Disclaimer: I have looked for maybe 2 or 3 hours at the code up till
Also, I answer different people in this post.
>>> To answer Gene's question: The code looks like a train wreck.
Hm. I have a different first impression.
>> There are vast
>>> sections of code that are commented out. That's a first glance,
Yes, there are. However, IMHO that is a matter of taste. Actually, it
is a bit
of an argument between me and another developer at work. He says
commented-out lines decrease the readability. When I read code, I
simply ignore commented-out lines :-). Therefore, I keep them when I
think I will need them again later on.
Obviuously, at Rowan, "DEADCODE" is a sort of keyword. For example, I
saw code marked DEADCODE where a comment before it said this was code
> Yes. It also appears that multiple aircraft are hardcoded into it,
True. That is of course one of the bad things about the code from an
OSS standpoint. OTOH, we should not really hold this against the Rowan
people: It was a commercial product that depicts the Battle of
Britain, just those few month and nothing else. There were no other
planes involved than those they already have implemented and even
counting planes that were not involved, but were around at the time,
there are not that many more.
> and the FDM is very tightly integrated into the rest of the code.
Ok - I haven't seen that yet.
> At first glance it doesn't appear they really used C++ to their
Hm. Interesting. I see what you mean, they don't even use exception
handling. OTOH, I up till now saw no code piece where I thought - Ah,
this was written under C and now they simply use it with the C++
>400k source files?
>I still haven't found the main makefile yet.
Aeh, this is for MSVC only. This is also something that would have to
be changed to port it. The "build system" is quite complicated, with
numerous 4Dos/4NT batch files called automagically by MSVC, pre- and
post-operations, empty projects to automagically catch new files etc.
BTW, I could compile it almost straight away. My compiled Bob.exe runs,
my compiled lib3d.dll not. AFAICT, I am the first person to compile it
under Win2k. Therefore, I am very impressed with that bit. Remember,
when I get 100 different errors compiling FlightGear with MSVC under
Windows, Geoff gets none or 5 and vice versa ;-).