[fusion-users] Speed of Fusion vs MapFish (OpenLayers)
Paul Deschamps
pdeschamps at dmsolutions.ca
Thu Oct 30 10:17:24 EDT 2008
This single file build process will basically get the application into a
initialized / loaded state much quicker. but have no real affect on the time
of which the map window is refreshed / paned / zoomed or first loaded.
Cheers
Paul D
On Thu, Oct 30, 2008 at 9:57 AM, Andrew Parker <andrew at source3.com> wrote:
> Paul D:
>
> I see what you are saying about the perceived perception. I was fooled
> myself. All I had to do is look at the status bar and count how many
> seconds it took to download the data. The total time was approximately 13
> seconds +/- 1 second to account for map size. I will play around with Mike
> Adair's suggestion:
>
> "Unrelated to the tiling issue, you can optimize the initial loading time
> of Fusion by creating a singleFile build as per
> http://trac.osgeo.org/fusion/wiki/ANTBuildSystem "
>
> to see if I can have Fusion draw the map (tiles?) while the data is being
> downloaded to make the users think that the UI is fast.
>
> ~andrew.
>
>
>
> Paul Deschamps wrote:
>
>> Andrew,
>>
>> Unfortunately, in regards to Fusion / Mapserver and it's tiling
>> capabilities this is not something that I am aware of. not to say that it
>> doesn't exist but I have no experience with it. Perhaps someone on the list
>> can weigh in here ;)
>>
>> Again as far as the "performance" issue:
>>
>> All three demos should have the same performance (openlayers / mapfish /
>> fusion) because they are all openlayers. I think what we are seeing here is
>> a perceived performance issue instead of an actual one:
>>
>> I believe is the fact that when fusion is requesting an image it blanks to
>> a white page as an image is requested this gives you a noticeable cue and or
>> illusion of a performance issue were as with a tiled map window, you can
>> see the tiles coming in so you think it's faster.
>>
>> Playing around with the map file's size attributes have no effect as they
>> are indeed being overridden by fusion. The map image being requested from
>> the mapserv binary "CGI".
>>
>> As far as contributions they are indeed welcome. However lets see if this
>> functionality is not already scheduled to be accomplished already. Perhaps
>> there is a milestone for this in fusion 2.0.
>>
>> Anyone what to comment here?
>>
>> Cheers
>>
>> Paul D.
>>
>> On Thu, Oct 30, 2008 at 8:40 AM, Andrew Parker <andrew at source3.com<mailto:
>> andrew at source3.com>> wrote:
>>
>> Paul D:
>>
>> From your post, it is my understanding that Fusion does not have
>> the capability to use tiles. Is this true? If needed, I have a
>> programmer that can dedicate his time to add tile functionality. I
>> set the SingleTile tag to "true" then "false" in the MapSet
>> (file ApplicationDefinition.xml) to see if there was any change on
>> the way Fusion rendered the map. I did not notice any difference.
>> I also played around with the map size in the *.map file from
>> "800 640" to "200 100" to "2000 1000" to see if it made any
>> difference (curiosity, what else can I say). I did not notice any
>> difference with respect to speed; I am sure Fusion is overwriting
>> this setting. I set the final size to "300 300" for no reason at all.
>>
>> At this time, I am most concerned with fast rendering of the USGS
>> topo maps and the Geologic Map. Google is not that important.
>> But, if need be, I can have the programmer start working on the
>> OpenLayers.Layers.Google class; I will need a little help to point
>> him in the right direction.
>>
>> ~andrew
>>
>>
>> Paul Deschamps wrote:
>>
>> Hi Andrew,
>>
>> hehe yes or no answer ;)
>>
>> What you are doing here is not a "true" comparison of the
>> three different technologies.
>>
>> Your mapfish and your openlayers application are using tiles
>> where your fusion app is not. You see in a tiled mapwindow
>> tiles outside of the mapwindow can be cached so that they
>> display when they are dragged into position. Another
>> consideration about tiled mapviewers is the size of each tile.
>>
>> If the size of the tile requested from the server is requested
>> at the same native size of the tile on the server then the
>> server does not need to stitch tiles together.
