The Agile Delivery Blanket Purchase Agreement is almost here | Hacker News Reader

The Agile Delivery Blanket Purchase Agreement is almost here

18f.gsa.gov

31 pointsrburhum17 comments

This seems interesting, but I am having a hard time parsing all of the government contractor-specific terminology in order to understand what the actual opportunity is here.

Can someone explain this in layman's terms? Let's say I am a regular programmer / consultant, and I can even hire / manage / scale teams of programmers.

I would be happy to submit a working prototype as part of an RFQ process, in exchange for landing a lucrative and stable government contract.

What exactly is being offered here for someone like me?

You can get a government contract. Normally this would not be available to you even if you could handle the work because of how contracting for the government works. What I will find interesting is if they solve the contract vehicle/payments problem. That is, the U.S. government isn't great about prompt payment. For a large contractor, they can afford to wait 3-6 months between payments. However for the small vendors being targeted with this, that kind of pay delay may kill their business.

From the requirements for the demo:

"used at least three modern and open source frontend or client side web technologies" ... “Modern” is to be understood as any technology or standard released, created, initiated or finalized in the 5 years preceding the release of this RFQ."

That's an amusing requirement. Do they really want vast amounts of browser-side crap? From the list of Javascript client-side libraries in Wikipedia, almost all of them are more than 5 years old.[1] JQuery is over 10 now.

There's also "used at least three “human-centered design” techniques or tools". Whatever that's supposed to mean. Here's IDEO's presentation on "human centered design."[2] You can take a seven-week course on it.

This document is classic Government procurement in hipster terminology.

[1] https://en.wikipedia.org/wiki/List_of_JavaScript_libraries [2] http://www.designkit.org/human-centered-design

>>That's an amusing requirement. Do they really want vast amounts of browser-side crap? From the list of Javascript client-side libraries in Wikipedia, almost all of them are more than 5 years old.[1] JQuery is over 10 now.

I don't understand your criticism. At all. It comes across as intentionally obtuse.

Yes, JQuery is over 10 years old. But there have been many releases since then. Going by its release history [1], using version 1.5+ as of this writing would fulfill the requirement you quoted. Any earlier version would not qualify as "modern" (which is a definition I agree with considering how quickly this particular landscape moves).

[1]https://en.wikipedia.org/wiki/JQuery#Release_history

And you are making exactly the same assumption he did. It is not clear whether they meant libraries first released in the past 5 years or libraries that have seen new releases in the past 5 years. He assumed the former, you're assuming the latter. He has a valid concern.

Under that interpretation, COBOL 2014 [1] would qualify.

[1] https://en.wikipedia.org/wiki/COBOL#COBOL_2014

Since the title was edited somehow, this is the interesting part: "We’re requiring vendors to submit a working prototype based on a public dataset and show their work in a publicly available git repository. "

It is my understanding that the title on Hacker News is meant to match the actual title of the article.

We edited the title in accordance with the HN guidelines: https://news.ycombinator.com/newsguidelines.html.

If you want to highlight the interesting part, the place to do it is by adding a comment to the thread, as you just did.

ok. sounds good to me.

Happy that they chose an open source technology, but is it necessary to force git on people that would prefer to use something else?

There are benefits for a large organization to standardize on a technology.

Make a working prototype AND make it public?

https://news.ycombinator.com/item?id=9797482

I am not going to say that I agree or disagree with this methodology, but let me give you some context. Traditionally, getting into the GSA schedule is a lot of work and it takes a lot of time. So much indeed, that it doesn't make sense for most startups to do this.

For some startups in the Govtech space this makes more sense. It is a way to hack the procurement process. I would happily get one of our devs to put something together that uses our Open Source codebase and solves this "challenge" than to spend months navigating all the bureaucracy to get in the schedule.

Somewhere around $50b/year go through the GSA schedule (http://www.gsa.gov/portal/content/198473) so, IMHO, anything that makes this process easier is a step in the right direction against traditional vendor lock-in.

Unfortunately the schedule 70 provision in this RFQ requires that you both get your devs to put something together using your open source codebase, AND spend months navigating all the bureaucracy to get on GSA schedule 70.

This is not open to anyone, you still have to be gsa sched 70.

I spent some considerable time and effort trying to hack my way into the UK equivalent, and so far have mortally failed. The most frustrating was a failure to use the correct phrasing in the 42 box of the RFQ for getting onto the digital framework for LOcal authorities.

I say go for it - the field is being tilted in your favour a bit - but I am not convinced it's an easy win for a one or two person biz.

Reply on news.ycombinator.com