See below, please. regards,
Richard Erlacher ----- Original Message ----- From: "Dave McGuire" <[EMAIL PROTECTED]> To: <sdcc-user@lists.sourceforge.net> Sent: Friday, September 05, 2008 1:18 PM Subject: Re: [Sdcc-user] documentation & open source generally > On Sep 5, 2008, at 3:00 PM, Richard Erlacher wrote: >>>>> The point, however, is to document what the software suite is >>>> SUPPOSED to >>>> do, and not necessarily what it is OBSERVED to do. That's why >>>> it's so >>>> unfortunate that the doc's weren't written before the first line of >>>> code, >>>> rather than as an afterthought. >>> >>> That's nice in theory, but in the real world, it just doesn't >>> work. That you're having such difficulties, and that this thread has >>> gone on for days, illustrates this fact. >>> >> Well, that's just not true. If the doc's had been generated BEFORE >> the code >> was started, there'd be clear evidence of what it was intended to >> do, and of >> whether it does it. > > Uhh yeah...but that's not how software is written in the real > world. It's not a very practical approach, which is why it's rarely > done that way. Overspecification (which is where that approach goes > if any tie-wearers are involved) is what kills software development > projects before they even get off the ground. > >> That's one reason programmers don't like spec's to >> precede programming. If there had been documented spec's, it would >> be easy >> to show that their work product doesn't work. > > If there had been documented specs, the software would never be > released, due to pie-in-the-sky wish lists and unrealistic goals. > That's what I was talking about above. > >> The reason it's uncommon in the "real" world, that this highly >> desirable >> condition doesn't exist, is that programmers like to do what they >> like to >> do, and leave what they don't like to someone else. > > Well, programmers are PROGRAMMERS, not paper-pushers. > >> With no documented >> specifications, there's no way to show that their work is >> defective, so they >> like doing things that way. > > That's an awfully big stretch. I, as a programmer, write damn > good code...and terrible documentation. That's because I'm a > PROGRAMMER, not a writer. > Do you program to a set of requirements, or do you simply imagine what you want to write and then, when you think you've finished, accept what you have, regardless of what it does? How do you know that you've done what you set out to do? In fact, how do you know you've finished? After that, how do you verify that it works as it should? There's a huge difference between trial and testing, doncha know! > >> This points out that management in software development is weak, first >> because software managers are often former and often incompentent >> programmers, kicked upstairs because they were in the way, and >> secondly >> because they're hopelessly incompentent as managers of software >> development, >> possibly both. > > Richard, you talk like someone who has been "suitified" for far > too long. There was a big push in the 1970s to over-formalize > software development, which is where this obsession with "specs" came > from. It died in the 1980s. Go write some code! > As I recall, in the early '70's, at which time I did write code (mostly FORTRAN) for a living, at least part of the time, things were VERY formal. It was going into the '80's that things became "fast and loose." The result was that folks wrote software that wasn't needed, incorporated features, and the associated failures, that weren't required, and generated huge schedule and budget overruns. BTW, Last time I wore a suit was to go to a funeral, and, in fact, that was the last 3 times. Next time, I hope it will be to my own. I've never liked neckties and the associated "appearance over substance" culture. I don't need any more code right now. I could use some tools, though. I'm building hardware, and the other three boxes are running simulations. If I've done everything right, one will be finished in about an hour and I'll be done with that portion of the work. This one's web-connected, so I don't use it for "useful work." It's too risky. I'd point out that you've gone on and on but suggested nothing that would lead to better documentation. I'm still reading ... and hoping someone comes up with something I can use to further that goal. > > -Dave > > -- > Dave McGuire > Port Charlotte, FL > > ------------------------------------------------------------------------- ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ Sdcc-user mailing list Sdcc-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sdcc-user
