Map Application - GSoC 19

classic Classic list List threaded Threaded
133 messages Options
1234567
Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Fawad Ali
Hi Caty,

The last thing to be wary of is to have the Leaflet dependency installed.
If I remove my Leaflet webjar, I get this set of errors.

GET http://ginpachi-pc:8080/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.css
net::ERR_ABORTED 404 (Not Found)
require.min.js?r=1:34 GET
http://ginpachi-pc:8080/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.js?r=1
net::ERR_ABORTED 404 (Not Found)
require.min.js?r=1:7 Uncaught Error: Script error for "leaflet", needed by:
leafletSearch
http://requirejs.org/docs/errors.html#scripterror
    at F (require.min.js?r=1:7)
    at HTMLScriptElement.onScriptError (require.min.js?r=1:30)

I think Its the same problem your side.
For installing Leaflet, search for Map Macro in extensions, from its
dependencies tab in details go to Leaflet and install it.
That ought to work. :)

Best,
Fawad


On Thu, Jun 6, 2019 at 2:24 PM Ecaterina Moraru (Valica) <[hidden email]>
wrote:

> Hi Fawad,
>
> I still have the issue: no maps loaded, leaflet.css error in console.
> Tested with the most recent snapshot of 11.5 and with the build from
> https://github.com/xwiki-contrib/application-interactive-maps
> I have the same problem with Firefox, Chrome, Safari and Opera. I'm on
> Mac.
> https://up1.xwikisas.com/#mXONZtF2qcMeftvnyJOSjw
> The script from “
> http://localhost:8084/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.js?r=1”
> was loaded even though its MIME type (“text/html”) is not a valid
> JavaScript MIME type.
>
> Also Maps.MapTesting.Points.Islamabad.WebHome page seems to have error
> when importing (with the "Replace the page history with the history from
> the package" option) https://up1.xwikisas.com/#ag8AIsPHgcFYqVod2RMJQQ
>
> Thanks,
> Caty
>
>
> On Wed, Jun 5, 2019 at 8:50 PM Fawad Ali <[hidden email]> wrote:
>
>> I just built and imported the application from my own repo (
>> https://github.com/9inpachi/interactive-maps-new) and everything seems
>> fine.
>>
>> There was that error in the more earlier builds but it was fixed, may be
>> some of the source files (especially Leaflet.xml) are old?
>> Try building the source files anew with "mvn clean install". May be that
>> will help.
>>
>> Thanks,
>> Fawad
>>
>> On Wed, Jun 5, 2019, 10:37 PM Fawad Ali <[hidden email] wrote:
>>
>>> We have shifted the repo to xwiki-contrib again. You may try that. I
>>> will also check my own repo for any errors ASAP.
>>>
>>> Best,
>>> Fawad
>>>
>>> On Wed, Jun 5, 2019, 10:35 PM Ecaterina Moraru (Valica) <
>>> [hidden email] wrote:
>>>
>>>> I've used the latest build from
>>>> https://github.com/9inpachi/interactive-maps-new
>>>> and I have the error both on 11.4-rc-1 and some 11.4-snapshot.
>>>>
>>>> Thanks,
>>>> Caty
>>>>
>>>> On Wed, Jun 5, 2019 at 8:30 PM Fawad Ali <[hidden email]>
>>>> wrote:
>>>>
>>>>> Hi,
>>>>>
>>>>> Caty, do you have the error with the latest git repo as well?
>>>>> Actually the leaflet-commons require leaflet but the functions are not
>>>>> actually called anywhere without the leaflet dependency used in require.
>>>>>
>>>>> There is no error on my or Stephane side.
>>>>> I have the 11.3-rc version of XWiki.
>>>>> You can try Ctrl+F5 for a complete new load of the resources.
>>>>>
>>>>> Best,
>>>>> Fawad
>>>>>
>>>>> On Wed, Jun 5, 2019, 10:22 PM Ecaterina Moraru (Valica) <
>>>>> [hidden email] wrote:
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> Some notes:
>>>>>> - We don't have guidelines regarding the singular / plural thing. I'm
>>>>>> glad that on the new sources we don't have the Maps/Map anymore. I'm fine
>>>>>> with Maps. In practice we have a mix of singular (like Diagram, Calendar,
>>>>>> Meeting) and plural (like Ideas, Forums). I prefer the plural version,
>>>>>> although in practice I think we have more with singular. There was a
>>>>>> tentative old draft for having such guidelines
>>>>>> https://design.xwiki.org/xwiki/bin/view/Proposal/ApplicationGuidelines
>>>>>> but we didn't worked on it for some time.
>>>>>>
>>>>>> - Regarding the new Git repository. Since you've committed the
>>>>>> initial commits in issues, you should do a release with the initial
>>>>>> version, and than just release a new version for the interactive-maps-new .
>>>>>> It's normal in an application's development flow that changes happen,
>>>>>> that's why versioning schemes are all about.
>>>>>>
>>>>>> - I still have the error I've mentioned before :
>>>>>> Uncaught Error: Script error for "leaflet", needed by: leafletSearch
>>>>>> http://requirejs.org/docs/errors.html#scripterror
>>>>>>     at F (require.min.js?r=1:7)
>>>>>>     at HTMLScriptElement.onScriptError (require.min.js?r=1:30)
>>>>>>
>>>>>> leaflet.css:1 Failed to load resource: the server responded with a
>>>>>> status of 404 (Not Found)
>>>>>>
>>>>>> so I cannot actually test the build, since I don't see the maps. I
>>>>>> have this both on Chrome and Firefox. Do I need to do something?
>>>>>>
>>>>>> Thanks,
>>>>>> Caty
>>>>>>
>>>>>> On Tue, Jun 4, 2019 at 2:31 PM Fawad Ali <[hidden email]>
>>>>>> wrote:
>>>>>>
>>>>>>> Also, I forgot to mention it before but we will need a better and
>>>>>>> more expressive way to show popups. We need something that can accomodate
>>>>>>> sufficient amount of text with a scroll if the information exceeds the page.
>>>>>>> I will prepare a mockup for this once I am done with some of the
>>>>>>> next steps.
>>>>>>>
>>>>>>> And I think we should use the colortheme colors for our map controls
>>>>>>> and consequently for the popups. I will update you on that as well.
>>>>>>>
>>>>>>> Best,
>>>>>>> Fawad
>>>>>>>
>>>>>>>
>>>>>>> On Tue, Jun 4, 2019 at 3:31 PM Fawad Ali <[hidden email]>
>>>>>>> wrote:
>>>>>>>
>>>>>>>> Hi Stephane, Caty and all,
>>>>>>>> Hope you are doing fine.
>>>>>>>>
>>>>>>>> I am glad you brought up the topic of custom marker icon. I am well
>>>>>>>> aware of the issue. Actually there are two problems with custom markers.
>>>>>>>> - The icon offset
>>>>>>>> - The document attachment
>>>>>>>>
>>>>>>>> For the icon offset, when I tried to fix it initially it seemed
>>>>>>>> that I can overcome the offset either by height or width which means that
>>>>>>>> the offset still exists from a single side so I had that postponed since I
>>>>>>>> thought solr query tasks take priority.
>>>>>>>>
>>>>>>>> For the attachment, for now I am getting the first attachment (0th
>>>>>>>> index) from the Point page which is not very reliable. For example if we
>>>>>>>> have images on the page, it could be that the marker takes one of the
>>>>>>>> attachments even if the user did not want a custom icon or an image
>>>>>>>> different from what the user wanted to choose is selected as the marker
>>>>>>>> icon.
>>>>>>>>
>>>>>>>> What I have in mind is that we define categories for marker icons
>>>>>>>> dynamically.
>>>>>>>> We could make a separate dedicated page "MarkerIcons" and attach
>>>>>>>> multiple images to it. Then these images could appear in a list as one of
>>>>>>>> the properties in the Point object where we can choose the icon from. WDYT?
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> Fawad
>>>>>>>>
>>>>>>>> On Tue, Jun 4, 2019, 11:31 AM Stéphane Laurière <
>>>>>>>> [hidden email] wrote:
>>>>>>>>
>>>>>>>>> Fawad, Thanks for letting us know, I could install the new app
>>>>>>>>> version, I confirm that all the changes you added to the progress file
>>>>>>>>> (very handy) work for me, and the refactoring is ok. I noticed a minor
>>>>>>>>> issue that you're certainly aware of already: it seems there's a small
>>>>>>>>> offset between the custom marker position (with the Islamabad point) and
>>>>>>>>> the popup position.
>>>>>>>>>
>>>>>>>>> Talk to you soon,
>>>>>>>>>
>>>>>>>>> Stéphane
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Fawad Ali:
>>>>>>>>> > Hi all,
>>>>>>>>> >
>>>>>>>>> > Thanks for the detailed review, Stephane. I have made the
>>>>>>>>> changes you suggested with some next steps also done.
>>>>>>>>> >
>>>>>>>>> > Furthermore, I will make changes to the application space once
>>>>>>>>> we have confirmed response from Caty or other developers.
>>>>>>>>> > I have started to work on the other next steps and will provide
>>>>>>>>> with updates soon.
>>>>>>>>> >
>>>>>>>>> > The original github repo is also updated, so future updates will
>>>>>>>>> be available at
>>>>>>>>> https://github.com/xwiki-contrib/application-interactive-maps.
>>>>>>>>> >
>>>>>>>>> > Thanks,
>>>>>>>>> > Fawad
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> Stéphane Laurière
>>>>>>>>> XWiki – https://xwiki.com
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Ecaterina Moraru (Valica)
That most likely is the problem, since I've just imported the XAR, no real
install. I will try like that, thanks for helping.

Caty

On Thu, Jun 6, 2019 at 1:03 PM Fawad Ali <[hidden email]> wrote:

