Writing perfect software

[Mark0]

It's all about the process, as per the on-board shuttle group, the team that writes the software that control the STS.
Very interesting article here:
FastCompany.com - They Write the Right Stuff

Bye!

[brotherS]

Great article, great site! Since it's a long article, I used Zap Reader (https://www.donation...dex.php?topic=4869.0).

Btw, I subscribe to the FastCompany (paper) magazine, I like the content and the style!

[JavaJones]

Definitely a very good article. I wonder how many people, how many coders, can really work like that though. Documentation in particular is anathema to many of the "best" coders I know. Not necessarily in terms of commenting your code, but often as far as end-user documentation, and certainly as far as documenting the coding *process* is concerned. You may make simple notes and keep a basic log - a change log is important of course. But when you're talking about 40,000 pages of planning documents and who knows how many more of change documentation, it seems a bit mind-numbing.

So yeah, sounds like the way to write "perfect" software. But is that really what we need elsewhere? And if so, who is willing to write software this way? Coders - does this seem like an appealing environment to you?

- Oshyan

[Renegade]

And money is not the critical constraint: the groups $35 million per year budget is a trivial slice of the NASA pie, but on a dollars-per-line basis, it makes the group among the nation's most expensive software organizations.

And that's why. Money is no object for them. In the "real world", money matters. People need to get paid because eating is good, and sleeping with a roof over your house is good, and having clothes to wear is good, and...

That's a completely unrealistic and utopian idea about software development. Some of the practices can certainly be used (and should be), but their level of detail is extremely extreme.

At the end of the day you've got to ship a product - bugs and all. A friend of mine is the CEO of a well known software company and says ship at 80% and get the other 20% from feedback. It works for him and the company does well. I think he's right.

It was an interesting article, but not very informative. If anything, it's damaging to software development because it gives unrealistic expectations that can only be met in a glass bubble - the kind of bubble you get from an exclusive government contract and lots of money.

No matter what, at the end of the day shipping the product is where you're going to start making money. That's going to put food on people's tables and keep them developing. There are no coders with empty plates. Developing software like that for most companies is a great way to ensure products don't ever get shipped. I've seen them before - ventures that spend all their money on development and zero on marketing. They never ship and end up in bankruptcy.

Cheers,

Ryan

[yanjinpeng]

good views!

[Eóin]

Reviving an old thread I know but this is interesting.

http://news.yahoo.co...m/space_shuttle_dc_1

[If] the launch is delayed for any reason beyond December 17 or 18, the flight likely would be postponed until next year, officials at the U.S. space agency said on Monday.

To build in added cushion, NASA may move up the take off to December 6.

"The shuttle computers were never envisioned to fly through a year-end changeover" space shuttle program manager Wayne Hale told a briefing. (my emphasis)