>>
>> All three of these examples are using Openlayers. So from a
>> strictly fetching of the map image and presenting it on the
>> screen the performance would be identical as long as the
>> requested image is the same size across all three.
>>
>> And lastly, Your Openlayers app is using Google's Server Farm
>> directly for it's imagery where the mapfish and fusion are
>> requesting it. so it's an extra step for every map draw.
>>
>> Fusion is a toolkit that provides common set of map "widgets"
>> and UI controls that enable rapid application development
>> (RAD) there's a buzz wo from the past. :)
>>
>> If you really want fusion to preform in the same fashion that
>> your openlayers app is. You can:
>>
>> Add the OpenLayers.Layers.Google class to your openlayers.js
>> and get fusion to work with it. ( may or may not be an easy task)
>> Or wait for it to be added eventually ;)
>>
>> Cheers
>>
>> Paul D.
>>
>> On Wed, Oct 29, 2008 at 5:47 PM, Andrew Parker
>> <andrew at source3.com <mailto:andrew at source3.com>
>> <mailto:andrew at source3.com <mailto:andrew at source3.com>>> wrote:
>>
>> I am not sure if there is a way to make Fusion render maps
>> as fast
>> as OpenLayers. At this time, I just need a yes or no answer.
>>
>> For example, it would be great if I could make my current
>> Fusion
>> application (http://216.93.173.156/testing/) draw the map
>> as fast
>> as my current OpenLayers application
>> (http//:webgis.source3.com <http://webgis.source3.com>
>> <http://webgis.source3.com>) and my MapFish test at
>> (http://http://216.93.173.156/MFtest/).
>>
>> What I am using for a comparison is the Topo Map Layer with the
>> USGS wells (the data is limited to New Mexico, USA) at a
>> scale of
>> about 1:100,000. When I zoom and pan, I like how fast
>> OpenLayers/MapFish redraws.
>> I went ahead and turned on the two layers that I am using for a
>> comparison in MapFish test and my Fusion test.
>>
>> The topo is from TerraServer. The USGS data is in PostGIS.
>>
>> thanks,
>>
>> ~andrew
>> _______________________________________________
>> fusion-users mailing list
>> fusion-users at lists.osgeo.org
>> <mailto:fusion-users at lists.osgeo.org>
>> <mailto:fusion-users at lists.osgeo.org
>> <mailto:fusion-users at lists.osgeo.org>>
>>
>> http://lists.osgeo.org/mailman/listinfo/fusion-users
>>
>>
>>
>>
>> -- Paul Deschamps
>> Applications Specialist
>> DM Solutions Group Inc.
>>
>> Office: (613) 565-5056 x28
>> pdeschamps at dmsolutions.ca <mailto:pdeschamps at dmsolutions.ca>
>> <mailto:pdeschamps at dmsolutions.ca
>> <mailto:pdeschamps at dmsolutions.ca>>
>>
>> http://www.dmsolutions.ca
>> http://research.dmsolutions.ca
>>
>> _______________________________________________
>> fusion-users mailing list
>> fusion-users at lists.osgeo.org <mailto:fusion-users at lists.osgeo.org>
>> http://lists.osgeo.org/mailman/listinfo/fusion-users
>>
>>
>>
>>
>> --
>> Paul Deschamps
>> Applications Specialist
>> DM Solutions Group Inc.
>>
>> Office: (613) 565-5056 x28
>> pdeschamps at dmsolutions.ca <mailto:pdeschamps at dmsolutions.ca>
>> http://www.dmsolutions.ca
>> http://research.dmsolutions.ca
>>
>>
> _______________________________________________
> fusion-users mailing list
> fusion-users at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/fusion-users
>
--
Paul Deschamps
Applications Specialist
DM Solutions Group Inc.
Office: (613) 565-5056 x28
pdeschamps at dmsolutions.ca
http://www.dmsolutions.ca
http://research.dmsolutions.ca
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/fusion-users/attachments/20081030/f69a5e7d/attachment.html
More information about the fusion-users
mailing list