In such a case the output depends on the device, the audience is currently using.
It is not only layout, the content can change as well due to server sided user-agent-sniffing.
For example a shop could detect, that a visitor uses some Apple device - and surprise, within the output the prices for products are increased - or other products are displayed.
Concerning EPUB-Books, the idea could be, that your server sends another book, depending on the user-agent-sniffing.
I think, within the book your options to do such manipulations and paternalism is limited to java-script and DOM manipulartion, depending on the information, you can extract from the user-agent with DOM methods - because EPUB 3 requires, that always a non scripted alternative is available, either your book is damaged or users have a simple option to switch off such manipulation - this is different with server sniffing, the audience has to explore the responses of a server with different devices to discover such manipulations and frauds.
Nivin17 what you describe is more or lesse responsive layout, not adaptive layout.
You do not necessarily need CSS media queries for a responsive layout, but it is a possible part of it - for example if you need to scale embedded SVG fragments within XHTML with a font-size scaling flexible with the preferences of the reader at the same time - still a practical problem due to bugs of some user-agents and due to the problem, that EPUB3 does not allow some helpful new CSS 3 units and calculation methods.
Therefore even responsive design is somehow for EPUB 3 much more restricted and limited than for normal web pages.
As described, the options for adaptive layout are limited much more for EPUB 3, fortunately in this case.
In such a case the output depends on the device, the audience is currently using.
It is not only layout, the content can change as well due to server sided user-agent-sniffing.
For example a shop could detect, that a visitor uses some Apple device - and surprise, within the output the prices for products are increased - or other products are displayed.
Concerning EPUB-Books, the idea could be, that your server sends another book, depending on the user-agent-sniffing.
I think, within the book your options to do such manipulations and paternalism is limited to java-script and DOM manipulartion, depending on the information, you can extract from the user-agent with DOM methods - because EPUB 3 requires, that always a non scripted alternative is available, either your book is damaged or users have a simple option to switch off such manipulation - this is different with server sniffing, the audience has to explore the responses of a server with different devices to discover such manipulations and frauds.
A figment.
Nivin17 what you describe is more or lesse responsive layout, not adaptive layout.
You do not necessarily need CSS media queries for a responsive layout, but it is a possible part of it - for example if you need to scale embedded SVG fragments within XHTML with a font-size scaling flexible with the preferences of the reader at the same time - still a practical problem due to bugs of some user-agents and due to the problem, that EPUB3 does not allow some helpful new CSS 3 units and calculation methods.
Therefore even responsive design is somehow for EPUB 3 much more restricted and limited than for normal web pages.
As described, the options for adaptive layout are limited much more for EPUB 3, fortunately in this case.