Mom”Is this okay?
Exactly the issues I care about. We want business owners to be able to (easily) pull data from many sources to create a real-time credit profile they share on their own terms. It’s anti-black box, re-imagining credit bureau. If permissions are opaque or misleading that is exactly the wrong message. Still, it has to be super-easy and not require creating new data. Would be interested in followup by Skype (assuming you’re the Ohio Timothy Meade). I’m [email protected] or lavonnereimer if that’s of interest.
Yes I am, I’ll add you on Skype.
http://bits.blogs.nytimes.c…i was wondering what a newly minted entrepreneur does with his or her money.
It’s interesting how the campaigns, which are highly private and built by the most elite technologists, contrast with these more public efforts. A sign of our times.
I have spent most of my career (about 10 years) in this part of the IT-industry. I.e. building in-house systems for public and private organizations (albeit smaller). I have also done a fair bit of project evals and turnarounds of failuers.Evals always start with a few interviews with people who blame everything on technology and then quickly moves on to blame the development team. You listen (they are also the ones paying the bill). “No skills”. “Doens’t get what we are doing”. “Zero ability to communicate”. “Doesnt understand me”. It is always the same story. And it is always (almost) not the problem.In my experience there are really only 4 things to look at:Project ManagementDo they have a project manager? Is he/she able to make decisions? Is there a steering group comittee to handle changes and things with bigger impact? What does the project plan look like? Do they have a plan? Is the project manager able to control the client.Product ManagementDo they handle requirements in a reasonable way? Do they translate them into something “buildable”, e.g. user stories, PRDs, etc. Does the product manager control requirements or is the client able to sneak in things on the side? Is the product manager able to explain what gets built and when.Technology LeadershipAre the software products they are using suitable to the task? Do they have reasonable infrastructure? Are they designing their solution in a reasonable way? Do they have a toolset that is suited for building the system?ProcessDo they have some kind of process for the team to follow? E.g. Scrum, waterfall, kan-ban, etc. You can use almost whatever as long as you have some kind of process and stick to it reasonably well.If one of the above fails, the project will usually workout anyway. A good process makes up for bad product management. A great product manager makes up for bad project management. Etc. If two or more of the above fails, the project fails one way or the other.Note that I didn’t even include the development team in the above. Coding is this type of projects is a mundane task and there are normally lots of testers to help ironing out bugs. That is why outsourcing works in this part of the IT-industry. Building innovative software in a startup is something completely different and requires a different skillset.I would argue that it is also why we have commenters here saying that they can build something better in less time for 0.0000X% of the cost. Coding it is easy. It’s the other things that are hard.So I think you can open-source all you want, and it won’t matter one bit. Actually, as per Fred link in the tweet, adding more people (volunteers) to code things may just make it worse.Instead, they need to do a project audit from a few competing firms. Find out which of the above that isn’t working. Get a few really, really good people to fix those issues and have the rest of the current team work day and night to fix things.Finally, software is truly eating the world. The President of the United States is under pressure for a failing IT-project…
Absolutely yes. Open source it (what they can). Although I was tied up most of the day and just signed the petition now, I’m shocked it only has 2,050 signatures?
Interesting article from Atlantic in june mentioning “Open by design, open by default” http://www.theatlantic.com/…Its not like they haven’t been thinking about open source, scalability etc.
The best thing is that the open source is very important now a days. The OS are also good in mobile. http://www.mobilejury.com/
A good example how government can work on open source bases succesfully is http://gov.uk (https://github.com/alphagov )
And here’s a good video of a talk the Deputy Director of GDS (that run gov.uk) gave on their process – open, transparent, user-led and agile. http://www.mindtheproduct.c…
The government has actually become both a putz and a yutz.
