Server Event Log Dashboard
ExportTablesToCSV
Part of FunctionRunner
My logging template for Functions
(that automatically gathers all inputs and outputs for a function and logs them if needed)
Probably several other things both shared and not shared
Feels like it lol.
You may have broken my heart Olga, but you will forever be loved.
Do you think they could be reworked under the new restrictions, or would they be impossible going forward under these new restrictions?
Is that an eagle flying out of a dumpster fire…or a phoenix rising from the ashes?
Server Event Log Dashboard → Impossible
ExportTablesToCSV → Can be redone, with a buttload of work
Logging template → Can be worked around, but now will require manual work with each function.
Previously it was cut and paste.
Its Kevin sabotaging generative AI by feeding it its own dogfood.
Since you can just download the server logs now, could you write a utility that made em pretty?
Does it generate any meta data in the database on the function parameters (like for swagger or something?) that you could tap into?
Started last night… but that is not a replacement for what is lost
Yes, but I would need to run it through a utility for each function to generate the code, while previously it was generic. (The only non generic thing was settings, like exclusions etc)
You didn’t use reflection to generate code, couldn’t you query and to it dynamicly ? Too slow?
Half of @jgiese.wci and @josecgomez stuff is Reflection. Uh oh 
But also we do read the AuthToken Key to make our own Tokens for special things like Impersonation on-prem.
Maybe you need to block it on SaaS but allow people on-prem who pay higher licensing fees to continue to do as they please.
For those that ruined it for everyone, I hope Epicor makes everyone take a 12 months development class before they allow you to check customization mode! angry 
@Epic_Santiago
Thanks I appreciate the candour and information. Perhaps we are looking at things slightly incorrectly. Give On-prem users all the flexibility, add restrictions to Cloud (if required), @klincecums’ comment about containers certainly goes a long way resolving impacting other tenants in a cloud environment. Perhaps a certification scheme that allows enablement on a customer Id by customer Id basis for cloud may allow some flexibility…
I appreciate there are always costs involved in these types of things. To the Customer and to Epicor…It’s a matter of perspective and trying to get the balance right.
Perhaps the approach is something more detailed at the upgrade level to highlight poor code and potential issues. It’s not as if there a not tools that could be rolled into the process about that and the data is there (if everyone has telemetry enabled) I’m sure that’s where the analysis has come from or the upgrade analyser.
As far as reflection is concerned, perhaps the use cases need to be looked at in more detail, for example building a generic function that accepted the name of an object as a parameter (eg a UD table name). Using reflection would saves repetitious code, all a sudden you have a function that allows you to write to any ud table just by calling passing the UD table name and the key fields…Just an example. Perhaps this could be achieved another way, and happy to be schooled on that.
Are we getting to the point that Epicor need to start conducting a more detailed course series in “Developing with the Epicor Environment”. I hear the crowd…“Just use Widgets, problem solved”…
Anyway some thoughts and suggestions, and thanks. 
And communicated what, that we are losing it? It’s not as if we’ve been offered a solution, just a situation to figure out how to solve own our own.
I’d venture to say On Prem may have just garnered more “why I am not going to the Cloud” votes; But I can clearly see the writing on the wall, and I’ll take bets, on prem’s days are numbered because there is not enough control for Epicor (yet coincidentally, that’s why most who are, are still on prem eh?)
@Epic_Santiago the below issue must be resolved before reflection is no longer an option
The need for this utility isn’t required if logging didn’t only go to the Windows Server Log. 
In a more secure world, as @Bart_Elia pointed out years ago, keeps devs off the servers. Why log there, especially in a multi-tenant world? 
Why not enable modern logging, trace, and metric capabilities?
Maintain two code bases? Frankly, this was the mistake of multi-tenant (MT). It appears to be easier said than done I’m afraid.
EDIT: I meant two code paths in one base! See below.
I’m not a fan of Simon’s proposed solution, but no. Just two configurations. Options set on the backend.
Actually, one codebase with options was the mess we called MT.
And Epicor is not alone in this move. Listening to a security podcast, I just learned about SAP’s Clean Core initiative.
SAP cloud is wildly different from Kinetic though, keep that in mind.
No instances in SAP, it’s a true cloud product.
