Slashdot Mirror
How To Make Software Projects Fail
Bob Abooey writes: "SoftwareMarketSolution has an interesting interview of Joel Spolsky, of Joel on Software fame. Joel, a former programmer at Microsoft, discusses some of the reasons he thinks some very popular software companies or projects fail, including Netscape, Lotus 123, Borland, etc." This interview brings out some mild boiler-room stories which sound like they could be the basis of a good book, along the lines of Soul of a New Machine .
I imagine the interview goes something like:
Joel: We drove them all out of business.
pooptruck
So, will it make the main page next week when he talks about how "bad" Linux and Mozilla are? As the backseat editors like to say, "Rob, start spell-checking that title now; you know it's coming."
For that matter, i wonder what he'll have to say about Linux. Ten years and they're still on version 2.[4,5]. That might actually make him happy.
One might ask the same about birds. What ARE birds? We just don't know.
- Hire a bunch of people who don't share fluency any one language in common. Throw them all together and make no provisions to improve communication.
This is happening at my comany -- we're roughly 50% Chinese and 50% Americans/Canadians (not all white, but all native english speakers). We can't communicate very easily because the Chinese don't speak much english and the rest of the company doesn't speak much Chinese.
So, we end up just trying to work around each other -- since we only have two people who speak both languages fluently, we have to pick and choose how we use 'em. Pretty dumb.
Every year during my review, I just pray the words "slashdot.org" aren't mentioned.
Cringely is a moron. And Java isn't buggy.
And as far as speed goes, just like use is better than reuse, done is better than fast.
pooptruck
How fucking dense are you? Windows XP is a couple of interface changes on top of the Windows 2000 codebase which itself goes back to the Windows NT 3/4 codebase. The NT design paradigm is one where if you want to add features you just add system services. Want to do some funky .NET strategy? Just add a set of services to handle SOAP requests for DCOM objects. You don't just jam a Unix kernel onto some hardware and suddenly the system serves web pages, crunches d.net keys, and makes coffee.
As for the points you bring up, you can't possibly understand writing software meant to be sold. Bugs are a part of anything, do you think your mom's care rolled out of the factory absolutely bug free? Features do drive sales, sales provide a means for continued development and the feeding of one's family. Not everyone lives with mommy and daddy. Completely trashing old code is often times retarded, clean up dirty patches and whatnot but you don't scrap working code entirely. Writing ultra efficient software is often a waste of time since you're hammered by schedules. Today's screamer is a POS in 18 months, product life cycles are often only a little bit longer than that especially in business environments. People running Windows 95B on old 166 Pentiums are probably still using Office 95 or 97, they don't give a shit about new features. New features are the concern of people who really rely on new dodads and whistles. Customers know shit about development most of the time and you often times know what they are going to say. Read the fucking article man, he goes into why customer's suggestions mean shit.
I'm a loner Dottie, a Rebel.
Well, anyway, while this guy might not know a lot about making good software, he might know a few things about making software stick in the marketplace. Those things usually have nothing to do with each other.
Employee of Inrupt, Project Release Manager and Community Manager for Solid