> Hi Caty,
>
> The last thing to be wary of is to have the Leaflet dependency installed.
> If I remove my Leaflet webjar, I get this set of errors.
>
> GET http://ginpachi-pc:8080/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.css
> net::ERR_ABORTED 404 (Not Found)
> require.min.js?r=1:34 GET
> http://ginpachi-pc:8080/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.js?r=1
> net::ERR_ABORTED 404 (Not Found)
> require.min.js?r=1:7 Uncaught Error: Script error for "leaflet", needed
> by: leafletSearch
> http://requirejs.org/docs/errors.html#scripterror
>     at F (require.min.js?r=1:7)
>     at HTMLScriptElement.onScriptError (require.min.js?r=1:30)
>
> I think Its the same problem your side.
> For installing Leaflet, search for Map Macro in extensions, from its
> dependencies tab in details go to Leaflet and install it.
> That ought to work. :)
>
> Best,
> Fawad
>
>
> On Thu, Jun 6, 2019 at 2:24 PM Ecaterina Moraru (Valica) <
> [hidden email]> wrote:
>
>> Hi Fawad,
>>
>> I still have the issue: no maps loaded, leaflet.css error in console.
>> Tested with the most recent snapshot of 11.5 and with the build from
>> https://github.com/xwiki-contrib/application-interactive-maps
>> I have the same problem with Firefox, Chrome, Safari and Opera. I'm on
>> Mac.
>> https://up1.xwikisas.com/#mXONZtF2qcMeftvnyJOSjw
>> The script from “
>> http://localhost:8084/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.js?r=1”
>> was loaded even though its MIME type (“text/html”) is not a valid
>> JavaScript MIME type.
>>
>> Also Maps.MapTesting.Points.Islamabad.WebHome page seems to have error
>> when importing (with the "Replace the page history with the history from
>> the package" option) https://up1.xwikisas.com/#ag8AIsPHgcFYqVod2RMJQQ
>>
>> Thanks,
>> Caty
>>
>>
>> On Wed, Jun 5, 2019 at 8:50 PM Fawad Ali <[hidden email]> wrote:
>>
>>> I just built and imported the application from my own repo (
>>> https://github.com/9inpachi/interactive-maps-new) and everything seems
>>> fine.
>>>
>>> There was that error in the more earlier builds but it was fixed, may be
>>> some of the source files (especially Leaflet.xml) are old?
>>> Try building the source files anew with "mvn clean install". May be that
>>> will help.
>>>
>>> Thanks,
>>> Fawad
>>>
>>> On Wed, Jun 5, 2019, 10:37 PM Fawad Ali <[hidden email] wrote:
>>>
>>>> We have shifted the repo to xwiki-contrib again. You may try that. I
>>>> will also check my own repo for any errors ASAP.
>>>>
>>>> Best,
>>>> Fawad
>>>>
>>>> On Wed, Jun 5, 2019, 10:35 PM Ecaterina Moraru (Valica) <
>>>> [hidden email] wrote:
>>>>
>>>>> I've used the latest build from
>>>>> https://github.com/9inpachi/interactive-maps-new
>>>>> and I have the error both on 11.4-rc-1 and some 11.4-snapshot.
>>>>>
>>>>> Thanks,
>>>>> Caty
>>>>>
>>>>> On Wed, Jun 5, 2019 at 8:30 PM Fawad Ali <[hidden email]>
>>>>> wrote:
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> Caty, do you have the error with the latest git repo as well?
>>>>>> Actually the leaflet-commons require leaflet but the functions are
>>>>>> not actually called anywhere without the leaflet dependency used in require.
>>>>>>
>>>>>> There is no error on my or Stephane side.
>>>>>> I have the 11.3-rc version of XWiki.
>>>>>> You can try Ctrl+F5 for a complete new load of the resources.
>>>>>>
>>>>>> Best,
>>>>>> Fawad
>>>>>>
>>>>>> On Wed, Jun 5, 2019, 10:22 PM Ecaterina Moraru (Valica) <
>>>>>> [hidden email] wrote:
>>>>>>
>>>>>>> Hi,
>>>>>>>
>>>>>>> Some notes:
>>>>>>> - We don't have guidelines regarding the singular / plural thing.
>>>>>>> I'm glad that on the new sources we don't have the Maps/Map anymore. I'm
>>>>>>> fine with Maps. In practice we have a mix of singular (like Diagram,
>>>>>>> Calendar, Meeting) and plural (like Ideas, Forums). I prefer the plural
>>>>>>> version, although in practice I think we have more with singular. There was
>>>>>>> a tentative old draft for having such guidelines
>>>>>>> https://design.xwiki.org/xwiki/bin/view/Proposal/ApplicationGuidelines
>>>>>>> but we didn't worked on it for some time.
>>>>>>>
>>>>>>> - Regarding the new Git repository. Since you've committed the
>>>>>>> initial commits in issues, you should do a release with the initial
>>>>>>> version, and than just release a new version for the interactive-maps-new .
>>>>>>> It's normal in an application's development flow that changes happen,
>>>>>>> that's why versioning schemes are all about.
>>>>>>>
>>>>>>> - I still have the error I've mentioned before :
>>>>>>> Uncaught Error: Script error for "leaflet", needed by: leafletSearch
>>>>>>> http://requirejs.org/docs/errors.html#scripterror
>>>>>>>     at F (require.min.js?r=1:7)
>>>>>>>     at HTMLScriptElement.onScriptError (require.min.js?r=1:30)
>>>>>>>
>>>>>>> leaflet.css:1 Failed to load resource: the server responded with a
>>>>>>> status of 404 (Not Found)
>>>>>>>
>>>>>>> so I cannot actually test the build, since I don't see the maps. I
>>>>>>> have this both on Chrome and Firefox. Do I need to do something?
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Caty
>>>>>>>
>>>>>>> On Tue, Jun 4, 2019 at 2:31 PM Fawad Ali <[hidden email]>
>>>>>>> wrote:
>>>>>>>
>>>>>>>> Also, I forgot to mention it before but we will need a better and
>>>>>>>> more expressive way to show popups. We need something that can accomodate
>>>>>>>> sufficient amount of text with a scroll if the information exceeds the page.
>>>>>>>> I will prepare a mockup for this once I am done with some of the
>>>>>>>> next steps.
>>>>>>>>
>>>>>>>> And I think we should use the colortheme colors for our map
>>>>>>>> controls and consequently for the popups. I will update you on that as well.
>>>>>>>>
>>>>>>>> Best,
>>>>>>>> Fawad
>>>>>>>>
>>>>>>>>
>>>>>>>> On Tue, Jun 4, 2019 at 3:31 PM Fawad Ali <[hidden email]>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> Hi Stephane, Caty and all,
>>>>>>>>> Hope you are doing fine.
>>>>>>>>>
>>>>>>>>> I am glad you brought up the topic of custom marker icon. I am
>>>>>>>>> well aware of the issue. Actually there are two problems with custom
>>>>>>>>> markers.
>>>>>>>>> - The icon offset
>>>>>>>>> - The document attachment
>>>>>>>>>
>>>>>>>>> For the icon offset, when I tried to fix it initially it seemed
>>>>>>>>> that I can overcome the offset either by height or width which means that
>>>>>>>>> the offset still exists from a single side so I had that postponed since I
>>>>>>>>> thought solr query tasks take priority.
>>>>>>>>>
>>>>>>>>> For the attachment, for now I am getting the first attachment (0th
>>>>>>>>> index) from the Point page which is not very reliable. For example if we
>>>>>>>>> have images on the page, it could be that the marker takes one of the
>>>>>>>>> attachments even if the user did not want a custom icon or an image
>>>>>>>>> different from what the user wanted to choose is selected as the marker
>>>>>>>>> icon.
>>>>>>>>>
>>>>>>>>> What I have in mind is that we define categories for marker icons
>>>>>>>>> dynamically.
>>>>>>>>> We could make a separate dedicated page "MarkerIcons" and attach
>>>>>>>>> multiple images to it. Then these images could appear in a list as one of
>>>>>>>>> the properties in the Point object where we can choose the icon from. WDYT?
>>>>>>>>>
>>>>>>>>> Thanks,
>>>>>>>>> Fawad
>>>>>>>>>
>>>>>>>>> On Tue, Jun 4, 2019, 11:31 AM Stéphane Laurière <
>>>>>>>>> [hidden email] wrote:
>>>>>>>>>
>>>>>>>>>> Fawad, Thanks for letting us know, I could install the new app
>>>>>>>>>> version, I confirm that all the changes you added to the progress file
>>>>>>>>>> (very handy) work for me, and the refactoring is ok. I noticed a minor
>>>>>>>>>> issue that you're certainly aware of already: it seems there's a small
>>>>>>>>>> offset between the custom marker position (with the Islamabad point) and
>>>>>>>>>> the popup position.
>>>>>>>>>>
>>>>>>>>>> Talk to you soon,
>>>>>>>>>>
>>>>>>>>>> Stéphane
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Fawad Ali:
>>>>>>>>>> > Hi all,
>>>>>>>>>> >
>>>>>>>>>> > Thanks for the detailed review, Stephane. I have made the
>>>>>>>>>> changes you suggested with some next steps also done.
>>>>>>>>>> >
>>>>>>>>>> > Furthermore, I will make changes to the application space once
>>>>>>>>>> we have confirmed response from Caty or other developers.
>>>>>>>>>> > I have started to work on the other next steps and will provide
>>>>>>>>>> with updates soon.
>>>>>>>>>> >
>>>>>>>>>> > The original github repo is also updated, so future updates
>>>>>>>>>> will be available at
>>>>>>>>>> https://github.com/xwiki-contrib/application-interactive-maps.
>>>>>>>>>> >
>>>>>>>>>> > Thanks,
>>>>>>>>>> > Fawad
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> --
>>>>>>>>>> Stéphane Laurière
>>>>>>>>>> XWiki – https://xwiki.com
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Fawad Ali
Hi all,
Hope you all are doing wonderfully.

We still have the issue for the facets rendering but its up on the forum,
so I think we will have some fixes for it in due time (thanks to Stephane).
In the meantime, I will try making the facets more specific to the map. For
now I am considering the following options for facets:
- Tags
- Search field (searches inside map item pages)
- Map item type (PointClass for now)
- Map item icon (?)
- Map item space
I will need some suggestions on which other options should be included.

Next, I will
-  Be working on making a dedicated space for popups inside or besides the
map. Something similar to https://abc.gogocarto.fr
- Make use of theme styles and colors for the map itself and the items
inside
- Be making the query search asynchronous

Also, I will be working part time for the next week and I have exams the
week after that. So I would like to let you know that I would not be
available for the exams period for the most part. :(
I will try to finish most of the work and get the application in a stable
state by then. :)
Ecaterina also mentioned that we release the application, I propose we do
that next week as soon as the facets are in a stable state.

Thanks,
Fawad


On Thu, Jun 6, 2019 at 6:42 PM Ecaterina Moraru (Valica) <[hidden email]>
wrote:

> That most likely is the problem, since I've just imported the XAR, no real
> install. I will try like that, thanks for helping.
>
> Caty
>
> On Thu, Jun 6, 2019 at 1:03 PM Fawad Ali <[hidden email]> wrote:
>
>> Hi Caty,
>>
>> The last thing to be wary of is to have the Leaflet dependency installed.
>> If I remove my Leaflet webjar, I get this set of errors.
>>
>> GET
>> http://ginpachi-pc:8080/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.css
>> net::ERR_ABORTED 404 (Not Found)
>> require.min.js?r=1:34 GET
>> http://ginpachi-pc:8080/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.js?r=1
>> net::ERR_ABORTED 404 (Not Found)
>> require.min.js?r=1:7 Uncaught Error: Script error for "leaflet", needed
>> by: leafletSearch
>> http://requirejs.org/docs/errors.html#scripterror
>>     at F (require.min.js?r=1:7)
>>     at HTMLScriptElement.onScriptError (require.min.js?r=1:30)
>>
>> I think Its the same problem your side.
>> For installing Leaflet, search for Map Macro in extensions, from its
>> dependencies tab in details go to Leaflet and install it.
>> That ought to work. :)
>>
>> Best,
>> Fawad
>>
>>
>> On Thu, Jun 6, 2019 at 2:24 PM Ecaterina Moraru (Valica) <
>> [hidden email]> wrote:
>>
>>> Hi Fawad,
>>>
>>> I still have the issue: no maps loaded, leaflet.css error in console.
>>> Tested with the most recent snapshot of 11.5 and with the build from
>>> https://github.com/xwiki-contrib/application-interactive-maps
>>> I have the same problem with Firefox, Chrome, Safari and Opera. I'm on
>>> Mac.
>>> https://up1.xwikisas.com/#mXONZtF2qcMeftvnyJOSjw
>>> The script from “
>>> http://localhost:8084/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.js?r=1”
>>> was loaded even though its MIME type (“text/html”) is not a valid
>>> JavaScript MIME type.
>>>
>>> Also Maps.MapTesting.Points.Islamabad.WebHome page seems to have error
>>> when importing (with the "Replace the page history with the history
>>> from the package" option)
>>> https://up1.xwikisas.com/#ag8AIsPHgcFYqVod2RMJQQ
>>>
>>> Thanks,
>>> Caty
>>>
>>>
>>> On Wed, Jun 5, 2019 at 8:50 PM Fawad Ali <[hidden email]>
>>> wrote:
>>>
>>>> I just built and imported the application from my own repo (
>>>> https://github.com/9inpachi/interactive-maps-new) and everything seems
>>>> fine.
>>>>
>>>> There was that error in the more earlier builds but it was fixed, may
>>>> be some of the source files (especially Leaflet.xml) are old?
>>>> Try building the source files anew with "mvn clean install". May be
>>>> that will help.
>>>>
>>>> Thanks,
>>>> Fawad
>>>>
>>>> On Wed, Jun 5, 2019, 10:37 PM Fawad Ali <[hidden email] wrote:
>>>>
>>>>> We have shifted the repo to xwiki-contrib again. You may try that. I
>>>>> will also check my own repo for any errors ASAP.
>>>>>
>>>>> Best,
>>>>> Fawad
>>>>>
>>>>> On Wed, Jun 5, 2019, 10:35 PM Ecaterina Moraru (Valica) <
>>>>> [hidden email] wrote:
>>>>>
>>>>>> I've used the latest build from
>>>>>> https://github.com/9inpachi/interactive-maps-new
>>>>>> and I have the error both on 11.4-rc-1 and some 11.4-snapshot.
>>>>>>
>>>>>> Thanks,
>>>>>> Caty
>>>>>>
>>>>>> On Wed, Jun 5, 2019 at 8:30 PM Fawad Ali <[hidden email]>
>>>>>> wrote:
>>>>>>
>>>>>>> Hi,
>>>>>>>
>>>>>>> Caty, do you have the error with the latest git repo as well?
>>>>>>> Actually the leaflet-commons require leaflet but the functions are
>>>>>>> not actually called anywhere without the leaflet dependency used in require.
>>>>>>>
>>>>>>> There is no error on my or Stephane side.
>>>>>>> I have the 11.3-rc version of XWiki.
>>>>>>> You can try Ctrl+F5 for a complete new load of the resources.
>>>>>>>
>>>>>>> Best,
>>>>>>> Fawad
>>>>>>>
>>>>>>> On Wed, Jun 5, 2019, 10:22 PM Ecaterina Moraru (Valica) <
>>>>>>> [hidden email] wrote:
>>>>>>>
>>>>>>>> Hi,
>>>>>>>>
>>>>>>>> Some notes:
>>>>>>>> - We don't have guidelines regarding the singular / plural thing.
>>>>>>>> I'm glad that on the new sources we don't have the Maps/Map anymore. I'm
>>>>>>>> fine with Maps. In practice we have a mix of singular (like Diagram,
>>>>>>>> Calendar, Meeting) and plural (like Ideas, Forums). I prefer the plural
>>>>>>>> version, although in practice I think we have more with singular. There was
>>>>>>>> a tentative old draft for having such guidelines
>>>>>>>> https://design.xwiki.org/xwiki/bin/view/Proposal/ApplicationGuidelines
>>>>>>>> but we didn't worked on it for some time.
>>>>>>>>
>>>>>>>> - Regarding the new Git repository. Since you've committed the
>>>>>>>> initial commits in issues, you should do a release with the initial
>>>>>>>> version, and than just release a new version for the interactive-maps-new .
>>>>>>>> It's normal in an application's development flow that changes happen,
>>>>>>>> that's why versioning schemes are all about.
>>>>>>>>
>>>>>>>> - I still have the error I've mentioned before :
>>>>>>>> Uncaught Error: Script error for "leaflet", needed by: leafletSearch
>>>>>>>> http://requirejs.org/docs/errors.html#scripterror
>>>>>>>>     at F (require.min.js?r=1:7)
>>>>>>>>     at HTMLScriptElement.onScriptError (require.min.js?r=1:30)
>>>>>>>>
>>>>>>>> leaflet.css:1 Failed to load resource: the server responded with a
>>>>>>>> status of 404 (Not Found)
>>>>>>>>
>>>>>>>> so I cannot actually test the build, since I don't see the maps. I
>>>>>>>> have this both on Chrome and Firefox. Do I need to do something?
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> Caty
>>>>>>>>
>>>>>>>> On Tue, Jun 4, 2019 at 2:31 PM Fawad Ali <[hidden email]>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> Also, I forgot to mention it before but we will need a better and
>>>>>>>>> more expressive way to show popups. We need something that can accomodate
>>>>>>>>> sufficient amount of text with a scroll if the information exceeds the page.
>>>>>>>>> I will prepare a mockup for this once I am done with some of the
>>>>>>>>> next steps.
>>>>>>>>>
>>>>>>>>> And I think we should use the colortheme colors for our map
>>>>>>>>> controls and consequently for the popups. I will update you on that as well.
>>>>>>>>>
>>>>>>>>> Best,
>>>>>>>>> Fawad
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Tue, Jun 4, 2019 at 3:31 PM Fawad Ali <[hidden email]>
>>>>>>>>> wrote:
>>>>>>>>>
>>>>>>>>>> Hi Stephane, Caty and all,
>>>>>>>>>> Hope you are doing fine.
>>>>>>>>>>
>>>>>>>>>> I am glad you brought up the topic of custom marker icon. I am
>>>>>>>>>> well aware of the issue. Actually there are two problems with custom
>>>>>>>>>> markers.
>>>>>>>>>> - The icon offset
>>>>>>>>>> - The document attachment
>>>>>>>>>>
>>>>>>>>>> For the icon offset, when I tried to fix it initially it seemed
>>>>>>>>>> that I can overcome the offset either by height or width which means that
>>>>>>>>>> the offset still exists from a single side so I had that postponed since I
>>>>>>>>>> thought solr query tasks take priority.
>>>>>>>>>>
>>>>>>>>>> For the attachment, for now I am getting the first attachment
>>>>>>>>>> (0th index) from the Point page which is not very reliable. For example if
>>>>>>>>>> we have images on the page, it could be that the marker takes one of the
>>>>>>>>>> attachments even if the user did not want a custom icon or an image
>>>>>>>>>> different from what the user wanted to choose is selected as the marker
>>>>>>>>>> icon.
>>>>>>>>>>
>>>>>>>>>> What I have in mind is that we define categories for marker icons
>>>>>>>>>> dynamically.
>>>>>>>>>> We could make a separate dedicated page "MarkerIcons" and attach
>>>>>>>>>> multiple images to it. Then these images could appear in a list as one of
>>>>>>>>>> the properties in the Point object where we can choose the icon from. WDYT?
>>>>>>>>>>
>>>>>>>>>> Thanks,
>>>>>>>>>> Fawad
>>>>>>>>>>
>>>>>>>>>> On Tue, Jun 4, 2019, 11:31 AM Stéphane Laurière <
>>>>>>>>>> [hidden email] wrote:
>>>>>>>>>>
>>>>>>>>>>> Fawad, Thanks for letting us know, I could install the new app
>>>>>>>>>>> version, I confirm that all the changes you added to the progress file
>>>>>>>>>>> (very handy) work for me, and the refactoring is ok. I noticed a minor
>>>>>>>>>>> issue that you're certainly aware of already: it seems there's a small
>>>>>>>>>>> offset between the custom marker position (with the Islamabad point) and
>>>>>>>>>>> the popup position.
>>>>>>>>>>>
>>>>>>>>>>> Talk to you soon,
>>>>>>>>>>>
>>>>>>>>>>> Stéphane
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> Fawad Ali:
>>>>>>>>>>> > Hi all,
>>>>>>>>>>> >
>>>>>>>>>>> > Thanks for the detailed review, Stephane. I have made the
>>>>>>>>>>> changes you suggested with some next steps also done.
>>>>>>>>>>> >
>>>>>>>>>>> > Furthermore, I will make changes to the application space once
>>>>>>>>>>> we have confirmed response from Caty or other developers.
>>>>>>>>>>> > I have started to work on the other next steps and will
>>>>>>>>>>> provide with updates soon.
>>>>>>>>>>> >
>>>>>>>>>>> > The original github repo is also updated, so future updates
>>>>>>>>>>> will be available at
>>>>>>>>>>> https://github.com/xwiki-contrib/application-interactive-maps.
>>>>>>>>>>> >
>>>>>>>>>>> > Thanks,
>>>>>>>>>>> > Fawad
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> --
>>>>>>>>>>> Stéphane Laurière
>>>>>>>>>>> XWiki – https://xwiki.com
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Ecaterina Moraru (Valica)
Hi,

