←back to thread

285 points ridruejo | 1 comments | | HN request time: 0.223s | source
Show context
stackskipton ◴[] No.45893105[source]
As someone who has some familiarity with this process, just like safety regulations are written in blood, Federal Acquisition rules are written in misuse of money, sometimes criminally.

Yes, we have swung too much towards the bureaucrats but I'm not sure throwing out everything is solution to the issue.

Move fast works great when it's B2B software and failures means stock price does not go up. It's not so great when brand new jet acts up and results in crashes.

Oh yea, F-35 was built with move fast, they rolled models off the production line quickly, so Lockheed could get more money, but it looks like whole "We will fix busted models later" might have been more expensive. Time will tell.

replies(21): >>45893777 #>>45893843 #>>45893847 #>>45893934 #>>45894255 #>>45894410 #>>45894990 #>>45895591 #>>45895700 #>>45895838 #>>45896005 #>>45896219 #>>45896396 #>>45897182 #>>45897650 #>>45897842 #>>45899571 #>>45899715 #>>45899941 #>>45901076 #>>45902745 #
fpoling ◴[] No.45902745[source]
There is an observation that any organization or a process becomes totally bureaucratized over span of several decades. The only solution to this is to make things explicitly time-bound like 20-30 years and dismantle the whole thing at the end replacing it with something new and new people at management positions.
replies(1): >>45902809 #
1. jvanderbot ◴[] No.45902809[source]
I somewhat agree with this. It can be true that all the overwatch was a good idea. And it can be true that the process becomes the point, which is bad. It's hard to tell the difference, but I've personally been bogged down by review processes for months, just to get 2FTE from the government itself while working on the government's approved and funded projects while sitting at the government's desk. It's really silly how bad these things are sometimes. It'll take years to do the thing that even the "fast" programs want you to do.

And the solution to date has been to "start another program". That program promises to move fast, and often does, but it will eventually metastasize with process and review.

I just don't think we can add any more "same but faster" programs. It is time to cut back a lot of process, and thereby bring these programs back to parity so we can then cut down the number of programs.