what do you mean? clicking on a button causes perf issues more than real time refresh?
- Romain 2012/5/29 dsh <[email protected]> > I like your proposed approach of first keep it simple and just have a > manual refresh button and then go with WebSocket, Comet [1] or similar > server-side push techniques. On the other hand I would avoid > client-side polling cause it might be causing performance issues... > > [1] http://en.wikipedia.org/wiki/Comet_%28programming%29 > > PS: IIRC RAP uses qooxdoo internally. > > Cheers > Daniel > > On Tue, May 29, 2012 at 2:50 PM, Thiago Veronezi <[email protected]> > wrote: > > There is no secret with long-polling. It is just an ajax call that hangs > > longer than the usual ones. Websockets are different: the browser should > be > > able to handle it. > > To make it simple we can just add a "refresh button", so we can load the > > log with regular ajax calls. Once the first version of this new interface > > is done we can add the autorefresh functionality. What do you think? > > > >>> Eclipse RAP > > First time a hear about it. It looks very cool!!! I will definitely take > a > > look. Thanks! > > > > Thiago. > > > > > > On Tue, May 29, 2012 at 8:40 AM, dsh <[email protected]> > wrote: > > > >> Concerning long-polling (Comet, WebSockets etc. pp), depending on the > >> framework you use the actual polling implementation is getting > >> abstracted. So the user would get the impl depending on the > >> capabilities provided by the browser and you on the other hand would > >> not need to worrie about such technical details. I know Dojo does such > >> abstraction... not sure about jQuery... > >> > >> PS: The Mockup reminds me very much about the Eclipse perspective/view > >> concept. Probably using Eclipse RAP for the UI would be an > >> overkill/unsuitable but there you would get such UI > >> lay-outing/structuring mechanisms for free :) > >> > >> Cheers > >> Daniel > >> > >> On Tue, May 29, 2012 at 2:27 PM, Thiago Veronezi <[email protected]> > >> wrote: > >> > Yes, It makes more sense, otherwise it would be very hard to read the > >> > log... which doesn't make sense. :O) > >> > > >> >>> Will it be a live log viewer where new log entries are floating in > >> > similar to digg.com? > >> > Yes. For that I plan to use long-polling (something I have experience > >> > with), but websockets are the perfect alternative if we have this > feature > >> > available at the server side. > >> > > >> > I created another mockup with a new position for the log panel and two > >> > extra panels (MDB and Webservices): > >> > http://dl.dropbox.com/u/1459144/tomee_v2.png > >> > > >> > What you guys think? > >> > > >> > Thanks for the feedback! > >> > []s, > >> > Thiago. > >> > > >> > On Tue, May 29, 2012 at 5:27 AM, dsh <[email protected]> > >> wrote: > >> > > >> >> One thought concerning the log view - maybe it would make sense to > >> >> have it at the bottom where it could span the entire screen so log > >> >> messages can be digested easier. Will it be a live log viewer where > >> >> new log entries are floating in similar to digg.com? > >> >> > >> >> Cheers > >> >> Daniel > >> >> > >> >> On Tue, May 29, 2012 at 1:19 AM, Thiago Veronezi < > [email protected]> > >> >> wrote: > >> >> > Guys, > >> >> > I need your feedback about what I have in mind for the new webapp. > >> >> > Basically I want to replace all the pages we have by a new one > which > >> will > >> >> > give the user the access to a script console, the jndi tree and the > >> tomee > >> >> > log file. > >> >> > Please check the mockup I created for it: > >> >> > http://dl.dropbox.com/u/1459144/tomee_webapp.png > >> >> > > >> >> > What do you think? > >> >> > > >> >> > Thanks, > >> >> > Thiago. > >> >> > >> >
