Banks and Social Security are not using #COBOL because it’s so great, but because it’s so hard to migrate to something else.

They should and do migrate away … very slowly and carefully.

(I‘m in banking.)

I don’t dislike COBOL. I program daily in #ABAP, which looks like a COBOL dialect. I don’t mind verbosity, verbosity increases readability and debuggability. No devs? Train some. But COBOL developers tell me it has serious limitations. It’s just old.

What could possibly go wrong? #DOGE to rapidly rebuild #SocialSecurity codebase.
A safe and proper rewrite should take years not months.
The project organized by #Musk lieutenant #SteveDavis, multiple sources who were not given permission to talktell WIRED, and aims to migrate all #SSA systems off #COBOL, one of the first common business-oriented programming languages, and onto a more modern replacement like #Java within a scheduled tight timeframe of a few months.
https://arstechnica.com/tech-policy/2025/03/what-could-possibly-go-wrong-doge-to-rapidly-rebuild-social-security-codebase/

Am I the only one who's thinking of learning #COBOL just to see whether the idea of replacing the US Social Security codebase is *exactly* as delusional as I think or whether it is far more delusional than that?

https://gizmodo.com/doge-plans-to-rewrite-entire-social-security-codebase-in-just-a-few-months-report-2000582062

There’s [intentionally] so much to keep track of right now as #DOGE et al flood the zone with shit.

I’ve enjoyed chuckling at the doge-bags’ foolish / impossible shenanigans, like the rewriting of the Social Security database.

However, this wise commenter reminds us of the over-arching goal:

> The point of this isn't to rewrite SSA, it's to funnel SSA through Musk's burgeoning American WeChat and, as a stretch goal, break SSA so badly that they can just do away with it entirely.

> We need to stop sanewashing and positing that these twerps' intent is something noble but their methods are misguided.

> They aren't.

> They want to break the state and make money on the pieces. Everything else is window dressing.

Hat tip to @arstechnica for all of your great reporting.

https://arstechnica.com/tech-policy/2025/03/what-could-possibly-go-wrong-doge-to-rapidly-rebuild-social-security-codebase/?comments=1&post=43623499

I have no doubt that at least some of the legacy COBOL code at the SSA needs a rewrite.

Having had to maintain 20+ year old legacy COBOL code, it can get pretty ugly.

Having had to maintain 20+year old COBOL code that calculated various kinds of benefits for 80,000+ people located in different states with all the various odd situations that exist with the complexities that are humans, thinking a complete rewrite of the SSA system can be successfully coded, tested, and implemented in a few months . . . is to be expected from someone who doesn't really have a functional understanding of how computers or computer systems work. You know, like someone who would think that there were people 150+ years old in the SSA system who were receiving benefits. #SocialSecurity #USPol #COBOL

Oh dear.
"DOGE Plans to Rebuild SSA Code Base in Months, Risking Benefits and System Collapse.
Social Security systems contain tens of millions of lines of code written in #COBOL, an archaic programming language. Safely rewriting that code would take years—#DOGE wants it done in months."
https://www.wired.com/story/doge-rebuild-social-security-administration-cobol-benefits/
All the world's computer science professors just sat up like meerkats, sensing the imminent arrival of a new high-profile disaster case study in how not to do a major software project.

Your periodic reminder that old #COBOL isn't only old COBOL. It is IBM assembler with some COBOL for the easier parts. Business apps written in assembler because the COBOL wasn't fast enough.

I can get the gist of code in most languages, but when I look at edin (assembler!) I can't figure out what is going on despite profuse comments.

@GossiTheDog

I asked #Grok: "What are the chances of success for Elon Musk's / DOGE's plan of rewriting (in two month's time) the SSA's COBOL codebase using AI?"

"[…] the chances of success […] are extremely low, likely in the single-digit percentage range (e.g., 5-10% at best).
This estimate assumes DOGE has exceptional resources and AI capabilities beyond what’s publicly known […] The stakes are high, and the margin for error is razor-thin."
https://x.com/i/grok/share/n9svocmPv28oPUXlCbbJruxTS

DOGE plans to migrate all Social Security Administration systems off COBOL in months, risking the integrity of SSA infrastructure. #DOGE #SocialSecurity #TechNews #COBOL #GovernmentTech #SSA #Infrastructure #ITNews #TechChallenges

#DOGE plans to rebuild #SSA's antiquated #COBOL code base in months instead of the years it'd take to do it safely, risking benefits & system collapse.

But, hey, when did Musk/Trump/DOGE ever give a crap about doing anything safely? Or right?

#SocialSecurity

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

I think the system fails if they try to do this. I understand the desire to convert off of Cobol, but this doesn't seem feasible to me, especially if you are performing unit testing and qa before the release to make sure the code actually works.

Any thoughts on if this is possible to accomplish?

#SSA #CodebaseConverion #cobol #Doge #Musk

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

DOGE is putting together a team to migrate the Social Security Administration’s computer systems entirely off one of the oldest programming languages in a matter of months, potentially putting the integrity of the system—and the benefits on which tens of millions of Americans rely—at risk. Read more at @WIRED (subscription may be required). #SocialSecurity #DOGE #ElonMusk #SSA #COBOL #Tech #Technology https://flip.it/98qM1j

Anybody that suggests a decades old COBOL codebase be replaced with Java is someone you don’t have to take seriously anymore.