Commons talk:Photo scavenger hunts

Return to "Photo scavenger hunts" page.

ThemesEdit

Suggestion to add more example themes or objectives. I was in Washington when Clinton was elected and there was a fantastic atmosphere. Try to hook on existing events or manifestations (ecology, war, food, education, ...). --Foroa (talk) 17:32, 17 October 2008 (UTC)

Thanks, I've now added something on this.--Pharos (talk) 12:35, 20 October 2008 (UTC)

Comments made on the village pumpEdit

To keep the discussion in one place I have copied over the points I made earlier- with a few tweaks so it reads better.

Think positive.
If you can get so many folks on the streets taking these image, then you must have a plan for post-processing them.
Could you just prewarn the rest of us- post a notice somewhere(?village pump?) to alert the community.
  • How are you ensuring that each image correctly tagged with date, photographer, permissions and details of what it actually is.
  • What is your time scale?
  • When are you tagging your admin shots for deletion- there seems to be a management issue with the Spring photos.
  • Have you recorded camera position, so you can add the {{Location}} template, what is the time scale for getting them all tagged?
  • Categories seems to have been a problem in the past, have you got a list of the categories you plan to use, or have you got a member of your team assigned to checking that these categories are Commons format not :en:wiki. What is the time scale for getting all the images ?
  • Are you doing anything about quality control- are you assigning a team to retake images that are badly cropped or blurred?
Its a great idea, but you can appreciate why some of us have reservations but would be happy to suggest solutions for some of points above- that would be easy and fun for a large group of people to do- but mind numbingly boring for one of us to do alone in our humble garret.

ClemRutter (talk) 21:27, 5 October 2008 (UTC)

Thanks, ClemRutter. We have decided not to delete the context photos, because they document the creation of free content for Wikimedia, in a rather more direct way even than, say, Image:Wikipedians strolling througth New York.jpg, or the many other meetup and conference photos we have on Commons. With maderrn storage technology, space is not an issue. We should be done with the uploading and tagging in at most a month. Many of the locations have geocoded articles, but no, we don't have a GPS reading for each camera shot. The locations that were poorly taken (or that were not taken at all), we will keep on the list for next year, and certainly a number of local Wikipedians will be working from the list just at their leisure over the next few months.--Pharos (talk) 18:02, 6 October 2008 (UTC)
Hell, I don't use a GPS, just Google Maps and a clever software trick. Firstly you show you team how to put this bit of code into their Bookmark Tool bar
  • javascript:void(prompt('',%22{{location%20dec|%22%20+%20gApplication.getMap().getCenter().lat().toFixed(4)%20+%20%22|%22%20+%20gApplication.getMap().getCenter().lng().toFixed(4)%20+%20%22}}%22));
(On Firefox- you enter this into the Location field:-- and the single character + into the Name: field. and save it.).
Now with their upload software open in one window, and a map of NYC open in another, they move around the street map till you are at the very spot where the crime was committed (or image taken). Hit the bookmark button- and a pop up window explodes- There is the location tag they need. Ctrl-C and then Cancel- move over to your upload page and Ctrl-V. For the uncommitted, you take them to a page at random and work out from the map where other peoples images were take. You use shadows and the exif data to workout the approximate directions from a church spire, Moosburg, Kärnten would be a great place to start. Your images are so much easier as you have all this meta data. And a side effect will be that we have another 23 teams of Geotaggers.ClemRutter (talk) 19:36, 6 October 2008 (UTC)

ClemRutter (talk) 09:36, 22 October 2008 (UTC)

Generalized template for team attributionsEdit

I have now created {{Team}}, designed to be a universal template for attributions to teams in photo scavenger hunt events. This should obviate the need for individualized team templates for the different events, i.e. {{WTM}}, {{WTN}}, {{WSTM}}. Thanks.--Pharos (talk) 18:54, 22 October 2008 (UTC)

Preparation and Post processingEdit

This section seems to be a little slow in getting written. So to get it started I am putting down a few ideas- bear in mind that the last scavanger hunt I organised was some thirty years ago

Site of Scavenger hunt c1975

and in those days cameras used the wet glass process.

Advanced technology

.

  • Compiling the list

Select the area you are targetting. In the past it has been urban areas like Manhatten, an area well served by public transport but it could be a series of villages within 50 km of the starting point. (I am thinking of a university town- such as Reading:UK, Lancaster:UK, Montpellier:France). The hunt would then be a type of car rally.

previous photo hunt on the A23

.

To compile the list you need to consult Google Maps, Commons and en:Wikipedia and the local:Wikipedia.

