what's the current state of things with regard to ...
# lucee
j
what's the current state of things with regard to lucee step debugging? i think this is a partial list: • coldbug, which has its own site, but which i gather is a work in progress alpha: https://dev.lucee.org/t/lucee-coldbug-step-debugger/8187 • fusionreactor: i think i tried it once and had issues, but i can't remember what turned me off to it • hook up the lucee source with an IDE. i've done this several times in the last several years but each time, i've hacked the configuration from scratch because i failed to document it. i think i based it on an ancient writeup from @Mark Drew (he/him) what's the latest/greatest?
z
j
thanks, @zackster. reading.
it gives info on vscode. will it not work in other ides?
z
doubt it
j
happen to know if more than one javaagent can be registered on a jvm?
z
duunno, that is on my list of things to explore, but i'm slammed with core lucee stuff at the moment. micha has a encoder for webp working btw, will be published next week probably
j
writeable webp!?
z
yup
j
šŸ˜
that will be so convenient if the timing works out. thanks for mentioning it!
m
What happened to the CodeRaven Project that was planned for railo? Is this still a Thing or already burried?
🤷 1
j
i'm having some trouble with luceedebug, maybe since i'm not a vs code user:
@David Rogers , i submitted a PR and and an issue, btw.
šŸ‘€ 1
z
@David Rogers can you add some hashtags to the project about description to it's more discoverable ? i.e. like https://github.com/zspitzer/lucee-performance-analyzer
āœ… 1
j
@zackster, have you used this debugger yet, BTW?
z
nah, my code all works perfectly the first time, why would i need a step debugger?
šŸ˜› 1
j
just wondering if i'm the first person besides @David Rogers to give it a spin. we've been making progress as he's fixes the snags i come across but i haven't gotten it connected yet.
z
I'm glad you're diving in, I'd love to but i simply have to much other stuff on my hands am keeping an eye on it tho šŸ˜‰
j
@zackster, did you recommend it because it's the solution that you (and/or the lucee team) would like to come to fruition? just wondered how you came to recommend it when i asked about latest/greatest. i wondered about the context.
z
The Lucee team reckon this is a really interesting project
j
Okay, that answers my question, thanks.
am keeping an eye on it tho šŸ˜‰
thanks for that, too. i'm glad you're chiming in on the tickets when you can.
it's working!!! @David Rogers and i have flushed out some bugs and some extra documentation and after one more (docs-related) PR gets merged, i think it's in a good place for the community to give it a try.
šŸŽ‰ 1
z
Time for an introductory post / announcement on Dev.lucee.org then
j
happen to know if more than one javaagent can be registered on a jvm?
as I was just walking a colleague through using the step debugger, we saw that lucee didn't start up. it turned out that she had our fusionreactor-enabling environment variable set, which would have tried to run both agents at once. there may be some syntax or trick in using multiple agents, but it didn't work right out of the box.
not a big deal. just circling back.
on the plus side, i think she may be the third person to use luceedebug. i'm going to be giving an orientation to my team next week so they can start using it!
d
I think multiple agents can run alongisde each other, so long as they don't conflict with what they're doing to the JVM. But it does appear there are conflicts. I took a quick look and it seems that FusionReactor takes JVMTI capabilities that are "solo-capabilities" (i.e. one agent only -- fusion reactor sets breakpoints on its own without jdwp yes?), so then jdwp can't get the capabilities it needs and luceedebug can't use jdwp and all is lost. There might be some classloading/conflicting-instrumentation issues too, but that first thing seems like a blocker.
z
have you seen all the FD*.java classes in lucee?
d
yeah, I think there's a (closed source) native agent portion that works alongside them though? not sure how to glom onto it.