Hi Everyone,
I would like to invite your feedback on [1] prototype of Project-'UploadWizard:OSM embedding' [ user/password is mediawiki/mw ].
This project is about enhancing the image upload process for Wikimedia Commons.The main goal of the OSM embedding project is to provide map interface to the UploadWizard. We can not expect each and every uploader to know exact location where image being uploaded was taken, or to use location enabled camera/phone to take the image. With introduction of map interface, uploader can easily choose location coordinates for uploaded images.
Detailed project description can be found at [2].
Functioning of embedded map :
- On valid coordinates/location inputs widget first loads a static map which on 'click' will initialize leaflet powered dynamic map - Address search - ' type' or 'importance' specific zoom for various kinds of locations on address search - .Dynamic updation of coordinate fields and location field on leaflet map interaction and vice versa
Kindly give your suggestions for map widget- design as it needs to be improved, also propose features and/or improvements in current features. Please share your insights at [3].
[1]http://uploadwizard-osm.bitnamiapp.com/mediawiki/Special:UploadWizard [2] https://www.mediawiki.org/wiki/UploadWizard_OSM_map_embedding [3] https://www.mediawiki.org/wiki/Talk:UploadWizard_OSM_map_embedding
Best regards Anu G Enchackal
Ack, I wish I had known about this earlier. Most of these features had already been implemented for UploadWizard, but the patches were never polished enough to be merged: https://gerrit.wikimedia.org/r/#/c/12352/ https://gerrit.wikimedia.org/r/#/c/12353/
I would be good to at least compare against your implementation and see if there are any ideas that would be useful to incorporate.
Ryan Kaldari
On Wed, Mar 12, 2014 at 2:11 PM, Anu George Enchackal < inchikutty13@gmail.com> wrote:
Hi Everyone,
I would like to invite your feedback on [1] prototype of Project-'UploadWizard:OSM embedding' [ user/password is mediawiki/mw ].
This project is about enhancing the image upload process for Wikimedia Commons.The main goal of the OSM embedding project is to provide map interface to the UploadWizard. We can not expect each and every uploader to know exact location where image being uploaded was taken, or to use location enabled camera/phone to take the image. With introduction of map interface, uploader can easily choose location coordinates for uploaded images.
Detailed project description can be found at [2].
Functioning of embedded map :
- On valid coordinates/location inputs widget first loads a static map
which on 'click' will initialize leaflet powered dynamic map
- Address search
- ' type' or 'importance' specific zoom for various kinds of locations
on address search
- .Dynamic updation of coordinate fields and location field on leaflet
map interaction and vice versa
Kindly give your suggestions for map widget- design as it needs to be improved, also propose features and/or improvements in current features. Please share your insights at [3].
[1]http://uploadwizard-osm.bitnamiapp.com/mediawiki/Special:UploadWizard [2] https://www.mediawiki.org/wiki/UploadWizard_OSM_map_embedding [3] https://www.mediawiki.org/wiki/Talk:UploadWizard_OSM_map_embedding
Best regards Anu G Enchackal
Design mailing list Design@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/design
Anu,
Also much of the future work for Image upload is on the roadmap for the Multimedia group, people sync up with fabrice (PM) and Pau (design) both cc'd. I want to make sure any contributions you'd like to help out with don't run counter to anything being planned!
*Jared Zimmerman * \ Director of User Experience \ Wikimedia Foundation
M : +1 415 609 4043 | : @JaredZimmermanhttps://twitter.com/JaredZimmerman
On Wed, Mar 12, 2014 at 2:42 PM, Ryan Kaldari rkaldari@wikimedia.orgwrote:
Ack, I wish I had known about this earlier. Most of these features had already been implemented for UploadWizard, but the patches were never polished enough to be merged: https://gerrit.wikimedia.org/r/#/c/12352/ https://gerrit.wikimedia.org/r/#/c/12353/
I would be good to at least compare against your implementation and see if there are any ideas that would be useful to incorporate.
Ryan Kaldari
On Wed, Mar 12, 2014 at 2:11 PM, Anu George Enchackal < inchikutty13@gmail.com> wrote:
Hi Everyone,
I would like to invite your feedback on [1] prototype of Project-'UploadWizard:OSM embedding' [ user/password is mediawiki/mw ].
