On Tue, 2005-06-28 at 08:22 +0100, Lee Carroll wrote: > > Well, honestly I didn't spend too much time on thinking which way would > > be better, sync or async. > > I think I used sync requests because the hourglass appears and the user > > know they have to wait a little. If I recall correctly I also had some > > problems getting async requests to work correctly on some browsers. > > Lee, do you think async requests are better? Can you explain why? > > > > Thanks, Josias > > In ie it seems that the lenya menu bar at the top of the page disappears and > then "distorts" briefly during any subsequent httpxmlrequests. I'm guessing > that async would help prevent this.
I never noticed that before... it looks like yet another ie 'feature'. > > Subsequent requests after the initial tree population may not need any > visual cue to the user as its speedy enough to be related to the user action > of clicking the + image on the menu. They will "allow" a small time delay. Actually there can be a considerable delay if one opens a node with a large number of subnodes. In that case it makes sense to display some kind of hint to the user. > > The w3c site claims that asnc is preferable as on error or transport delays > won't lock the browser. (Although I think they are asuming that the request > is doing background stuff and displaying the site tree might not be > classified as this ??) I agree that in case of an error it's not good to "lock" the browser. By using async requests, one could just display a meaningful error message after a timeout. > > Instead of the hour class an alternative could be a menu status area which > could read (loading.... etc). > > More importantly well done on writing the most impressive rocket science > javascript I've ever saw. The core part of the tree library actually was written by Guido Wesdorp, one of the developers of Kupu. I you want to improve something about the xmlhttp implementation, feel free to submit patches ;-) Cheers, Josias > > Lee C > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
