Possible bugs in man.cgi?

Hello Bob and all.

There is some strange behaviour of the man pages web interface, which I 
originally thought is specific to one platform, but it can be reproduced 
with any platform. Under some circumstances (described below) the web 
interface to man pages does not display a device driver man page if 
section 4 is explicitly selected. It applies to other sections too (like 
i386_vm86(2) for i386). After finding out a rule when it happens, it is 
not an issue for me anymore - I am not complaining, just reporting an 
unexpected behaviour (or a 'bug' if you consider it to be).

I am describing my findings on a socppc platform example:

When I enter a socppc specific device as a keyword (like tsec, socpcic,
sociic, or obio) and select section '4 - Special Files' (and socppc as
Architecture), I always get

   Sorry, no data found for `socpcic(4)'.     (or tsec(4), or whatever).

This does not happen if I select 'All Sections' - then I get the man page.
On the other hand, man pages of devices not specific to socppc only are
always found and displayed correctly for socppc architecture.


And here is even more curious and dangerous thing, apparently related:

If you ask for biosboot/All Sections/i386, you get.... yes, the man page
for biosboot(8) - i386-specific first-stage system bootstrap.

On the other hand, if you ask explicitly for biosboot/8 - Maintenance
Commands/i386, you get... guess what? The man page for biosboot(8) -
amd64-specific first-stage system bootstrap. If you do not spot it in
time, you are reading a wrong man page! This one seems to be a real
problem.

Sorry for not providing a solution, but I think this report is the only 
way I can help for now. I hope an e-mail like this is the proper way in 
which web-related bugs should be reported.

Thanks for otherwise perfectly working website for a great OS.

Regards,
David