Re: [Pasdoc-main] PasDoc2 finished

[Michalis K. <mic...@gm...>]

Hans-Peter Diettrich wrote:
> After fixing an annoying bug during destruction of the components, I 
> think that PasDoc2 is ready for use. One of the last major changes was a 
> single-component approach, for every kind of documentation.
> 
> Since no response came for the last beta versions, I consider my work on 
> PasDoc finished. Now it's up to you to decide about the future of the 
> PasDoc2 branch - leave it as it is, or merge it into the trunk.
> 

Don't worry about the lack of responses, it only means that no one else 
(including myself, admittedly) didn't have much time to play with 
development of PasDoc lately. Just to be clear: I *do* want to merge 
your work, after carefully checking and fixing it.

In fact, if all the automatic tests would succeed, I would probably just 
merge it, without much reading. I don't have much time, and I don't want 
to become a bottleneck here.

Unfortunately, many tests fail --- this is something that must be fixed 
before merging, there's no way we can apply code that breaks something 
that previously worked. Which brings me to the most important question: 
Can you reproduce these failures? I understand that you compiled PasDoc 
usually with Delphi, occasionally trying also FPC/Windows and FPC/Linux? 
Do all the tests currently succeed with Delphi?

Under FPC/Linux I get quite a lot of errors from simple "make html" 
followed by comparing "diff -wur html/ correct_output/html" inside 
tests/ subdirectory. There is "Fatal Error: Range check error" produced 
when pasdoc was run on the base tests (first mk_test call in 
tests/mk_tests.sh), and many other problems/differences later.

So my trouble is that I don't have much time for PasDoc, but I also 
can't merge blindly PasDoc2 in current state. Which means that either 
someone will fix it, or I'll be (unfortunately slowly) merging and 
fixing it along the way...

Michalis