I have them, including an aftermeeting discussion on mouse events
Steven
Gertjan van Oosten wrote:
Hi Andreas,
As quoted from Andreas Hocevar <[EMAIL PROTECTED]>:
in response to the discussion of today's team meeting,
Is this discussion archived somewhere? It's not on:
http://uwyn.com/drone/log/bevinbot/mapbuilder/20070628
Regards,
12:02 < jsb123123> hi all
12:02 < ahocevar> hi everyone!
12:02 < oterral> hi
12:04 < crschmidt> hi
12:05 < stvn> hi
12:07 < stvn> Cameron mailed that he wasn't comming
12:07 < stvn> so I'd suggest we start
12:07 < lindaD> hi
12:07 < ahocevar> ok
12:07 -!- Irssi: Pasting 7 lines to #mapbuilder. Press Ctrl-K if you wish to do
this or Ctrl-C to cancel.
12:07 < stvn> Suggested agenda:
12:07 < stvn> 00 Introductions, roll call
12:07 < stvn> 05 Progress to next release
12:07 < stvn> 10 Setting release dates
12:07 < stvn> 15 Technical Discussions
12:07 < stvn> 40 General Q&A
12:07 < stvn> 59 Topics for next meeting
12:07 < stvn> 00 Close
12:07 < stvn> anyone anything to add/remove to this?
12:08 < ahocevar> just one short remark on cscs
12:08 < oterral> yep
12:08 < stvn> k
12:08 < ahocevar> i found out that conversion of x-coordinate -90 to
gauss-kruger fails.
12:09 < ahocevar> sorry, y-coordinate
12:09 < oterral> Ok i could have a look next week what is youy epsg?
12:09 < oterral> youy=your
12:10 < ahocevar> it's in the complete example, the 3rd context in the context
collection is the target srs
12:10 < oterral> ok
12:11 < stvn> ok, progress to next release:
12:11 < jsb123123> can i ask some question about loading javascript file on
safari 3 ,in the end of meeting ?
12:11 < stvn> jsb123123: yep
12:12 < stvn> the progress seems good: I've tested quite a few functions and
only getfeatureinfo wasn't working
12:12 < stvn> but that's solved now
12:12 < stvn> I haven't tested GMLrendering/WFS-T functionality
12:12 < ahocevar> everything should work here
12:13 < stvn> ok, I noticed you started closing issues as well
12:13 < stvn> good work
12:14 < stvn> did you include URLs to demos as well?
12:14 < stvn> ahocevar: ^
12:15 < ahocevar> the wfs-t demo contains all edniting functio
12:15 < ahocevar> the shipTracks and the GeoBliki examples also use gml
renderiing
12:15 < stvn> ok, is the demo.cmb.org example working?
12:15 < ahocevar> renderiing -> rendering
12:16 < stvn> hm, there seems to be a configuration issue, I'll discuss it with
matt
12:17 < ahocevar> the proxy configuration does not work on
demo.communitymapbuilder.org
12:18 < stvn> If WFS-T and GML rendering is working I believe all functionality
is working on top of OL
12:19 < stvn> so I'm in favour of putting out an RC
12:19 < stvn> what's your opinion on this?
12:19 < ahocevar> the flickr example is still missing, but i am planning to do
that within one week
12:20 < ahocevar> and the complete example seems to have some errors related to
the legend widget, but i have not further investigated that
12:20 < stvn> ahocevar: I just found a bug: if you zoom in on the shiptrack it
dissapears you need to hide/show it to make it appear
12:20 < ahocevar> i think we can put out an rc if all examples that we want to
publish are running and looking good
12:21 < ahocevar> before that, we have to merge back to the trunk, but i think
there are only a dozen conflicting files or so
12:22 < ahocevar> stvn: might be that you are not looking at the current
version. this was fixed a week ago
12:22 < stvn> ahocevar: ok, might be the nightlies failing again
12:22 < stvn> will discuss it with matt
12:22 < stvn> is there any reason why we shouldn't merge back the comming week?
