Keep it simple.
That was the philosophy. Don't bother reading the article, I just summed up the only info that wasn't just there to take up space.
Disappointing that this was light on any real details like who the contractor was, or if it was done by state employees. What the integration story was? Technology? Staffing count? Budget? You know, all those things which would speak to 'How' Kentucky built the site.
[1] http://www.courier-journal.com/article/20131021/PRIME07/3102...
This is the right way to do things. Even if they took 2 years and had elite agencies working on it... there would be a list of little complaints purists would have.
The point here is... theirs works... because they kept it simple.
http://builtwith.com/healthcare.gov says the Federal site uses Akamai ... fat lot of good it does them efficiently propagating gubble....
Still, getting the job done is not optional and, when it involves a government, not trivial either. They get my respect.
Even if I disagree with their taste for technology.
</sarcasm>
This is why I really hope they use the government's failure as a means to educate the lay person on how difficult it is to build a technology system and equally how important testing it is: https://news.ycombinator.com/item?id=6586499
ADDED: to the initial responses to this: no, not even close. E.g. explain where the Federal government has a role in this beyond the FDA's standards establish what can be sold, which are in part an economy of scale.
E.g. the Naval Observatory and NIST each maintain independent ultra-precise time bases for the nation, which strikes me as just right. At the national level, Daylight Savings Time is promulgated, but states like Arizona (minus the big NE Indian Reservation) and Hawaii where it's not a good idea don't observe it.
But most things should be left to the "50 Petri dishes" of the states (or groups of them, e.g. the Multi-State Lottery Association http://en.wikipedia.org/wiki/Multi-State_Lottery_Association ). Romneycare? Sure, why not try it? Obamacare? I think you know my answer to that.
And please don't make replies that don't factor in the Federal government's interference in this starting with WWII.
You mean, set national standards and leave the implementation to the states within those standards, with feds only stepping in if the states refuse or fail?
That's exactly what PPACA does for most things -- including the exchanges.
No, that's a modern bastardization of the idea, where the Federal government claims to be following the letter of the idea but steps in with all sorts of central mandates and requirements even so. They've got some pretty good rackets going on with "optional" compliance to programs tied to truckloads of money, too.
Federalism would be... let the states handle it. No Federal requirements. No Federal interests. No Federal laws. If Vermont wants full on state health care with no private option and Texas wants full private healthcare with little more than disability assistance, let them. Washington DC interferes with neither.
The PPACA hasn't got a drop of Federalism in it.
Whether you consider that a problem is up to you. I'm just explaining that it doesn't have any.
Why did I reply to this.
Mercury through Apollo were political stunts. The sequence goes like this: Eisenhower let the Soviets orbit the earth first to let them establish an "Open Space" president; prior to that they were saying any satellites orbiting over their territory were illegitimate.
However at the same time we were competing for allies among the Third World, and our claims of technological superiority took a big hit. So the "space race" began. First was of course Mercury, put a man in orbit, then a couple of years later JFK committed us to "landing a man on the Moon and returning him safely to the Earth"
But it was just a stunt, leaving no infrastructure in place, heck, even finished Saturn Vs were left outside to rot (due to the inevitable delays, Nixon had to cancel the last planned Moon missions because of the sunspot cycle making them too dangerous).
The Interstate system was based on the German Autobahns, and President Eisenhower's experience trying to move on troops on our old system. However we've had a long and very mixed history of Federal involvement in transportation systems, such as the transcontinental railroad scams (see e.g. http://en.wikipedia.org/wiki/Cr%C3%A9dit_Mobilier_of_America... or which went bankrupt (hint: the land grants were the real prize)).
The Internet was developed by/for/with the Defense Advanced Research Projects Agency (DARPA) to allow the sharing of the then very expensive computers.
I trust we all know what pushed the initial development of atomic energy in the US, and as I understand it our first reactors developed to specifically produce power were for submarines, where they made a massive qualitative difference.
