This looks brilliant! Before downloading the imagery, will the script tell you the size of the download, so you know you wonāt run out of disk space?
Thanks,
Chris
Mmmm, given I have 2 million images I might need to get a few extra terrabytes of disc space!
the script does not specify the size of the files. I calculate about 3 TB for 1,000,000 images.
I am pretty sure we have all the rights to do that under GDPR ā letās try to request to provide all our personal data and content. If I read the law right, Mapillary will have 30 days to do that. Moreover, under the very same GDPR we can request to remove all data uploaded by us, tooā¦
Not a lawyer, but GDPR covers personally identifiable data. Contributions to a crowdsourced dataset are governed by an agreement, usually licensing.
Quoting Terms of Use :
You and each Mapillary User hereby grant to us the perpetual, irrevocable, worldwide, royalty-free, fully paid-up, sub-licensable, non-exclusive right and license to use, reproduce, modify, create derivative works of, perform, display and distribute without restriction (including via any medium) Content, in whole or in part, for commercial and non-commercial purposes, including (a) for purposes of providing, developing, improving, making available (including via public search engines), promoting and marketing the Mapillary Solution and (b) making available Mapillary User Content under the Outbound Terms.
I somehow view possible deletion of contributed content similar to somebody deciding to delete their contributions to OSM. Other users based their contributions on what was in there already (like not surveying - or photographing - a particular area). Ability to delete contributions at will harms other contributors.
Granted, OSM and Mapillary are significantly different, given the highly open and decentralised nature of OSM, contrasted with very centralised nature of Mapillary.
I can sympathise a lot with people having concerns, and I hope Mapillary team can work on legal issues, commitment to openness and communication to address concerns in a diplomatic way instead of a legal hammer (as in saying āyou licensed it, too badā).
I agree with @Richlv, I strongly doubt you can demand Facebook deletes the pictures you uploaded, as they are your contributions under their TOS, not your personal data.
Deleting your photos through the site or a tool, will only delete them from the Mapillary website, not from any backup and Facebook will still own them. This actually makes things easier for Facebook, since Facebook has less hosting costs with your photos gone, and worse for other users, since they canāt use your contributions anymore.
I also agree with @Richlv: Iāve been using a lot of Mapillary images for my OSM editing, of course not only my photos, but from lots of other people. It would be a shame if these photos are all gone because everyone is deleting theirs.
As Iāve written before, Iāll probably just stop contributing more images, but am happy to use the āoldā onesā¦
Exactl!
The problem I also see is that there is no other platform, since OpenStreetCam in particular seems to have gone the way of the Dodo since it was transferred to Grab.
I would be helpful if Facebook responded what their plans are with Mapillary. Personly I do not believe they bought Mapillary just for the developers and not for the content.
I tried that on Windows but it gives me an error because it wants to name the file in the following format:
X:\mapillary takeout/2017-08-21/2017-08-21T13:28:00.714Z_2017-09-18T16:17:22.350Z_0001.jpg
and Windows doesnt accept ā:ā in file names. Is there a possibility to change the output format? Otherwise Ill power on my Linux machine (probably the easier solution I guess)
And do the file names look different for you in the end when using the date filter? I did manage to get it to start downloading but it would fail at the first download due to the colon in the file name which Windows doesnāt support.
The space in the path is definitely not the problem, I wrap the whole path in quotes and it created the first folder in the desired location too before getting stuck on the first image.
ah, I use linux, now I understand you. Maybe create an issue at https://github.com/gitouche-sur-osm/mapillary_takeout/issues
Or guess you can edit the mapillary_takeout.py just below line 155 where sequence_name is set. (you may do a colon replace python command I think)
Thanks!
I donāt know how to code that replacement but Iāve just started up my Linux PC and Iām doing it there. Thatās probably the easiest fix 
Do you perhaps know how to rename all files afterwards in one step, including all subfolders etc? Because I want to use the files in Windows after downloading but I assume it wonāt display them if they donāt meet Windowsā naming criteria.
I just tested, just add that last replace line (just below line 155):
sequence_name = (
sequence["properties"]["captured_at"]
+ "_"
+ sequence["properties"]["created_at"]
)
sequence_name = sequence_name.replace(":", "_")
Since Iāve actually started downloading already Iām now using krename to rename them, but thanks
As far as I can see the images get downloaded without any geolocation, right? Is there any way to get the corresponding locations as well?
Iāve taken most of my imagery with the Mapillary app and never saved the gpx tracks that it stores temporarily.
That would be very bad.
I do not intend to delete them, but I would like them safe in an archive.
the downloads contains the Exif data with geolocation. Just try to upload some of those downloaded files with the webuploader for review and see them placed correctly.
I receive publicity from all extremist parties in my FB account. But what worries me more is that I also receive publicity for diapers.
Alright, Iāll check it out later, I didnāt see them when checking yesterday. Thanks again!