This project is about enhancing the image upload process for Wikimedia Commons.The main goal of the OSM embedding project is to provide map interface to the UploadWizard. We can not expect each and every uploader to know exact location where image being uploaded was taken, or to use location enabled camera/phone to take the image. With introduction of map interface, uploader can easily choose location coordinates for uploaded images.
Detailed project description can be found at [2].
Functioning of embedded map :
- On valid coordinates/location inputs widget first loads a static
map which on 'click' will initialize leaflet powered dynamic map
- Address search
- ' type' or 'importance' specific zoom for various kinds of
locations on address search
- .Dynamic updation of coordinate fields and location field on
leaflet map interaction and vice versa
Kindly give your suggestions for map widget- design as it needs to be improved, also propose features and/or improvements in current features. Please share your insights at [3].
[1]http://uploadwizard-osm.bitnamiapp.com/mediawiki/Special:UploadWizard [2] https://www.mediawiki.org/wiki/UploadWizard_OSM_map_embedding [3] https://www.mediawiki.org/wiki/Talk:UploadWizard_OSM_map_embedding
Best regards Anu G Enchackal
Design mailing list Design@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/design
Design mailing list Design@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/design
Hey there, Here's some things I'd do differently.
1. Make the map full-width and put the "Latitude and Longitude" on top of it. That would make it way easier to figure out where you are on the map and find the place you're looking for.
2. Why can't I drag the map? Is that just not supported by OSM, or did you for whatever reason disable that feature? Either way, seems super-counter-intuitive for me.
3. Just show the map on pageload, not just when I enter an adress. There's no reason to put in an extra step like this in the process. Let's make it as streamlined as possible.
Other than that, great work. Best, max.
Hi Anu,
First, congratulations for your work, making it easy to provide geographic information to media is quite important. I planned to send some feedback earlier but didn't found the time, hope it is still helpful:
- *Coordinates vs map as main input for location.* Currently coordinate fields and the map are provided side-by-side. It would be good to think how to simplify this since I guess that both are not intended to be used with the same frequency. Several options came to my mind: (a) Remove coordinate fields, and add a coordinates action or icon inside the map search bar. In that way users can click the coordinates icon to turn the search bar into the coordinate fields, (b) in addition to the previous one allow flexible input so that the same bar can be used both ways, (c) present coordinates as part of the map with output as a main purpose but suggesting that they can be also edited (similar to what Max proposed). - *Anticipating the needs of the user.* The searching process can be benefited by some strategies to save the user from typing. For example, as soon as the user indicates their intent of setting a location, the map could load some area nearby the user location (or maybe keep the previous are the user looked for). In addition, updating the map as the user types with autocompletion could also help in that regard. - *Initial status.* Initially, for me, the map was a grey box. It would be preferred it being the world map (to give context) or a usual/current location to save time in some cases. - *Continuity between map types.* The change between the two kinds of maps is not fluent. According to what Gergo said, the dynamic map can be used if only one image is being uploaded (or make the switch when the user clicks n the search box). For the cases where such transition cannot be avoided, some visual cue will be needed to indicate that you are expected to click: the cursor could be a hand, as you hover, the location indicator can follow the cursor to communicate that you can place it in a more detailed position, and the label that indicates this could be more prominent and refer to "pick a specific location" instead that referring to technical terms such as "leaflet map".
Making it possible to make the location input compact (e.g., integrating coordinates with the map) is something really useful. Just to share some initial design explorations about commons, I was thinking on making extra information such as location more accessible form the description page (see very early mockup http://i.imgur.com/LlbcXwV.png). in the same way I illustrated for categories (see another mockuphttp://i.imgur.com/MAbLFi8.png), clicking on the globe can show your map-driven input.
Hope these ideas are helpful.
Pau
On Thu, Mar 13, 2014 at 6:53 AM, Max max@koehler-kn.de wrote:
Hey there, Here's some things I'd do differently.
- Make the map full-width and put the "Latitude and Longitude" on top of
it. That would make it way easier to figure out where you are on the map and find the place you're looking for.
- Why can't I drag the map? Is that just not supported by OSM, or did you
for whatever reason disable that feature? Either way, seems super-counter-intuitive for me.
- Just show the map on pageload, not just when I enter an adress. There's
no reason to put in an extra step like this in the process. Let's make it as streamlined as possible.
Other than that, great work. Best, max.
