DOGE Plans to Rebuild SSA Codebase in Months, Risking Benefits and System Collapse

Jokes aside, nothing wrong with rewriting in Java. It is well-suited for this kind of thing.

Rewriting it in anything without fully understanding the original code (the fact they think 150yo are collecting benefits tells me they don't) is the biggest mistake here. I own codebases much smaller than the SSA code and there are still things I don't fully understand about it AND I've caused outages because of it.

I've worked on these "cost saving" government rewrites before. The problem is getting decades of domain logic and behavior down to where people can be productive. It takes a lot of care and nuance to do this well.

Since these nazi pea brains can't even secure a db properly I have my doubts they'll do this successfully.

Functional, yes. But rarely are these sorts of things efficient. They’re covered in decades of cruft and workarounds.

Which just makes them that much harder to port to a different language. Especially by some 19 year old who goes by “Big Balls”

screams in quality assurance

It's worth noting that one of those organizations is IBM. Mostly relevant because they're the ones that originally built a lot of that cobol, the mainframes it runs on, and even the compilers that compiled it.
They're basically the people you would expect to be able to do it, and they pretty quickly determined that the cost of a rewrite and handling all the downstream bugs and quirks would exceed the ongoing maintenance cost of just training new cobol developers.

My dad was a cobol developer (rather, a pascal developer using a compiler that transpiled to cobol which was then linked with the cobol libraries and recompiled for the mainframe), and before he retired they decided to try to replace everything with c#. Evidently a year later their system still took a week to run the nightly reports and they had rehired his former coworkers at exorbitant contractor rates.

I bet they'll do it in Waterfall too.

Nah B. This will be Extreme Agile XP with testing exclusively in Prod. Xitter will be the code repository.

What's "vibe programming"?

Bold of you to assume they'll use Java and not some obscure language picked based on the need to pad their resumes.

They're not going to use Java, it's going to be typescript.

Also : it's very complex and it happens to work fine for decades.

If one day i write a code project and manage to make it work without any major issues for several decades, there is no way i attemptto rewrite it.

But dude, bro, we could put the entire system on the blockchain man, and make it super efficient with an AI backend that will remove all errors bro.

Dude it's not even written in Rust bro. WTF is this dinosaur shit?

It’s so important that the last major bug was squashed 15 years ago

There are no such systems. What instead happens is that the surrounding business process gets distorted to work around the unfixed major bugs. And then, everyone involved retires and nobody knows anymore why things are done that way.

We know. Those of us who are paying attention, anyway.

🌍🧑‍🚀🔫🧑‍🚀

AI is going to write it.

Yep, months is a joke, doubly so when talking about tens of millions of lines of code and also COBOL specifically.

This is going to be a hilarious disaster but not so hilarious when people who need the benefits need them and won't be able to get them.

Okay but have you ever tried just throwing genAI at the problem and not caring about the consequences?

They have an experienced team of teenagers don't worry.

That is the mother load of all code bases. Probably still some COBOL if not mostly cobol.

By rebuild I don't think they mean it's going to function the same. ...just torn apart and replaced.

I mean this is a great example of what happens when you put conservative men in power who think they know what they are doing but are just going to loudly, incompetently and incorrectly re-invent the wheel while everyone else suffers from not having an actual practical solution.

Yep, this is it. Show how “broken” it is by breaking it, and enough of the population won’t even notice when it’s “fixed” and they’re only getting 2/3 of what they were before (and are entitled to). Plus grift, etc.

AT BEST it's gonna be some ridiculous npm svalbard worth of projects in one tree, require all new hardware, and declare bankruptcy on the way. Canada did this with the Phoenix Pay System, except didn't have 'efficient' funding so it only sucked but didn't die.

It's not a case of "seeing the code isn't perfect" but rather, not understanding the myriad problems the code is solving or mitigating.

I'm reminded of this shitshow:

https://en.m.wikipedia.org/wiki/2010_Queensland_Health_payroll_system_implementation

Queensland is a state of about 3m people in Australia. Their health service employs about 100k people. They ended up spending about 900m USD to develop their payroll software and fix the fuck ups it caused.

I'm an accountant by trade, there's a classic "techbro does accounting" style of development we see a lot. Like if you hadn't spent a career learning how complex accounting can be, it would be easy to look at a payroll system and conclude "it's just a database with some rules".

I'm sure the doge boys are expert grock vibe coders, it will be fine, they've got big ballz on the team, what could possibly go wrong? /s

I did such a thing, but I had a big advantage: the codebase had been done by people who had never really learned to code, and I was a seasoned programmer with 20 years of experience.