On Fri, Jun 7, 2019 at 2:01 PM Fawad Ali <[hidden email]> wrote:

> Hi all,
> Hope you all are doing wonderfully.
>
> We still have the issue for the facets rendering but its up on the forum,
> so I think we will have some fixes for it in due time (thanks to Stephane).
> In the meantime, I will try making the facets more specific to the map.
> For now I am considering the following options for facets:
> - Tags
> - Search field (searches inside map item pages)
> - Map item type (PointClass for now)
> - Map item icon (?)
> - Map item space
> I will need some suggestions on which other options should be included.
>
> Next, I will
> -  Be working on making a dedicated space for popups inside or besides the
> map. Something similar to https://abc.gogocarto.fr
> - Make use of theme styles and colors for the map itself and the items
> inside
> - Be making the query search asynchronous
>
> Also, I will be working part time for the next week and I have exams the
> week after that. So I would like to let you know that I would not be
> available for the exams period for the most part. :(
> I will try to finish most of the work and get the application in a stable
> state by then. :)
> Ecaterina also mentioned that we release the application, I propose we do
> that next week as soon as the facets are in a stable state.
>

That would be great in order for people to faster install it and test it.
You could also request feedback on the forum if you want.
When releasing you also need to write the documentation for the existing
features, mentioning how to use them, etc. + list the fixed issues.

I managed to get the maps working. Thanks for your help.
Some comments on the current state:
- Maps/MapTesting/Maps/TestMap - I find it strange that the Maps space is
duplicated
- MapTesting, Maps, Points - spaces don't have homepages. The users will
navigate to them, since they are present in breadcrumb. So what is the
plan? Simpler paths? or create Homepages for these types of entry?
- Lots of pages that are not hidden. All technical pages needs to be
hidden.
- A bit confusing that there are 2 search boxes for the maps, see
https://up1.xwikisas.com/#XngcOAZexsKE4DryH2i6zA
- The Search input is not really working. It kind of works (with a strange
reload effect) for queries like "Paris" or "Moscow", but if you enter
anything else it defaults on Paris. If that search is supposed to filter
only the existing points than the placeholder should state that somehow +
treat the non existing point with a "No point of interest found" or
something. Also when it doesn't find the location, the search facets area
is empty, so kind of hard for an user to recover it's track.
- Regarding the facets, we need some more user friendly translations and
customizations for this kind of UI, see
https://up1.xwikisas.com/#TYj_9oLn84Mfp87VnwNgkw

Thanks,
Caty


> Thanks,
> Fawad
>
>
> On Thu, Jun 6, 2019 at 6:42 PM Ecaterina Moraru (Valica) <
> [hidden email]> wrote:
>
>> That most likely is the problem, since I've just imported the XAR, no
>> real install. I will try like that, thanks for helping.
>>
>> Caty
>>
>> On Thu, Jun 6, 2019 at 1:03 PM Fawad Ali <[hidden email]> wrote:
>>
>>> Hi Caty,
>>>
>>> The last thing to be wary of is to have the Leaflet dependency installed.
>>> If I remove my Leaflet webjar, I get this set of errors.
>>>
>>> GET
>>> http://ginpachi-pc:8080/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.css
>>> net::ERR_ABORTED 404 (Not Found)
>>> require.min.js?r=1:34 GET
>>> http://ginpachi-pc:8080/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.js?r=1
>>> net::ERR_ABORTED 404 (Not Found)
>>> require.min.js?r=1:7 Uncaught Error: Script error for "leaflet", needed
>>> by: leafletSearch
>>> http://requirejs.org/docs/errors.html#scripterror
>>>     at F (require.min.js?r=1:7)
>>>     at HTMLScriptElement.onScriptError (require.min.js?r=1:30)
>>>
>>> I think Its the same problem your side.
>>> For installing Leaflet, search for Map Macro in extensions, from its
>>> dependencies tab in details go to Leaflet and install it.
>>> That ought to work. :)
>>>
>>> Best,
>>> Fawad
>>>
>>>
>>> On Thu, Jun 6, 2019 at 2:24 PM Ecaterina Moraru (Valica) <
>>> [hidden email]> wrote:
>>>
>>>> Hi Fawad,
>>>>
>>>> I still have the issue: no maps loaded, leaflet.css error in console.
>>>> Tested with the most recent snapshot of 11.5 and with the build from
>>>> https://github.com/xwiki-contrib/application-interactive-maps
>>>> I have the same problem with Firefox, Chrome, Safari and Opera. I'm on
>>>> Mac.
>>>> https://up1.xwikisas.com/#mXONZtF2qcMeftvnyJOSjw
>>>> The script from “
>>>> http://localhost:8084/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.js?r=1”
>>>> was loaded even though its MIME type (“text/html”) is not a valid
>>>> JavaScript MIME type.
>>>>
>>>> Also Maps.MapTesting.Points.Islamabad.WebHome page seems to have error
>>>> when importing (with the "Replace the page history with the history
>>>> from the package" option)
>>>> https://up1.xwikisas.com/#ag8AIsPHgcFYqVod2RMJQQ
>>>>
>>>> Thanks,
>>>> Caty
>>>>
>>>>
>>>> On Wed, Jun 5, 2019 at 8:50 PM Fawad Ali <[hidden email]>
>>>> wrote:
>>>>
>>>>> I just built and imported the application from my own repo (
>>>>> https://github.com/9inpachi/interactive-maps-new) and everything
>>>>> seems fine.
>>>>>
>>>>> There was that error in the more earlier builds but it was fixed, may
>>>>> be some of the source files (especially Leaflet.xml) are old?
>>>>> Try building the source files anew with "mvn clean install". May be
>>>>> that will help.
>>>>>
>>>>> Thanks,
>>>>> Fawad
>>>>>
>>>>> On Wed, Jun 5, 2019, 10:37 PM Fawad Ali <[hidden email]
>>>>> wrote:
>>>>>
>>>>>> We have shifted the repo to xwiki-contrib again. You may try that. I
>>>>>> will also check my own repo for any errors ASAP.
>>>>>>
>>>>>> Best,
>>>>>> Fawad
>>>>>>
>>>>>> On Wed, Jun 5, 2019, 10:35 PM Ecaterina Moraru (Valica) <
>>>>>> [hidden email] wrote:
>>>>>>
>>>>>>> I've used the latest build from
>>>>>>> https://github.com/9inpachi/interactive-maps-new
>>>>>>> and I have the error both on 11.4-rc-1 and some 11.4-snapshot.
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Caty
>>>>>>>
>>>>>>> On Wed, Jun 5, 2019 at 8:30 PM Fawad Ali <[hidden email]>
>>>>>>> wrote:
>>>>>>>
>>>>>>>> Hi,
>>>>>>>>
>>>>>>>> Caty, do you have the error with the latest git repo as well?
>>>>>>>> Actually the leaflet-commons require leaflet but the functions are
>>>>>>>> not actually called anywhere without the leaflet dependency used in require.
>>>>>>>>
>>>>>>>> There is no error on my or Stephane side.
>>>>>>>> I have the 11.3-rc version of XWiki.
>>>>>>>> You can try Ctrl+F5 for a complete new load of the resources.
>>>>>>>>
>>>>>>>> Best,
>>>>>>>> Fawad
>>>>>>>>
>>>>>>>> On Wed, Jun 5, 2019, 10:22 PM Ecaterina Moraru (Valica) <
>>>>>>>> [hidden email] wrote:
>>>>>>>>
>>>>>>>>> Hi,
>>>>>>>>>
>>>>>>>>> Some notes:
>>>>>>>>> - We don't have guidelines regarding the singular / plural thing.
>>>>>>>>> I'm glad that on the new sources we don't have the Maps/Map anymore. I'm
>>>>>>>>> fine with Maps. In practice we have a mix of singular (like Diagram,
>>>>>>>>> Calendar, Meeting) and plural (like Ideas, Forums). I prefer the plural
>>>>>>>>> version, although in practice I think we have more with singular. There was
>>>>>>>>> a tentative old draft for having such guidelines
>>>>>>>>> https://design.xwiki.org/xwiki/bin/view/Proposal/ApplicationGuidelines
>>>>>>>>> but we didn't worked on it for some time.
>>>>>>>>>
>>>>>>>>> - Regarding the new Git repository. Since you've committed the
>>>>>>>>> initial commits in issues, you should do a release with the initial
>>>>>>>>> version, and than just release a new version for the interactive-maps-new .
>>>>>>>>> It's normal in an application's development flow that changes happen,
>>>>>>>>> that's why versioning schemes are all about.
>>>>>>>>>
>>>>>>>>> - I still have the error I've mentioned before :
>>>>>>>>> Uncaught Error: Script error for "leaflet", needed by:
>>>>>>>>> leafletSearch
>>>>>>>>> http://requirejs.org/docs/errors.html#scripterror
>>>>>>>>>     at F (require.min.js?r=1:7)
>>>>>>>>>     at HTMLScriptElement.onScriptError (require.min.js?r=1:30)
>>>>>>>>>
>>>>>>>>> leaflet.css:1 Failed to load resource: the server responded with a
>>>>>>>>> status of 404 (Not Found)
>>>>>>>>>
>>>>>>>>> so I cannot actually test the build, since I don't see the maps. I
>>>>>>>>> have this both on Chrome and Firefox. Do I need to do something?
>>>>>>>>>
>>>>>>>>> Thanks,
>>>>>>>>> Caty
>>>>>>>>>
>>>>>>>>> On Tue, Jun 4, 2019 at 2:31 PM Fawad Ali <[hidden email]>
>>>>>>>>> wrote:
>>>>>>>>>
>>>>>>>>>> Also, I forgot to mention it before but we will need a better and
>>>>>>>>>> more expressive way to show popups. We need something that can accomodate
>>>>>>>>>> sufficient amount of text with a scroll if the information exceeds the page.
>>>>>>>>>> I will prepare a mockup for this once I am done with some of the
>>>>>>>>>> next steps.
>>>>>>>>>>
>>>>>>>>>> And I think we should use the colortheme colors for our map
>>>>>>>>>> controls and consequently for the popups. I will update you on that as well.
>>>>>>>>>>
>>>>>>>>>> Best,
>>>>>>>>>> Fawad
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> On Tue, Jun 4, 2019 at 3:31 PM Fawad Ali <[hidden email]>
>>>>>>>>>> wrote:
>>>>>>>>>>
>>>>>>>>>>> Hi Stephane, Caty and all,
>>>>>>>>>>> Hope you are doing fine.
>>>>>>>>>>>
>>>>>>>>>>> I am glad you brought up the topic of custom marker icon. I am
>>>>>>>>>>> well aware of the issue. Actually there are two problems with custom
>>>>>>>>>>> markers.
>>>>>>>>>>> - The icon offset
>>>>>>>>>>> - The document attachment
>>>>>>>>>>>
>>>>>>>>>>> For the icon offset, when I tried to fix it initially it seemed
>>>>>>>>>>> that I can overcome the offset either by height or width which means that
>>>>>>>>>>> the offset still exists from a single side so I had that postponed since I
>>>>>>>>>>> thought solr query tasks take priority.
>>>>>>>>>>>
>>>>>>>>>>> For the attachment, for now I am getting the first attachment
>>>>>>>>>>> (0th index) from the Point page which is not very reliable. For example if
>>>>>>>>>>> we have images on the page, it could be that the marker takes one of the
>>>>>>>>>>> attachments even if the user did not want a custom icon or an image
>>>>>>>>>>> different from what the user wanted to choose is selected as the marker
>>>>>>>>>>> icon.
>>>>>>>>>>>
>>>>>>>>>>> What I have in mind is that we define categories for marker
>>>>>>>>>>> icons dynamically.
>>>>>>>>>>> We could make a separate dedicated page "MarkerIcons" and attach
>>>>>>>>>>> multiple images to it. Then these images could appear in a list as one of
>>>>>>>>>>> the properties in the Point object where we can choose the icon from. WDYT?
>>>>>>>>>>>
>>>>>>>>>>> Thanks,
>>>>>>>>>>> Fawad
>>>>>>>>>>>
>>>>>>>>>>> On Tue, Jun 4, 2019, 11:31 AM Stéphane Laurière <
>>>>>>>>>>> [hidden email] wrote:
>>>>>>>>>>>
>>>>>>>>>>>> Fawad, Thanks for letting us know, I could install the new app
>>>>>>>>>>>> version, I confirm that all the changes you added to the progress file
>>>>>>>>>>>> (very handy) work for me, and the refactoring is ok. I noticed a minor
>>>>>>>>>>>> issue that you're certainly aware of already: it seems there's a small
>>>>>>>>>>>> offset between the custom marker position (with the Islamabad point) and
>>>>>>>>>>>> the popup position.
>>>>>>>>>>>>
>>>>>>>>>>>> Talk to you soon,
>>>>>>>>>>>>
>>>>>>>>>>>> Stéphane
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> Fawad Ali:
>>>>>>>>>>>> > Hi all,
>>>>>>>>>>>> >
>>>>>>>>>>>> > Thanks for the detailed review, Stephane. I have made the
>>>>>>>>>>>> changes you suggested with some next steps also done.
>>>>>>>>>>>> >
>>>>>>>>>>>> > Furthermore, I will make changes to the application space
>>>>>>>>>>>> once we have confirmed response from Caty or other developers.
>>>>>>>>>>>> > I have started to work on the other next steps and will
>>>>>>>>>>>> provide with updates soon.
>>>>>>>>>>>> >
>>>>>>>>>>>> > The original github repo is also updated, so future updates
>>>>>>>>>>>> will be available at
>>>>>>>>>>>> https://github.com/xwiki-contrib/application-interactive-maps.
>>>>>>>>>>>> >
>>>>>>>>>>>> > Thanks,
>>>>>>>>>>>> > Fawad
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> --
>>>>>>>>>>>> Stéphane Laurière
>>>>>>>>>>>> XWiki – https://xwiki.com
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Fawad Ali
Hi Caty,
Thanks for the review.


