On Mon, Mar 15, 1999 at 11:33:01PM +0100, Matthias Welwarsky wrote:
> Jonathan NAYLOR wrote:
>
> > What is interesting having seen the open source become established, I don't think
> > John had mush to worry about. Relatively few people play with the code, and those
> > that do, tend to know what they are doing.
>
> Well other, bad examples also exist. "Digiware" for the rather unknown
> "Falcon"-TNC is one of them. People made slight modifications within the
> software and refused to install bugfixed versions of "Digiware" since
> they didn't wanted to redo their changes. As a result, the German
> network still suffers from frequent, nasty hiccoughs once one of these
> nodes starts to feel unhappy again...
>
> BayComNode (ceased) is another example. Someone took the source and
> created a clone called SNET, iirc. The original author, Florian, DL8MBT
> then decided that searching for other peoples bugs was not worth the
> effort and kept the source of sequel releases for himself.
On the internet (the real `The Net'), there is a well established tradition
that is called free software or open source. One of the most used programs
around the net, is perhaps the sendmail utility. It's source code is freely
available on the net. Don't You think that internet would long ago been
swamped shut by `nasty hiccoughs' if the before mentioned was true?
As a rule, most programs, especially widely used, evolve much more rapidly
after the source code is published.
Let's take the two examples from above. First the "Falcon"-TNC. The problem,
deduced from the description, seems to be lack of programming knowledge.
Obviously the people who did the ` slight modifications within the software'
never heard off diff and patch utilities. Usually one's modifications can
be applied to newer versions of a program with few or no modifications at all.
The second example raises another, far more profound point (warning! this is
pure speculation, based solely on the quoted article. I don't know if this
the case, because I know neither the software (BayComNode and SNET) nor
the author (DL8MBT)) of too big an ego. People don't want to release their
code to public, because they've worked hard on it and fear that someone makes
a better or a different, perhaps more popular, program based on their work,
thus rendering the original author unknown. This is of course no problem if
the original author has written and published the program solely to help
him/herself and others. Then any improvements on the code benefit the author
further. If on the other hand the program was made to make oneself known or
earn fame as programmer the previous is of, course, unacceptable. Even in
this case something can be saved by clever use of licenses. Make the license
such that every derivative work must have Your copyright message displayed
somewhere (it's been done several times), so no-one forgets Your efforts.
Other solution is to make the program so good, that You're automatically
remembered :D Third option is to be actively involved in the (public)
development process: collect all the patches and issue `official'
releases. This can be done (If someone doesn't believe Me, I suggest
learning about a bloke from Finland called Linus Torvalds.)
--
BSc. Pertti Karppinen <[EMAIL PROTECTED]> |'Bridge Players |
Systems Designer, University of Jyvaskyla, Finland | Do |
http://www.iki.fi/~pjka/ | Office : +358 14 602 088 | It |
HAM: OH6KTR QTH: KP22UF | Cellular: +358 40 564 0786 | on the Table' |
