On 3/28/2015 3:21 PM, Shawn Heisey wrote:
> SOLR-7316 seems like a reasonable candidate for the work, but I can
> understand any desire to start with a clean slate on a new issue where
> there's no baggage.

Some of Mark Miller's words on the "webapp" vote are brought back by
this discussion: "but honestly, I think of 5 new reasons every month
this webapp situation is just a limitation"

Solr will never be fully aware of its entire directory structure as long
as there's an external servlet container sitting between it and the rest
of the system.  That's a major impediment to the desired outcome on
SOLR-7316.

As things are written now, Solr as a webapp can really only be aware of
what's under the solr home and the current working directory, and the
solr home strikes me as a terrible place to put jars, logs, and default
configs, at least out of the box.  Some users may be perfectly fine with
it, but I don't think that would be a prevailing view among experienced
admins.  You never know what the current working directory is unless you
are in strict control of the startup, but until we truly eliminate the
war, we can't be sure we're in control of that.

I've begun writing down some ideas for how to deal with these problems
in the bin/solr context, even while we are still a webapp.

Thanks,
Shawn


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to