Firstly select any photo in target area that has a geotag. Find it on commons and click on Location of Photo:Google Maps. This will show each of the images in the area (that has been properly geotagged) showing where images are missing.
Still on the map, click on the more tag and then Wikipedia, this shows you all the articles on Wikipedia within that area.
Now the fun starts. Read each article and on your texteditor (gedit/notepad) list any notable item that needs a piccy. At the same time, you can give an approximate {{coord}} of the item. This can be done by creating a W:bookmarklet containing the code:
javascript:void(prompt(,%22{{location%20dec|%22%20+%20gApplication.getMap().getCenter().lat().toFixed(4)%20+%20%22|%22%20+%20gApplication.getMap().getCenter().lng().toFixed(4)%20+%20%22}}%22));
A little thought now saves time later. For each item- cut and paste details from the wikipedia article into your text file- you can use this in the photos descriptions later when you are uploading them.
Time penalty!
Other items can be included, for instance villages/ building that look great but don't yet have an article. There can be fun generic items (jokers) - like a gas meter, a gasometer, a taximeter showing exactly £5.60, a traffic camera flashing a speeding car other than your own..... all that may have use in a future article.
It you plan for your teams to take photos of ten items and two jokers, a little thought is needed to ensure that it is possible within the time limit. If you are hitting an area by car, remember to get a driver to help at this stage. bThey know about one way systems, the time it takes to park- the time it takes to get from the car to the target etc. In urban areas- note bus and tube routes and the time it takes to cross the streets ( in some countries there is a quaint convention about using pedestrian crossings and waiting till the little man turns green! )
When the list is prepared, it is worth trawling through commons to discover what Categories may be appropriate for your target photos. Obviously the name of the locality, but do check- a church in Castrop-Rauxel does not go in Category:Castrop-Rauxel but in the sub category Category:Churches in Castrop-Rauxel!
  • Scoring

There are three principles here.

  1. It must be unambiguous
  2. Easy to understand
  3. Weighted to achieve the desired result.

One system.

Joker red 02.svg
1 points for a photo of the target
2 points for a photo of the target that is uploaded successfully
3 points for a photo of the target that is uploaded successfully with the prepared description and meaningful name.
5 points for a photo of the target that is uploaded with description and a correct {{location}}
A bonus of 10 for photograph of one joker, and 25 for two or more jokers.

Post processingEdit

  • Download

Download each cameras treasures into a separate directory on one of you master machines- by hook, crook, USB stick or cable. On the master machine you will have the upload software usually Commonist

  • Renaming
Where but no what and no geotag

Using a Mac- Using Linux- mv command Using Windows XP there is a convenient program called Irfan View Thumbnails that allows for Batch Renaming of selected files in a directory. You use this to change all your filenames from DSCF2366.jpg into something more meaningful like Winchester cathedral South Door WTW176 2366.jpg . That is Where, What, Scavenger Hunt code, image number

  • Templates
  • Uploading and adding the geotag

A quick way of uploading a directory of images is to use Commonist. This shobuld be loaded onto each of your master machines. Also you need to have a bookmarklet

javascript:void(prompt(,%22{{location%20dec|%22%20+%20gApplication.getMap().getCenter().lat().toFixed(4)%20+%20%22|%22%20+%20gApplication.getMap().getCenter().lng().toFixed(4)%20+%20%22}}%22));
This is different from the one you use on en:wikipedia, because they use a {{coord}} and commons uses {{location}}
Open Commonist, and find the teams directory. All the images will be displayed on the right side. The general stuff that applies to all images is placed on the left side, and the stuff specific to each image on the rightside. It is here you put the geotag and the description you have saved in your gedit/Notepad file.
Commonism in action
Now with their upload software open in one window, and a map of your target city open in another, you move around the street map till you are at the very spot where the crime was committed (or image taken). Hit the bookmark button- and a pop up window explodes- There is the location tag they need. Ctrl-C and then Cancel- move over to your image on the upload page (in Commonist)and Ctrl-V.
  • Category

All images must be in a category. There are several tools to help.Cat-a-lot,HotCat and Commonsense.

  • Gallery

Galleries are optional, but to finish the job each team will want to display their photos so they can be scored.

