The reason this capability is useless, is that it limits the usage of SSE connection way too much... There is no reason I would have to stick my entire code for dealing with updates to a section of my app, within a single action... It makes no sense at all...
If I have an action that updates some data for a "related" section of my application, then using this mechanism I would have to copy it in it's entirety into the while-loop... This is unacceptable... I could have docents of such actions, spread across my application, each dealing with a different "aspect" within the component that is presented in the SSE-enabled view... There is a relational-database in the back - the entire application is segregated in accordance with these relations. I can have many views that deal with related pieces of information, each belonging to a different controller, and has it's own dedicated update-action within that controller. Then I might have a view, that I wan't to make SSE-enabled, which I want to be updated "in reaction to" changes in "other views".... I am not talking about "chat" only... We have a complex business-logic and a varied business-application views. We need more than a simple stand-alone controller that can only be called by the client, and can talk to (and be called from) no one else in web2py... I suggest that a very serious look be done on 0MQ for inter-process/inter-thread-communication - it could be the solution for web2py to become SSE capable - maybe even Web-Sockets capable... -- --- You received this message because you are subscribed to the Google Groups "web2py-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to web2py+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