12:23 < ahocevar> stvn: you are right, there is still a bug in shipTracks. I
will fix that today.
12:24 < ahocevar> i think merging back next week is ok.
12:24 < stvn> ok, I'm +1 on merging next week
12:25 < stvn> others?
12:25 < ahocevar> +1
12:26 < lindaD> +1
12:27 < stvn> ok, next problem: who's volunteering?
12:28 < stvn> hehe
12:29 < stvn> i'll discuss it on the list then
12:29 < stvn> next item: technical discussions
12:29 < stvn> jsb123123: you wanted to discuss safari?
12:29 < jsb123123> yep
12:29 < jsb123123> i work on mapbuilder for safari 3 and i found a problem the
same problem that old safari,
12:29 < jsb123123> is for dynamicaly loading javascript files, in
mapbuilder.js, indeed the files don't load correctly and it
makes that safari 3 doesn't work. so i do some research and
i found a link to fix it.
12:30 < lindaD> right have to use the compressed file
12:30 < jsb123123>
http://pnomolos.com/article/5/dynamic-include-of-javascript-in-safari
12:32 < jsb123123> ?
12:32 < crschmidt> This is known behavior in OpenLayers
12:32 < ahocevar> i think it is enough if it works with the compressed file
12:32 < lindaD> The compressed file is created by the build process so all
mapbuilder files are in a single file
12:33 < oterral> it works for ff and opera and ie too
12:33 < oterral> ?
12:34 < stvn> jsb123123: does this mean that it creates an iframe for each js
file?
12:34 < jsb123123> yes i think
12:35 < lindaD> that is the way I'm reading the code as well
12:36 < jsb123123> yes i think , but it is the only way to fix it on safari
12:36 < jsb123123> that i found
12:36 < crschmidt> jsb123123: just use the compressed file version
12:36 < ahocevar> jsb123123: i would only agree with that if it is only used
when safari is detected. for other browsers, this is overkill.
12:36 < ahocevar> crschmidt: i totally agree
12:37 < jsb123123> me too
12:37 < stvn> the function does differentiate between IE,FF and safari
12:37 < oterral> +1
12:38 < stvn> I think it's a useful function to include for those who are
developing on safari
12:38 < stvn> a single compressed file is the proper way to build a
production-webapp anyway
12:38 < lindaD> It sure is a PITA to try to debug the compressed file
12:39 < oterral> PITA?
12:39 < lindaD> PITA = Pain in the A$$
12:39 < oterral> ahaha
12:39 < stvn> lindaD: indeed, so I like to be able to develop on seperate files
and build a single file for normal use
12:39 < crschmidt> who the hell develops on safari?
12:39 < lindaD> +11
12:39 < stvn> crschmidt: lol
12:39 < crschmidt> I mean, not to be rude or anything
12:40 < stvn> s/develop/test
12:40 < lindaD> no but we might need to debug on it
12:40 < crschmidt> yes, but in testing, you can use the compressed version
12:40 * crschmidt shrugs
12:40 < stvn> is possible
12:40 < lindaD> so really it's for us
12:40 < stvn> yep
12:40 < crschmidt> Is mapbuilder including the non-single file version of
OpenLayers?
12:40 < lindaD> end user will be using compressed file
12:40 < crschmidt> or is the OpenLayers build always a single file build?
12:41 < stvn> jsb123123: can you include it in mapbuilder so it only is used
for a non-single file use?
12:41 < stvn> crschmidt: by default it's a single file
12:41 < stvn> it is possible to use a multi-file OL if needed
12:41 < crschmidt> Okay
12:41 < crschmidt> So this discussion doesn't affect OpenLayers strongly
12:42 < stvn> nope
12:42 < crschmidt> Okay, then i'll shut up :)
12:42 < crschmidt> I was going to pipe up and say "OpenLayers is never going to
go the iframe route'
12:42 < stvn> it is mainly to make the life of developers with safari issues
easier
12:42 < crschmidt> but it sounds like it's not neccesary
12:42 < jsb123123> stvn maybe
12:43 < stvn> If it works only in multi-file setup with safari, I'm in favor of
using it
12:43 < stvn> but if it interferes with other browsers/single file setup I'd
say leave it, since singlefile works with safari
12:44 < stvn> jsb123123: does this answer your question?
