I ain't no big computer-talking guy, but I'm thinking this plan might have a few of them there 'bugs' in it

Mar 29, 2025 6:58 AM

Pally01

Views

814

Likes

37

Dislikes

1

What could go wrong?

The source - https://www.wired.com/story/doge-rebuild-social-security-administration-cobol-benefits/

dystopia

doge

politics

technology

elon_musk

Java, really...

1 year ago | Likes 4 Dislikes 0

Old Tech doesn't mean bad tech. Wheels are as Old Tech as anything but they still work fine.

1 year ago | Likes 2 Dislikes 1

Like Java!?
Businesses have been abandoning Java for over a decade because it is so incurably insecure

1 year ago | Likes 4 Dislikes 0

I’m sure that they build in “bugs”, stuff which would give the broligarchs access later on. If I were the US, I’d consider every system Musk and the DOGEbags touched compromised.

1 year ago | Likes 2 Dislikes 0

11 months ago | Likes 2 Dislikes 0

As someone who has been doing this type of work for 25yrs, I will bet every single penny I have that this will be a collosal failure. Like, nobody with any experience at all would ever even suggest it would be done in even a year. People are going to die here.

10 months ago | Likes 2 Dislikes 0

Chesterton's Fence is gonna get knocked the fuck down with disastrous results!

1 year ago | Likes 2 Dislikes 0

He can’t even use the right glue on his wankpanzers.

1 year ago | Likes 4 Dislikes 0

Want that their goal all along. Remember,

By September 2000, while he [Musk]was away on a trip, the board of directors fired him. The main reason? His reckless decisions, including trying to switch PayPal's entire infrastructure from Unix to Microsoft Windows, sparking outrage from the tech team.

11 months ago | Likes 2 Dislikes 0

Don't forget DOGE, you can build a replacement database twice as fast with two developers on each keyboard.

1 year ago | Likes 6 Dislikes 0

I mean, there are some great arguments for pair programming, but pretty sure this isn't it

1 year ago | Likes 1 Dislikes 0

anyone that knows anything about cobol:

1 year ago | Likes 4 Dislikes 0

The truth is COBOL is easy because so much of it is done for you. Rewriting it in something else means you have to know what COBOL was doing behind the covers of the original COBOL code. Trust me, it's not as straight foreword as they want you to believe.

1 year ago | Likes 4 Dislikes 0

if they actually going to do this its going to be hella fun to watch these dumbasses pouring more fuel on theit shiny new (probaby java) dumpster fire every day.

1 year ago | Likes 2 Dislikes 0

Absolutely uneccessary and contrived to disable or destroy the ability of the government to deliver social security. This will kill people.

1 year ago | Likes 17 Dislikes 1

Rebuilding the code is absolutely needed and has been delayed and understaffed and underfunded for decades. This is about the opposite of what a sane reimplementation would be however.

1 year ago | Likes 3 Dislikes 0

Their playbook is to use a grain of truth to give their lies the illusion of legitimacy. It’s equally frustrating and effective manipulation

1 year ago | Likes 2 Dislikes 0

they should share the street number of the building in which they'll be doing this. The innocent have nothing to fear, right?

1 year ago | Likes 7 Dislikes 0

Probably one teenager's laptop is where. He made a website once, so he's a real 'whizz kid' [sic]

1 year ago | Likes 3 Dislikes 0

if matthew broderick is suddenly a Trump Supporter, I'm gonna be really, really pissed.

1 year ago | Likes 3 Dislikes 0

oh damn, I certainly hope not!

1 year ago | Likes 1 Dislikes 0

This falls in my professional ballpark! Migration off of COBOL is a reasonable goal.

Doing any migration in a hurry is always a bad idea - 2-3 years would be admirably fast, honestly, given the magnitude of the task and potential failure-mode downside. 5 years is probably still optimistic.

Their first best chance to do it right sailed out the door with firing the old-timers who know the current system.

1 year ago | Likes 15 Dislikes 0

I am not an IT guy, but I was working at a large bank when it purchased another bank and tried to integrate two account systems into one. It cost millions, took 2 years and about 98% of the data migrated without issue. That 2% resulted in national news stories, thousands of people being unable to access their money for weeks, 12-hour rolling shifts in branches and call centres for all the front-line staff and millions in compensation payouts. That's what success looks like, I'm told.

1 year ago | Likes 3 Dislikes 0

As a 61y old seasoned IT program mgt. That can be managed but requires some dedication, (lotsa) time, $$€€ and thus the will of project board to manage it.
Doing some extra dry runs to find these type of issues very likely van mitigate these issue.
But at projects I was able to convince the board to such actions we are talking 4-6 months extra thru time plus the budget for that period. And in a project ‘burning’ €1M a month that results in serious budgets.
But understimating the RISK is easy!

1 year ago | Likes 2 Dislikes 0

2 years is actually pretty damn good. Banks move verrrrry slowly. On purpose.

1 year ago | Likes 3 Dislikes 0

So, if I understand you right, knocking up an Excel Spreadsheet probably isn't going to cut it?

1 year ago | Likes 3 Dislikes 0

Probably not.

Having said that, I've seen Excel do some amazing things. Things it shouldn't do, but did anyway. Unholy things.

I worked my first IT job for a guy who believed you could use Lotus 123 for anything, and proceeded to do so with much success.

1 year ago | Likes 2 Dislikes 0

I still use Lotus Organizer...

1 year ago | Likes 1 Dislikes 0

No it’s not feasible as systems tend to be much much more complex than people think.
As a 61y old seasoned IT program manager that started beginning of 80’s with IBM.
That will NOT work at all!
I have been doing BIG project for companies that thought ‘our business processes are very simple & easy’. So replacing this >30y old IBM COBOL/RPG ERP system is gonna be easypeasy.
To end up with a MULTI €10M project lasting >4y.

1 year ago | Likes 3 Dislikes 0

Maybe.

Thing is - this is essentially banking. Keeping track of money and cutting checks isn't a new concept. There is EXTENSIVE research and best practice and expertise available to do the job in a reasonable (sub 5 year) timeframe and competent manner, even at this scale.

None of which Elon and his boy gang have. They'll fuck it up righteously. Which I suspect is their goal.

1 year ago | Likes 2 Dislikes 0

I have done a few banking project and can tell you that even those can be devilishly difficult. But IMHO banking in US is very old fashioned with (compared to EU and EU-NL) a much lower level of automation.
I can wire money here and it’s on your account in 10s.
From a distance the systems in US look very archaic!

1 year ago | Likes 1 Dislikes 0

Heh - they look that way because they are! At least the social security system is.

But having been part of recent (the last decade) banking upgrades - a few years is reasonable for even the biggest job =if the team doing it is competent=.

That last bit is the trick here. Elon's boys are far less competent than they think they are. God only knows how long they will take to succeed, if ever.

1 year ago | Likes 2 Dislikes 0

Ha, fun thing is that finance projects, ERP and banking is what I’ve been doing last 30y.
But biggest issues ALWAYS are level of commitment from the organisation itself and underestimating grossly the complexity of BIG systems.
I am a bit autistic so I kinda can dream these systems. But always rather far in the project something stupid pops up and I say that’s not gonna work or yeah possible but costs a LOT extra (€€ and thru time).
“Well we thought it was just a minor change”!
Yup, NO!

1 year ago | Likes 1 Dislikes 0