The Americans didn't need the same capacity as the Soviets in rockets for several reasons. First, US nuclear weapons were supposedly half the size of the Soviets'. Second, the Americans had air superiority. Third, the Americans had land next to the Soviet Union (thus didn't need to fire as far). Fourth, the R7 that launched Sputnik took hours to load. And could not be on standby indefinitely.
Both space programs were stunts in the early days. But using a Saturn V to deploy nuclear warheads is like using a flame thrower to start a barbecue. It could have been done, but there are easier/cheaper ways.
However, you do have to be somewhat blind to see that a large segment of the population, it's the Republican platform for instance, genuinely believe the government can do nothing right. Every failure, no matter how small, supports that position. No amount of facts change that position.
In a more general way, less and less peoples views seem to be swayed by any facts that disagree with their worldview. Everything has become an us vs them, my team is better, right and perfect view of the world.
It's like purchasing a car, never doing any maintenance, and then when it dies saying: "whelp, all Fords are shit! I should have bought a Chevy"
Not necessarily to my taste either (https://news.ycombinator.com/item?id=6627673), but if it gets the job done, who really cares?
Unless you are talking about licensing SQL Server per processor, the cost of licenses is unimportant.
But I was more thinking of the shoestring startup shooting initially for ramen profitability, where every unnecessary, can't instantly scale dollar spent is one too many.
This is an effort with a serious budget, where today's back end costs of all sorts are I'm pretty a very small part of it all, probably even bandwidth.
Yes. That's exactly what I think. So, what I'm about to say isn't in any kind of defense. I'm just going to document my personal bias for all to read and hope to be enlightened:
I don't think anyone really chooses to run a Microsoft Server, other than Exchange-Server[1]. I think people are forced into it because of at least one of the following: 1) Their corporation demands it, 2)They don't have the skills to use Linux or OSX, or 3) That individual may have the skills but the other people on the team that need to be able to help out, do not have the skills.
Anecdotal: I once interviewed in a place that had Microsoft servers and SQL Servers, I asked them if they'd consider using Linux. They said "Oh, we're a Microsoft-Gold-Partner, so no Linux here." This was in 2007 for a small-to-mid company somewhere in San Francisco.
So IFF I'm correct about the above, then anytime someone chooses a Microsoft server(not counting anything related to Exchange)... I assume the whole shop is on MS-stack. I might be horribly wrong; I'm posting this to have other HN people correct my terrible thought-process.
1. Because Exchange really, truly, works and it's hard to get everyone to agree on a tech to handle emails+meetings. Exchange seems to be the most widely supported choice.
As for Windows running the server, it has been a very long time since anyone tried to claim it inferior as a web server - about 5 years ago when Scott Guthrie's team made IIS modular the story changed and it is now every bit as capable as Apache.
Don't worry though, I wont try to defend Sharepoint.
The mix is there though, I have clients who run Windows on the server and Oracle as the DB. Also MySQL, Raven and one using Redis as a backing store. I have a client who runs a rails shops and use SQL Server as the backend because its BI stack is so strong and even have a client who is a 'Microsoft shop' who is considering ASP.NET MVC running on Linux pointing at a MongoDB repo.
What I recommend is based on the feature set they need and their access to talent, the cost of server licensing can sometimes play in the discussion but only at the lower end of the market (startups, small business).
It feels like there was no one overseeing engineering quality across any of these government sites. Quality doesn't have to be QA but best practice standards in both software design and testing. Few unit & automation tests and peer reviews would have gone a long way to releasing decent software.
> DEVELOPED BY SURESH PALDIA
A quick search tells me that someone with that name works for Deloitte, so if I had guess, that's who KY contracted.
IMO, it would be inappropriate to single out state employees or the contractors and appreciate only one of them. Great execution or bad one is mostly an outcome of how capable and driven the executive team is.
Sounds easy. But it requires a person with deep understanding of software projects and with real political weight within the project to keep creeping requirements at bay.