12:44 < jsb123123> yep
12:45 < stvn> ok, I have an issue:
12:45 < stvn> it seems to me that the new mapbuilder is slow
12:45 < oterral> that's true
12:46 < stvn> try dragging this map around:
12:46 < stvn> http://www1.edugis.nl/nieuwestijl/
12:46 < stvn> it uses the panzoombar so one can just click/drag on the map
12:46 < stvn> it has redraw issues and feels sluggish
12:46 < crschmidt> The old MapBuilder didn't have Google Maps in it
12:47 < crschmidt> or is that not google maps?
12:47 < stvn> crschmidt: it is, but OL does it faster iirc
12:47 < crschmidt> i doubt it
12:47 < crschmidt> google integration into openlayers is sucky at best
12:47 < stvn> heh
12:48 < stvn> hm, that's a bummer
12:48 < crschmidt> I don't know, maybe i'm just being overly negative
12:48 < crschmidt> but I'd say you should take google out of the picture and
investigate
12:49 < stvn> even without google it is flickery
12:49 < stvn> but ir's faster
12:49 < stvn> it
12:50 < stvn> crschmidt: are yahoo maps/MS maps faster or is it generally an
issue?
12:51 < crschmidt> it's just generally an issue
12:51 < stvn> k
12:51 < crschmidt> Working with the commercial apis sucks.
12:52 < stvn> pity
12:52 < stvn> ok, apart from the google slowness, MapPaneOL has some issues
with dragging and redrawing
12:52 < stvn> It's nothing big, but would be nice to solve it at some point
12:52 < stvn> I'll open an issue on it
12:54 < stvn> Are there any other issues which people want to address?
12:54 < jsb123123> yep
12:54 < jsb123123> it just for a new
12:55 < jsb123123> sarrisa will be avaible with the xslprocessor/xpath for
safari soon, on cvs Head
12:56 < jsb123123> avaible = available
12:56 < jsb123123> it just for say that
12:56 < stvn> great
12:57 * stvn retracts his redrawing issue, it's only with google as far as I
can tell
12:58 < stvn> ok, anything people want to discuss next week? (apart from the
merge)
12:59 < oterral> nope
13:00 < stvn> ok, thanks everyone and till next time
13:00 < lindaD> cya
13:00 < oterral> bye
13:01 < stvn> oterral/ahocevar do you have time to hang around
13:01 < oterral> yep
13:01 < stvn> I'm working on a getfeatureinfo together with panzoombar and some
help would be appreciated
13:02 < oterral> ahah you love the difficulty
13:02 < oterral> events conflict I suppose
13:03 < stvn> well it is working, except I use the old way and there's no way
to disable it ;)
13:04 < stvn> so I was planning on triggering it with an OL event
13:04 < stvn> but not sure which one
13:04 < stvn> mouseup isn't a good one
13:04 < stvn> since I want to trigger it only on a single click, not on a
double click or drag
13:05 < stvn> brb
13:05 < ahocevar> stvn: sorry, was away for a minute. yes, I am hanging around
:-)
13:06 < ahocevar> ok, yesterday I added the mouseup event to mapPaneOL. So if
you want to do something after a point has been clicked, you
could register for the Mapbuilder mouseup event. You get an
object as parameter with the evpl property, like in the old
Mapbuilder mouseup event
13:07 < stvn> ahocevar: I saw it, but mouseup is too tricky as I expleined
13:07 < ahocevar> oh, sorry. was just catching up with the rest of the
conversation.
13:07 < stvn> np
13:08 < ahocevar> so what exactly is your use case?
13:09 < stvn> I want getfeatureinfo in this setting:
http://www1.edugis.nl/nieuwestijl/
13:09 < stvn> preferably like this:
13:09 < stvn> double click/scrollwheel = zoom in
13:09 < stvn> drag = move
13:09 < stvn> single click is getfeatureinfo with popup window
13:10 < oterral> so where is the problem?