While I agree the code should be open sourced I believe the government should have it’s own hands-on tech team that own the core and manage the development. Government services online should first and foremost be user-led, then agile and lean, and they can only be those things with an in-house team leading the work.A great case study is the Government Digital Service in the UK which is building gov.uk as a single-domain touchpoint for all government services and transactions. They’re a crack team of product people, designers, UXers and developers, then outsource/opensource the chunks they need more resources on.Take a look at these two great talks given by Tom Loosemore who runs GDS;http://www.mindtheproduct.c…http://www.mindtheproduct.c…
I’ve run many large-scale successful software projects, including turnarounds, both for Gov as a public servant in the Bloomberg administration, and for Wall St and the tech industry as a business owner. Open source models are a terrific opportunity for government, but don’t kid yourself: 90% of open source projects go nowhere and are used by noone. Another, totally different approach is public/private partnerships like this one we’re doing here in NYC, as reported in Newsday: NYPD crime-fighting system to be licensed to 2 cities (http://bit.ly/1auwkid)
Another plausible argument point I heard is that there is disconnect between different departments agreeing on how things are done – bureaucracy.I could see some conflict with say, NSA wanting full real-time open access to health data from people, vs. perhaps other groups that may not want to allow that?That’s easiest example – though maybe NSA gets it by default – regardless of anyone having to authorize anything. There could be other partnership holdups though.I wonder if a technical problem is even the issue or an easy way to offset / displace responsibility from all political parties and decision makers – to techie scape goats.
Actually Charlie… when the pet project of a politician gets passed and implemented by his party, without a single supporting vote by the opposing party…. is implemented according to the schemes set out in that law, on the seemingly unlimited budget and timetable set out by that party and leadership….is sold continuously to the citizens as something great that we would literally have to wait and see what was in it but believe them it’s great!….and then fails this spectacularly (though not surprisingly, to most rational thinkers)…… well….that kinda is a left/right political fact, ain’t it? Though I do understand why you’d want the conversation to steer clear of assigning actual responsibility in this case.
Good luck on the rants request. I’m signing off to do some real work.
It’s a damned tragedy.
HHS has had several successes developing apps (i participated in one called heart health) using a hackathon challengle model. Healthcare.gov could have been an easy layup using this approach. For example, HHS has had contests with awards in excess of $1 million dollars for systems for scheduling VA appoinents. Why HHS didn’t use this model to start healthcare.gov probably is the work of a DC lobbyist. Mendacity.
The most complicated thing might be immediate scale.
I couldn’t agree with you more on this.We’ve been discussing this a lot internally and I’m certain my team could have built this out in a few months with remarkable improvements in both functionality and UX with infinite savings to taxpayers.We’ve already built a product for a client that is a white label, stand alone app that provides info on insurance rates, penalties, tax credits, etc on a state by state basis. This product is a significant improvement over the state exchanges and was built by a small team in a matter of weeks.More expensive contractors are not the answer. We need talented people who can execute. We joked about starting an kickstarter campaign so that we could rebuild the entire healthcare.gov from scratch, but open sourcing would seem to be a better answer.
I really want to see the bid process for this – I want to know why mainstream web development agencies were not involved.Booz Allen doesn’t build this kind of software…
Supposedly one of the complications was thatthey wanted the site to get a lot of info fromthe user and then use that info to go, serverto server, to several other servers with variousdatabases on people, with some authenticationfrom some Oracle software, to collect enoughdata to calculate subsidies and, thus, final costto the user. The other servers were not reallypart of Obamacare and likely had interfaceand security issues for the peer to peerinteractions. That still doesn’t explain howto spend $500+ million, but the whole thingmight be beyond three RoR programmers withMacbooks.
Its partially tech but mostly people and process, as with any project this large. The real root, I believe, is the fragmented back end which is the result of poor legacy tech and a super fragmented data ownership.Last year I was at the federal Medicaid conference and attended a session with CGI and reps from each state. It was the largest group of cynics I’ve ever been around. In August 2012 they all predicted they weren’t going to be ready and were petitioning for a delay.
You’re onto something. Maybe this isn’t a technical problem…maybe this is an organizational problem….and maybe that organization is far too big, far too intrusive, incapable of efficiency. Maybe that organization should be strictly limited in scope…maybe leave these kinds of projects to smaller organizations that are naturally incented and aligned with the people with whom they interact.Just a thought. Not sure if anyone from 1776 would agree.
