[Lognorm] last call for next liblognorm release

[Rainer Gerhards]

2015-03-09 14:48 GMT+01:00 Tomas Heinrich <theinric at redhat.com>:

> On 03/09/15 14:09, Rainer Gerhards wrote:
>
>> Maybe a bug in libestr?
>>
>
> Turns out that it is a bug in the packaging. Filed as rhbz 1199995.
>
>  ... but I also get some unexpected warnings, for which I did not yet
>> find a cure. It's about json_object_object_get being depricated [0].
>>
>> This is (kind of;)) being checked for in configure.ac
>> <http://configure.ac> as well. I thought the include file was missing
>> here, too, but this definition doesn't make any difference:
>>
>> AC_CHECK_FUNCS(json_object_object_get_ex,,,[#include <json.h>])
>>
>> Any idea?
>>
>
> Uhm, doesn't it check for json_object_object_get_ex()? Maybe I don't get
> how those two are related.
>
> Either way, is the warning a real bug? It just warns that the function
> should be avoided in the future. The code in json_compatibility.c only
> handles the nonexistence of json_object_object_get_ex(), but explicitly
> references json_object_object_get().


It's a warning, and it's not a real problem. But the detection of
json_object_object_get_ex() does not work on el6. On other platforms,
configure detects it, and so the old one is never used. On el6, it looks
like it doesn't detect the function, and so the "json compatiblitly layer"
in liblognorm is invoked, which calls the old function. A little bit of
overhead involved, but mostly the issue is what's going on.

In any case, I'll release with more or less cosmetic issue. We can always
"fix" it later.

Rainer

>
>
>  Rainer
>>
>> [0]
>> enc_syslog.c: En la función ‘ln_fmtEventToRFC5424’:
>>
>
> What locale are you using? :]
>
> Tomas
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.adiscon.net/pipermail/lognorm/attachments/20150310/2a3c197d/attachment-0001.html>