> Maps/MapTesting/Maps/TestMap - I find it strange that the Maps space is
> duplicated
>

This space exists only for testing. It won't be there for the real
application. I named them so that its easier to know which type of object
pages are located in them (for myself).

MapTesting, Maps, Points - spaces don't have homepages. The users will
> navigate to them, since they are present in breadcrumb. So what is the
> plan? Simpler paths? or create Homepages for these types of entry?
>

Since we are in the beta stage now, the whole MapTesting space exists for
testing for developers. It would not be there once we have a stable version
ready for release.

Lots of pages that are not hidden. All technical pages needs to be
> hidden.
>

Again, these pages are not technical and exist only for testing purposes.

A bit confusing that there are 2 search boxes for the maps, see
> https://up1.xwikisas.com/#XngcOAZexsKE4DryH2i6zA


 Yes, thanks for pointing out. We need to move the search function directly
inside the map. I will look into it once I am done with the facets.

 For the search input, its in extremely beta stage. I am still trying to
figure out the macros I have borrowed from SolrSearchMacros. So it will
take time for me to make a more stable version of the facets. I hope I can
do that in due time. :)

Regarding the facets, we need some more user friendly translations and
> customizations for this kind of UI, see
> https://up1.xwikisas.com/#TYj_9oLn84Mfp87VnwNgkw


As discussed earlier with Stephane, we are still having issues using the
normal $facetDisplayer so we are using a workaround for testing purposes
that's why it looks like this. More precisely, we are using
the #displaySearchFacetValues($facetValues) macro for now.

Best,
Fawad


On Fri, Jun 7, 2019 at 5:18 PM Ecaterina Moraru (Valica) <[hidden email]>
wrote:

> Hi,
>
> On Fri, Jun 7, 2019 at 2:01 PM Fawad Ali <[hidden email]> wrote:
>
>> Hi all,
>> Hope you all are doing wonderfully.
>>
>> We still have the issue for the facets rendering but its up on the forum,
>> so I think we will have some fixes for it in due time (thanks to Stephane).
>> In the meantime, I will try making the facets more specific to the map.
>> For now I am considering the following options for facets:
>> - Tags
>> - Search field (searches inside map item pages)
>> - Map item type (PointClass for now)
>> - Map item icon (?)
>> - Map item space
>> I will need some suggestions on which other options should be included.
>>
>> Next, I will
>> -  Be working on making a dedicated space for popups inside or besides
>> the map. Something similar to https://abc.gogocarto.fr
>> - Make use of theme styles and colors for the map itself and the items
>> inside
>> - Be making the query search asynchronous
>>
>> Also, I will be working part time for the next week and I have exams the
>> week after that. So I would like to let you know that I would not be
>> available for the exams period for the most part. :(
>> I will try to finish most of the work and get the application in a stable
>> state by then. :)
>> Ecaterina also mentioned that we release the application, I propose we do
>> that next week as soon as the facets are in a stable state.
>>
>
> That would be great in order for people to faster install it and test it.
> You could also request feedback on the forum if you want.
> When releasing you also need to write the documentation for the existing
> features, mentioning how to use them, etc. + list the fixed issues.
>
> I managed to get the maps working. Thanks for your help.
> Some comments on the current state:
> - Maps/MapTesting/Maps/TestMap - I find it strange that the Maps space is
> duplicated
> - MapTesting, Maps, Points - spaces don't have homepages. The users will
> navigate to them, since they are present in breadcrumb. So what is the
> plan? Simpler paths? or create Homepages for these types of entry?
> - Lots of pages that are not hidden. All technical pages needs to be
> hidden.
> - A bit confusing that there are 2 search boxes for the maps, see
> https://up1.xwikisas.com/#XngcOAZexsKE4DryH2i6zA
> - The Search input is not really working. It kind of works (with a strange
> reload effect) for queries like "Paris" or "Moscow", but if you enter
> anything else it defaults on Paris. If that search is supposed to filter
> only the existing points than the placeholder should state that somehow +
> treat the non existing point with a "No point of interest found" or
> something. Also when it doesn't find the location, the search facets area
> is empty, so kind of hard for an user to recover it's track.
> - Regarding the facets, we need some more user friendly translations and
> customizations for this kind of UI, see
> https://up1.xwikisas.com/#TYj_9oLn84Mfp87VnwNgkw
>
> Thanks,
> Caty
>
>
>> Thanks,
>> Fawad
>>
>>
>> On Thu, Jun 6, 2019 at 6:42 PM Ecaterina Moraru (Valica) <
>> [hidden email]> wrote:
>>
>>> That most likely is the problem, since I've just imported the XAR, no
>>> real install. I will try like that, thanks for helping.
>>>
>>> Caty
>>>
>>> On Thu, Jun 6, 2019 at 1:03 PM Fawad Ali <[hidden email]>
>>> wrote:
>>>
>>>> Hi Caty,
>>>>
>>>> The last thing to be wary of is to have the Leaflet dependency
>>>> installed.
>>>> If I remove my Leaflet webjar, I get this set of errors.
>>>>
>>>> GET
>>>> http://ginpachi-pc:8080/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.css
>>>> net::ERR_ABORTED 404 (Not Found)
>>>> require.min.js?r=1:34 GET
>>>> http://ginpachi-pc:8080/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.js?r=1
>>>> net::ERR_ABORTED 404 (Not Found)
>>>> require.min.js?r=1:7 Uncaught Error: Script error for "leaflet", needed
>>>> by: leafletSearch
>>>> http://requirejs.org/docs/errors.html#scripterror
>>>>     at F (require.min.js?r=1:7)
>>>>     at HTMLScriptElement.onScriptError (require.min.js?r=1:30)
>>>>
>>>> I think Its the same problem your side.
>>>> For installing Leaflet, search for Map Macro in extensions, from its
>>>> dependencies tab in details go to Leaflet and install it.
>>>> That ought to work. :)
>>>>
>>>> Best,
>>>> Fawad
>>>>
>>>>
>>>> On Thu, Jun 6, 2019 at 2:24 PM Ecaterina Moraru (Valica) <
>>>> [hidden email]> wrote:
>>>>
>>>>> Hi Fawad,
>>>>>
>>>>> I still have the issue: no maps loaded, leaflet.css error in console.
>>>>> Tested with the most recent snapshot of 11.5 and with the build from
>>>>> https://github.com/xwiki-contrib/application-interactive-maps
>>>>> I have the same problem with Firefox, Chrome, Safari and Opera. I'm on
>>>>> Mac.
>>>>> https://up1.xwikisas.com/#mXONZtF2qcMeftvnyJOSjw
>>>>> The script from “
>>>>> http://localhost:8084/xwiki/webjars/wiki%3Axwiki/leaflet/leaflet.js?r=1”
>>>>> was loaded even though its MIME type (“text/html”) is not a valid
>>>>> JavaScript MIME type.
>>>>>
>>>>> Also Maps.MapTesting.Points.Islamabad.WebHome page seems to have
>>>>> error when importing (with the "Replace the page history with the
>>>>> history from the package" option)
>>>>> https://up1.xwikisas.com/#ag8AIsPHgcFYqVod2RMJQQ
>>>>>
>>>>> Thanks,
>>>>> Caty
>>>>>
>>>>>
>>>>> On Wed, Jun 5, 2019 at 8:50 PM Fawad Ali <[hidden email]>
>>>>> wrote:
>>>>>
>>>>>> I just built and imported the application from my own repo (
>>>>>> https://github.com/9inpachi/interactive-maps-new) and everything
>>>>>> seems fine.
>>>>>>
>>>>>> There was that error in the more earlier builds but it was fixed, may
>>>>>> be some of the source files (especially Leaflet.xml) are old?
>>>>>> Try building the source files anew with "mvn clean install". May be
>>>>>> that will help.
>>>>>>
>>>>>> Thanks,
>>>>>> Fawad
>>>>>>
>>>>>> On Wed, Jun 5, 2019, 10:37 PM Fawad Ali <[hidden email]
>>>>>> wrote:
>>>>>>
>>>>>>> We have shifted the repo to xwiki-contrib again. You may try that. I
>>>>>>> will also check my own repo for any errors ASAP.
>>>>>>>
>>>>>>> Best,
>>>>>>> Fawad
>>>>>>>
>>>>>>> On Wed, Jun 5, 2019, 10:35 PM Ecaterina Moraru (Valica) <
>>>>>>> [hidden email] wrote:
>>>>>>>
>>>>>>>> I've used the latest build from
>>>>>>>> https://github.com/9inpachi/interactive-maps-new
>>>>>>>> and I have the error both on 11.4-rc-1 and some 11.4-snapshot.
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> Caty
>>>>>>>>
>>>>>>>> On Wed, Jun 5, 2019 at 8:30 PM Fawad Ali <[hidden email]>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> Hi,
>>>>>>>>>
>>>>>>>>> Caty, do you have the error with the latest git repo as well?
>>>>>>>>> Actually the leaflet-commons require leaflet but the functions are
>>>>>>>>> not actually called anywhere without the leaflet dependency used in require.
>>>>>>>>>
>>>>>>>>> There is no error on my or Stephane side.
>>>>>>>>> I have the 11.3-rc version of XWiki.
>>>>>>>>> You can try Ctrl+F5 for a complete new load of the resources.
>>>>>>>>>
>>>>>>>>> Best,
>>>>>>>>> Fawad
>>>>>>>>>
>>>>>>>>> On Wed, Jun 5, 2019, 10:22 PM Ecaterina Moraru (Valica) <
>>>>>>>>> [hidden email] wrote:
>>>>>>>>>
>>>>>>>>>> Hi,
>>>>>>>>>>
>>>>>>>>>> Some notes:
>>>>>>>>>> - We don't have guidelines regarding the singular / plural thing.
>>>>>>>>>> I'm glad that on the new sources we don't have the Maps/Map anymore. I'm
>>>>>>>>>> fine with Maps. In practice we have a mix of singular (like Diagram,
>>>>>>>>>> Calendar, Meeting) and plural (like Ideas, Forums). I prefer the plural
>>>>>>>>>> version, although in practice I think we have more with singular. There was
>>>>>>>>>> a tentative old draft for having such guidelines
>>>>>>>>>> https://design.xwiki.org/xwiki/bin/view/Proposal/ApplicationGuidelines
>>>>>>>>>> but we didn't worked on it for some time.
>>>>>>>>>>
>>>>>>>>>> - Regarding the new Git repository. Since you've committed the
>>>>>>>>>> initial commits in issues, you should do a release with the initial
>>>>>>>>>> version, and than just release a new version for the interactive-maps-new .
>>>>>>>>>> It's normal in an application's development flow that changes happen,
>>>>>>>>>> that's why versioning schemes are all about.
>>>>>>>>>>
>>>>>>>>>> - I still have the error I've mentioned before :
>>>>>>>>>> Uncaught Error: Script error for "leaflet", needed by:
>>>>>>>>>> leafletSearch
>>>>>>>>>> http://requirejs.org/docs/errors.html#scripterror
>>>>>>>>>>     at F (require.min.js?r=1:7)
>>>>>>>>>>     at HTMLScriptElement.onScriptError (require.min.js?r=1:30)
>>>>>>>>>>
>>>>>>>>>> leaflet.css:1 Failed to load resource: the server responded with
>>>>>>>>>> a status of 404 (Not Found)
>>>>>>>>>>
>>>>>>>>>> so I cannot actually test the build, since I don't see the maps.
>>>>>>>>>> I have this both on Chrome and Firefox. Do I need to do something?
>>>>>>>>>>
>>>>>>>>>> Thanks,
>>>>>>>>>> Caty
>>>>>>>>>>
>>>>>>>>>> On Tue, Jun 4, 2019 at 2:31 PM Fawad Ali <[hidden email]>
>>>>>>>>>> wrote:
>>>>>>>>>>
>>>>>>>>>>> Also, I forgot to mention it before but we will need a better
>>>>>>>>>>> and more expressive way to show popups. We need something that can
>>>>>>>>>>> accomodate sufficient amount of text with a scroll if the information
>>>>>>>>>>> exceeds the page.
>>>>>>>>>>> I will prepare a mockup for this once I am done with some of the
>>>>>>>>>>> next steps.
>>>>>>>>>>>
>>>>>>>>>>> And I think we should use the colortheme colors for our map
>>>>>>>>>>> controls and consequently for the popups. I will update you on that as well.
>>>>>>>>>>>
>>>>>>>>>>> Best,
>>>>>>>>>>> Fawad
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> On Tue, Jun 4, 2019 at 3:31 PM Fawad Ali <
>>>>>>>>>>> [hidden email]> wrote:
>>>>>>>>>>>
>>>>>>>>>>>> Hi Stephane, Caty and all,
>>>>>>>>>>>> Hope you are doing fine.
>>>>>>>>>>>>
>>>>>>>>>>>> I am glad you brought up the topic of custom marker icon. I am
>>>>>>>>>>>> well aware of the issue. Actually there are two problems with custom
>>>>>>>>>>>> markers.
>>>>>>>>>>>> - The icon offset
>>>>>>>>>>>> - The document attachment
>>>>>>>>>>>>
>>>>>>>>>>>> For the icon offset, when I tried to fix it initially it seemed
>>>>>>>>>>>> that I can overcome the offset either by height or width which means that
>>>>>>>>>>>> the offset still exists from a single side so I had that postponed since I
>>>>>>>>>>>> thought solr query tasks take priority.
>>>>>>>>>>>>
>>>>>>>>>>>> For the attachment, for now I am getting the first attachment
>>>>>>>>>>>> (0th index) from the Point page which is not very reliable. For example if
>>>>>>>>>>>> we have images on the page, it could be that the marker takes one of the
>>>>>>>>>>>> attachments even if the user did not want a custom icon or an image
>>>>>>>>>>>> different from what the user wanted to choose is selected as the marker
>>>>>>>>>>>> icon.
>>>>>>>>>>>>
>>>>>>>>>>>> What I have in mind is that we define categories for marker
>>>>>>>>>>>> icons dynamically.
>>>>>>>>>>>> We could make a separate dedicated page "MarkerIcons" and
>>>>>>>>>>>> attach multiple images to it. Then these images could appear in a list as
>>>>>>>>>>>> one of the properties in the Point object where we can choose the icon
>>>>>>>>>>>> from. WDYT?
>>>>>>>>>>>>
>>>>>>>>>>>> Thanks,
>>>>>>>>>>>> Fawad
>>>>>>>>>>>>
>>>>>>>>>>>> On Tue, Jun 4, 2019, 11:31 AM Stéphane Laurière <
>>>>>>>>>>>> [hidden email] wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>> Fawad, Thanks for letting us know, I could install the new app
>>>>>>>>>>>>> version, I confirm that all the changes you added to the progress file
>>>>>>>>>>>>> (very handy) work for me, and the refactoring is ok. I noticed a minor
>>>>>>>>>>>>> issue that you're certainly aware of already: it seems there's a small
>>>>>>>>>>>>> offset between the custom marker position (with the Islamabad point) and
>>>>>>>>>>>>> the popup position.
>>>>>>>>>>>>>
>>>>>>>>>>>>> Talk to you soon,
>>>>>>>>>>>>>
>>>>>>>>>>>>> Stéphane
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> Fawad Ali:
>>>>>>>>>>>>> > Hi all,
>>>>>>>>>>>>> >
>>>>>>>>>>>>> > Thanks for the detailed review, Stephane. I have made the
>>>>>>>>>>>>> changes you suggested with some next steps also done.
>>>>>>>>>>>>> >
>>>>>>>>>>>>> > Furthermore, I will make changes to the application space
>>>>>>>>>>>>> once we have confirmed response from Caty or other developers.
>>>>>>>>>>>>> > I have started to work on the other next steps and will
>>>>>>>>>>>>> provide with updates soon.
>>>>>>>>>>>>> >
>>>>>>>>>>>>> > The original github repo is also updated, so future updates
>>>>>>>>>>>>> will be available at
>>>>>>>>>>>>> https://github.com/xwiki-contrib/application-interactive-maps.
>>>>>>>>>>>>> >
>>>>>>>>>>>>> > Thanks,
>>>>>>>>>>>>> > Fawad
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> --
>>>>>>>>>>>>> Stéphane Laurière
>>>>>>>>>>>>> XWiki – https://xwiki.com
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>
Reply | Threaded
Open this post in threaded view
|

Map Application - GSoC 19

Stéphane Laurière-6
In reply to this post by Fawad Ali
Hi Fawad, Caty, all,

I just submitted some code to the interactive-maps repository that aims at a sample map with real data and facets (as a pull request, just to make sure we're in tune first, let me know). The code consists of the following:

- A Museum class with a single "country" property (string)

- A DataImporter script for converting Wikidata exported in JSON via a SPARQL query (that you can find in page Museums.WebHome) into XWiki pages consisting of a Museum object and a Point object.

- A simple map configuration for displaying the imported data

In order to configure the facets for this map, you'll have to add the following code to the top of the CommonMacros (if we go this way, we'll have to make this configurable at each map level):

{{velocity output="false"}}
#set ($solrConfig = {
   'filterQuery': [
     'type:DOCUMENT',
     'class:Maps.Code.PointClass'
   ],
   'facetFields': [
     'property.Maps.MapTesting.Museums.Code.MuseumClass.country_string'
   ]
})
{{/velocity}}

Here's a capture of what I get on my side when selecting the "Spain" facet:

   https://up1.xwikisas.com/#ZTkIQ8K1fXLrY5dgvp6BvA

Hoping such sample data can help developing further and think in greater details about real case scenarios. When displaying such a map, that'd be great if the available facets would get displayed directly, without the need to enter a query in the input field (you probably have this in mind already as well).

Cheers

Stéphane

Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Stéphane Laurière-6
In reply to this post by Fawad Ali
Fawad, Caty, all,

I have a short comment about the tests:

> Hi Caty,
> Thanks for the review.
>
>     Maps/MapTesting/Maps/TestMap - I find it strange that the Maps space is duplicated
>
>
> This space exists only for testing. It won't be there for the real application. I named them so that its easier to know which type of object pages are located in them (for myself).
>
>     MapTesting, Maps, Points - spaces don't have homepages. The users will navigate to them, since they are present in breadcrumb. So what is the plan? Simpler paths? or create Homepages for these types of entry?
>
>
> Since we are in the beta stage now, the whole MapTesting space exists for testing for developers. It would not be there once we have a stable version ready for release.

Actually this raises a question, all the more as we also discussed the importance of having automated functional tests earlier today on #xwiki with Vincent. For automated testing, we will need sample data, and I'm wondering where we should store this sample data (and possible scripts or code for obtaining it). How do other projects deal with test data in such a context? Is the test data stored in the same repository or in a distinct one? I was looking for some Solr application test data but could not find it yet. Note that we may consider the testing area as a set of demos instead in some way, couldn't we? It would make sense to keep it (just like if it's real test data), and to provide a navigation across these pages as you suggest it, Caty.

Cheers

Stéphane


>     Lots of pages that are not hidden. All technical pages needs to be hidden.
>
>
> Again, these pages are not technical and exist only for testing purposes.
>
>     A bit confusing that there are 2 search boxes for the maps, see https://up1.xwikisas.com/#XngcOAZexsKE4DryH2i6zA 
>
>
>   Yes, thanks for pointing out. We need to move the search function directly inside the map. I will look into it once I am done with the facets.

>   For the search input, its in extremely beta stage. I am still trying to figure out the macros I have borrowed from SolrSearchMacros. So it will take time for me to make a more stable version of the facets. I hope I can do that in due time. :)
>
>     Regarding the facets, we need some more user friendly translations and customizations for this kind of UI, see https://up1.xwikisas.com/#TYj_9oLn84Mfp87VnwNgkw 
>
> As discussed earlier with Stephane, we are still having issues using the normal $facetDisplayer so we are using a workaround for testing purposes that's why it looks like this. More precisely, we are using the #displaySearchFacetValues($facetValues) macro for now.
>
> Best,
> Fawad


Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Stéphane Laurière-6
In reply to this post by Fawad Ali
Fawad,

> Hi all,
> Hope you all are doing wonderfully.
>
> We still have the issue for the facets rendering but its up on the forum, so I think we will have some fixes for it in due time (thanks to Stephane).
> In the meantime, I will try making the facets more specific to the map. For now I am considering the following options for facets:
> - Tags
> - Search field (searches inside map item pages)
> - Map item type (PointClass for now)
> - Map item icon (?)
> - Map item space
> I will need some suggestions on which other options should be included.

I created the museum-map-test branch in order to progress on this path, I hope this helps. In particular, imho, facets will be provided typically by objects from different classes than the geographical ones, such as a the Museum class (which very basic for now, it's just to expose the idea). My suggestion would be to progress toward an application that works well with a single sample facet + a full text input field + popups indeed + list + configurable icons and colors (typically one per facet value, I would say, so as for instance to represent all archeologia museums in a given color - even though we don't have that data yet, but we could use the countries all the same), and in parallel to gather more sample data with more facets and to progress toward having several configurable facets, what do you think?
 
> Next, I will
> -  Be working on making a dedicated space for popups inside or besides the map. Something similar to https://abc.gogocarto.fr <https://abc.gogocarto.fr/>
> - Make use of theme styles and colors for the map itself and the items inside
> - Be making the query search asynchronous
>
> Also, I will be working part time for the next week and I have exams the week after that. So I would like to let you know that I would not be available for the exams period for the most part. :(
> I will try to finish most of the work and get the application in a stable state by then. :)
> Ecaterina also mentioned that we release the application, I propose we do that next week as soon as the facets are in a stable state.

Great

Cheers

Stéphane

> Thanks,
> Fawad
 


Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Fawad Ali
In reply to this post by Stéphane Laurière-6
Hi,

Thanks for the PR Stephane! It really helps in knowing how I should do
things.

I will accept the PR now and make changes later if required. I tested it
and everything seems fine so far except for some country facets the map
goes blank (bad data handling on my part maybe).

Hoping such sample data can help developing further and think in greater
> details about real case scenarios.


It helps a lot, thanks.

When displaying such a map, that'd be great if the available facets would
> get displayed directly, without the need to enter a query in the input
> field (you probably have this in mind already as well).
>

This has been done and will be included in today's commit.

I was actually writing my custom solr configurations for the map. Now that
I know a use case, I can include better options.

