Tomcat Regression still haunts projects with URL Encoding?

[Marc Mercer <mm...@ct-unlimited.com>]

Lo folks;

I am emailing today to hopefully get an update or perhaps re-ignite a fire
somewhere that can fix a bug that should be impacting many different
projects at this point.  The bug in question is:

https://bz.apache.org/bugzilla/show_bug.cgi?id=59317

It has been discussed briefly by Vincent Massol and Mark Thomas going back
as far as July 8, 2016 ( http://markmail.org/message/
jwm5ip245empcghi#query:+page:1+mid:jwm5ip245empcghi+state:results ), and it
appears that the latest known update to the situation was that it had been
fixed in 9.0 branch, as seen here:

https://github.com/apache/tomcat/commit/eb195bebac8239b994fa921aeedb13
6a93e4ccaf


It would be nice to get input from the someone on the tomcat side of things
and work to find a way of resolving this regression permanently across the
board, or if it is actually resolved, having a valid update for that as
well.

Anyone willing to help get this looked into, updated and fixed/validated
once and for all (across the board)?

Thanks,

Marc Mercer

Re: Tomcat Regression still haunts projects with URL Encoding?

[Rémy Maucherat <re...@apache.org>]

2016-11-18 23:51 GMT+01:00 Marc Mercer <mm...@ct-unlimited.com>:

> Anyone willing to help get this looked into, updated and fixed/validated
> once and for all (across the board)?
>
> Very funny. There's no good way to do it (although personally, I would
have said dispatcher paths are not encoded), so the best advice if you
don't want problems is: don't use URLs that need encoding. If you still
want to do it, there are two possibilities: encode/decode or don't. So
there's now a setting for that as you can see in the commit you linked.

Rémy