Look at Linux, the operating system that is an universal platform

It didn't continue its promotional campaign they started to introduce Linux to the world. Whatever the reason was, in my opinion, IBM made the second biggest mistake since the first of not pursuing OS2.

Could it be that MS has some influence at IBM?

IBM writes software like Apple writes software, to support it's hardware business.

Pushing OS2 would have cost a fortune, while people bought Windows instead.

I'm right there with you about the Linux advertising. Those ads were GLORIOUS.

"Never give up. Never surrender!"

By Grapthar's hammer...

No, IBM's biggest mistake was contracting with MS for MS-DOS. What a disaster that turned out to be. Can you imagine how different computing would be if they had made a better choice back then?

The "Prodigy" ad by IBM was perhaps the best tech ad ever (sorry "1984"). When John Wooden died a while back, I wrote about the ad because he was in it -- "A player that makes a team great is better than a great player." A lot of great stuff there, and incidentally most of what the folks tell the kid applies to FOSS as well.

In fact, in a new users class recently, I showed it, along with the "Truth Happens" short film by Red Hat.

And there's always Muhammad Ali: "Speak your mind. Don't back down."

Great stuff

And Penny Marshall: "Everything's about timing, kid." Which the SCOracle crowd obviously ignored.

khamul: one interesting question though is: was there a better choice? i mean technically it doesn't really matter. it is prettly clear that linux/unix and windows are the two major systems left today. no sign of ms-dos here. yes, windows was initially written on top of dos and it still can run dos programs, but who is to say that building on top of eg cp/m would have turned out any better. sure it would have turned out different. but better? hindsight is 20/20 as they say. i'd call choosing ms-dos unfortunate but i'd not blame IBM for the outcome today.

in fact even if IBM had managed somehow to get a unix-like system for their 8 bit machines, all that would have lead to is UNIX commonly available for PCs, which would not have compelled the creation of linux. not a total loss, we'd still have BSD and maybe hurd would have been pushed forward more, but they would have a much harder stance against a non-free desktop unix controlled by IBM or even microsoft (remember xenix?)

greetings, eMBee.

CP/M would not have been a better choice. DRI moved aggravatingly slowly unless pushed into a competitive corner. Microsoft led with directory trees and some semblance of access-control bits in MSDOS while CP/Mers (and MP/M users) were stuck with user--areas (one byte in the directory entry specifying what user out of 16 was supposed to see that file). The MSDOS FAT system was technically better than CP/M's extents IMO. I worked with a CP/M-80 machine equipped with a 5 meg HD; the reduction in TPA (transient program area, the space left over to run programs in) because of the added buffering was problematic, and trying to manage a flat sea of files was ludicrous.

When DRI came out with DRDOS it was strictly as a comeback, because MS pushed them. They would not have come up with such an OS on their own. The disconnect between simple single-tasking "disk monitor" OSes and a Unix-style true multiuser/multitasking OS would have made the very idea incomprehensible to most users, in part because even the simple singletasking OS would have been balky and repellent. Par of the success of the IBM PC/XT/AT was MSDOS; it made stuff doable, thus inviting users, and paved the way for more powerful machines running more powerful OSes.

Understand, ever since the antitrust trials and what they revealed, I despise Microsoft, but... Credit where due.

--crb3

Quoting:even the simple singletasking OS would have been balky and repellent.

I submit Geoworks Ensemble as a counterpoint to that. I'd rather do assignments and reports in that environment than be stuck with DOS's edlin or edit.

Geos/Geoworks was ahead of it's time, and could've been a real contender. After reading the Wiki, I'm not surprised to learn why it wasn't.

Surely IBM could have created an internal team to make their own OS, instead of outsourcing it to MS and losing control of it. They were a huge company at the time, they had the manpower.

Expediency was the name of the game at the time. They wanted something to market as decentralized computing, but they considered the centralized multi-user system their core business.

The point of the PC was merely to say "yes, we have that, too", but they weren't intending to support it aggressively enough to threaten their established business model.

IBM & other large computer manufacturers didn't want to bother with the headaches of little PC sales. They didn't believe in "Sell cheap, sell many" model. They preferred the "sell big, more profit" model.

Plus they were wary of antitrust issues and of deep commitments to internally-competitive products IIRC. Boca Raton was a bit of a skunkworks, staying small and flying low to stay under the radar of execs who would cripple the project (as PC Jr actually was) to avoid letting it cut into sales of bigger and more profitable iron; outsourcing the OS was a survival tactic. The outstanding success of the PC was a real tail-wags-dog event to Armonk.

@gus3: No argument on Geoworks, but look at the timeframes. The C64 (and even the C128) had too much plastic to be accepted as a business machine (Apple II had a molded plastic body but it looked beige-box); Ensemble for PC didn't come until after the wave had passed.

Edlin? Ed? If you were on a PC, you used WordStar or WordPerfect; I brought the WordStar habit with me from CP/M. Other editors were out there but those two dominated IIRC until Word showed up.

In 1990, IBM was still selling "dumb terminals" to connect to their mainframes. They did make add-on cards for PCs that allowed them to _act_ like dumb terminals, but even after 10 years and the PS/2 they still treated PCs as toys.

The last big project I had on the mainframe was changing all the cluster controllers (each supporting 32 dumb terminals) over to Tokenring from SNA.

Twisted pair ethernet and TCP/IP were far more disruptive than many people realize.

And then the EFF broke RSA-64 (I think) with a $10,000 cluster of cheap 386 motherboards that blew the doors off of multi-million dollar "supercomputers".

The period from 1990 to 1995 was truly revolutionary. What has happened since has been little more than fleshing out the changes seen during that time.

It was DES that the EFF cracked.

So much for the security of government "standards".

Speaking of which...

http://www.wired.com/threatlevel/2012/03/ff_nsadatacenter/al...

"The breakthrough was enormous, says the former official, and soon afterward the agency pulled the shade down tight on the project, even within the intelligence community and Congress. “Only the chairman and vice chairman and the two staff directors of each intelligence committee were told about it,” he says. The reason? “They were thinking that this computing breakthrough was going to give them the ability to crack current public encryption.”"

Anyone else thinking "Sneakers"?