Best,
Fawad


On Fri, Jun 7, 2019 at 9:59 PM Stéphane Laurière <[hidden email]>
wrote:

> Fawad, Caty, all,
>
> I have a short comment about the tests:
>
> > Hi Caty,
> > Thanks for the review.
> >
> >     Maps/MapTesting/Maps/TestMap - I find it strange that the Maps space
> is duplicated
> >
> >
> > This space exists only for testing. It won't be there for the real
> application. I named them so that its easier to know which type of object
> pages are located in them (for myself).
> >
> >     MapTesting, Maps, Points - spaces don't have homepages. The users
> will navigate to them, since they are present in breadcrumb. So what is the
> plan? Simpler paths? or create Homepages for these types of entry?
> >
> >
> > Since we are in the beta stage now, the whole MapTesting space exists
> for testing for developers. It would not be there once we have a stable
> version ready for release.
>
> Actually this raises a question, all the more as we also discussed the
> importance of having automated functional tests earlier today on #xwiki
> with Vincent. For automated testing, we will need sample data, and I'm
> wondering where we should store this sample data (and possible scripts or
> code for obtaining it). How do other projects deal with test data in such a
> context? Is the test data stored in the same repository or in a distinct
> one? I was looking for some Solr application test data but could not find
> it yet. Note that we may consider the testing area as a set of demos
> instead in some way, couldn't we? It would make sense to keep it (just like
> if it's real test data), and to provide a navigation across these pages as
> you suggest it, Caty.
>
> Cheers
>
> Stéphane
>
>
> >     Lots of pages that are not hidden. All technical pages needs to be
> hidden.
> >
> >
> > Again, these pages are not technical and exist only for testing purposes.
> >
> >     A bit confusing that there are 2 search boxes for the maps, see
> https://up1.xwikisas.com/#XngcOAZexsKE4DryH2i6zA
> >
> >
> >   Yes, thanks for pointing out. We need to move the search function
> directly inside the map. I will look into it once I am done with the facets.
>
> >   For the search input, its in extremely beta stage. I am still trying
> to figure out the macros I have borrowed from SolrSearchMacros. So it will
> take time for me to make a more stable version of the facets. I hope I can
> do that in due time. :)
> >
> >     Regarding the facets, we need some more user friendly translations
> and customizations for this kind of UI, see
> https://up1.xwikisas.com/#TYj_9oLn84Mfp87VnwNgkw
> >
> > As discussed earlier with Stephane, we are still having issues using the
> normal $facetDisplayer so we are using a workaround for testing purposes
> that's why it looks like this. More precisely, we are using
> the #displaySearchFacetValues($facetValues) macro for now.
> >
> > Best,
> > Fawad
>
>
>
Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

vmassol
Administrator
In reply to this post by Stéphane Laurière-6
Hi,

> On 7 Jun 2019, at 18:59, Stéphane Laurière <[hidden email]> wrote:
>
> Fawad, Caty, all,
>
> I have a short comment about the tests:
>
>> Hi Caty,
>> Thanks for the review.
>>    Maps/MapTesting/Maps/TestMap - I find it strange that the Maps space is duplicated
>> This space exists only for testing. It won't be there for the real application. I named them so that its easier to know which type of object pages are located in them (for myself).
>>    MapTesting, Maps, Points - spaces don't have homepages. The users will navigate to them, since they are present in breadcrumb. So what is the plan? Simpler paths? or create Homepages for these types of entry?
>> Since we are in the beta stage now, the whole MapTesting space exists for testing for developers. It would not be there once we have a stable version ready for release.
>
> Actually this raises a question, all the more as we also discussed the importance of having automated functional tests earlier today on #xwiki with Vincent. For automated testing, we will need sample data, and I'm wondering where we should store this sample data (and possible scripts or code for obtaining it). How do other projects deal with test data in such a context? Is the test data stored in the same repository or in a distinct one? I was looking for some Solr application test data but could not find it yet. Note that we may consider the testing area as a set of demos instead in some way, couldn't we? It would make sense to keep it (just like if it's real test data), and to provide a navigation across these pages as you suggest it, Caty.

For the Release Notes app, I also have some data for the tests. See the demo module in https://github.com/xwiki-contrib/application-releasenotes

Thanks
-Vincent

>
> Cheers
>
> Stéphane
>
>
>>    Lots of pages that are not hidden. All technical pages needs to be hidden.
>> Again, these pages are not technical and exist only for testing purposes.
>>    A bit confusing that there are 2 search boxes for the maps, see https://up1.xwikisas.com/#XngcOAZexsKE4DryH2i6zA   Yes, thanks for pointing out. We need to move the search function directly inside the map. I will look into it once I am done with the facets.
>
>>  For the search input, its in extremely beta stage. I am still trying to figure out the macros I have borrowed from SolrSearchMacros. So it will take time for me to make a more stable version of the facets. I hope I can do that in due time. :)
>>    Regarding the facets, we need some more user friendly translations and customizations for this kind of UI, see https://up1.xwikisas.com/#TYj_9oLn84Mfp87VnwNgkw As discussed earlier with Stephane, we are still having issues using the normal $facetDisplayer so we are using a workaround for testing purposes that's why it looks like this. More precisely, we are using the #displaySearchFacetValues($facetValues) macro for now.
>> Best,
>> Fawad
>
>

Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Fawad Ali
Hi everyone,
Hope all are well.

Ecaterina, Stephane,
For the search, do you think that we have to keep both the filter search
and the search inside the map? I feel like its an important use case for
the users to be able to search a location/place but that is not possible
with the query search. One approach is to have a single form with select
options if the the user wants to query the data or make a location search.
WDYT?
Here's a categorical single search form I made about a year ago:
https://jsfiddle.net/9inpachi/e428fLgr/

 I took a look at the application-releasenotes and what I understand is
that there are sample demo pages instead of functional tests. I personally
think our Interactive Maps Application aligns well with that approach and
we can have the same type of tests/demos. WDYT?

I would start preparing for a release for now and implement the tests once
we have coordinated on how we do it.

Best,
Fawad


On Fri, Jun 7, 2019 at 10:29 PM Vincent Massol <[hidden email]> wrote:

> Hi,
>
> > On 7 Jun 2019, at 18:59, Stéphane Laurière <[hidden email]> wrote:
> >
> > Fawad, Caty, all,
> >
> > I have a short comment about the tests:
> >
> >> Hi Caty,
> >> Thanks for the review.
> >>    Maps/MapTesting/Maps/TestMap - I find it strange that the Maps space
> is duplicated
> >> This space exists only for testing. It won't be there for the real
> application. I named them so that its easier to know which type of object
> pages are located in them (for myself).
> >>    MapTesting, Maps, Points - spaces don't have homepages. The users
> will navigate to them, since they are present in breadcrumb. So what is the
> plan? Simpler paths? or create Homepages for these types of entry?
> >> Since we are in the beta stage now, the whole MapTesting space exists
> for testing for developers. It would not be there once we have a stable
> version ready for release.
> >
> > Actually this raises a question, all the more as we also discussed the
> importance of having automated functional tests earlier today on #xwiki
> with Vincent. For automated testing, we will need sample data, and I'm
> wondering where we should store this sample data (and possible scripts or
> code for obtaining it). How do other projects deal with test data in such a
> context? Is the test data stored in the same repository or in a distinct
> one? I was looking for some Solr application test data but could not find
> it yet. Note that we may consider the testing area as a set of demos
> instead in some way, couldn't we? It would make sense to keep it (just like
> if it's real test data), and to provide a navigation across these pages as
> you suggest it, Caty.
>
> For the Release Notes app, I also have some data for the tests. See the
> demo module in https://github.com/xwiki-contrib/application-releasenotes
>
> Thanks
> -Vincent
>
> >
> > Cheers
> >
> > Stéphane
> >
> >
> >>    Lots of pages that are not hidden. All technical pages needs to be
> hidden.
> >> Again, these pages are not technical and exist only for testing
> purposes.
> >>    A bit confusing that there are 2 search boxes for the maps, see
> https://up1.xwikisas.com/#XngcOAZexsKE4DryH2i6zA   Yes, thanks for
> pointing out. We need to move the search function directly inside the map.
> I will look into it once I am done with the facets.
> >
> >>  For the search input, its in extremely beta stage. I am still trying
> to figure out the macros I have borrowed from SolrSearchMacros. So it will
> take time for me to make a more stable version of the facets. I hope I can
> do that in due time. :)
> >>    Regarding the facets, we need some more user friendly translations
> and customizations for this kind of UI, see
> https://up1.xwikisas.com/#TYj_9oLn84Mfp87VnwNgkw As discussed earlier
> with Stephane, we are still having issues using the normal $facetDisplayer
> so we are using a workaround for testing purposes that's why it looks like
> this. More precisely, we are using the
> #displaySearchFacetValues($facetValues) macro for now.
> >> Best,
> >> Fawad
> >
> >
>
>
Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

vmassol
Administrator
Hi Fawad,

> On 12 Jun 2019, at 18:37, Fawad Ali <[hidden email]> wrote:

[snip]

> I took a look at the application-releasenotes and what I understand is
> that there are sample demo pages instead of functional tests.

This is not 100% correct. There are both, and the demo pages are used in the functional tests.

* Demo pages: https://github.com/xwiki-contrib/application-releasenotes/tree/master/application-releasenotes-demo
* Func tests: https://github.com/xwiki-contrib/application-releasenotes/tree/master/application-releasenotes-test
* Demo dep in the func test module: https://github.com/xwiki-contrib/application-releasenotes/blob/master/application-releasenotes-test/application-releasenotes-test-tests/pom.xml#L43

Thanks
-Vincent

> I personally
> think our Interactive Maps Application aligns well with that approach and
> we can have the same type of tests/demos. WDYT?
>
> I would start preparing for a release for now and implement the tests once
> we have coordinated on how we do it.
>
> Best,
> Fawad
>
>
> On Fri, Jun 7, 2019 at 10:29 PM Vincent Massol <[hidden email]> wrote:
>
>> Hi,
>>
>>> On 7 Jun 2019, at 18:59, Stéphane Laurière <[hidden email]> wrote:
>>>
>>> Fawad, Caty, all,
>>>
>>> I have a short comment about the tests:
>>>
>>>> Hi Caty,
>>>> Thanks for the review.
>>>>   Maps/MapTesting/Maps/TestMap - I find it strange that the Maps space
>> is duplicated
>>>> This space exists only for testing. It won't be there for the real
>> application. I named them so that its easier to know which type of object
>> pages are located in them (for myself).
>>>>   MapTesting, Maps, Points - spaces don't have homepages. The users
>> will navigate to them, since they are present in breadcrumb. So what is the
>> plan? Simpler paths? or create Homepages for these types of entry?
>>>> Since we are in the beta stage now, the whole MapTesting space exists
>> for testing for developers. It would not be there once we have a stable
>> version ready for release.
>>>
>>> Actually this raises a question, all the more as we also discussed the
>> importance of having automated functional tests earlier today on #xwiki
>> with Vincent. For automated testing, we will need sample data, and I'm
>> wondering where we should store this sample data (and possible scripts or
>> code for obtaining it). How do other projects deal with test data in such a
>> context? Is the test data stored in the same repository or in a distinct
>> one? I was looking for some Solr application test data but could not find
>> it yet. Note that we may consider the testing area as a set of demos
>> instead in some way, couldn't we? It would make sense to keep it (just like
>> if it's real test data), and to provide a navigation across these pages as
>> you suggest it, Caty.
>>
>> For the Release Notes app, I also have some data for the tests. See the
>> demo module in https://github.com/xwiki-contrib/application-releasenotes
>>
>> Thanks
>> -Vincent
>>
>>>
>>> Cheers
>>>
>>> Stéphane
>>>
>>>
>>>>   Lots of pages that are not hidden. All technical pages needs to be
>> hidden.
>>>> Again, these pages are not technical and exist only for testing
>> purposes.
>>>>   A bit confusing that there are 2 search boxes for the maps, see
>> https://up1.xwikisas.com/#XngcOAZexsKE4DryH2i6zA   Yes, thanks for
>> pointing out. We need to move the search function directly inside the map.
>> I will look into it once I am done with the facets.
>>>
>>>> For the search input, its in extremely beta stage. I am still trying
>> to figure out the macros I have borrowed from SolrSearchMacros. So it will
>> take time for me to make a more stable version of the facets. I hope I can
>> do that in due time. :)
>>>>   Regarding the facets, we need some more user friendly translations
>> and customizations for this kind of UI, see
>> https://up1.xwikisas.com/#TYj_9oLn84Mfp87VnwNgkw As discussed earlier
>> with Stephane, we are still having issues using the normal $facetDisplayer
>> so we are using a workaround for testing purposes that's why it looks like
>> this. More precisely, we are using the
>> #displaySearchFacetValues($facetValues) macro for now.
>>>> Best,
>>>> Fawad
>>>
>>>
>>
>>

Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Stéphane Laurière-6
In reply to this post by Fawad Ali
Hi Fawad, Hi all,

> Hi everyone,
> Hope all are well.
>
> Ecaterina, Stephane,
> For the search, do you think that we have to keep both the filter search
> and the search inside the map? I feel like its an important use case for
> the users to be able to search a location/place but that is not possible
> with the query search. One approach is to have a single form with select
> options if the the user wants to query the data or make a location search.
> WDYT?
> Here's a categorical single search form I made about a year ago:
> https://jsfiddle.net/9inpachi/e428fLgr/

Imho the ability to search for a location should be an option, because not all users will need it, I think. In case it is activated, I would suggest a user experience inspired by the one exposed by GoGoCarto, the one of Google Maps, the CCI Map and the widget you mention indeed, that is made of the following elements:

1) a search input

2) only if the location search option is activated: two radio buttons for defining the scope: either the map data or a location

3) a panel that shows up upon explicit activation by the user for displaying and applying facets (on this aspect, it differs from GoGoCarto where the category filter is always present when the search widget is visible)

4) a list of search results that show up under the search widget and that can be hidden

Examples:

- https://abc.gogocarto.fr/annuaire#/carte/recherche/abc?cat=all
- http://carte.preference-commerce.fr/cci-fr
- https://www.google.com/maps/search/museums/@48.8761955,2.3004558,13z/data=!3m1!4b1

What do you think about this approach in terms of UX?
 
