I agree with Josh(and also Tim, of course) as programmers/managers/etc. we should try to overthink about what the user is trying to obtain. Users are quite miopic when expressing their needs and even that particular mental process to attain that info. I have some users that ask pretty specifc things and I end drilling them with questions to see if the process is a one time thing or if it looks it will be something that will spread for more cases.
That being said, the fact is that sometimes I have prepared “nice” configurable solutions for problems that end changing their parameters and render the solution useless. That is a risk also and we could argue that then I didn’t do a good job at obtaining the actual requirement and developing the solution, but I am pretty sure that we all have been there.
Sometimes(depending) it is better to deliver a first entry development so the users end up defining the final and “real” permanent solution.
Sometimes, done is better than perfect. As so eloquently put in Agile terms.
Sounds like you need to farm some of this work out to that REST pro contact.
Which is what @jgiese.wci was saying. MVP…
I used to be horrible about documentation, hard coding etc… it always haunted me down the road. If anything, i have learned so much over the last two years since i started Agile Scrum project management, not only from iterative development, but gathering requirements that can be somewhat nebulous from USERS (four letter word), architecture a bit upfront pays dividends for everyone, users, other developers (future self) and lends credence to IQ/OQ/PQ which is all about delivering something that works and is scaleable as business needs always evolve and sooner or later someone will be revisiting that work down the road. All good thoughts here…
THIS
A post was merged into an existing topic: RANT: I almost lost faith in E10 today
Reminds the to ask Epicor to remove the 36 Check Limit Hardcoded and 10000 rows per Multi-Company Run Hardcoded. 
When you load in 3million parts, you can wait a few days for all the MC stuff to propogate properly from Master Company.
When you hardcode Volume:
Hey… hardcoding of things like, “MFGSYS” Site id, and other features when you create a company are my favorite complaints… you can’t have that as a Rant either… it is my rant… hahahhaah
(Glad everyone is feeling better now that they have ranted a bit).
Oh, that’s right. This was Tim’s rant about hardcoding values…
Yeah, it doesn’t take much time to code a User-Code entry lookup for many hard-coded situations. 
It’s interesting that I read this this morning, because I’ve come up against this in Kinetic, and just solved one today.
We have two companies in our installation, and some users are in both companies, but need differing permissions in the different companies. So I had to completely change the securities to segregate the companies, leaving the original securities no longer in use.
The problem I’ve run up against twice now is that certain business objects have certain securities hard coded in them.
Example: In order to print AR invoices from AR Invoice Tracker, a user needs permissions on Mass Print AR invoices. The user had permissions on the new security in my menu structure, but the business object is checking against the original (now unused) security, which they didn’t have permissions on. So they couldn’t print, until I gave them permissions on the old unused security. I’ve run up against this in two different scenarios (the other being the Order Job Wizard) in about 6 months, so I imagine there are others.
On that note, I have to keep a printout of this XKCD pinned in my cubicle to help remind me not to over-engineer sometimes 
My only caveat to this XKCD, is that it takes my IT guy 5 minutes - because he’s the only one who knows how to do it. It will take the rest of us weeks to find the document to do it - which is often out of date, more time to understand why it’s not working, and more money to hire someone who really knows how to do it. Combining documentation with automation ensures that it will be reproducible.
This thread just popped up on my radar…started reading it and got queasy, thinking this was specifically about OUR implementation. And then I saw the 2019 date - well before we started looking at Epicor. Suddenly felt like the Pepcid AC kicked in…
Can’t argue with the accuracy though. Sometimes hardcoding in BPMs can’t be avoided (and in those cases make sure you use CustNum 
)…especially if Customer “W” and Customer 
are high-visibility and high-volume (and high-anxiety) situations. Over the years and regardless of ERP platform - I’ve tried to build as much logic in each partner’s EDI maps as I can…then you don’t have to touch the core applications.