At 11:23 AM 4/11/2002 -0700, you wrote: >--- Dustin Puryear <[EMAIL PROTECTED]> wrote: > > At 07:05 PM 4/10/2002 -0700, you wrote: > > >Because open source means open libraries, open > > >documentation and open standards, whereas closed > > >source means that not everything is disclosed. > > > > > >And what I said initially was that it is easier to > > >screw up closed source because the libraries/APIs > > are > > >not fully disclosed. At least in M$'s case. > > > > That might have been implied, but wasn't stated. > >Yeah, yeah. Whatever.
Hey, you said "what I said initially," not me. > > Regardless, I'm not sure > > that I fully agree with this train of thought. Some > > products do fully > > disclose their APIs. > >If it's closed source, how do know if the API is fully >disclosed or not? You can't know. It comes down to >trust. And I trust open source more than I trust >closed source (commercial) software, but at least the >software can be compared to the API. Open standards do not make good software; open standards promote interoperable software. Now, that being said, well documented APIs do assist programmers, obviously. However, documented APIs are not only available in the world of open software. Even Microsoft releases its API for Windows. Now, it is true that they reserved some of the information that could have been used to increase performance and the ability to integrate into Windows and Office, but the information required to make good software is there. Since this isn't a question of trust, or at least not originally, but a question about which development model promotes reduced errors when writing code (that was the original subject [you know what I mean John Beamon!]), then as long as an API is documented and exposes the required tools needed for a stable program, then that's what you really need. This can be done by opening the source code, or by documenting the APIs. > > In that situation is it still > > the case that it is > > easier to introduce errors into code in close source > > projects? > >Situation does not exist, see above. Your trust in >commercial software's disclosed API/documentation is >my measure of how much advertising dollars they have >thrown at you. Great, if I disagree then I'm a communists. Where is McCarthy when you need him? > > All in all though, I do agree that more > > documentation does provide better > > results. > >Duh. But this doesn't mean that writing or relying upon open source code reduces the amount of errors you inject into the code. > > I'm not sure though if we can use a blanket > > statement that open > > source software is better documented in a [form that > > is immediately > > usable]. > >It's not. It can be. You can't verify that closed >source is fully documented. Therein lies the rub. > > > However, the fact that you can view open > > source code gives you the > > potential for more documentation, but it also > > requires more work. So that > > is a trade-off if you follow that line of reasoning. > > > >Wow, did I really just hold my own in an argument with >Dustin? I'm going to Disneyland! Actually, I just deleted this message in the site's spool so that it won't make it into the mailing list archive. Regards, Dustin --- Dustin Puryear <[EMAIL PROTECTED]> UNIX and Network Consultant http://members.telocity.com/~dpuryear PGP Key available at http://www.us.pgp.net In the beginning the Universe was created. This has been widely regarded as a bad move. - Douglas Adams ================================================ BRLUG - The Baton Rouge Linux User Group Visit http://www.brlug.net for more information. Send email to [EMAIL PROTECTED] to change your subscription information. ================================================