>   I took a look at the application-releasenotes and what I understand is
> that there are sample demo pages instead of functional tests. I personally
> think our Interactive Maps Application aligns well with that approach and
> we can have the same type of tests/demos. WDYT?
>
> I would start preparing for a release for now and implement the tests once
> we have coordinated on how we do it.

That looks like a great plan.

Cheers

Stéphane

 
> Best,
> Fawad


--
Stéphane Laurière
XWiki – https://xwiki.com

Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Fawad Ali
Hi all,
Hoping that everything is going well.

Stephane, I was able to do implement most of your suggestions except the
search results list. I am not too sure where I should place it. As per the
latest build, the filter widget appears to the left. Do you think it is
practical that we replace this widget with the search results when the user
wants to see the search results and show the widget back again when the
user clicks on the widget control? I am not too sure what approach I should
choose here in terms of UX.
Ecaterina, your views on this would help a lot. Thanks. :)

Also, do you foresee the map to take full browser height and width as is
seen in all the examples you gave me?

Regarding the release, I will try preparing everything by tonight so that
it is available for your review, including the demo tests that Vincent
suggested.

Best,
Fawad


On Thu, Jun 13, 2019 at 12:59 PM Stéphane Laurière <[hidden email]>
wrote:

> Hi Fawad, Hi all,
>
> > Hi everyone,
> > Hope all are well.
> >
> > Ecaterina, Stephane,
> > For the search, do you think that we have to keep both the filter search
> > and the search inside the map? I feel like its an important use case for
> > the users to be able to search a location/place but that is not possible
> > with the query search. One approach is to have a single form with select
> > options if the the user wants to query the data or make a location
> search.
> > WDYT?
> > Here's a categorical single search form I made about a year ago:
> > https://jsfiddle.net/9inpachi/e428fLgr/
>
> Imho the ability to search for a location should be an option, because not
> all users will need it, I think. In case it is activated, I would suggest a
> user experience inspired by the one exposed by GoGoCarto, the one of Google
> Maps, the CCI Map and the widget you mention indeed, that is made of the
> following elements:
>
> 1) a search input
>
> 2) only if the location search option is activated: two radio buttons for
> defining the scope: either the map data or a location
>
> 3) a panel that shows up upon explicit activation by the user for
> displaying and applying facets (on this aspect, it differs from GoGoCarto
> where the category filter is always present when the search widget is
> visible)
>
> 4) a list of search results that show up under the search widget and that
> can be hidden
>
> Examples:
>
> - https://abc.gogocarto.fr/annuaire#/carte/recherche/abc?cat=all
> - http://carte.preference-commerce.fr/cci-fr
> -
> https://www.google.com/maps/search/museums/@48.8761955,2.3004558,13z/data=!3m1!4b1
>
> What do you think about this approach in terms of UX?
>
> >   I took a look at the application-releasenotes and what I understand is
> > that there are sample demo pages instead of functional tests. I
> personally
> > think our Interactive Maps Application aligns well with that approach and
> > we can have the same type of tests/demos. WDYT?
> >
> > I would start preparing for a release for now and implement the tests
> once
> > we have coordinated on how we do it.
>
> That looks like a great plan.
>
> Cheers
>
> Stéphane
>
>
> > Best,
> > Fawad
>
>
> --
> Stéphane Laurière
> XWiki – https://xwiki.com
>
>
Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Fawad Ali
Hi all,

I am still a little confused regarding tests. What I get from analyzing the
application-release-notes is that I have to make demo pages containing maps
first, then just include test (containing demo pages) and ui folders and
change the pom files to include the tests, is that correct?

Thanks,
Fawad


On Sat, Jun 15, 2019 at 5:42 PM Fawad Ali <[hidden email]> wrote:

> Hi all,
> Hoping that everything is going well.
>
> Stephane, I was able to do implement most of your suggestions except the
> search results list. I am not too sure where I should place it. As per the
> latest build, the filter widget appears to the left. Do you think it is
> practical that we replace this widget with the search results when the user
> wants to see the search results and show the widget back again when the
> user clicks on the widget control? I am not too sure what approach I should
> choose here in terms of UX.
> Ecaterina, your views on this would help a lot. Thanks. :)
>
> Also, do you foresee the map to take full browser height and width as is
> seen in all the examples you gave me?
>
> Regarding the release, I will try preparing everything by tonight so that
> it is available for your review, including the demo tests that Vincent
> suggested.
>
> Best,
> Fawad
>
>
> On Thu, Jun 13, 2019 at 12:59 PM Stéphane Laurière <[hidden email]>
> wrote:
>
>> Hi Fawad, Hi all,
>>
>> > Hi everyone,
>> > Hope all are well.
>> >
>> > Ecaterina, Stephane,
>> > For the search, do you think that we have to keep both the filter search
>> > and the search inside the map? I feel like its an important use case for
>> > the users to be able to search a location/place but that is not possible
>> > with the query search. One approach is to have a single form with select
>> > options if the the user wants to query the data or make a location
>> search.
>> > WDYT?
>> > Here's a categorical single search form I made about a year ago:
>> > https://jsfiddle.net/9inpachi/e428fLgr/
>>
>> Imho the ability to search for a location should be an option, because
>> not all users will need it, I think. In case it is activated, I would
>> suggest a user experience inspired by the one exposed by GoGoCarto, the one
>> of Google Maps, the CCI Map and the widget you mention indeed, that is made
>> of the following elements:
>>
>> 1) a search input
>>
>> 2) only if the location search option is activated: two radio buttons for
>> defining the scope: either the map data or a location
>>
>> 3) a panel that shows up upon explicit activation by the user for
>> displaying and applying facets (on this aspect, it differs from GoGoCarto
>> where the category filter is always present when the search widget is
>> visible)
>>
>> 4) a list of search results that show up under the search widget and that
>> can be hidden
>>
>> Examples:
>>
>> - https://abc.gogocarto.fr/annuaire#/carte/recherche/abc?cat=all
>> - http://carte.preference-commerce.fr/cci-fr
>> -
>> https://www.google.com/maps/search/museums/@48.8761955,2.3004558,13z/data=!3m1!4b1
>>
>> What do you think about this approach in terms of UX?
>>
>> >   I took a look at the application-releasenotes and what I understand is
>> > that there are sample demo pages instead of functional tests. I
>> personally
>> > think our Interactive Maps Application aligns well with that approach
>> and
>> > we can have the same type of tests/demos. WDYT?
>> >
>> > I would start preparing for a release for now and implement the tests
>> once
>> > we have coordinated on how we do it.
>>
>> That looks like a great plan.
>>
>> Cheers
>>
>> Stéphane
>>
>>
>> > Best,
>> > Fawad
>>
>>
>> --
>> Stéphane Laurière
>> XWiki – https://xwiki.com
>>
>>
Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Stéphane Laurière-6
In reply to this post by Fawad Ali
Hi Fawad, Caty, all,

> Hi all,
> Hoping that everything is going well.
>
> Stephane, I was able to do implement most of your suggestions except the search results list.

I just gave a try to the latest code, well done with the Solr queries and the search integration in the user interface, that's cool!

> I am not too sure where I should place it. As per the latest build, the filter widget appears to the left. Do you think it is practical that we replace this widget with the search results when the user wants to see the search results and show the widget back again when the user clicks on the widget control? I am not too sure what approach I should choose here in terms of UX.
> Ecaterina, your views on this would help a lot. Thanks. :)

Imho the most convenient UX is the following for these widgets, something similar to this map:

   http://carte.preference-commerce.fr/cci-fr/

That is:

- The facets can get activated from a button. When they get activated, they show up in an overlay panel on top of the map, without hiding the list.
- The list gets displayed under the search input, in an overlay as well, and can be completely hidden on request

It's rather close to what Google Maps proposes as well, except that the facets replace the list in that case (when hitting "Autres filtres"), which is a bit less convenient in my opinion, but not a big deal:

   https://www.google.com/maps/search/Restaurants/@48.865957,2.352974,16z

What do you think?

> Also, do you foresee the map to take full browser height and width as is seen in all the examples you gave me?

That would be a nice feature to have a button on the map to turn it full-screen indeed, similarly to what happens with the XWiki text editor.

Cheers

Stéphane
 
> Regarding the release, I will try preparing everything by tonight so that it is available for your review, including the demo tests that Vincent suggested.
>
> Best,
> Fawad
>
>
> On Thu, Jun 13, 2019 at 12:59 PM Stéphane Laurière <[hidden email] <mailto:[hidden email]>> wrote:

Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Fawad Ali
Hi all,

I just gave a try to the latest code, well done with the Solr queries and
> the search integration in the user interface, that's cool!


Thanks Stephane. :)

Imho the most convenient UX is the following for these widgets, something

> similar to this map:
>
>    http://carte.preference-commerce.fr/cci-fr/
>
> That is:
>
> - The facets can get activated from a button. When they get activated,
> they show up in an overlay panel on top of the map, without hiding the list.
> - The list gets displayed under the search input, in an overlay as well,
> and can be completely hidden on request


Just to clarify things, the facets here refer to the "Refine your search"
area, the search input refers to the "Search in map" text input and the
list refers to the map item search results. Is that right?
If so, I believe, given the nature of XWiki's design with widgets to both
the right and left (for the default flavor), the space is a little cramped
for overlaying anything on the map. For the implementation of full screen
maps, we can overlay search and facets but for the normal view, the map
will become very small.

Here is a mockup I prepared based on my understanding of your suggestions:
https://up1.xwikisas.com/#SB7B5mLNnfnUVAogWTLarw
Let me know what you think?

I have also come up with an idea to generalize the museum maps import and
export you created, Stephane. We could have a standard form of wikidata
query with some extra custom parameters for each location like the
MuseumClass in the Museums' case. These extra parameter will be checked in
JSON and classes will be created if required so that objects can be
associated to each map item and then later facets can be used based upon
these newly created classes.
WDYT?

Also, due to unprecedented circumstances within college, my exams have been
delayed for this week and will start from next week. So I will be working
this whole week on the project as opposed to what I told earlier.

Best,
Fawad


On Mon, Jun 17, 2019 at 11:04 PM Stéphane Laurière <[hidden email]>
wrote:

> Hi Fawad, Caty, all,
>
> > Hi all,
> > Hoping that everything is going well.
> >
> > Stephane, I was able to do implement most of your suggestions except the
> search results list.
>
> I just gave a try to the latest code, well done with the Solr queries and
> the search integration in the user interface, that's cool!
>
> > I am not too sure where I should place it. As per the latest build, the
> filter widget appears to the left. Do you think it is practical that we
> replace this widget with the search results when the user wants to see the
> search results and show the widget back again when the user clicks on the
> widget control? I am not too sure what approach I should choose here in
> terms of UX.
> > Ecaterina, your views on this would help a lot. Thanks. :)
>
> Imho the most convenient UX is the following for these widgets, something
> similar to this map:
>
>    http://carte.preference-commerce.fr/cci-fr/
>
> That is:
>
> - The facets can get activated from a button. When they get activated,
> they show up in an overlay panel on top of the map, without hiding the list.
> - The list gets displayed under the search input, in an overlay as well,
> and can be completely hidden on request
>
> It's rather close to what Google Maps proposes as well, except that the
> facets replace the list in that case (when hitting "Autres filtres"), which
> is a bit less convenient in my opinion, but not a big deal:
>
>    https://www.google.com/maps/search/Restaurants/@48.865957,2.352974,16z
>
> What do you think?
>
> > Also, do you foresee the map to take full browser height and width as is
> seen in all the examples you gave me?
>
> That would be a nice feature to have a button on the map to turn it
> full-screen indeed, similarly to what happens with the XWiki text editor.
>
> Cheers
>
> Stéphane
>
> > Regarding the release, I will try preparing everything by tonight so
> that it is available for your review, including the demo tests that Vincent
> suggested.
> >
> > Best,
> > Fawad
> >
> >
> > On Thu, Jun 13, 2019 at 12:59 PM Stéphane Laurière <[hidden email]
> <mailto:[hidden email]>> wrote:
>
>
Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Fawad Ali
Hi all,

I have prepared demos and have set up the repository after analyzing
application-releasenotes. Please check to see if the tests and demos are
working fine so we can move on to releasing the application.
https://github.com/xwiki-contrib/application-interactive-maps

Best,
Fawad


On Tue, Jun 18, 2019 at 8:25 PM Fawad Ali <[hidden email]> wrote:

