[openstack-dev] Less of coding standard for i18n of LOG messages

Eric Windisch eric at cloudscaling.com
Wed Nov 7 17:26:28 UTC 2012



On Wednesday, November 7, 2012 at 11:57 AM, Sean Dague wrote:

> I think that assumes that everyone uses the same search engines and user 
> forums for their content. I think if you look at China in particular 
> this isn't going to be the case. 
> http://chineseseoshifu.com/china-search-engine-market-share/
> 
> 


As previously discussed, there is the option to add universal identifiers for logs. Then, logs could be searched by either their universal identifier, or their native text.  It could also be used to translate messages post-defacto. Unfortunately, this does introduce some complex problems to making the numbers unique, relatively static between releases, etc.

One advantage is that we could potentially use such identifiers to allow searching across differences in the text within the same language, across releases. If a log message were changed between releases, but the triggering event has not changed, the universal identifier could potentially remain unchanged.  Again, that might not be easy to implement, but it is a nice concept.

In fact, now that I think abut it, is there a reason we use _() from callers instead of just having the log formatter wrap this for us?

Regards,
Eric Windisch


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20121107/31ed7bb2/attachment.html>


More information about the OpenStack-dev mailing list