Yeah, I've cleaned up the messes that idiots like that have left.

Yeah, this is going to end in disaster.

That happens. Even if said new programmer had seen before that IRL the important part of that codebase consists of specific domain area quirks, scarcely documented and understood. They have an advantage in doing something good for the specific stage of that system's evolution, but a huge disadvantage in knowing what the hell it really does.

This is my suspicion as well.

If they planned a 1 month migration of a small component, 6 months to complete would be pretty lucky imo. Refactoring Legacy Code mentions the 2.0 approach they're taking. Spoiler alert, it doesn't work...

They are trying to break the government beyond all repair. At that point they'll say it's the Democrats that broke it.

Their cult members will swallow the lie hook line and sinker, and continue to keep them in power. (Side note, this will be made easier by gutting all election oversight as part of the package.)

Meanwhile, all that tax money we paid into Social Security, SNAP, Medicaid and Medicare, Unemployment insurance... basically any program meant to help people, will flow directly into billionaire's pockets.

I hope they all fly around in Boeing aircraft together.

I'm on SSDI and VA disability. I can't work. I'm terrified every time they even breath in the direction of either place. I'd be super fucked. Even losing SSDI would destroy us. That's about 3/4 of our mortgage payment. So many other vets are in the same boat.

how do you do this? probably good for all of us to know

No they’d just blame the liberals and maintain 100% loyalty lol.

Nah, they'll just say the Democrats are still worse because reasons

Yes. They need to move quickly. Public opinion is already shifting against Trump and Musk, and right now they are vulnerable.

Musk would probably think that's just fine.

Server-side javascript is an abomination, but there's more of it around than you might think.

they will make chatGPT do it and then not verify.

I know how to code, how to manage programs, how to architect huge safety-critical systems, and quite a few other things, and I know that you are right. I'd give it 5 to 7 years if it were adequately resourced, there was political commitment, and the stakeholders could be made to agree a set of requirements, then not change them unless there's a really convincing reason (conflicts with other requirements, impossible to implement, breaks everything, etc).

And the validation and verification of such a system could itself take a year or more, if it's well-planned and correctly executed.

I think the hope is that it fails, they don’t want social security to work.

If Trump is smart, he'll let Musk do all the unpopular project 2025 stuff, then throw him in prison at the end and escape the blame personally. This way he gets to keep popularity with his base while telling his donors they got everything they asked for. It's what all dictators do, really.

So the way things already were?

Nah I think it will just be

 undefined

    
const benefits = false;

  

Bro. Check it out bro, we're gonna like make it this dope Electron app, bro. It'll interface with X, bro and everyone will have to login there to get their money, bro. Don't worry tho, you'll get paid in recession-proof Trumpbux crypto currency as long as you claim it in time. But X gets a fee of 60% bro.

Seriously bro we like hired a bunch of grads that took a one week X created code boot camp that like you know revolved around a language big balls created called "cyber coin purse++". On second thought bro we're rewriting it in that. Should be like 2 weeks to rewrite it cuz old people wrote the current code and they're like old or whatever bro. Like I live in an old person's basement and they're just like old, bro.

In theory it is a horrible idea. No port like this ever works out. An incremental approach has much higher chance of success but will take long.

in theory yeah it's not a bad idea but it's also 60 million lines of a cobol

If these people are good enough to redo federal codebases I should apply for a coding job already

How many bugs? How will they secretly siphon money to their accounts? How much access will the Russians have? Who's gonna get discriminated against?

Sooo only gonna blow up the test databases?

Over 70 million including many retirees, orphans, and disabled workers. The people most in need of help and the reason that trying to run a government like a capitalist business is one of the dumbest forms of government organization ever. A quick way to radicalize someone against you is to harm their family or take their money.

Suicide bombers on mobility scooters. Be very afraid.

COBOL is perfectly suitable for financial purposes for which it was designed.

Nobody uses COBOL for greenfield projects, even in the banking and financial sectors. And, as people with COBOL expertise die of old age, it becomes increasingly unmaintainable.

DARVO is all you need to avoid blame. Deny. Reverse Victim and Offender. Incredibly effective with everyone except the genuinely principled.

When he suggested PayPal convert its data centers from Linux to Microsoft Server, I knew he was a wrong'un. That was one of the reasons he got booted out of PayPal, by the way. Even the other bullshitter tech bros knew how stupid that idea was.

Is it a “risk” if it’s the desired outcome?