We prefer the term “glitch”
Ah, being tied up in contracts has a good probability.
Wait I have to hear 10 years of whining about the Bush tax cuts but then when it comes to the Obamacare Tax I’m just supposed to “get over it”?Endless complaining about the Bush v Gore decision, but the Obamacare decision is beyond reproach?This is a partisan law and a partisan failure. There’s no way to spin out of that.We agree that the most efficient way to reach the desired end would have been to put IRS guns to people’s heads and force them to sign up for ehealthinsurance.com and put the bills on the AMEX cards of high earners— but the truth is, they chose the less transparent, more complicated path because it was the only way to make it palatable to the bare minimum of partisan votes that they did get.And this was the logical result. This isn’t a technology problem, as we will continue to see….
I will voluntarily monitor “Swan Week” with the government. But will horde all the bourbon for myself, and tax the rest of the world who wants it. Heh.
Why limit the # of efforts? A for-profit, incentive-based open market sounds fantastic.Loser-pays tort reform, open borders bidding and absolutely no rules on what coverage must include. Let’s do this. In fact, I’ll give away my biz model here: Term catastrophic health insurance.and p.s. a week with Swan might do them some good….historically, it has happened before believe it or not lol
.This is how they develop fighter planes right now.Competing designs.Fighter planes are just a bit complicated.JLM.
It’s a shame that two weeks ago when we were sitting together in person you weren’t able to say out loud how punishing it was to be around me as articulately as you’ve typed it into your computer today….
And that it will become a poster child for the haters of the AHA 🙁
Or they should have just made a choice not influenced by for-profit businesses and made it open source?The problem is that current “politicians” are in it as a career, where they get into power by leveraging for-profit businesses to spout certain messaging and pass certain laws and policies (or prevent others from passing). It’s the same problem that patents create, it slows down innovation and efficiencies to a near-halt – creating a ton of waste/inefficiencies, and including wasted time – and we’re talking about people’s lives and the environment of the world – the health of society, kind of all of the most important things to humanity.If for-profit business big company lobbyists weren’t so powerful / have so much influence with their pooled resources – and the election system wasn’t rigged to have only 2 leading parties – then many of these problems wouldn’t exist.Accountability needs to be taken more seriously as well. I like some of JLM’s ideas for how to do this – with perhaps a bit of modification but the right idea overall.And the only scope that should be set in governance is how much support political parties or politicians can get from for-profit businesses (or whatever monetary channels they’ve setup).
i think the founding fathers would be terrified of MRI machines…
Weren’t the tax cuts mainly benefiting big business? Not the individual, which is what you seem to boast as the main benefit – an individual getting better representation by not having a big federal governing body?
“This is a partisan law and a partisan failure.”But aren’t we talking about a website design and implementation failure here ?
i’m not whining. Tax reform is the one issue I am angry at Schumer about…
Sounds like bad management – or under pressure – and someone who maybe won a huge contract by promising something impossible.What I’d want to know then is1) who voted for them to win the contract, and2) what lies were they told about how quickly this could be done?3) Is there anything in the contract penalty wise for not completing the work within X amount of time?
If you’re talking about the Bush tax cuts, they primarily benefitted the individual…as all tax cuts do. Yes, the individual is who benefits from more localized government and increased choice of governance via mobility.
And once again we come down to the bottom line: They did not choose that path. Of all the paths open to them, they chose this one, which landed on a 404.How did we libertarians predict such a disaster so accurately? Think about it.Government is by definition the application of force. Once you realize that, you will have no choice but to conclude that its operations should be severely limited.
It is curious that in the country most famous for its legendary creative efficiencies that improving the key pivotal processes of democratic governance have been left out of that creative tour-de-force circle for so long !From: 7 Habits of Highly Effective People”first things first”Focus on fixing the damn S-OS Kernel !
Yeah, but without booze? What’s with that?
Yep, just a bit. 🙂
would it make website design faster?