Design mailing list Design@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/design
Date: Thu, 13 Mar 2014 16:52:01 +0100 From: Pau Giner pginer@wikimedia.org To: "A list for the design team." design@lists.wikimedia.org Subject: Re: [Design] Requesting feedback on Project- UploadWizard: Openstreetmap embedding Message-ID: <CALRPo1DDhkVZfGV4wu7Xda7sSCMpGtmpuddMYvyQiB4+GGFC= Q@mail.gmail.com> Content-Type: text/plain; charset="iso-8859-1"
Hi Anu,
First, congratulations for your work, making it easy to provide geographic information to media is quite important. I planned to send some feedback earlier but didn't found the time, hope it is still helpful:
- *Coordinates vs map as main input for location.* Currently coordinate
fields and the map are provided side-by-side. It would be good to think how to simplify this since I guess that both are not intended to be used with the same frequency. Several options came to my mind: (a) Remove coordinate fields, and add a coordinates action or icon inside the map search bar. In that way users can click the coordinates icon to turn the search bar into the coordinate fields, (b) in addition to the previous one allow flexible input so that the same bar can be used both ways, (c) present coordinates as part of the map with output as a main purpose but suggesting that they can be also edited (similar to what Max proposed).
- *Anticipating the needs of the user.* The searching process can be
benefited by some strategies to save the user from typing. For example, as soon as the user indicates their intent of setting a location, the map could load some area nearby the user location (or maybe keep the previous are the user looked for). In addition, updating the map as the user types with autocompletion could also help in that regard.
- *Initial status.* Initially, for me, the map was a grey box. It would
be preferred it being the world map (to give context) or a usual/current location to save time in some cases.
- *Continuity between map types.* The change between the two kinds of
maps is not fluent. According to what Gergo said, the dynamic map can be used if only one image is being uploaded (or make the switch when the user clicks n the search box). For the cases where such transition cannot be avoided, some visual cue will be needed to indicate that you are expected to click: the cursor could be a hand, as you hover, the location indicator can follow the cursor to communicate that you can place it in a more detailed position, and the label that indicates this could be more prominent and refer to "pick a specific location" instead that referring to technical terms such as "leaflet map".
Making it possible to make the location input compact (e.g., integrating coordinates with the map) is something really useful. Just to share some initial design explorations about commons, I was thinking on making extra information such as location more accessible form the description page (see very early mockup http://i.imgur.com/LlbcXwV.png). in the same way I illustrated for categories (see another mockuphttp://i.imgur.com/MAbLFi8.png), clicking on the globe can show your map-driven input.
Hope these ideas are helpful.
Pau
Thank you very much, Pau, Your ideas are really helpful.Can you share some information about implementing something like "map of Canada and some bird pictures" as illustrated in http://blog.wikimedia.org/2014/01/09/multimedia-vision-2016/ ?
On Thu, Mar 13, 2014 at 6:53 AM, Max max@koehler-kn.de wrote:
Hey there, Here's some things I'd do differently.
- Make the map full-width and put the "Latitude and Longitude" on top of
it. That would make it way easier to figure out where you are on the map and find the place you're looking for.
- Why can't I drag the map? Is that just not supported by OSM, or did
you
for whatever reason disable that feature? Either way, seems super-counter-intuitive for me.
- Just show the map on pageload, not just when I enter an adress.
There's
no reason to put in an extra step like this in the process. Let's make it as streamlined as possible.
Other than that, great work. Best, max.
Thank you Max for your feedback. I have temporarily disabled drag feature due to leaflet library's bug https://github.com/Leaflet/Leaflet/issues/872 . But next version will address all your concerns.
Anu
Thank you very much, Pau, Your ideas are really helpful.Can you share some information about implementing something like "map of Canada and some bird pictures" as illustrated in http://blog.wikimedia.org/2014/01/09/multimedia-vision-2016/ ?
The basic idea there was that objects (such as images) could have geographic data associate to them. This data can represent a point (e.g., the place where the bird picture was taken) or an area (e.g., the coordinates that an image of a map of Canada covers). Once you have this info, you can visually arrange information based on it. In the particular example of the Multimedia Vision, when preparing some slides, a user selects a group of pictures of birds and a map and simply by selecting "arrange by location", all pictures are placed at the corresponding locations in the map.
This was an example of how geographic data can be surfaced when organising information only for illustration purposes, but (a) there may be other/better ways, and (b) there may be cases not considered and technical difficulties not reflected in the example. So further analysis should be made for the specific cases we want to solve.
