very cool!
Launch HN: Hypercubic (YC F25) – AI for COBOL and Mainframes
41–50 of 60 posts
Re: Launch HN: Hypercubic (YC F25) – AI for COBOL and Mainframes
#42Mainframe COBOL to POSIX native machine language or Java is largely a solved problem. 40 year old code that hasn't been touched in 30 years "ain't broke" and likely doesn't need fixing IMHO. Our take is get it to POSIX/Java so you can sort out all the OTHER STUFF and then rewrite/re-architect at your leisure. {disclosure: we're the GCC COBOL team}.
Re: Launch HN: Hypercubic (YC F25) – AI for COBOL and Mainframes
#43I heard the story once on how you migrate these old systems: 1 you get a super extensive test suite of input - output pairs 2 you do a "line by line" reimplementation in Java (well banks like it). 3 you run the test suite and track your progress 4 when you get to 100 percent, you send the same traffic to both systems and shadow run the new implementation. Depending on how that goes you either give up, go back to work…
Yeah, we've heard the same "big bang" story a bunch of times as well. However, step 1 (extensive test suite) is often missing and is something you'd have to do as part of the modernization effort as well. Overall, it is a pretty hairy and difficult problem.
Sort of related to what bloop is trying to do too right?
Re: Launch HN: Hypercubic (YC F25) – AI for COBOL and Mainframes
#44> rapidly increasing
Surely not at a rate faster than one year per year?
Re: Launch HN: Hypercubic (YC F25) – AI for COBOL and Mainframes
#45There is a large graveyard of people trying to escape the mainframe, replace COBOL, etc.
Absolutely true, and the challenge is that a large portion of modernization projects fail (around 70%). The main reasons are the loss of institutional knowledge, the difficulty of untangling 20–30-year-old code that few understand, and, most importantly, ensuring the new system is a true 1:1 functional replica of the original via testing. Modernization is an incredibly expensive process involving numerous SMEs, movin…
IMO, not the best use case for LLMs.
COBOL projects have millions of lines of code. Any prompt/reasoning will rapidly fill the context window of any model.
And you'll probably have a better luck if you had tokenization understands COBOL keywords.
You probably have better luck implementing a data miner that slowly digests all the code and requirements into a proprietary information retrieval solution or ontology that can help answer questions...
What an engineer tells you can be inaccurate, incomplete, outdated, etc.
Re: Launch HN: Hypercubic (YC F25) – AI for COBOL and Mainframes
#46> What’s left behind are opaque, black box systems with almost no one who understands how they work. Maybe 50-year-old COBOL programs are the original neural networks.
Really good money in it for people who want to learn COBOL though.
> Really good money in it for people who want to learn COBOL though.
False, all those jobs were outsource and offshored long ago.
Re: Launch HN: Hypercubic (YC F25) – AI for COBOL and Mainframes
#47Earlier quoted context omitted.
Really good money in it for people who want to learn COBOL though.
> Really good money in it for people who want to learn COBOL though. False, all those jobs were outsource and offshored long ago.
A lot of them are domestic, either in the public sector directly (so probably not really good money, but usually pretty decent benefits and job security) or with government contractors which require the work to be domestically (so maybe good money.)
Re: Launch HN: Hypercubic (YC F25) – AI for COBOL and Mainframes
#48Mainframe COBOL to POSIX native machine language or Java is largely a solved problem. 40 year old code that hasn't been touched in 30 years "ain't broke" and likely doesn't need fixing IMHO. Our take is get it to POSIX/Java so you can sort out all the OTHER STUFF and then rewrite/re-architect at your leisure. {disclosure: we're the GCC COBOL team}.
[flagged]
Re: Launch HN: Hypercubic (YC F25) – AI for COBOL and Mainframes
#49Earlier quoted context omitted.
Yeah, we've heard the same "big bang" story a bunch of times as well. However, step 1 (extensive test suite) is often missing and is something you'd have to do as part of the modernization effort as well. Overall, it is a pretty hairy and difficult problem.
Sort of related to what bloop is trying to do too right?
Like Bloop, we’re also focused on modernization, but our approach extends beyond code to include the people behind these systems and capturing the institutional knowledge they hold.
Re: Launch HN: Hypercubic (YC F25) – AI for COBOL and Mainframes
#50I submitted this [0] story a few weeks ago, which led to some discussion and then being flagged since (I think) people were unsure of how verifiable the stats around COBOL were (the submitted page had more than a tinge of self-promotional language). I was curious to ask you, as domain experts, if you could talk more to the "70% of the Fortune 500 still run on mainframes" stat you mentioned. Where do these numbers com…
Rule of thumb — if a bank already was there and dealing with things 30 years ago, it likely has some cobol left. Generalizing — if the company had enough need for it 30 years ago, was big enough to but a mainframe and the thing they used it for barely changes — chances are it’s still there, if the company is still there. Banks absolutely do have it in house, in a dedicated secure site with a fence and a moat
This is accurate to what we've seen in the market.
If they were large enough to need compute 30-40+ years ago, they certainly have some mainframes running today. Think Walmart, United Airlines, JPMC, Geico, Coca Cola and so on.