Longhorn Gutted: An example of regular software dev processes

Scoble explains a bit of at least his understanding as to why Longhorn was gutted.

It makes sense in many ways. MS (and virtually every other tech company) is always big on "what's coming down the pipe". I remember getting a letter from BillG when Windows 3.1 was released, trying to get people to switch from DOS (and then again with Win95). In many cases, it's about providing a better solution to existing problems. (as an example, remember Apple's Navigator?)

However, in the case of LongHorn, MS seemed to be going after one type of user (corporate dev) and fighting off another tech attack (from Linux users) instead of addressing the bigger issues.

When Bill G originally became the "architect" for MS (instead of CEO), it was supposed to be because he was going to refocus the team on fixing the critical issues he saw in the product. Two years later, unfortunately, MS is more of a target than ever before from tech attacks and little security problems are like holes in the dam, needing to be filled in.

Software developers go through this on a regular basis. You want to build the next greatest application but you have to deal with fixing your existing one first. That's not to say that you screwed up big time on the first one but that as it became used more, you try to find solutions in your existing application. Where does that place your "next big thing"? On the backburner. The ideal is to provide a mounting strategy so that the steps are incremental and result in a better overall tool. Can that happen with an OS? MS has tried it with XP, moving users from DOS to Win 3x to Win95 and now finally to XP. But the idea of being able to simply hit a switch and have all the users jump ship to a new OS?

It's a difficult challenge just as getting your entire company wrapped around Security and the "little things", such as pop-ups and spam.

MS has done it before, turning the entire company on a dime to be Internet-aware (back in 94/95) and they were , arguably, successful at it. But now, they are the primary focus of hackers, security pundits, anti-trust lawyers (and anyone else you can drop into this bucket), and the challenge is harder.

Are there great tools coming down the pipe? Absolutely but MS needs to provide an easy way to get those people on board. Otherwise they will suffer the same fate as Apple with declining marketshare. The difference here is that there is NO real single alternative. Don't get me started on Linux - yes, it's an alternative. But there are at least 6 different mainstream variations so it reminds me of the Sun Open Unix days, when there was NeXT, Sun, IBM AIX, etc. And many of the Linux camps simply try to COPY the existing Windows interface.

So I see it as a good thing that MS pulls back a little bit. Otherwise when Longhorn debuted it would be like the original Infinity car ads - everyone would be left wondering "and this is for .....?"

I would love nothing more than to have a new OS come in that revolutionizes the way my life in computers is done but I'm also a realist. Don't preach "this is the better way" and then change it every few years. Unless it fundamentally changes even the concept of computing, the slow, steady, evolutionary approach is the not the only, but the most logical way to go about it. It may not bring in 40 billion dollars overnight but then again, what does?

Scobleizer: Microsoft Geek Blogger


Popular Posts