The expandable subfile is pretty much same as the loadall subfile, both having a buffer limit of 9999. The only difference being in case of expandable we have to take care of the page down.
My query is if both of them are nearly same then what is the advantage of using an expandable subfile over a loadall subfile?
On
Back in the day with older hardware, possibly talking to remote terminals over a slow comm link, expandable subfiles made a lot of sense.
You're only sending pages as needed, and compared to a page-at-a-time subfile, you're only sending them once.
Nowadays, there's very little need for them. Especially if you're not actually talking to a 5250 terminal. Even if you're not using a GUI interface toolset, modern 5250 emulators support some GUI elements such as radio buttons and scroll bars.
My preference, given that we use Profound UI, is to do a load all subfile as much as possible. Possibly with feedback to the user that more than 9999 records are available and that they need to include more filtering.
The screen is displayed faster when a request is processed by the application. No need to wait until all records are loaded and then the screen will be sent to the display. One page (aka 13 records, ±) are loaded very quickly, so even scrolling down is just minimally slower.
The older the machine, the more apparent the effect.