<headius>
the frameless way to do this would be to always super from Set here, since we know this method is defined in Set, but we have not done that before
<headius>
this is the traditional way we have been doing super from Java code
<enebo[m]>
This solution makes sense to me even if it forces some extra overhead
<enebo[m]>
What I am now thinking is how important this is
<enebo[m]>
I don't like people not getting a fix but I am questioning whether we release this in a week in case we get some other problems
<headius>
the main problem with always supering from Set is acquiring the Set class... we don't have an easy reference to it from here
<headius>
I guess this is why we do security releases
<enebo[m]>
It is clearly critical for the person who reported it but they are also seemingly using an earlier version
<headius>
9.2.20 is still affected in any case but nobody reported it there
<headius>
so it has been in the wild for over a month
<headius>
perhaps this means people are moving off of 9.2 😀
<enebo[m]>
maybe. I guess it also depends on how many people are using concurrent set
<headius>
enebo: so it is up to you... this is a pretty standard way to do super dispatch, I was just overzealous in trying to optimize the frame away
<headius>
it works in several other core classes... because many of them can't be extended
<headius>
the long term right fix is to pass frame class and method name into the java method
<headius>
I am annoyed by this as well but it looks like a showstopper for the reporter
<enebo[m]>
yeah I don't really have an opinion on which fix
<enebo[m]>
I am more concerned how when we push the fix
<headius>
that's what I meant
<enebo[m]>
oh maybe I will think about this differently in a couple of hours but I think there is some merit to seeing what else falls out over the next few days
<headius>
yeah I agree
<headius>
if we merge the fix and push some snapshots this guy could possibly use that but a release on Monday is pretty quick
<enebo[m]>
So I think we plan on releasing a point for 9.2/9.3 when I get back from my week off
<enebo[m]>
yeah that is a good idea
<enebo[m]>
he has reported other problems in the last few months
<enebo[m]>
If I remember he is the person who hit the gem push IOError roo