[Oberon] Math.Mod / compile rrror (to many digits)

[Jan Verhoeven]

Chris Burrows wrote:
>> BW,
>> Volkert
>> PS: With an "bug or issue tracker" and we all can  see, what bugs are new,
>> open, fixed ...
> Where's the fun in that? It's like doing a crossword puzzle and having
> somebody looking over your shoulder and giving you all the answers.


And now you know, dear Volkert, why Oberon is still in the same place of 
development as it has been since 1994. Oberon is about fun. No practical 
use, just fun. Worldwide there are 40 enthusiasts. And each one has to 
find his own bugs and create his own fixes and should keep his mouth shut.
No need to spoil the fun and tell others about bugs, fixes and new features.


> The whole point of Project Oberon is to help you to learn how to understand, maintain etc. an Operating system and compiler yourself and become self-sufficient. It can be done and it has been done.


That was the original goal. But sooner or later some one has to break 
that tradition and make Oberon something for the world. For the people. 
Some stable branch to hit the market.


> I use a comparison tool to see what has changed and then try to understand
> what has been fixed and how it has been fixed. After a very short time of
> this sort of exercise you learn how to fix these things yourself,


Unless the fixing part has completely reformatted the source files and 
then issued his fixes. Many of the NW sources are spaghetti code. The 
code is used without (m)any rules. Indenting, structure of IF/THEN's 
etc. It's a drama getting through many of the sources.
Many of the NW produced sources are a direct violation of all the things 
taught in Data Structures and in the other books. Many statements on one 
line, which are very hard to understand,

I reformatted ORG.Mod and it went from 1125 lines to 2000 lines of code 
with lots of whitespace so the commands stand out and control structures 
are easily identified and understood. Now I can see and understand 
what's going on.

Volkert's proposal poses the next question to face: do we want Oberon to 
remain the never ending academic OS? Or do we want to give it a shot and 
try to give it an audience?

As I see things, Paul Reed and Peter de Wachter have opted for the 
second option. And if you do that, excellent documentation and a bug/fix 
tracking system are inevitable.


-- 
Groetjes

Jan Verhoeven