[Cosmo-dev] Re: [commits-cosmo] (br) [4557] the ui uses a diffrent id system from the id's that come back from the server.

Matthew Eernisse mde at osafoundation.org
Thu May 31 18:14:57 PDT 2007


Bobby,

There's already been some previous action on this front. I foolishly 
sent my message as a reply to commits-cosmo, so it's nowhere in the 
mailing list archive. I'll repost here.

According to BCM, the set of possible characters for the UID is 
compatible, so we should be able to fix this at some point.

Multiple lozenge pieces for multi-day events don't share the same DOM 
id, so that should not cause whatever problem you saw -- the extra divs 
appearing off to the right of the main one get named like this:

eventDiv__evgh2Q2AbFBXhm (main)
eventDiv__aux1__evgh2Q2AbFBXhm
eventDiv__aux2__evgh2Q2AbFBXhm
... Etc.

If you swap out the id being used, it *should* all still work.

I can see where one of the sub-pieces of the divs in the multi-day 
lozenge array are actually getting a dupe id 
(eventDivContent__evgh2Q2AbFBXhm) for the corresponding one in the main 
div, but that's in the extisting implementation, and wouldn't cause 
stuff to break just because of the change you made.

I know the randomized string thing is a big irritation for everybody. 
Let's get a bug filed for this and get it on the list of stuff to fix.


Matthew


-------- Original Message --------
Subject: Re: [commits-cosmo] (travis) [4400] Reverting unnecessary 
change	in	generating event registry ids.
Date: Thu, 10 May 2007 17:45:43 -0500
From: Matthew Eernisse <mde at osafoundation.org>
To: commits-cosmo at osafoundation.org
References: <20070510172816.F2782F101D at tutu.osafoundation.org>

Travis,

The temp id strings are used in the DOM node ids to identify the
components of a particular cal event lozenge. I am working under the
assumption there might be characters in the UID that are not legal in
DOM node ids.

Since we're serving the content as text/html still, that means a those
characters defined as legal for the id attribute in HTML 4 -- "only
strings matching the pattern [A-Za-z][A-Za-z0-9:_.-]* should be used."

Here's the spec:

http://www.w3.org/TR/xhtml1/#C_8

A better alternative might be to do some sort of transformation on the
actual UID string of the item to ensure it's always legal for use in the
id attribute. Then we wouldn't have a different id for the same event
lozenge each time the data reloads.

Hope that helps clear up any confusion.


Matthew



svncheckin at osafoundation.org wrote:
> Revision
>     4400 <http://cvs.osafoundation.org/viewcvs.cgi?rev=4400&view=rev>
> Author
>     travis
> Date
>     2007-05-10 10:28:16 -0700 (Thu, 10 May 2007)
> 
> 
>       Log Message
> 
> Reverting unnecessary change in generating event registry ids. At some 
> point, we should revisit this, if only so I understand why we're using 
> these temporary ids.
> 
> 
>       Modified Paths
> 
>     * cosmo/trunk/cosmo/src/main/webapp/js/cosmo/view/cal.js
>       <#cosmotrunkcosmosrcmainwebappjscosmoviewcaljs>
> 
> 
>       Diff
> 
> 
>         Modified: cosmo/trunk/cosmo/src/main/webapp/js/cosmo/view/cal.js
>         (4399 => 4400)
> 
> --- cosmo/trunk/cosmo/src/main/webapp/js/cosmo/view/cal.js	2007-05-10 16:55:31 UTC (rev 4399)
> +++ cosmo/trunk/cosmo/src/main/webapp/js/cosmo/view/cal.js	2007-05-10 17:28:16 UTC (rev 4400)
> @@ -980,8 +980,7 @@
>          for (var i = 0; i < arr.length; i++) {
>              var note = arr[i];
>              var eventStamp = note.getEventStamp();
> -
> -            var id = note.getUid();
> +            var id = self.generateTempId();
>              var ev = new CalEvent(id, null);
>              ev.data = note;
>              h.setItem(id, ev);
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Commits-Cosmo mailing list
> Commits-Cosmo at osafoundation.org
> http://lists.osafoundation.org/cgi-bin/mailman/listinfo/commits-cosmo

_______________________________________________
Commits-Cosmo mailing list
Commits-Cosmo at osafoundation.org
http://lists.osafoundation.org/cgi-bin/mailman/listinfo/commits-cosmo



More information about the cosmo-dev mailing list