Pau
On Fri, Mar 14, 2014 at 1:08 PM, Anu George Enchackal < inchikutty13@gmail.com> wrote:
Date: Thu, 13 Mar 2014 16:52:01 +0100
From: Pau Giner pginer@wikimedia.org To: "A list for the design team." design@lists.wikimedia.org Subject: Re: [Design] Requesting feedback on Project- UploadWizard: Openstreetmap embedding Message-ID: <CALRPo1DDhkVZfGV4wu7Xda7sSCMpGtmpuddMYvyQiB4+GGFC= Q@mail.gmail.com> Content-Type: text/plain; charset="iso-8859-1"
Hi Anu,
First, congratulations for your work, making it easy to provide geographic information to media is quite important. I planned to send some feedback earlier but didn't found the time, hope it is still helpful:
- *Coordinates vs map as main input for location.* Currently coordinate
fields and the map are provided side-by-side. It would be good to think how to simplify this since I guess that both are not intended to be used with the same frequency. Several options came to my mind: (a) Remove coordinate fields, and add a coordinates action or icon inside the map search bar. In that way users can click the coordinates icon to turn the search bar into the coordinate fields, (b) in addition to the previous one allow flexible input so that the same bar can be used both ways, (c) present coordinates as part of the map with output as a main purpose but suggesting that they can be also edited (similar to what Max proposed).
- *Anticipating the needs of the user.* The searching process can be
benefited by some strategies to save the user from typing. For example, as soon as the user indicates their intent of setting a location, the map could load some area nearby the user location (or maybe keep the previous are the user looked for). In addition, updating the map as the user types with autocompletion could also help in that regard.
- *Initial status.* Initially, for me, the map was a grey box. It would
be preferred it being the world map (to give context) or a usual/current location to save time in some cases.
- *Continuity between map types.* The change between the two kinds of
maps is not fluent. According to what Gergo said, the dynamic map can be used if only one image is being uploaded (or make the switch when the user clicks n the search box). For the cases where such transition cannot be avoided, some visual cue will be needed to indicate that you are expected to click: the cursor could be a hand, as you hover, the location indicator can follow the cursor to communicate that you can place it in a more detailed position, and the label that indicates this could be more prominent and refer to "pick a specific location" instead that referring to technical terms such as "leaflet map".
Making it possible to make the location input compact (e.g., integrating coordinates with the map) is something really useful. Just to share some initial design explorations about commons, I was thinking on making extra information such as location more accessible form the description page (see very early mockup http://i.imgur.com/LlbcXwV.png). in the same way I
illustrated for categories (see another mockuphttp://i.imgur.com/MAbLFi8.png),
clicking on the globe can show your map-driven input.
Hope these ideas are helpful.
Pau
Thank you very much, Pau, Your ideas are really helpful.Can you share some information about implementing something like "map of Canada and some bird pictures" as illustrated in http://blog.wikimedia.org/2014/01/09/multimedia-vision-2016/ ?
On Thu, Mar 13, 2014 at 6:53 AM, Max max@koehler-kn.de wrote:
Hey there, Here's some things I'd do differently.
- Make the map full-width and put the "Latitude and Longitude" on top
of
it. That would make it way easier to figure out where you are on the map and find the place you're looking for.
- Why can't I drag the map? Is that just not supported by OSM, or did
you
for whatever reason disable that feature? Either way, seems super-counter-intuitive for me.
- Just show the map on pageload, not just when I enter an adress.
There's
no reason to put in an extra step like this in the process. Let's make
it
as streamlined as possible.
Other than that, great work. Best, max.
Thank you Max for your feedback. I have temporarily disabled drag feature due to leaflet library's bug https://github.com/Leaflet/Leaflet/issues/872 . But next version will address all your concerns.
Anu
Great to see this work done.
Is there a way to use a service other than leaflet? Have you tried out the Maps extension? Note that when selecting places in China for example, the name will have some characters from other languages. Is that handled correctly? After a file has been uploaded only the coordinates end up in the page's wikitext. It would be cool to see the address as well.
But anyways, this is awesome!
On Fri, Mar 14, 2014 at 7:20 PM, Pau Giner pginer@wikimedia.org wrote:
Thank you very much, Pau, Your ideas are really helpful.Can you share some
information about implementing something like "map of Canada and some bird pictures" as illustrated in http://blog.wikimedia.org/2014/01/09/multimedia-vision-2016/ ?
