Does anyone else find it humorous that the majority of top-level comments in this thread start with, "Cool! I built something just like this <insert link to example> because <commercial solution> sucks!"
This tells me that (a) it is a common problem, and (b) it is not solved well. I'm guilty too: I manage a large MySQL DB and the admin panels are essentially CRUD UI rendered from the schema. The alt solutions I've clicked on here try to add more functionality, but it seems that's where things break, because every solution is different beyond the 1st-order CRUD UI. This seems to indicate there is no generic solution beyond that?
Feel free to laugh at me but I often miss MS Access layout tools.
MS Access is totally unscalable and due to it's proprietary licence I can't recommend it to anyone... but it was super freaking easy to stich together a graphical UI usable by non expert users.
I'm a SQL/GIS/Dev based on PostgreSQL/PostGis and I would totally harass my boss to buy such a solution if it were to exist. I have no time to learn latest front-end flavor so being able to quickly deploy any customizable AND simple GUI between my SQL illiterate colleagues an our database would by priceless!
If such solution does exist please have a laugh at me for being ignorant and I would really appreciate if you could send some links to that product!
> Feel free to laugh at me but I often miss MS Access layout tools
Hell, I miss Paradox. In DOS. There was a direct relationship between the tables and the UI so you got CRUD as a side effect.
I built several real application using it, but the best was a true "enterprise" application used across three facilities. The "backend" was a Netware file system and the key to making it scale was to cache necessary data locally and batch writes to shared tables. Contention had to be avoided, not because Paradox couldn't multiplex readers and writers, but because the performance was so poor.
It was still spinning when I moved on. Have no idea how long they maintained it, but I'll bet whatever they replaced it with cost an order of magnitude more. It was a part of every dollar of revenue that outfit earned employing ~1000 people.
There are decent top-to-bottom solutions like Quickbase and Google's Appmaker. I think "generic" is just hard to mix well into other code. Works better if you control the whole stack.
This tells me that (a) it is a common problem, and (b) it is not solved well. I'm guilty too: I manage a large MySQL DB and the admin panels are essentially CRUD UI rendered from the schema. The alt solutions I've clicked on here try to add more functionality, but it seems that's where things break, because every solution is different beyond the 1st-order CRUD UI. This seems to indicate there is no generic solution beyond that?