Okay folks- there we are- savage away. I have left in a few deliberate errors to test who has been paying attention. ClemRutter (talk) 22:10, 24 October 2008 (UTC)

Use http://toolserver.org/~multichill/coordinates.php to get the coordinates! Multichill (talk) 22:56, 24 October 2008 (UTC)
@Multichill- Never tried using it- so I tried it out on a twenty or so images- I congratulate you on an excellent tool which is far easier for the task here. Have you considered adding the :en coord template to join the :commons locate and the :nl Koördinaten. The reason I would like it is that when you are researching the link to :en Wikipedia that you put in the image description, you often find articles that have the wrong coordinate, and want to make the correction.
Back to scavanger hunts: though Commonist is good- I was considering whether a dedicated upload tool may be easier for the hunters to use, and produce a higher quality upload. Still, I wait a little longer then transfer this write up over to the article.
ClemRutter (talk) 20:18, 28 October 2008 (UTC)
Thanks for your work on this. One quibble, though. I agree that geotagging is a good thing, and this is a fine method for it, but I'm not so sure I see geotagging as being so central to this process. Maybe we should have the geotagging part in its own section? Thanks.--Pharos (talk) 20:38, 15 November 2008 (UTC)
We all have our obsessions, and finding illustrations is one of mine.
Site of Scavenger hunt c1975
. Look at the image to the left. If you were writing an article on the South Tyne, would you have found this image. Do a search for South Tyne and it occurs as 140 out of 147 - while if you had Google maps open you would have found the splodge. But take a closer look and you will see that it has been incorrectly Geotagged at>>Location dec|54.938102|-2.508799<< instead of >>location dec|54.9413|-2.5109<<, If geotagging had been included in the process- then we would have got better quality metadata.
It is up to the guys running a project what they choose to do or not do- so I include it as a process to aspire to. ( Now here's a thought- we could have stub scavenger hunts- good scavenger hunts- and featured scavenger hunts.... !)
Let me know if there is anything else I can do to assist.-ClemRutter (talk) 21:55, 15 November 2008 (UTC)

Generalized template for description noticesEdit

I've now created {{Hunt}}, which can hopefully serve as a generalized template for these projects. Thanks.--Pharos (talk) 01:12, 28 October 2008 (UTC)

Technical considerations - the need to avoid uploading non-educational imagesEdit

Full support for this project. I am concerned, though, that unless the technical considerations section is fully worked out in advance there will be the same contentious issues that came up following the Wiki takes Manhattan project. In particular, I see no reason why sorting, labelling and categorization should not all be done off-line, with the aim being to upload fully Commons-ready useful images. The signal to noise ratio is otherwise much too small, and what does concern me is the statement made above by Pharos that "We have decided not to delete the context photos" such as Image:WTM NewYorkDolls 025.jpg. This is wrong in my view, as the so-called context images do not in the long term help Commons' aims (no-one is ever going to do a search for them). It should be possible to do without them (or at least not upload them) next time. This is not intended as a criticism of the idea, just a suggestion for the mechanism by which it should be carried out. --MichaelMaggs (talk) 22:47, 8 November 2008 (UTC)

We are looking into ways to improve preuploading sorting of images, particularly by possibly developing an alternative to Commonist, which is a bit clunky. It should be clear though, that we did categorize the images from the recent project preuploading (well, at least most of them). The thing about the context photos is that they are not part of the "encyclopedic" part of Commons, i.e. noone is going to want to use them on Wikipedia or Wiktionary, but they are part of scope in the same sense that meetup photos are part of scope. And in that way, although they will not be going in an encyclopedia or dictionary, I feel they have a value in documenting a free culture project, and in promoting the free culture concept to the general world. Also, I do agree that in future we should develop more sophisticated methods for assessing what should be treated as an "encyclopedic" photo and what should be treated as a context photo. Thanks.--Pharos (talk) 10:24, 14 November 2008 (UTC)

Dont leave to much templates behindEdit

One thing i don't like about the current system is the templates left behind. Take for example Image:WSTM Team Dustizeff 0074.jpg. The description is one big template. If i want to move the image to another category this is impossible because the category is included in the template. This should change and it's possible (already made a quick proof of concept). I'll describe it in steps.

Make initial list of subjectsEdit

Make a list of subjects teams have to take images of. For each subject, try to think of a category to add. You might want to keep these list offline until the day of the hunt.

Teams go huntingEdit

Lot's and lots of images

Teams upload their imagesEdit

Each upload should contain only a template with the following information:

  • Team
  • Subject id

For example :{{User:Multichill/Scavenger hunt|team|code}} This is easy for the teams and keeps it flexible.

Update category listEdit

Update the category list so that all the images have good categories

Substitute templateEdit

When an image is checked substitute the template


This way we are flexible, but we don't create template mania. See User:Multichill/Scavenger hunt, User:Multichill/Scavenger hunt/description and User:Multichill/Scavenger hunt/category for a quick proof of concept. Multichill (talk) 13:45, 14 November 2008 (UTC)

Last modified on 15 November 2008, at 22:00