Zoomer reporting in. Fill me in about IBM, I only know that they used to be too big to fail and no one ever got fired for buying an IBM.
1. Why didn't the IBM boomers just clone a *nix instead of hiring shill gates or later, writing OS/2? 2. Why didn't they stop paying programmers ny KLoC before it's too late and what's the current standard now? 3. Apart from the ani-trust BS they faced in 1969-82, are they responsible for other actia; anti-trust practices that opposed freedom? 4. What were the final nails to IBM's coffin? How did it fuck up so bad after the PC junior? 5. Should we _actually_ be concerned that IBM controls Linux now?
Fellow zoomer here, I know a little bit of things about stuff.
>Why didn't the IBM boomers just clone a *nix instead of hiring shill gates or later, writing OS/2? Do you mean a Unix clone like Linux or Minix, or a Unix distro based on System V? Because they've always had AIX, and it's still around today. But these days it runs headless on servers. Back when AIX was available on workstations it ran CDE. >Why didn't they stop paying programmers ny KLoC before it's too late and what's the current standard now? I don't know what you're asking. >Apart from the ani-trust BS they faced in 1969-82, are they responsible for other actia; anti-trust practices that opposed freedom? I'm not sure. >What were the final nails to IBM's coffin? How did it fuck up so bad after the PC junior? They stopped making personal computers like their laptops and workstations and moved completely to cloud shit. They still have their POWER architecture but the last time that was widely used in consumer PCs was in PowerPC Macs, and I think those chips were made by Motorola. I'm not sure how their licensing agreements worked. There are POWER9 workstations like the Talos II being made so there's obviously some market for POWER based workstations. But they're made by Raptor Engineering, which is a third part why just buys the parts and assembles them in a certain form factor. >Should we _actually_ be concerned that IBM controls Linux now? No. Hopefully this will light a fire under the asses of other kernel projects that have far superior technologies. I hope that seL4 will be worked on more in the future and eventually replace Linux all together, or even run a stripped down version of Linux as a kernel module for legacy compatibility. At least 2/3 of the GNU utilities need to be rewritten from scratch and Gnome shit needs to be trashed completely.
Chase Thomas
>I don't know what you're asking. IBM used to pay programmers a certain amount per 100 LoC. This resulted in bloated and inefficient codebase. Back in the day, the compilers weren't smart enough to optimize them all out and they and their customers had to pay premium for the storage and memory that shouldn't have in the first place. Apparently one of the microshaft engineers once made a joke out of IBM fags by rewriting a 33,000 characters IBM program within 200 characters (yeah, it was THAT bad)
Chase Clark
>IBM used to pay programmers a certain amount per 100 LoC That's fucking dumb. They should be salaried and given yearly performance reviews based on the end results of their work.
Josiah Hill
ibm isn't in a coffin. they've been a profitable company for a long time and they just bought red hat
they didn't stop making computers because they went bankrupt, they just didn't decide it was worth it anymore.
they make their money in supercomputers, academic, and enterprise computing
Angel Lopez
1000*
Justin Powell
>>IBM used to pay programmers a certain amount per 100 LoC This was common practice in the 80's. The "big bosses," who didn't understand code needed a way to evaluate performance. Obviously, this WAS NOT the right way.
Ryan Robinson
IBM made the machines that Hilter used to number the Jews before extermination. so any Jewish person who uses IBM kit is a traitor and literally shitting on the graves of their ancestors.
William Cox
The problem is that the reviewer is often more clueless as to what is going on than the reviewee.
Christopher Clark
It wasn't dumb, it was just that before the 70s companies tried not to jew employees too much, and vice versa. So there was a common agreement that programmers weren't just going to write useless code to artificially inflate their metrics. And given that, making LOC comparisons between code written in the same language tends to work just fine. But it wasn't the 60s anymore, it was the era of hippies and easy money, so everybody began jewing everybody and LOC became an useless metric.
Justin Anderson
wtf I love IBM now!
Daniel Hall
Oy vey! Rememba da 600 gorillion punchcard machines!
Chase Howard
What the fuck is a "Zoomer?" And as for your questions:
1 - IBM has AIX and has had for a while. Microsoft actually had Xenix long before it had DOS (it paid for a guy to develop DOS for them, and DOS in itself was a clone of an older Digital operating system). But back then, UNIX was encumbered by literally dozens or even hundreds of competing variants, often incompatible with most hardware or each other, and nearly all of them encumbered by bizarre licensing agreements and copyright - battles over which rage to this day, and at one point even posed a serious risk to Linux.
UNIX also wasn't particularly friendly. It had a lot more quirks and issues than Linux does these days, and OS/2 (and Windows) was intended to be an operating system that could be targeted at the middle-management class - people who just wanted a word processor and calculator, basically. Which is the reason why Windows and OS/2 abstract so much away from the users.
2 - Lines of Code was just some random metric invented by an un-technical management when the idea of "computer science" was a twinkle in someone's eye. The same way journalistic outlets will pay based on certain word counts, etc. Ironically, we still roughly audit programs by "lines of code per hour", despite the fact it's a terrible metric for OO and other non-procedural paradigms (and arguably also those).
3 - I won't speak about anti-trust stuff, because back in those days everyone was doing it, but IBM keeps very tight control over source code, doesn't disclose the content of security patches for a large majority of their current operating systems, and engages in other practices that I would quite frankly consider archaic in the world of software.
