Igor

As per @m4dcoder suggestion, I am adding my 2 cents to publishing vars. Instead of having them as part of the transition model, I'd have it as part of the...

@m4dcoder I don't know about those hypothetical scenarios where you'd want to publish different variables, but in my experience I've rarely needed it (if at all, to my recollection). I've...

@m4dcoder well, I expected this kind of response, thus was a bit hesitant to ignite this sort of debate :) Of course, if I needed to get rid of True/False,...

@m4dcoder I like YAQL, it allows to be pretty flexible when needed. Wish it could also let your handle exceptions :) Regardless, could you consider the compromise suggested? Let users...

Well, the compromise where users would be able to publish in two places introduces no changes to YAQL/Jinja processing. Only that if `publish:` is defined within the task itself, run...

P.S. As much as you don't want to rush make changes, I also don't want to use lengthy and bulky workarounds just to later re-do it the right way (if...

In light of @blag comment, I believe something like `local().var_name` would work for everyone. It'd be in line with global `ctx().var_name` style/approach.

@blag, yeah, right, we're talking about declaration here, my bad. As to using `:` in variables declaration I believe yaml wouldn't be too happy about it, unless you wrap it...

@farthinder I managed to make it work for a VM with one interface and two openvpn tunnels. Had to patch `Balance.pm` a little bit, though. I didn't test the use-case...

Yeah, that's something that could be really useful. Just please make sure that the files with custom filters placed in the st2/st2common/st2common/jinja/filters/ directory are not removed during upgrade of Stackstorm...