Hacker News new | past | comments | ask | show | jobs | submit login
Gary Kildall (wikipedia.org)
87 points by shawndumas on June 5, 2015 | hide | past | favorite | 61 comments

The failure of Digital Research and IBM to reach a deal to bundle CP/M with the original PC provides one of the great "what-ifs" in computing history.

If that deal had been made, Microsoft wouldn't have gotten the big infusion of capital that turned it from a niche vendor into the behemoth it eventually became. Which means Windows, Office, and IE (if they were ever developed at all!) would have had to compete in the marketplace without MS being able to use their monopoly over PC operating systems as a lever to boost them up over the competition. Which in turn means a whole host of products whose names are just footnotes today -- Netscape, OS/2, GEM, TopView, DESQView, DeskMate, GEOS, Lotus 1-2-3, WordPerfect, Ami Pro, and on and on and on -- would have had access to the oxygen that could have turned them into the Next Big Thing. Oxygen that Microsoft was only able to deny them because of the muscle the failure of that original deal had given them.

Who knows what the computing world would look like today if that deal had been reached?

A (the?) big reason Microsoft got the IBM contract was because Bill Gate's mom (Mary Gates) was on the board of United Way with the CEO of IBM at the time (John Opel.)

Bill Gate's parents are impressive on their own rights.

http://en.wikipedia.org/wiki/Mary_Maxwell_Gates "In 1980, she discussed with John Opel, a fellow committee member who was the chairman of the International Business Machines Corporation, her son's company. Mr. Opel, by some accounts, mentioned Mrs. Gates to other I.B.M. executives.

A few weeks later, I.B.M. took a chance by hiring Microsoft, then a small software firm, to develop an operating system for its first personal computer."

IBM approached Microsoft to provide various software for their PC. Microsoft said they could provide BASIC, but not an Operating System, and helpfully directed IBM to Digital Research. Digital Research passed on the onerous terms (close to free).

IBM went back to Microsoft and said "now what?". Microsoft, fearing they would lose the BASIC deal as well, then purchased the rights to QDOS or 86-DOS (a CP/M clone) for $50,000 from Seattle Computer Products, and did the deal with IBM, agreeing to the onerous terms that Digital Research wouldn't. Which as we all know didn't actually turn out to be very onerous at all.

This is well documented in Triumph of the Nerds.

Mary Maxwell Gates may have provided an introduction, but it was the chutzpah, genius and desperation of Bill Gates that got the deal done.

Per Walter Isaacson [1], the influence of Mary Gates was more to endorse and confirm the deal, than to serve as an introduction. During a business trip, Mary Gates mentioned to John Opel that "My son is doing business with your company", but he answered that he wasn't aware of any deal, and never heard of any "Micro-soft". On her return, she joked with Bill that his deals with IBM shouldn't be that important.

Several weeks later, when deal was ready to be signed, IBM execs ran the agreement by John Opel, who then mentioned "oh, this must be Mary Gates' son. She's great. Yes, go ahead".

[1] http://smile.amazon.com/The-Innovators-Hackers-Geniuses-Revo...

But CPM/86 WAS shipped for the PC, by IBM. When I bought my IBM PC, I had a choice of PC-DOS or CPM/86. I chose the former because it was $40 and the latter was $180 or something like that.

It was the price that killed CPM/86, as nobody could give a reason why someone should pay so much more than for PC-DOS.

Don't forget UCSD Pascal! (Or maybe we should)

Turbo Pascal was considerably better - it was the non-assembler systems programming language of choice for PeeCees until 'C' took over.

Well yes, but it wasn't an alternative OS to MS-DOS and CPM :)

Indeed it wasn't!

That was supposed to be under the question "what about Turbo Pascal"? I musta goofed.

What was so bad about UCSD Pascal?


It was basically Java before machines got fast enough to make Java practical.

Definitely ahead of its time. For awhile it was possible to run the same applications on IBM, Apple II, TI-99/4A and more but it was cumbersome and it never had a "killer app".

It was also difficult to transfer files between disparate platforms in those days. If there had ever been a p-code applications market, distribution would have still required separate boxed copies for different platforms.

Then again you wouldn't have the initial JVM bytecode interpreter without p-code and the p-system, which were the bytecode and runtime environment used by UCSD Pascal. These were a direct inspiration for the Sun engineers developing Java.

As mentioned by others, it's pascal implementation and p-code concept were pretty cool. But it left a lot to be desired as an operating system. For example, IIRC, files had to be stored in continuous blocks on the disk and management of blocks were up to the user. If your file needed 10 blocks then you needed to find a hole on the disk with at least 10 blocks... which meant you might need to move other files around to make enough that hole.

Yeah, but it's not clear that given how IBM had cut a deal with Gates and pushed PC-DOS as the default, whether CP/M could realistically have continued as a legitimate alternative with a working business model.

I bought an IBM PC brand new. No software was bundled with it (besides ROM Basic). There was PC-DOS and CPM/86 available in IBM boxes. I picked the far cheaper one. So did about everyone else I knew at the time.

