TMS Webcore - load time, overhead, scalability

A few performance questions regarding TMS Webcore applications

What sort of load time are you experiencing for a small application ?

How scalable is it ? Can it support dozens of concurrent users ? What about hundreds, thousands and squillions ?

What is the CPU and memory load like as the number of concurrent users increases

Of course this will depend on the application, however I am interested in your gut feeling for best case scenarios for small applications

I dont have a specific application in mind. This is just a general discussion

Scott Hollows

Some of these, like the questions about concurrent users, are more dependent on your backend. Web Core is about your client, so that’s running on individual users machines.

Load time, definitely a web core thing. Concurrent users? that’s your server (rad server, datasnap, nodejs, aspnet, etc, etc) and your design.


In terms of load time, I know you can either:

a) have a single large js file that contains all your units and also all the rtl/framework
b) have individual js files for each unit.

a) is nice as it’s a single download, but if one unit changes, the whole thing gets re-downloaded.
b) is nice as a single change only triggers one small downloaded. The rest can be used from cache. However, you have potentially hundreds of downloads.

What I don’t yet know is whether you can have something in the middle: a single large js file with all the rtl/framework, and then maybe individual files for your app. The RTL/Framework one will probably change less frequently, so keeping that one a single file in cache sounds good. But individual files for your app, which might change more frequently, will also let us maximise cache hits/minimise downloads.