3 ms·
I was once a part of a team who's job it was to create a low/no code solution for our product. The issue we hit square in the face is that while it was easy fo
by cogman10 4y ago
I was once a part of a team who's job it was to create a low/no code solution for our product.
The issue we hit square in the face is that while it was easy for our C level person to say "Yeah, this is simple, let's just make a low code solution!" when rubber hit the road and we started asking "but, what should this DO. What sort of solution do you want?" we got nothing but "We want it to do everything, but we also want that in the next 3 months".
In other words, they wanted a full blown programming language and integration with our legacy system (with literally 1000s of datapoints) all done with like 5 devs and no requirements other than "everything" and a PM that couldn't give us a "we are successful if we can do x" sort of usecase.
Needless to say, it was a giant CF and failure.
On the flip side, the low/no code solution that I've seen work unreasonably well was doing a custom groovy dsl hooked into the necessary bits doing data transforms. It worked great for us in dev because we simply wired in an existing language and it worked great for the product because the people writing the groovy scripts were easier to hire (and often got moved over to dev if they were good enough!).