I’m very sad about this, especially because of all the friends I have made on Flickr, but with Verizon’s acquisition of Yahoo (and so Flickr) and the consequent sharing of Flickr user data with the new “Oath” “”””family”””” I feel like it’s time for me to admit just how shit Flickr has become and finally leave. I’ve been using it (and paying for it) for 10 years though, so I’ve a lot of pictures, about 13K in number, about 23G. I’ve got all my data and will delete my account tomorrow (which I think it’s their deadline, but they seem confused about it).
It’s been a busy week so I don’t know what I’ll replace it with yet, maybe something simple and static I’ll write myself like the thing I had 11 years ago, with some, I dunno, RSS feeds or something. But anyway, here’s the best way I’ve found to get my data back, and kudos to Flickr that the API is still there to make it possible. I tried a few things and flickrmirrorer looks best. It’s straightforward for pictures; some older videos need downloading by hand with it. As far as I can tell it gets all the metadata for each photo. No comments though, and no notes that I can see.
Because of the video issue I did images first, leaving it overnight (forgot to time it)
mkdir pictures cd pictures/ mkdir flickr cd flickr/ git clone https://github.com/markdoliner/flickrmirrorer cd flickrmirrorer/ sudo easy_install flickrapi mkdir data ./flickrmirrorer.py --ignore-videos --statistics /Users/libby/personal/pictures/flickr/flickrmirrorer/data/
New photos / videos: 12921 Deleted photos / videos: 0 Modified photos /videos: 0 Modified albums: 198 Modified collections: 0
Check it matches the volume of data from your stats page (roughly might be all you can hope for; there’s a problem with flickr’s reporting)
du -sh .
check a couple to make sure we’re actually getting some data
open ./data/photostream/72332206.jpg cat ./data/photostream/72332206.jpg.metadata
./flickrmirrorer.py --ignore-photos --statistics /Users/libby/personal/pictures/flickr/flickrmirrorer/data/
downloading about 50 by hand.
I was worried I didn’t have the metadata for some of them, so I hacked together a script that just got all the video metadata – which is here.
I also wanted a list of my favourites – I rolled my own script for that, here. I hardcoded the number of pages, sorry!
There doesn’t seem to be any way to get notes, which sucks.
To use these two scripts you need to get an api key from flickr here.
I’m really really annoyed about all the cool urls I’ll kill because of this. Oh well.
Update: Matthew in the comments thought that notes were available from this api, and he’s right. flickrmirrorer doesn’t get them (and actually doesn’t get as much metadata as I want) so I grabbed all the ids of my photos using the dump from flickrmirrorer as a starting point:
find . | grep "\.metadata" > list_of_photos.txt
and then use this script to get as much metadata as I can.
I also realised I didn’t have a list of my friends 😦 So, I wrote one more script to do that.