Imagine a world in which HyperCard had been open sourced 20 years ago?

[Peter Alcibiades]

Chipp rightly points out that there are very many open source projects which 
are started, and then wither.  He points to the 2,000 or so low activity 
projects on SourceForge.  He could also point to a high proportion of the 
distributions on DistroWatch.   

On the other hand, in programming environments, look at Perl, Python, Lua, 
Gtk, Qt, fltk, Fox.....   Look at their history.  They haven't fragmented 
into incompatible streams, they haven't withered.  

So what does the evidence prove?  Not much.  It might work, as some 
programming environments evidently do, or it might not, as probably many 
haven't.

Going open source is a bit like writing it in Perl.  People have written stuff 
like this in Perl and its worked.  Others have written stuff like this in 
Perl and failed.  Does this mean we should stick to writing it in C like we 
always have?  No.  It tells you nothing one way or the other.  Do you have 
problems with productivity?  Do you have the feeling that a lot of what you 
are doing would be so much simpler in Perl?  Well, think hard about moving, 
but think the whole thing through before you decide.

It just says, think through the open source question with the same rigor with 
which you would design a program.  Just because its business strategy does 
not mean its easy or doesn't require proper analysis.

For Rev, to go open source is not a simple well defined thing, and does not 
just mean lets have anarchy and give away the engine.  And it does not have a 
predictable defined outcome in terms of profitability that you can forecast 
by looking at other projects.  And, it might not work.  

I can tell you one thing for sure though.  It will not lead to a combination 
of multiple incompatible streams AND a total lack of development. There's no 
evidence this happens.  One or the other, but not both!

Peter