13:10 < oterral> :)
13:10 < stvn> :)
13:11 < ahocevar> ok, I also had this problem once with ol events. it can be
solved if you store the coordinates of the previous mousedown
event. if your mouseup event has the same coordinates, you
know that it was a click.
13:12 < stvn> but you don't know whether it's a click or a double click
13:13 < stvn> however, mappaneol/ol does know the difference
13:13 < stvn> sngle click doesn't do anything right now
13:13 < stvn> and double click zooms in
13:15 < ahocevar> ok, then you better use the OpenLayers ZoomBox handler.
13:16 < ahocevar> like in mapbuilder/lib/GetFeatureInfo.js
13:16 < ahocevar> it has the (dis)advantage of also letting the user drag a
box, but then you can get feature info for a larger area or
set a tolerance. or you just take the center of the position
13:17 < stvn> right
13:17 < ahocevar> Control.prototype = OpenLayers.Class.inherit(
OpenLayers.Control, {
13:17 < ahocevar> CLASS_NAME: 'mbControl.GetFeatureInfo',
13:17 < crschmidt> ol can't tell when you double clicked that you didn't single
click
13:17 < ahocevar> type: OpenLayers.Control.TYPE_TOOL,
13:17 < ahocevar> objRef: objRef,
13:17 < ahocevar> draw: function() {
13:17 < ahocevar> this.handler = new OpenLayers.Handler.Box( this,
13:17 < ahocevar> {done: this.zoomBox}, {keyMask: this.keyMask} );
13:17 < ahocevar> }
13:17 < ahocevar> }
13:17 < crschmidt> er
13:18 < crschmidt> you don't have a zoomBox function there
13:18 < ahocevar> ah sorry,OpenLayers.Handler.Box, not ZoomBox
13:18 < crschmidt> no, i mean:
13:18 < crschmidt> {done: this.zoomBox}
13:19 < stvn> crschmidt: hm, how do you propose to solve this then?
13:19 < ahocevar> yes, zoomBox is my custom function of performing the required
action
13:19 < ahocevar> zoomBox: function(position) {....}
13:19 < crschmidt> stvn: listen to ahocevar
13:19 < oterral> andrea has right ,I've used the same way in my project
13:20 < crschmidt> you're being told exactly what to do, I'm just pointing out
lack of context so that it can be corrected :)
13:20 < stvn> :)
13:20 < oterral> but , I'm not sur it resolve the problem of conflict
dblclick/singleclick
13:20 < stvn> was just about to point that out
13:21 < ahocevar> stvn: you might want to look at
mapbuilder/lib/widget/ZoomOut.js. It uses the Box handler in
the most simple way, and also extracts the coordinates
13:21 * stvn just did a test with gmaps and notices they can't do a different
function on double click/single click either
13:21 * stvn probably opt for removing double click if it can't be both
13:22 < ahocevar> stvn: if you really want to be sure that there was a single
click, you have to delay your click handler for a few ms and
then check the coordinates of the previous doubleclick event.
if they are the same, you do not perform your click action
13:22 < stvn> hm, that might be a nice alternative
13:22 < ahocevar> but this is ugly, and that's the reason why it is not
commonly used
13:22 < stvn> thanks for the input guys
13:23 < stvn> I'll play a bit
13:23 < stvn> meanwhile, me kicks browsers for a lack of doubleclick-event
13:23 < ahocevar> ok, then I'll be out. bye!
13:23 < stvn> ciao
13:23 < oterral> kill double click , it's your only choice to survive
13:24 < stvn> lol
13:24 < oterral> may be a right click
13:25 < oterral> I don't know if it exist in ol
13:43 < stvn> crschmidt: does ol have a right click event
13:45 < crschmidt> there's a click event, and an OpenLayers.Event.isLeftClick
13:45 < stvn> k
-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
mapbuilder-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mapbuilder-devel
