Since <@U01EY27APNH> is absolutely killing it :sli...
# adobe
m
Since @Mark Takata (Adobe) is absolutely killing it šŸ™‚ just wanted to surface two tickets by @lmajano from the bug tracker, regarding passing jar paths to
createObject
- the first was from 2013 and marked as "To Fix" 🤯 - it's been a long time, but it'd still be really useful: • https://tracker.adobe.com/#/view/CF-3592789 • https://tracker.adobe.com/#/view/CF-4087803
b
The other thing Lucee allows that would be very helpful is the ability to update
this.javaSettings
on the fly from within your app code using the
action="update"
attribute of the cfapplication tag.
The lack of these features is why we still have to use JavaLoader in any of our Box modules that we want to be able to load 3rd party jars to keep the
install xyz
and it "just works" flow.
ā˜šŸ¾ 1
I've kept meaning to put in a ticket for it, but it's a lot bigger than this.javasettings-- Lucee allows ANYTHING from the Application.cfc's
this
scope to be updated mid-request with this method
I tried to get Rakshith to adopt this back around CF 2018 to help ColdBox's behavior of dynamically adding CF mappings, but they didn't want to spend the time on it, which is why we have a hackaround for ACF to this day in ColdBox for CF mapping creation.
m
the need for javaloader in box modules is actually what caused me to raise this ticket. It makes it more complicated to support ACF when writing third-party modules.
m
Considering the new "inline java" stuff in ACF2021, seems like something to resurface for consideration. @Aditya Nema let's give this a chat later this week.
šŸ‘ 3
šŸ‘šŸ¾ 1
Something to note also is that I believe there is currently a bug with dynamically loading JARs which needs to be looked at. The details escape me outside of it being REALLY annoying and getting me and Ray ticked off last year lol. I wonder if, in fixing that issue, we could look at this as a feature addition, since it would likely live in the same codespace (I'm assuming).
b
There's a handfull of "killer" features CFML really needs to continue being relavent in the JVM space. The other ones IMO is a built in closure/lambda bridge and the ability to extend java classes and interfaces on the fly like java itself allows. Modern Java libs are full of that sort of stuff
ā˜šŸ¾ 1
šŸ’Æ 2
m
Also @mjclemente thank you, just out here trying my best. šŸ™‚
b
But instead, we got the spread operator šŸ˜•
I ran into this bug with a client last year but I don't know if it's what Ray hit https://tracker.adobe.com/#/view/CF-4212485
m
Brad, possibly. The gist of the issue was that attempting to use Java inline with CF using a 3rd party jar was impossible unless you put that jar in a very particular place in ACF. Something about the scope not loading properly or some other such whacky thing. It was damn annoying and its on the docket to fix, I just haven't poked my head into that particular alleyway in a long while.
I'm sure if Ray poked his head in here he'd be able to explain it much better. He was trying to do a talk on integrating the adobe.io PDF Java SDK w/ CF and it was blowing up in spectacular ways until we figured out, generally, that the "proper" way of doing was not the obvious way lol.
b
Ugh, I just remembered the "reported by" field in the adobe bug tracker search doesn't work at all
I typed in "Ray" and got dozens of results back with tickets reported by people with "ray" found nowhere in their name
Been like this for years I think
This is what I mean when I say Adobe doesn't need to replace their entire ticket tracker, just make simple crap like searching by reporter name actually work!
m
Well, Ray works for Adobe so he just emailed the engineers directly when we were looking at this, I don't think he dropped anything in the tracker (I think the engineers just make a JIRA for it, but honestly I was new at that point and don't remember what we did).
just make simple crap like searching by reporter name actually work!
If only it WERE simple, I would have made sure it got done.
b
I put in this ticket for it over a year ago https://tracker.adobe.com/#/view/TKR-223
Not a single reply from Adobe yet even acknowledging it.
I know that's not the same team as the CF team, but that's still my point when I say Adobe doesn't need a new tracker, they just need to take care of the one they have