The normal config.h is generated by ./configure through autoconf/automake.

I'm sure there is a way of adding version specific lines (and there is
already at least one case in ./configure for Darwin stuff you can use), but
I don't remember the macro offhand - that would need to be looked up in the
docs.

Of course, MinGW doesn't use the normal ./configure, so you can probably
just hack the Makefile.mingw.

Doing it right?  Don't bother with the test ... Since it's a MinGW only
file.  In fact, you should probably #define something and use that instead
of the #if defined(WIN32) && defined(__GNUC__) everywhere.  And again to
make it useful and portable, you should NOT change and WIN32 stuff, but
rather should live with it and just add what you need to make it work under
MinGW.

Once again, MinGW is a compatibility layer (shim) - the more you stress it
the more likely things will not work.  Windows support for POSIX threads is
minimal at best (different threading models), so that's likely a high stress
point.


-----Burton


-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf
Of Georger Araujo
Sent: Tuesday, August 30, 2005 1:40 PM
To: [email protected]
Subject: RE: [Ntop-dev] Patches and documentation for building ntop on MingW

Jac,
yes, that error occurs because although IPv6 is enabled, its structures and
their members (for example, s6_addr) are only declared if you have #included
the files in config.h. But from my experience, it doesn't hurt to compile
with IPv6 enabled. I just didn't submit config.h as-is because I thought it
could cause problems in the *nix world.
Burton, we could include config.h directly into the sources, but with the
following text:

---
/* config.h for MingW */
#if defined(WIN32) && defined(__GNUC__)
#include <winsock2.h>
#include <ws2tcpip.h>
#endif /* WIN32 && __GNUC__ */

---
That way, things would work out of the box for both platforms. What do you
think?

Jac, about the second problem, I really have no idea what's wrong. It seldom
happens in my environment, and doesn't give me any trouble at all.
I manage to run ntop on https, but it is flaky and unresponsive. I'm working
on other topics right now (upgrading the libs, rrdtool and fixing arbitrary
graphs) to look at this now, maybe later. Regards,

Georger

--- jacengel <[EMAIL PROTECTED]> escreveu:

> Georger,
> FYI
> Thanks for the files but I have problems to compile I got  Error:
> In file included from ntop.h:490,
>                  from address.c:21:
> globals-core.h:556: warning: `struct timespec'
> declared inside parameter
> list
> address.c: In function `queueAddress':
> address.c:553: structure has no member named `s6_addr'
> address.c: In function `dequeueAddress':
> address.c:637: structure has no member named `s6_addr'
> make: *** [MingW_obj/address.o] Error 1
> 
> When I compile with IPv6 support by renaming
> config.h.win32 in
>     C:\MinGW\ntop3.1\ntop to config.h
> it compiles OK !!!
> 
> But still the known error:
> - 08/30/05 19:28:59 [MSGID8962748] **ERROR** errno
> 73330508 during sending
> of page
>   to web client
> - and https  on port 3001 does not work
> 
> Cheers
> Jac

__________________________________________________
Converse com seus amigos em tempo real com o Yahoo! Messenger
http://br.download.yahoo.com/messenger/
_______________________________________________
Ntop-dev mailing list
[email protected]
http://listgateway.unipi.it/mailman/listinfo/ntop-dev

_______________________________________________
Ntop-dev mailing list
[email protected]
http://listgateway.unipi.it/mailman/listinfo/ntop-dev

Reply via email to