Joel Berger
Joel Berger
Yeah, I get your point. I was imagining that a user would create a new instance per message, in which case encode and decode make sense as they are, but...
The thought is still on my mind yes, I just haven't seen much need to actually do it. If you are interested in presenting a patch, I'd consider it.
Yes I should get around to that. Galileo doesn't get enough of my time lately :s.
the difficulty, sad as it is, is testing. With only one hash key by default, I know the hash ordering. With two, (ie if iat and some test data) were...
how does https://github.com/jberger/Mojo-JWT/compare/set_iat look?
Image uploads are planned for Galileo, I just haven't gotten there yet. I will be interested to see your approach. Cheers. (Proof that its on my mind: http://irclog.perlgeek.de/mojo/2013-02-06#i_6417984)
Here is the first implementation of the file upload mechanism that Galileo will eventually use: https://gist.github.com/jberger/4744482
My original design was to follow Wikipedia style flat structure. I'm not opposed to having a richer structure but it will take some doing.
The fork was originally necessary because the author felt that the hash ordering change in Perl 5.18 was incorrect and chose not to fix JSON::XS it for the entire 5.17...
reading the new minion query, it isn't obvious to me, does it handle deeper dependencies to me it looks like if A (finished) -> B (finished) -> C (active) then...