The basic idea there was that objects (such as images) could have geographic data associate to them. This data can represent a point (e.g., the place where the bird picture was taken) or an area (e.g., the coordinates that an image of a map of Canada covers). Once you have this info, you can visually arrange information based on it. In the particular example of the Multimedia Vision, when preparing some slides, a user selects a group of pictures of birds and a map and simply by selecting "arrange by location", all pictures are placed at the corresponding locations in the map.
This was an example of how geographic data can be surfaced when organising information only for illustration purposes, but (a) there may be other/better ways, and (b) there may be cases not considered and technical difficulties not reflected in the example. So further analysis should be made for the specific cases we want to solve.
Pau
On Fri, Mar 14, 2014 at 1:08 PM, Anu George Enchackal < inchikutty13@gmail.com> wrote:
Date: Thu, 13 Mar 2014 16:52:01 +0100
From: Pau Giner pginer@wikimedia.org To: "A list for the design team." design@lists.wikimedia.org Subject: Re: [Design] Requesting feedback on Project- UploadWizard: Openstreetmap embedding Message-ID: <CALRPo1DDhkVZfGV4wu7Xda7sSCMpGtmpuddMYvyQiB4+GGFC= Q@mail.gmail.com> Content-Type: text/plain; charset="iso-8859-1"
Hi Anu,
First, congratulations for your work, making it easy to provide geographic information to media is quite important. I planned to send some feedback earlier but didn't found the time, hope it is still helpful:
- *Coordinates vs map as main input for location.* Currently
coordinate
fields and the map are provided side-by-side. It would be good to think how to simplify this since I guess that both are not intended to be used with the same frequency. Several options came to my mind: (a) Remove coordinate fields, and add a coordinates action or icon inside the map search bar. In that way users can click the coordinates icon to turn the search bar into the coordinate fields, (b) in addition to the previous one allow flexible input so that the same bar can be used both ways, (c) present coordinates as part of the map with output as a main purpose but suggesting that they can be also edited (similar to what Max proposed).
- *Anticipating the needs of the user.* The searching process can be
benefited by some strategies to save the user from typing. For example, as soon as the user indicates their intent of setting a location, the map could load some area nearby the user location (or maybe keep the previous are the user looked for). In addition, updating the map as the user types with autocompletion could also help in that regard.
- *Initial status.* Initially, for me, the map was a grey box. It
would
be preferred it being the world map (to give context) or a usual/current location to save time in some cases.
- *Continuity between map types.* The change between the two kinds of
maps is not fluent. According to what Gergo said, the dynamic map can be used if only one image is being uploaded (or make the switch when the user clicks n the search box). For the cases where such transition cannot be avoided, some visual cue will be needed to indicate that you are expected to click: the cursor could be a hand, as you hover, the location indicator can follow the cursor to communicate that you can place it in a more detailed position, and the label that indicates this could be more prominent and refer to "pick a specific location" instead that referring to technical terms such as "leaflet map".
Making it possible to make the location input compact (e.g., integrating coordinates with the map) is something really useful. Just to share some initial design explorations about commons, I was thinking on making extra information such as location more accessible form the description page (see very early mockup http://i.imgur.com/LlbcXwV.png). in the same way I
illustrated for categories (see another mockuphttp://i.imgur.com/MAbLFi8.png),
clicking on the globe can show your map-driven input.
Hope these ideas are helpful.
Pau
Thank you very much, Pau, Your ideas are really helpful.Can you share some information about implementing something like "map of Canada and some bird pictures" as illustrated in http://blog.wikimedia.org/2014/01/09/multimedia-vision-2016/ ?
On Thu, Mar 13, 2014 at 6:53 AM, Max max@koehler-kn.de wrote:
Hey there, Here's some things I'd do differently.
- Make the map full-width and put the "Latitude and Longitude" on top
of
it. That would make it way easier to figure out where you are on the
map
and find the place you're looking for.
- Why can't I drag the map? Is that just not supported by OSM, or did
you
for whatever reason disable that feature? Either way, seems super-counter-intuitive for me.
- Just show the map on pageload, not just when I enter an adress.
There's
no reason to put in an extra step like this in the process. Let's make
it
as streamlined as possible.
Other than that, great work. Best, max.
Thank you Max for your feedback. I have temporarily disabled drag feature due to leaflet library's bug https://github.com/Leaflet/Leaflet/issues/872 . But next version will address all your concerns.
Anu
-- Pau Giner Interaction Designer Wikimedia Foundation
Design mailing list Design@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/design