(continuing)
Kayden Wilson
>1. Why didn't the IBM boomers just clone a *nix instead of hiring shill gates or later, writing OS/2? They did, AIX exists. AIX was big back in the day of commercial Unix, but everyone hated it because of autistic security.
Like all commercial Unix, AIX is pretty much dead, and is only maintained due to very old contracts and machines that still run it.
>4. What were the final nails to IBM's coffin? How did it fuck up so bad after the PC junior? A obsession with manipulating their own stocks and pleasing their investors. Their business for the most part depends on their aging mainframes, which they are only making a profit on from very old support contracts. They also sell some proprietary software, and these two things do keep them afloat, but they aren't a company that can keep up with market developments.
IBM is widely regarded as a terrible place to work and you will find hundreds of ex-IBM employees speaking out against the company. In fact, the website above is set up for that purpose.
This is why they bought Red Hat. However, because of IBM's love of patent trolling, restrictive licensing and mass layoffs you can expect all of Red Hat's software stack, including the parts which the Linux desktop depends on, to go to shit as IBM will eat Red Hat alive and kill it.
>5. Should we _actually_ be concerned that IBM controls Linux now? Yes, absolutely. The only thing the company makes money on is old mainframes, old software and suing for patent violations. They frequently do mass layoffs, they fucking invented Indian outsourcing to the point where people call them "Indian Business Machines" and they are generally a cancer that eats up other companies and kills everything good about them, leaving only the bad parts for absorption into the IBM.
IBM is genuinely a terrible company. They are dying a slow, painful death.
Cameron Evans
4 - IBM is fine? It's far from dead. It produces mainframes, minicomputers, supercomputers, servers, hardware, and so on. Mainframes probably sound like something out of the 70s to you, but they're actively developed and power the backend of the vast majority of the world's financial institutions (banks, insurance companies, etc) and also have a significant amount of use in the airline industry, and IBM has a huge lead over any rivals here. Minicomputers will be found at the heard of most companies that are older than a few years. IBM's Blue Gene supercomputers are some of the only ones that can compete with Cray on price, and they're more efficient. IBM is doing perfectly fine.
5 - Not particularly. If it was HP, I'd be worried. But IBM has worked relatively closely with the Linux world for a long time now (which sounds odd given my comment on their archaic practices above, but IBM is a big beast with many arms, and the left often doesn't give a shit what the right does etc), and Red Hat in particular. I suspect much won't change, either structurally within Red Hat, or with their practices, at least for a while.
Point out one company that got bought by IBM and ended up in a healthy state afterwards.
I'll be waiting.
Christian Bell
>It's far from dead. It produces mainframes, minicomputers, supercomputers, servers, hardware, and so on. But it *is* dying. All of what you mentioned are products where IBM's presence and market share are in decline, except for mainframes which are just generally in decline altogether and being replaced with clusters.
The only reason people buy IBM mainframes is because they have existing software that only runs on them.
Jaxson Foster
I'm going to make the argument that basically every company that gets bought ends up getting fucked up. This is not restricted to IBM. Look at the fate of everything Oracle touches, or HP. Mainframes are literally irreplaceable. Clusters (or god forbid, "cloud" computing) can't replace them - not just because of the bespoke software involved, like all the fancy CICS shit, but also because of their infinite uptime, functional immortality, and unique design benefits (LPARs and the RACF/Top Secret access controls being good examples). Certainly, younger companies won't have any frames, but they're everywhere in governments, and probably ensure your country's economy doesn't fall down overnight. The difference between Linux/WIndows and z/OS or z/TPF is utterly phenomenal in terms of capabilities. I've consulted on a number of projects where some exec has had the bright idea to switch to OTS software and operating systems, and it usually ends up with a few million dollars of loses and an IBM team being teleported onsite to clean up the mess and revert the whole clusterfuck as soon as possible. You don't hear about these mishaps unless they take out a bank for a few hours, but they happen very bloody regularly.
Zachary Parker
>Clusters (or god forbid, "cloud" computing) can't replace them They can in a huge amount of cases though, which is why they're doing it. Sometimes you just need a really fucking powerful machine with good reliability, and that's not a cluster, and that's fine. The point is that it doesn't need to be an IBM mainframe and that the market that needs this is shrinking by the day, and very much dying outside of heavily specialised contexts.
>like all the fancy CICS shit CICS hardly qualifies as "fancy". It's just old and hard to replace.
>Certainly, younger companies won't have any frames, but they're everywhere in governments, and probably ensure your country's economy doesn't fall down overnight. Because said governments and banks have a software stack that's older than both you and me most likely, and they have architected their entire computer network on it. It is VERY hard to untangle that sort of stuff, which is why you have so many failure stories.
However, people HAVE untangled it and it will continue to happen. It's just a question of time.
>LPAR Is not a unique technology to IBM mainframes and will work on any system that has a proper hardware hypervisor. This requires OS support but z/OS is not the only operating system that provides it.
>RACF/Top Secret access controls being good examples RCAF is software that doesn't require an actual IBM mainframe, and could be ported to other systems. It's just sold with restrictive licensing *to* IBM customers.
You're right, no alternative exists to that aside from other stuff like ACF2, Top Secret etc. but that's likely to change at some point and once it does you'll kill IBM.
>You don't hear about these mishaps unless they take out a bank for a few hours, but they happen very bloody regularly. Of course they do, but my point is that this is ALL IBM has and it's all IBM has had for over 2 decades now. They are not a growing company, and thus they are slowly dying.
