Comment by erikpukinskis | Hacker News Reader

Comment by erikpukinskis | Hacker News Reader

Parent Full thread

I've waved the "What did Microsoft ever invent?" flag plenty in my life, so let me try and advocate for the devil:

Microsoft basically invented the hardware-agnostic software platform. Before Microsoft, there were a variety of hardware platforms, each with their own libraries and languages... not unlike the situation today with phones. Developers had to choose what to target... not only Mac vs PC, but different video cards, different sound cards, different network stacks, etc.

Microsoft took that fragmented market and unified it under a set of standard APIs for both hardware makers and software developers to target. I would consider that an innovation: the loose coupling that they enabled allowed much faster innovation in both hardware and third-party software.

Interestingly, Netscape tried to play the same game on top of them (unifying Windows, Mac, and UNIX under one standardized API), and though Microsoft was able to destroying them, Webkit has succeeded where Netscape failed, and it is slowly making Microsoft irrelevant, just as Gates feared.

Other than that, I don't think Microsoft invented much. But they did a good job copying, were ruthless with business executation, and so their one innovation ended up being a multi-trillion dollar innovation.

> Microsoft basically invented the hardware-agnostic software platform

Digital Research did that. You could run a CP/M program on any CP/M-capable computer. Microsoft tried that with MSX (and failed) and then with MS-DOS (and succeeded on a different hardware platform). Only then, different graphics cards and printers became relevant.

> Other than that, I don't think Microsoft invented much

Visual Basic is one thing that comes to mind. I don't remember anything like it before. Hypercard and Toolbook were huge and heavy in comparison. VB allowed BASIC programmers to write compact Windows programs. That more or less doomed DOS.

Note that VB was an outside acquisition.

Oh... So was MS-DOS...

Well... Not very surprising. That company has been sick for decades.

To be more precise, MS-DOS (and 86-DOS before that) tried to copy CP/M on that area (the original platform for SCP 86-DOS which MS bought and renamed MS-DOS was SCP's S-100 8086 cards which existed before the IBM PC), but MS bought SCP in the first place because they needed an OS to sell to IBM for their 8088-based PC, and so of course it first shipped on the IBM PC, and what happened was that the DOS calls was too limited and slow, so apps used the BIOS and even direct hardware access. So in the US full IBM PC hardware and BIOS compatiblity ended up being needed. (In Japan it was a different story)

Microsoft basically invented the hardware-agnostic software platform

I'm trying hard to understand what you're talking about and its not working. Windows has run on the "PC clone" for almost its entire existence. IBM pioneered the original PC as a fairly easily clone-able platform, meaning it wasn't hard to reverse engineer the bios. It was one hardware platform, one hardware standard, that multiple manufacturers could target (and it was created by IBM despite repeating more of the benefits). But that kind of thing had happened at least since Boroughs (I think) created the first IBM compatible mainframe.

Windows runs on PCs. It doesn't run on Mac, on mainframes or anything else - contrast that Linux, which truly runs on a huge variety of platforms.

I think you're too focused on looking at it from a technical perspective. I think what erikpukinskis is referring to is "hardware-agnostic software platform" as a business model/industry environment.

Maybe IBM did invent that, but it was accidental. I'm not sure they did though, I think MS did.

If you said "based on standardized, commodity hardware-based" instead of "hardware agnostic", I'd agree.

I mean java was an effort to hardware agnostic. Windows, not-so-much. It supports versions of the "PC Standard". That's it.

Just as an aside to your final point:

I think a case can be made that huge software companies generate gigantic sums of cash. They have to do SOMETHING with this cash. They could give it back in dividends but most investors would prefer they instead used their internal cash-multiplier machine because it probably runs better than their own would should they have to reinvest dividends.

So what do you do? Acquire.

Look at Microsoft. Look at Google. Adwords, Android, Chome, Maps, Voice, YouTube, etc, etc, etc, all acquired.

It still takes a talented company to marshal the resources and assets of these acquisitions into a successful, larger conglomerate. And in the cases where the sum is more valuable than the parts, THAT'S innovation just the same as a homegrown product or language.

Execution is everything. For everything Microsoft copied from someone you know about, five things you know about were copied from someone you do not know about.

Interesting comment. IMO using a virtual machine to abstract away the hardware is what an operating system is all about. Microsoft has done a lot of work to prevent any competing middleware (Java, Internet browser, Flash, OpenGL, ...) from challenging their position.

http://www.ecis.eu/documents/Finalversion_Consumerchoicepape...

Reply on news.ycombinator.com