$18 Million for recovery.gov redesign. Could HN do it better and cheaper? | Hacker News Reader

$18 Million for recovery.gov redesign. Could HN do it better and cheaper?

blogs.abcnews.com

18 pointsjakewolf22 comments

This is going to get ugly very fast. Just last week we had people coming out of the woodwork to say they could clone StackOverflow over the holiday weekend.

There can't be a debate about this because the article doesn't mention any specific improvements.

There could potentially be a lot that goes into this. There could be new hardware costs, integration of existing systems so ease the publication of information, new payroll costs, office space, etc. For all we know they could be putting together a whole team of people to run this thing. They could be doing something that's basically like starting a new company. Or they could be paying way too much for a Wordpress theme. The point is we've got nowhere near enough information for people to make the argument that it could be done better and cheaper because we really don't know what "it" is.

There's a weekend, and then there's $18 million.

How much do you think recovery.org spent? And they're not 6 months late.

That's what happens when you violate the guidelines and add, "Could HN do it better and cheaper?" to the title.

Kinda like adding the expression "in bed" to the end of any fortune cookie to completely change the meaning.

Could HN do it better and cheaper in bed?

That's true, but whatever they are doing, it still sounds like they've got money to spend so they are spending it. 9.5 million in 6 months? That's a lot of servers and a lot of programmers working on a web app.

The Sunlight Foundation tried to put together a bid on this. Here's their recap of the process:

http://www.sunlightlabs.com/blog/2009/06/25/recoverygov-bid-...

Think you're better than a bunch of freetard hippies? That's a joke; I'm a big Sunlight Foundation fan, and they do some great work. The fact that they couldn't pull together a bid should give you pause.

As a government contractor, I can attest to the complication of process involved in getting work via FAR (Federal Aquisition Regulation) process. It, simply put, is very, very hard. Companies like Lockheed, Boeing, SRA and SAIC literally spend millions to write a contract. There are teams of very experienced people who specialize in FAR contracts and government bids, and if you don't have those people -- good ones bill $600 an hour -- you simply cannot possibly win a contract.

In addition to just being able to put together a team of people who can do the work, a team of contract writers who can summarize the work, a sales team who can establish the necessary relationships to sell the work and a contracts management team who can take all of the above input, translate the soft requirements to get what the customer really wants, rationalize that with what they really need, and win the work for the lowest bid, you also have to include unknown entities, You also need to have a certain portion of your team consist of minorities or disabled veterans and such. This is typically done by sub-contracting out portions of the work to Veteran owned or Minority owned businesses. In addition to that, you have to have past performance qualifications within the government sector. Unless you've previously developed a website for a government agency satisfactorily, and are able to get a good review for that work, you don't have the past performance quals to win work to do that work in the future.

And through this entire process, you're bidding against Lockheed, Boeing, Northrup Grumman, SAIC, EDS, SRA et al, and they do this every day. That Sunlight couldn't do it really isn't that surprising to me, having done it before as the least experienced member of a very experienced team. Without the necessary guidance, I'd say the process is damn near impossible.

Just to be clear, I wasn't trying to imply that the Sunlight Foundation would have been expected to succeed; I was pointing to a group of talented, motivated hackers (like you might find here) who took a shot and learned firsthand just how difficult and convoluted the process really is.

I was going to say something similar but this pretty much sums it up. I read the postmortem written by the Sunlight Foundation guys; while they were well-meaning and I don't want to knock their effort, they didn't stand a chance.

Getting Federal work isn't anything like working in the private sector. There's all the paperwork to go through, putting together a proposal and writing all the contracts (which is a huge amount of work and is basically done at-risk), and that's just to avoid getting your proposal thrown out. To actually win, you need connections.

A "good" government contracting officer can basically use the FAR to their agency's benefit, and in effect steer the contract to whomever they want to win. If you're not that person, you're done for. I'm not even implying anything illegal, just that the FAR is so complicated and has so much room for subjective judgment, it's pretty much guaranteed in my experience that contracts will be awarded to whomever the agency feels most comfortable working with.

It is not a system that is kind to outsiders. If you don't have a staff full of proposal writers and lots of experienced people with personal networks to draw on and extensive resumes, your best -- and IMO only -- bet is to cozy up to one of the big players and work as a subcontractor for a while.

I'm a little surprised they didn't know the language. It doesn't sound like it's too far off from Grant Speak and the Sunlight Foundation has obviously applied for a lot of grants.

It's unfortunate that a chunk of that $18 million must be the premium commanded by the organizations that speak the language, only because others can't.

Not just a redesign (where design means making the entire system), sounds like it's also maintenance until 2014. It's a complex project, just getting everyone to submit the data would be a huge chore. Actually doing something with it and keeping it all up to date for 5 years is a tall task.

Update: here's the RFP and it is indeed a huge job

http://www.scribd.com/doc/16515421/RAT-Board-Solicitation

This isn't exactly easy. The site will need a heavy backend that talks to a whole array of old legacy databases to pull the data out, meaning you will probably need to have people proficient in cobol, db2 and all sorts of other greybeard stuff. You'll need to be able to scale well with large amounts of data, and you need backups and redundancy of this. You will also have to be able to play the political game, expect pressure, and expect specifications to change often due to politics.

I think I could do it in a weekend.

I'm just wondering if the U.S. government will ever be Ramen-profitable... 'cause I'm sick of them always taking my Ramen.

Amusing snark aside, this is an erroneous view of how the government functions and of how it provides services.

There are a lot of things you can do efficiently and well that can't be done when working with the Gov. due to politics and a massive overhead imposed by red tape and design constraints.

Let's fix HN first, then we'll worry about another site.

http://validator.w3.org/check?uri=http%3A%2F%2Fnews.ycombina...

http://validator.w3.org/check?uri=http%3A%2F%2Fwww.google.co...

So, because Google is invalid as well, then we should all strive to write our pages with malformed and non-accessible HTML?

I think the point is that if a given site is no more busted than one of the most widely used sites on the internet, the fact that it doesn't validate is a minor concern.

This argument has been hashed, re-hashed, and re-re-hashed on here a million times -- well, variants of it, at least, such as "HN uses a table-based layout, which offends me".

You can find some here: http://www.google.com/search?hl=en&q=site%3Anews.ycombin...

In particular, pg responds somewhere in this thread: http://news.ycombinator.com/item?id=100460). I might be carefully building a straw man here, as you are speaking to valid xhtml, not necessarily semantic xhtml. I hope you'll forgive me. I think the two topics are related, and that some of the arguments overlap.

The bottom line seems to be "Who cares? It works fine."

Thanks for the links.

In any case, it doesn't really matter to me, so I apologize if I came off as snobbish. I was simply trying to point out the irony of asking HN if they could make a site better when HN itself has invalid xhtml.

the hn community didnt build this site

Reply on news.ycombinator.com