feel kinda silly...

id10t · 12-05-2025, 06:44 AM

Just submitted a pull request (change to master branch of our code base for back end services) for a 3 character change - adding ", 0" to the end of a sql string. Already built and deployed to prod (because I can) but a formal change request must be made....

Since it is Friday and we're feelin' kinda punchy around here, I fully expect it to turn into a series of change requests that will go into fine detail as to why the space between the comma and the 0 should be removed or re-added....

masraum · 12-05-2025, 06:47 AM

Yeah, but a mature org should have all changes and updates tracked, even small ones.

Do you guys have a peer review process where when you submit a change, someone has to give it a once over to see if they think it's good? Or are you "the man" and asking others to look at your stuff would be a waste of time because no one would have any clue?

oldE · 12-05-2025, 06:48 AM

The devil is in the details

As you pointed out, too many people feel the compulsion to contribute.
You should have submitted 10 minutes before everyone was set to leave for the weekend.

Alan A · 12-05-2025, 06:53 AM

and then it would have sat for a month.

masraum · 12-05-2025, 06:55 AM

We recently had an issue where an app that reviews device configs for drift from the "config standard" is monitored and reported. The app was telling us to delete a physical interface on the device. The problem was that the line that was being looked at wasn't anchored. One character ($) was added to one line and that fixed the issue. The code had been in place and working for years with no issues, but a new device was put in place that had a small difference in the config that exposed the issue which had existed without causing a problem for years.

Alan A · 12-05-2025, 07:10 AM

Since it's Friday and I'm not looking to install anything new today...

Try looking at 50 year old undocumented assembler, where the 40 characters per line of comments are the only source if information on whatever twisted logic the original developer thought they were doing.

Or if you really want an exercise in WTAF frustration - decompile some from the load module because someone deleted the source and try to follow the logic. And no, Opus isn't as much help here as you might hope it would be.

id10t · 12-05-2025, 07:12 AM

Quote:

Originally Posted by masraum

Yeah, but a mature org should have all changes and updates tracked, even small ones.

Do you guys have a peer review process where when you submit a change, someone has to give it a once over to see if they think it's good? Or are you "the man" and asking others to look at your stuff would be a waste of time because no one would have any clue?

I am fortunate to be surrounded by competent coworkers for the most part... and everyones changes gets reviewed like this, and must be approved by 2 people.

However, I am the only one that can deploy straight to prod since technically my job is linux admin not developer - I just spend 38 of my 40 hours writing code.

id10t · 12-05-2025, 07:16 AM

Quote:

Originally Posted by Alan A

Since it's Friday and I'm not looking to install anything new today...

Try looking at 50 year old undocumented assembler, where the 40 characters per line of comments are the only source if information on whatever twisted logic the original developer thought they were doing.

Or if you really want an exercise in WTAF frustration - decompile some from the load module because someone deleted the source and try to follow the logic. And no, Opus isn't as much help here as you might hope it would be.

We still have a large dependency on 25 year old code and db schemes written by former VSAM developers. Data duplicated everywhere since they couldn't grok a join, the languages (VG/EGL) were originally by IBM as a way to move towards Java, you write in EGL/VG and it generates Java+JSP or COBOL code. We can't get the installer any more, we're running it on Win10 vms made from ripping a physical Win7 install out and upgrading the OS on it. We can't even buy licenses, we tried and IBM said "we have no way of charging you for it, just ignore the license warning, here's a nice letter from our legal folk giving permission"

masraum · 12-05-2025, 07:36 AM

Quote:

Originally Posted by id10t

we still have a large dependency on 25 year old code and db schemes written by former vsam developers. Data duplicated everywhere since they couldn't grok a join, the languages (vg/egl) were originally by ibm as a way to move towards java, you write in egl/vg and it generates java+jsp or cobol code. We can't get the installer any more, we're running it on win10 vms made from ripping a physical win7 install out and upgrading the os on it. We can't even buy licenses, we tried and ibm said "we have no way of charging you for it, just ignore the license warning, here's a nice letter from our legal folk giving permission"

ROFLMAO!

Classic.

masraum · 12-05-2025, 07:38 AM

masraum · 12-05-2025, 07:41 AM

I work for one of the largest banks in the world, and we do have Excel spreadsheets being used for all sorts of stuff.

KFC911 · 12-05-2025, 07:42 AM

LOL ... I just don't understand you guys anymore

Yippee!