We had both available at work, and I tried CPM/86. I thought it was inferior and awkward compared to PC-DOS. Nobody could justify paying 4-5 times as much for it, even its adherents.

I don't know who set that high price on it, but it was sufficient to destroy CPM/86, no nefarious closed door maneuvering was necessary.

I wish I had a photo of that IBM CPM/86 box, few people believe it existed :-)

I believe you!

Some more context: In "Computer Chronicles" memorial episode, they too did not know who set the high price.

I remember coming upon a website of someone in the industry that asked Gary Kildall, "Who set the high price?". Gary said it was him.

I don't know if the page is up. I also don't know how to find it. I'm betting it's somewhere in the Internet Archive.

Even before OSes, Gates would argue with Allen: "If we give away our programming language for free, we will have high market share!" Allen would reply, "We would also go broke." Source: Andrews and Manes' "Gates": http://www.amazon.com/Gates-Microsofts-Reinvented-Industry-H...

Grabbing market share, low prices... combined w/ Scott Oki's strategic thinking (barriers of entry, legal contracts as competitive tool, from book "Microsoft Generation" http://www.amazon.com/Microsoft-First-Generation-Library-Edi...), and MSFT's motto, "We set the standard", it would make sense that MSFT would do it's best to price the OS as low as possible for long-term gain.

Also, there was a Pascal-based OS, but most people forget about that one.

As I understand it, PC-DOS was a quick and dirty copy of CP/M (Tim Paterson's QDOS).

It was bought from Tim and then further developed into DOS 1.0.

If PC-DOS didn't exist, nobody would have thought twice about the price of CP/M. It was still a small portion of the overall machine price. It's only in the direct comparison, when you realized you were getting less and spending more, that it fell short.

Could have something to do with Microsoft acquiring Seattle Computer DOS for $25,000 vs. Digital Research investing what I imagine is quite a bit more to port CP/M to the 8086.

A flip side of that: if the IBM/Microsoft deal had been made, the PC clone industry might not have succeeded, so PC might have meant exclusively "IBM PC". Forget about all the products that Microsoft crowded out; what would have happened to the hardware and software industry if a dozen competing "PC" manufacturers didn't spring up?

Maybe a more advanced architecture like the Amiga would have won out.

I was under the impression that most of the consumer hardware that has been dominating the scene over the past couple of decades has firmly been dictated by constraints pertaining to IBM PC compatibility. It's only with smartphones and tablets coming in to overtake the microcomputer/PC (though they still have a long way to go until self-hosting and everything) that this is being challenged.

I wouldn't go so far as "dictated", since that architecture has heavily evolved over time; PCI, AGP, PCIe, USB, VGA, and many other standards all occurred long after the dethroning of the IBM-branded PC.

And there's IBM's MicroChannel the ignoring of which greatly contributed to the continued growth of the PC industry as it shifted to 16 and 32 bits as standard.

It's possible that Microsoft would have "offered" to build the BIOS to get its basic in ROM and used control over the BIOS to put itself into the same gatekeeper position. Also Microsoft was a very resourceful, ambitious company, they would still have developed Office. I wouldn't even write off Windows, it could have easily been developed on top of CP/M.

Things could have been very different or they could have been very much the same.

I think the single behemoth is a bit of stable state for computing, especially in those days. So you would end up with just a different behemoth and it could have been worse.

> a whole host of products whose names are just footnotes today

Don't forget BeOS!

Apple's been doing a credible job of duplicating the features of BeOS in MacOS X, with the help of some key BeOS refugees (notably Dominic Giampaolo).

> notably Dominic Giampaolo

His book on file system design is still a must-read, imo.


This is very interesting. Could you give any specifics? Thanks!

HFS/HFS+ are very very similar to the BeOS filesystem.

Dominic Giampaolo developed that, so I assume that's what's being referred to.

There's a great writeup of the file system here: http://www.nobius.org/~dbg/practical-file-system-design.pdf

In particular, the labels and how they worked in BeOS are a big part of HFS, when Finder says it's "indexing" it's indexing all those extended attributes (among other things).

If Operating Systems had all been a couple of hundred dollars a pop, there wouldn't necessarily have been all the clones on disks and in homes that allowed those companies to be viable...and the actual story of Netscape is quite different from the myth. Netscape enjoyed a multi billion dollar exit and Andreessen is a leading venture capitalist.

Companies go bust. Microsoft didn't kill DEC.

There's no guarantee that CP/M would have maintained its lead. It could have been the case that MSFT didn't just curl up and die but fight and regain the monopoly. Say what you want about Microsoft, but Gates and Ballmer were/are smart and even smarter with business.

I imagine an alternate universe where you wrote :

"The failure of Microsoft and IBM to reach a deal to bundle DOS with the original PC provides one of the great "what-ifs" in computing history."

> Next Big Thing

WordStar and WordPerfect were pretty big. They were standard requirements for office jobs.

WordPerfect didn't fail because MS had a stranglehold on the industry but because they botched their transition to Windows.