I don’t want to appear to suck up to thebrass, but in my time in DC doing mostlyDoD work, with all the loud screaming about’waste’ in the DoD, I concluded that, net,the DoD did quite well, even ‘patriotically’well, and much better than nearly any otherpart of the Federal Government. E.g., for a while I wrote software for processingpassive sonar data in the group that did theorbit determination calculations for the Navy’sversion of GPS — before GPS from the USAF.So, I was along side the people who did thatNavy work and heard the stories. A brilliantmasterstroke.Then there was Kelly Johnson and the SR-71,magnificent. Brilliant. Astounding. A turbo-ram jet engine from a Florida lab ofPratt and Whitney and astounding airframedesign.Then, Keyhole, basically a Hubble telescopebut aimed at Russia.And there’s more. I could tell you butthen I’d have to …!
Sure, but Fred also brought up the point about cost, noting that the Internet and marketplaces tend to lower costs. Except in this case, regulations gave a huge impact on costs, and Obamacare regulations banning many high-deductible plans have raised health insurance costs for millions of Americans.That’s the other major cause of dissappointment with the ACA so far, but there’s a connection with the website fiasco: when a program is passed on entirely partisan lines, there’s less give & take and you’re liable to get the worst of government.
Why would I support a Republican gov’t expansion into healthcare? You have me confused with a party advocate.I try not to “whine” (feedback suggests I might be failing at that effort), but rather to inform and lead logical thinkers to examine the organization behind this grand failure…as well as their motives— a remote government that controls our health.
Speaking from middle, I have to agree with Andy. You can sa 100x it is a Republican plan, but we need to look at who is CEO. This is the THING for the current CEO and we already have the storyline starting his not knowing of the coming failure. No excuse and this is a little more than Apple’s maps because you had an alternative (Google).We need to stop treating uninsureds as numbers, though many are between jobs, and face the reality regarding confusion at this point.
I get it. Everyone else has to watch Andy drink.
They need to capture the lead first (for the IRS) Charlie.
“Sounds like bad leadership– or under pressure– and someone who maybe won a huge election by promising something impossible.”FTFY
I know nada about this stuff, so I was kinda hoping that the blog comments here on AVC would provide a more informative dive into these types of technical issues !
Do privacy issues on a site like this account for any significant increase in implementation difficulties ?
Democratic governance is by definition a tour-de-force that aspires to be driven by the citizens for the citizens.If at first you don’t succeed and all that stuff !We are not at the end of history as regards the process of perfecting democratic governance.Surly the distributive stakeholder-market-place power of the network-effect will ultimately invade the democratic-governance process with monumentally positive effect.Finally we have adequate enough tools on board, now is the time to focus on creative network-effect driven distributive-governance solutions.This is surly not the time to panic and throw out the democratic-governance-baby with the corrupted bathwater!
They haven’t been left out – it’s just been at its yin state, while the yang state of business does its thing but is reaching its tipping point of collapse. It’s yang energy is starting up now.
how do all tax custs benefit the individual, considering how messed up our tax code is.
Why? Many were inventors and tech leaders
we’re about 200 years from them – it is like a magic machine which makes an awful lot of sounds. They did bloodletting.
not a techie, but it’s hard to know what the root cause of the problem is. The administration hasn’t been transparent about it at all-which leads to using open source.
I know your frustrated also, didn’t mean it as personal.
Ah, didn’t see whole reply til nighttime. I agree. If we were to set a gov. Plan from birth that starts coverage at a high beginning point and did private on $0 to that point, we can get around the pre existing delimna that won’t go away as it stands now. Only logical way to do it.Was going to send that to you and some of the others on AVC figuring you’d respond with your already thinking that.
They may well be some good examples but there are plenty of counter ones too. I worked as consultant for “Big Defense Contractor” for a number of years and the waste, incompetence (of some) and sheer politicking to get (and most importantly *spend* ) all possible funds from the government was prevalent !If everywhere operated the same way as the areas I saw, I could easily cut the defense budget 25% and rather than being weaker, our military would probably actually be a good bit stronger !