On-demand Apps (ODAG) - how to automatize the generation?
Maybe I will start from taking a step back. I want to achive the following:
I want to show on external web site a few charts from my Qlik Sense application. It will be viewed anonymously (by salesman). The app itself contains data for all customers in the company. I want to show only one customer (or a group of customers for users with more access) at once. What more I want to make sure that data is safe - that noone without required permissions won't be able to look at the data of specific customers. So just mashup with preselected customer is maybe not the best idea, since You could always change the Javascipt calling the Qlik Sense object (as it cointains preselected values in URL).
I came up with an idea of generating On-demand Apps with only the slices of data specific users need. The problem is that i have to automatize the whole process. In a daily manner i need to maintain ~50 on-demand apps for such case. Or maybe this is wrong approach? Maybe i need to define on-demand app once, and since then data in it will be refreshed automatically every time the app reloads? Do on-demand apps even reloading?
ODAG apps allow section access. There is no need to generate more than 1 template app if you you have section access in place.
If you are not using section access - than yes, you would need to have more applications. Maybe document chaining here would be a better option for you? Not sure, as I don't know the full requirements of your project.
And yes - all on-demand apps reload normally; on-demand using the selections submitted in selection app or ad-hoc by accessing the script editor in each template app.
Can't really make use of section access, since Qlik App objects will be injected into webpage via one authorised user only. That means I have to cut-off data for them previously. Of course I could made a pre-selection, but there are some security issues here - anyone with half of a brain can change the selections in URL Thats why ODAG is better here - it simply doesnt conain any more data than its already needed...