[Net-snmp-patches] [ net-snmp-Patches-2431353 ] C++ support for various header files

[SourceForge.net <no...@so...>]

Patches item #2431353, was opened at 2008-12-15 09:21
Message generated for change (Comment added) made by hardaker
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=312694&aid=2431353&group_id=12694

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Hal Rosenstock (hnrose)
Assigned to: Nobody/Anonymous (nobody)
>Summary: C++ support for various header files

Initial Comment:
Allow agent/mibgroup/mibII/system_mib.h to be included by C++ program 

----------------------------------------------------------------------

>Comment By: Wes Hardaker (hardaker)
Date: 2008-12-16 09:04

Message:
Just attach it to this one; I've changed the summary line to match the "new
patch"

----------------------------------------------------------------------

Comment By: Hal Rosenstock (hnrose)
Date: 2008-12-16 07:50

Message:
Sure; I can create a single patch for all the files I'm interested in
changing. Should I just attach it to this or do I need to create a new
"artifact" for this ?

----------------------------------------------------------------------

Comment By: Wes Hardaker (hardaker)
Date: 2008-12-16 06:45

Message:
I don't have a problem accepting patches that apply C++ header checks into
the various headers and there are people that have successfully replaced
parts of the snmp agent with a full C++ wrapper to make better use of their
existing C++ infrastructure.  The patches to the headers should have 0
effect on a C compiler so there is no downside to applying them (unless of
course confusion results because people believe we're supporting a C++
infrastructure directly, but i suspect we can simply say "no we're not"
there).

However, it would be more convenient if you submitted one patch to fix all
the headers rather than a single patch for each header...  Can you generate
a diff for all of them that you've patched instead?

----------------------------------------------------------------------

Comment By: Hal Rosenstock (hnrose)
Date: 2008-12-15 15:02

Message:
[You wrote:]
Please do not try to use a C++ compiler as a C compiler - you will only
get annoyed. It is two different languages.

C++ can be used as a better C compiler. It is instructive in some of the
additional warnings it spits out as some of these are latent bugs IMO.

Even though these patches don't cause any harm, it sounds like you don't
want to move net-snmp in this direction.

----------------------------------------------------------------------

Comment By: Magnus Fromreide (magfr)
Date: 2008-12-15 13:25

Message:
While it certainly is true that good C++ can be tighter and faster than the
equivalent C I fail to see how this applies to the situation at hand -
building snmpd (that is written in C - not in the common subset of C and
C++)

The headers that do include wrappers are all for the snmplib interface,
not the agent interface. The snmplib library is written in the common
subset of C and C++. Yes, you can compile the library as C++, but since it
is C there aren't that much advantage of doing it.

Please do not try to use a C++ compiler as a C compiler - you will only
get annoyed. It is two different languages.

----------------------------------------------------------------------

Comment By: Hal Rosenstock (hnrose)
Date: 2008-12-15 12:07

Message:
C++ is a better compiler than the C compiler so we'd like to be able to use
that rather than C to build the snmp daemon. There are other headers that
are setup this way.

----------------------------------------------------------------------

Comment By: Magnus Fromreide (magfr)
Date: 2008-12-15 12:00

Message:
Why do you want to include those files in a C++ program, what is it that
you are trying to achieve?

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=312694&aid=2431353&group_id=12694