My favorite is actually the OS/2 2.0 fiasco now, given that it involved unethical tactics like Microsoft Munchkins etc., not to mention DR-DOS.

Highly recommended (though the Gary Kildall bit is very short):

- A movie: Pirates of Silicon Valley (https://youtu.be/BI-nzUIYIX4)

- A documentary: Triumph of the Nerds (https://www.youtube.com/playlist?list=PLudrw8Z7-gFa7Is4YZitO...)

- Not involving Gary Kildall, but about open source, gnu, and linux: Revolution OS: (https://youtu.be/fxjElWL8igo)

I too would highly recommend Trimuph of the Nerds Documentary (That Cringley Guy is involved in it). Lots of first hand accounts. Plus IBM Carols being sung!

Triumph of the Nerds has some first person accounts from Jack Sams of IBM and Bill Gates and Balmer (of the whole IBM/ Digital Research negotiation (or lack thereoff). Its in part 2 (11 minutes in).

Basically IBM wanting programming languages and an OS from Microsoft. MS didn't have an OS, so they sent IBM to Digital Research (Gary's Company). The deal to get CP/M fell through Microsoft said, we'll sell you (IBM) and OS too. Microsoft then bought the OS from a company across town. That os was basically adapted from CP/M (the author used the cp manual as a starting point)

That's not exactly how it happened. IBM already knew about CP/M because it was already powering the non-Apple PC market. Knowing IBM's desire for an OS as well as Basic he sent Steve Balmer across town to license 86-DOS, aka Seatle DOS which became his offering MSDOS.


IBM eventually offered three different OS on the original IBM PC but Gates shrewdly made sure his was the lowest priced one and the rest is history.

If I recall the story correctly, there was some confusion because IBM saw "Microsoft CP/M" running on an Apple II. (MS licensed it from DR and sold it with their "Z80 SoftCard". The Apple II+SoftCard was supposedly the most popular CP/M machine.)


Paul Allen (Microsoft co-founder) writes about the Z80 SoftCard and the early Microsoft history in his book "Idea Man": http://www.amazon.com/Idea-Man-Memoir-Cofounder-Microsoft/dp...

Thats it (according to Balmer, and Jack Sams who is the IBM exec that went to Digital Research to license it). IBM thought they could licence CP/M from microsoft. MS couldn't license CP/M and sent Jack to the Digital Research house. It was awkward. (Triumph of the Nerds Documentary episode 2 18 minutes in)

I second the recommendation although my least favourite was the Revolution OS; It just felt like a propaganda pamphlet to me but still was insighful and entertaining.

I love this video Computer Chronicles: Unix (1985) feat. Gary Kildall


Even today, years after that show has long since ended, talking about tech that isn't necessarily relevant any more, it is still entertaining and worth watching for anyone interested in the field or its history.

I was on Computer Chronicles back in 1987. He was a cool guy despite missing out on the whole DOS explosion.

If you want to know how was he IRL I recommend watching Computer chronicles. Gary Kildall Special:


Whole playlist of episodes, its like watching non-fiction 'Halt and Catch Fire':


You can watch many of the Computer Chronicles episodes Kildall was in at the Internet Archive. They're simply awesome snapshots of the state of the market at the time.


And the episode celebrating his life https://www.youtube.com/watch?v=OVqBokd3l2E

Before CP/M and such, he was a serious compiler guy who made contributions to data-flow analysis, etc. E.g., http://dl.acm.org/citation.cfm?id=512945

Oddly, I learned about Gary Kildall a few days ago after reading a wired article about the "Halt and Catch Fire" tv series ( http://www.wired.com/2014/05/halt-and-catch-fire/ )

"Easter eggs: Gordon and Donna Clarks’ first initials and back story echo those of Gary and Dorothy Kildall"

Gordon Eubanks worked with Kildall during this time period, he talks about the events surrounding the deal with IBM in this oral history. Fascinating stuff:


He was one of my childhood programming idols lo those many years ago; cemented after reading the original Programmers At Work.

If you've ever wondered why newlines are sometimes CR and other times CR+LF, I think Gary is part of that answer.

It's because on a terminal, CR moved the cursor to the left column and LF moved the cursor down one line. That was also true for dot matrix printers, and paper-based terminals like a modded IBM Selectric or the DecWriter.

So if you wanted your file to display properly when type'd to the console or pip'd to a printer, you needed CR,LF make it work. LF alone would give you text in a barber pole pattern, and CR alone would give you the last line of a file on a terminal, or a horrible mess of over strike text on a printer.

Ah, I remember those days. If someone made the mistake of just using CR alone in a big file they spooled to the lineprinter, it sometimes quite literally would halt and catch fire. There was a fire-extinguisher placed next to it for a reason.

It seems that IBM is in bad shape because of a non-disclosure agreement. Maybe the lesson should be that NDA are bad.

The article didn't ever mention the second listed spouse, Karen. Apparently he underwent a second divorce after beating her with a pool cue < http://issorp.com/site/umour/garykildall.htm >.

If nothing else, he was certainly one of the best looking of the early micro-computer pioneers.

Applications are open for YC Winter 2022

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact