Wes Hardaker wrote:
Thomas> a) introduce some config_XXX in the mib module header files for the
Thomas> *supported* platforms
Thomas> b) introduce some config_XXX in the mib module header files for the
Thomas> *unsupported* paltforms
We have mechanisms for doing these. config_error and config_warning
both allow you to mention problems. However, I think:
1) ifdefs is the right way to go; it's much more flexible than
anything we could dream up and isn't prone to new parsing errors.
2) As to whether to do a) or b), I'd tend to think it's likely a
case-by-case basis. Much of the time, for example, it might depend
on the existence of a test/header/something and should be treated
that way rather than using #if[n]def ARCH. If I was going to pick
one to lean toward, it would probably be ifdeffing out unsupported
platforms unless it's only known to work on 1-2.
As for the benefits of doing so, it's not only the mentioned "bail
early" factor. I could envision using e.g. --with-mib-module=ALL to
select all mib modules available on the current platform (which isn't
possible today, but handy e.g. for nightly builds). Also, we could get
rid of the current "uname" switching in configure.[in] for the "host"
module.
Any objections against moving forward this way? Only in MAIN?
+Thomas
--
Thomas Anders (thomas.anders at blue-cable.de)
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
_______________________________________________
Net-snmp-coders mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/net-snmp-coders