> Hi all,
>
> I just gave a try to the latest code, well done with the Solr queries and
>> the search integration in the user interface, that's cool!
>
>
> Thanks Stephane. :)
>
> Imho the most convenient UX is the following for these widgets, something
>> similar to this map:
>>
>>    http://carte.preference-commerce.fr/cci-fr/
>>
>> That is:
>>
>> - The facets can get activated from a button. When they get activated,
>> they show up in an overlay panel on top of the map, without hiding the list.
>> - The list gets displayed under the search input, in an overlay as well,
>> and can be completely hidden on request
>
>
> Just to clarify things, the facets here refer to the "Refine your search"
> area, the search input refers to the "Search in map" text input and the
> list refers to the map item search results. Is that right?
> If so, I believe, given the nature of XWiki's design with widgets to both
> the right and left (for the default flavor), the space is a little cramped
> for overlaying anything on the map. For the implementation of full screen
> maps, we can overlay search and facets but for the normal view, the map
> will become very small.
>
> Here is a mockup I prepared based on my understanding of your suggestions:
> https://up1.xwikisas.com/#SB7B5mLNnfnUVAogWTLarw
> Let me know what you think?
>
> I have also come up with an idea to generalize the museum maps import and
> export you created, Stephane. We could have a standard form of wikidata
> query with some extra custom parameters for each location like the
> MuseumClass in the Museums' case. These extra parameter will be checked in
> JSON and classes will be created if required so that objects can be
> associated to each map item and then later facets can be used based upon
> these newly created classes.
> WDYT?
>
> Also, due to unprecedented circumstances within college, my exams have
> been delayed for this week and will start from next week. So I will be
> working this whole week on the project as opposed to what I told earlier.
>
> Best,
> Fawad
>
>
> On Mon, Jun 17, 2019 at 11:04 PM Stéphane Laurière <[hidden email]>
> wrote:
>
>> Hi Fawad, Caty, all,
>>
>> > Hi all,
>> > Hoping that everything is going well.
>> >
>> > Stephane, I was able to do implement most of your suggestions except
>> the search results list.
>>
>> I just gave a try to the latest code, well done with the Solr queries and
>> the search integration in the user interface, that's cool!
>>
>> > I am not too sure where I should place it. As per the latest build, the
>> filter widget appears to the left. Do you think it is practical that we
>> replace this widget with the search results when the user wants to see the
>> search results and show the widget back again when the user clicks on the
>> widget control? I am not too sure what approach I should choose here in
>> terms of UX.
>> > Ecaterina, your views on this would help a lot. Thanks. :)
>>
>> Imho the most convenient UX is the following for these widgets, something
>> similar to this map:
>>
>>    http://carte.preference-commerce.fr/cci-fr/
>>
>> That is:
>>
>> - The facets can get activated from a button. When they get activated,
>> they show up in an overlay panel on top of the map, without hiding the list.
>> - The list gets displayed under the search input, in an overlay as well,
>> and can be completely hidden on request
>>
>> It's rather close to what Google Maps proposes as well, except that the
>> facets replace the list in that case (when hitting "Autres filtres"), which
>> is a bit less convenient in my opinion, but not a big deal:
>>
>>    https://www.google.com/maps/search/Restaurants/@48.865957,2.352974,16z
>>
>> What do you think?
>>
>> > Also, do you foresee the map to take full browser height and width as
>> is seen in all the examples you gave me?
>>
>> That would be a nice feature to have a button on the map to turn it
>> full-screen indeed, similarly to what happens with the XWiki text editor.
>>
>> Cheers
>>
>> Stéphane
>>
>> > Regarding the release, I will try preparing everything by tonight so
>> that it is available for your review, including the demo tests that Vincent
>> suggested.
>> >
>> > Best,
>> > Fawad
>> >
>> >
>> > On Thu, Jun 13, 2019 at 12:59 PM Stéphane Laurière <[hidden email]
>> <mailto:[hidden email]>> wrote:
>>
>>
Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

vmassol
Administrator
In reply to this post by vmassol
Hi,

> On 12 Jun 2019, at 20:46, Vincent Massol <[hidden email]> wrote:
>
> Hi Fawad,
>
>> On 12 Jun 2019, at 18:37, Fawad Ali <[hidden email]> wrote:
>
> [snip]
>
>> I took a look at the application-releasenotes and what I understand is
>> that there are sample demo pages instead of functional tests.
>
> This is not 100% correct. There are both, and the demo pages are used in the functional tests.
>
> * Demo pages: https://github.com/xwiki-contrib/application-releasenotes/tree/master/application-releasenotes-demo
> * Func tests: https://github.com/xwiki-contrib/application-releasenotes/tree/master/application-releasenotes-test
> * Demo dep in the func test module: https://github.com/xwiki-contrib/application-releasenotes/blob/master/application-releasenotes-test/application-releasenotes-test-tests/pom.xml#L43

Just realized that the release notes app is a bad example since it currently doesn’t contain any func test! They’re missing and a TODO.

You could check other app like the FAQ app for an example of a func test.

Sorry about that.
Thanks
-Vincent

>
> Thanks
> -Vincent
>
>> I personally
>> think our Interactive Maps Application aligns well with that approach and
>> we can have the same type of tests/demos. WDYT?
>>
>> I would start preparing for a release for now and implement the tests once
>> we have coordinated on how we do it.
>>
>> Best,
>> Fawad
>>
>>
>> On Fri, Jun 7, 2019 at 10:29 PM Vincent Massol <[hidden email]> wrote:
>>
>>> Hi,
>>>
>>>> On 7 Jun 2019, at 18:59, Stéphane Laurière <[hidden email]> wrote:
>>>>
>>>> Fawad, Caty, all,
>>>>
>>>> I have a short comment about the tests:
>>>>
>>>>> Hi Caty,
>>>>> Thanks for the review.
>>>>>  Maps/MapTesting/Maps/TestMap - I find it strange that the Maps space
>>> is duplicated
>>>>> This space exists only for testing. It won't be there for the real
>>> application. I named them so that its easier to know which type of object
>>> pages are located in them (for myself).
>>>>>  MapTesting, Maps, Points - spaces don't have homepages. The users
>>> will navigate to them, since they are present in breadcrumb. So what is the
>>> plan? Simpler paths? or create Homepages for these types of entry?
>>>>> Since we are in the beta stage now, the whole MapTesting space exists
>>> for testing for developers. It would not be there once we have a stable
>>> version ready for release.
>>>>
>>>> Actually this raises a question, all the more as we also discussed the
>>> importance of having automated functional tests earlier today on #xwiki
>>> with Vincent. For automated testing, we will need sample data, and I'm
>>> wondering where we should store this sample data (and possible scripts or
>>> code for obtaining it). How do other projects deal with test data in such a
>>> context? Is the test data stored in the same repository or in a distinct
>>> one? I was looking for some Solr application test data but could not find
>>> it yet. Note that we may consider the testing area as a set of demos
>>> instead in some way, couldn't we? It would make sense to keep it (just like
>>> if it's real test data), and to provide a navigation across these pages as
>>> you suggest it, Caty.
>>>
>>> For the Release Notes app, I also have some data for the tests. See the
>>> demo module in https://github.com/xwiki-contrib/application-releasenotes
>>>
>>> Thanks
>>> -Vincent
>>>
>>>>
>>>> Cheers
>>>>
>>>> Stéphane
>>>>
>>>>
>>>>>  Lots of pages that are not hidden. All technical pages needs to be
>>> hidden.
>>>>> Again, these pages are not technical and exist only for testing
>>> purposes.
>>>>>  A bit confusing that there are 2 search boxes for the maps, see
>>> https://up1.xwikisas.com/#XngcOAZexsKE4DryH2i6zA   Yes, thanks for
>>> pointing out. We need to move the search function directly inside the map.
>>> I will look into it once I am done with the facets.
>>>>
>>>>> For the search input, its in extremely beta stage. I am still trying
>>> to figure out the macros I have borrowed from SolrSearchMacros. So it will
>>> take time for me to make a more stable version of the facets. I hope I can
>>> do that in due time. :)
>>>>>  Regarding the facets, we need some more user friendly translations
>>> and customizations for this kind of UI, see
>>> https://up1.xwikisas.com/#TYj_9oLn84Mfp87VnwNgkw As discussed earlier
>>> with Stephane, we are still having issues using the normal $facetDisplayer
>>> so we are using a workaround for testing purposes that's why it looks like
>>> this. More precisely, we are using the
>>> #displaySearchFacetValues($facetValues) macro for now.
>>>>> Best,
>>>>> Fawad
>>>>
>>>>
>>>
>>>
>

Reply | Threaded
Open this post in threaded view
|

Re: Map Application - GSoC 19

Stéphane Laurière-6
In reply to this post by Fawad Ali
Hi Fawad, hi all,

> Hi all,
>
>     I just gave a try to the latest code, well done with the Solr queries and the search integration in the user interface, that's cool!
>
> Thanks Stephane. :)
>
>     Imho the most convenient UX is the following for these widgets, something similar to this map:
>
>     http://carte.preference-commerce.fr/cci-fr/
>
>     That is:
>
>     - The facets can get activated from a button. When they get activated, they show up in an overlay panel on top of the map, without hiding the list.
>     - The list gets displayed under the search input, in an overlay as well, and can be completely hidden on request
>
>
> Just to clarify things, the facets here refer to the "Refine your search" area, the search input refers to the "Search in map" text input and the list refers to the map item search results. Is that right?

Yes indeed,

> If so, I believe, given the nature of XWiki's design with widgets to both the right and left (for the default flavor), the space is a little cramped for overlaying anything on the map. For the implementation of full screen maps, we can overlay search and facets but for the normal view, the map will become very small.
>
> Here is a mockup I prepared based on my understanding of your suggestions: https://up1.xwikisas.com/#SB7B5mLNnfnUVAogWTLarw
> Let me know what you think?

I think it's an efficient way to layout the widgets indeed. I agree the map may look cramped in case there are panels on the left and on the right, but the user typically would have an option to hide the results, ideally. There's one change I would suggest, that would be to layout the filters either as an overlay or as a replacement of the list, and to move the filter button closer to the search input. Regarding the filter position: the Airbnb search illustrated below behaves quite nicely imho (when you hit "More filters", you get a new panel on top with all the filters), what do you think?

   https://up1.xwikisas.com/#zNABtIwH-Z-hSSUgFx-y4Q

As for the information associated with each point or area, we have several options: either place it in popups on top of the map like Airbnb, or have it in the search result area, like what Google Maps proposes, or display it in a lateral panel like GoGoCarto. I would opt for popups by default, and if possible later on, leave the option to display it in the search area in case of large content, what do you think? Ideally, the panel for displaying this information would be a template that could be easily styled and customized for each map?

> I have also come up with an idea to generalize the museum maps import and export you created, Stephane. We could have a standard form of wikidata query with some extra custom parameters for each location like the MuseumClass in the Museums' case. These extra parameter will be checked in JSON and classes will be created if required so that objects can be associated to each map item and then later facets can be used based upon these newly created classes.
> WDYT?

I think that'd be a nice feature generally speaking to ease the import of Wikidata into XWiki indeed, I have a side project on which I'm considering such developments as well, not sure yet about the outcome. If you feel like it will be useful for generating more demo maps, I'd say that'd be good indeed, but with the caveat of not digging too much in that direction for not hampering the map development of course.

> Also, due to unprecedented circumstances within college, my exams have been delayed for this week and will start from next week. So I will be working this whole week on the project as opposed to what I told earlier.

Ok, thank you for letting us know.

I'd say at this stage that'd be great to keep progressing on functional testing automation and to finalize and release 1.0, I guess we're in tune but, as you did since the beginning, don't hesitate to raise questions about the priorities and the next steps of course.

Wishing you good work days ahead,

Cheers

Stéphane


> Best,
> Fawad
>
>
> On Mon, Jun 17, 2019 at 11:04 PM Stéphane Laurière <[hidden email] <mailto:[hidden email]>> wrote:
>
>     Hi Fawad, Caty, all,
>
>      > Hi all,
>      > Hoping that everything is going well.
>      >
>      > Stephane, I was able to do implement most of your suggestions except the search results list.
>
>     I just gave a try to the latest code, well done with the Solr queries and the search integration in the user interface, that's cool!
>
>      > I am not too sure where I should place it. As per the latest build, the filter widget appears to the left. Do you think it is practical that we replace this widget with the search results when the user wants to see the search results and show the widget back again when the user clicks on the widget control? I am not too sure what approach I should choose here in terms of UX.
>      > Ecaterina, your views on this would help a lot. Thanks. :)
>
>     Imho the most convenient UX is the following for these widgets, something similar to this map:
>
>     http://carte.preference-commerce.fr/cci-fr/
>
>     That is:
>
>     - The facets can get activated from a button. When they get activated, they show up in an overlay panel on top of the map, without hiding the list.
>     - The list gets displayed under the search input, in an overlay as well, and can be completely hidden on request
>
>     It's rather close to what Google Maps proposes as well, except that the facets replace the list in that case (when hitting "Autres filtres"), which is a bit less convenient in my opinion, but not a big deal:
>
>     https://www.google.com/maps/search/Restaurants/@48.865957,2.352974,16z
>
>     What do you think?
>
>      > Also, do you foresee the map to take full browser height and width as is seen in all the examples you gave me?
>
>     That would be a nice feature to have a button on the map to turn it full-screen indeed, similarly to what happens with the XWiki text editor.
>
>     Cheers
>
>     Stéphane
>
>      > Regarding the release, I will try preparing everything by tonight so that it is available for your review, including the demo tests that Vincent suggested.
>      >
>      > Best,
>      > Fawad
>      >
>      >
>      > On Thu, Jun 13, 2019 at 12:59 PM Stéphane Laurière <[hidden email] <mailto:[hidden email]> <mailto:[hidden email] <mailto:[hidden email]>>> wrote:
>


--
Stéphane Laurière
XWiki – https://xwiki.com

1234567