Posted by harold at November 22nd, 2007

Okay so actually, Phil (my neighborhood crack dealer) mentioned that there are still a couple of tech companies out there. And I mean organizations that actually are results-oriented rather than wasting-peoples’-time-on-stupid-stuff oriented. Yes, I know it’s hard to believe. One of better performing (well, before this week’s stock disaster) firms is VMware. As anyone with half a brain knows, VMware is a software company with a focus on delivering virtualization solutions on x86 architecture PCs. Based off research done in Mendel Rosenblum’s research group at Stanford back in the late 90’s, VMware completely changed the face of machine virtualization by offering blazing fast speeds on quirky commodity hardware (interested in the details? Read this paper). But it’s not all fun and games and fairies and daises and lightly salted french fries. The x86 architecture with all its quirks is not naturally virtualizable under Popek’s and Goldberg’s canonical definition. To be naturally virtualizable, all privileged or sensitive instructions must generate a trap. On x86, there are at least 17 instructions in the base instruction set which violate this key property. As a result, trying to correctly virtualize the x86 is equivalent to attempting to drive a Yugo through Death Valley. Oh sure you can do it, but you’ll probably break down after about 10 feet. 11 and a half if you’re lucky. VMware claims their virtual machine is totally isolated, but hey, we all know it’s only a matter of time before someone breaks the isolation. Now I know legend has it that they convinced the NSA to use VMware… but I also happen to know that my old buddies at the NSA loved to claim they internally used products which in reality they knew exactly how to crack. Hehe.

Oh yeah so anyway, another one of these technology-oriented companies based in the Valley is Coverity. You might need three quarters of a brain to have heard of them, since they’re developer-oriented (half a brain and a quarter pounder with swiss would probably work too). This company produces tools which perform source code analysis in order to improve the software development process. Sounds like a pretty big deal to me… as we all know, bugs in software love to cause problems. However, your average VC didn’t see it that way. For example, rather than providing capital to Coverity, Sequoia decided to pour money into stupid little abominations like TokBox, which is a startup that combines the power of rounded corners and the built-in video and webcam support in Flash. Rather than thinking, Sequoia decided to throw good money at an inconsequential little trinket, and now they’re paying for it (I’ve heard the TokBox team is, ahem, “brilliant”). Coverity, on the other hand, has managed to grow to a respectable level of profitability based on, no not ads, certainly not VC funding… wait for it… innovations that deliver true value to customers. Hint to VCs: next time you see a company that’s developed innovative tools for finding ever more of those nasty bugs lurking in our code bases–give THEM the money! Not the guy who invented arsenic-flavored chalk candybars.

Well folks, we’ve gone through the bad, we’ve glimpsed at the good that remains. Now it’s up to all of us to decide whether the tech industry goes in the gutter or goes for the gold. It’s a simple decision, surely. But executing on it executing on it just seems to be a little too difficult given the apparent rampancy of zombies in the Valley today.