Alan A · 12-05-2025, 08:13 AM

Quote:

Originally Posted by id10t

We still have a large dependency on 25 year old code and db schemes written by former VSAM developers. Data duplicated everywhere since they couldn't grok a join, the languages (VG/EGL) were originally by IBM as a way to move towards Java, you write in EGL/VG and it generates Java+JSP or COBOL code. We can't get the installer any more, we're running it on Win10 vms made from ripping a physical Win7 install out and upgrading the OS on it. We can't even buy licenses, we tried and IBM said "we have no way of charging you for it, just ignore the license warning, here's a nice letter from our legal folk giving permission"

That's AS/400 no? Ours are a bit bigger.

If so you'll get a kick out of this - how many times do we need to read "it's the future":
https://www.itjungle.com/2007/09/17/tfh091707-story01/

The "can generate cobol" might give you a path forward.
Especially if you use that to then generate java.

As an aside - since you mentioned VSAM - we ended up writing our own file access code.
I (vaguely) remember than I did a bunch of work optimizing it back in the early 00s. VSAM was too slow and the file owning requirement was too onerous. EOF pointers in a dataspace FTW.

KFC911 · 12-05-2025, 08:19 AM

Maybe y'all were running VSAM on tiny blue boxes

?

I could tune a VSAM cluster that housed all of my humongous data bases, on hardware that only the very largest had ... like the biggliest.

I can fish

VSAM ... meh...

VTAM ...

Alan A · 12-05-2025, 08:29 AM

Quote:

Originally Posted by KFC911

Maybe y'all were running VSAM on tiny blue boxes

?

I could tune a VSAM cluster that housed all of my humongous data bases, on hardware that only the very largest had ... like the biggliest.

I can fish

VSAM ... meh...

VTAM ...

We needed to update from multiple address spaces without a CICS FOR. OLTP system didn't run CICS and we weren't rewriting to give big blue more $. The setup we wrote was similar to VSAM - but with a multi-layer CI and TI in memory and backwards pointers so you could do reverse order sequential reads. It replaced a system that replicated the an index on track 0 of each cylinder to reduce the seek time - that's how old it was.

And don't get me started on VTAM...
I still have the memory scars from switching to binary transfer and doing unsolicited writes to speed up data transfer to our VB apps with their own 3270 emulator. LZW compression for the win...

KFC911 · 08:44 AM

Limple-Ziev (sp

?) compression ... or compaction... I came outta communications r&d ...

VTAM sytems programmer for years (so I knew CICS, IMS, etc.) ... but ACF/NCP was my domain ... nobody, anywhere, was gonna look over my changes either .... and I could reach Larry R. at the Federal Reserve when "they" had issues ... that "I" had to fix ... with $$$ at stake.

.

Buncha geeks here I reckon ...

Deschodt · 12-10-2025, 08:41 AM

Quote:

Originally Posted by id10t

Since it is Friday and we're feelin' kinda punchy around here, I fully expect it to turn into a series of change requests that will go into fine detail as to why the space between the comma and the 0 should be removed or re-added....

I've sat on a change board that debated the color of a new icon for 3 months... In the end I suggested Brown because it was $%$%$% !
I have to open a change request for deploying a shortcut...
Or change a 0 to a 1 in registry - but in fairness the impact can be huge..
Change boards are rubber stampers that largely do not understand the job/consequences and just want to dot their "i" to cover their behind ;-)
Never in my career have they uncovered an impending catastrophy, but on the flip side when stuff goes south we can turn around and point to the approval !! LOL

masraum · 12-10-2025, 10:07 AM

We have a change review board for our small corner of the HUGE bank that I work for. I/my group supports a tiny corner (that deals in huge sums of money) of the Investment Banking group.

Our group reviews all of our changes weekly, and sometimes, in depth. We have very few rubber stamped changes except where they are SOP changes that are done on a weekly basis with very low risk and no, or at least, minimal history of issues. The folks in our group that approve are technical and understand what's being done. We occasionally have changes that are deferred due to risk or potential for problems. Once we've reviewed our changes, they are presented to our customers who then must also approve. Our customers are very smart technical folks, but they often don't really understand what we're presenting to them. In my experience, whether the business approves our changes is largely down to the person that presents the change. Some folks are able to present a change so they come off confident and the change seems very low risk. Other folks, even if very compentent and thorough don't instill the same confidence and are much more likely to have their changes heavily questioned. Because of the potential financial impact, the business will occasionally be overcautious and make us defer changes out of fear.

id10t · 12-10-2025, 10:23 AM

Another silly change... that I had to make in Java (Spring) for one of our about to retire people who refused to move on and learn Java 20 years ago...

masraum · 12-10-2025, 10:54 AM

Quote:

Originally Posted by id10t

Another silly change... that I had to make in Java (Spring) for one of our about to retire people who refused to move on and learn Java 20 years ago...

Sounds like everyone is ready for their retirement.

That looks so simple that I think even I could have done it, and I don't know java.