parallel-flickr is a tool for backing up your Flickr photos and generating a database backed website that honours the viewing permissions you've chosen on Flickr. parallel-flickr is not a replacement for Flickr.
parallel-flickr is still a work in progress. It ain't pretty or classy yet but it works. It almost certainly still contains bugs, some of them really stupid.
It still needs to be documented properly.
These instructions are incomplete. They'll probably work but I have tested them from scratch yet. The following assumes you're running on a brand new vanilla Ubuntu machine.
First, some basic OS-level setup:
$> sudo apt-get install git-core
$> git clone firstname.lastname@example.org:straup/parallel-flickr.git
$> cd parallel-flickr
$> sudo sh ./ubuntu/install.sh
$> sudo chown -R www-data templates_c
TO DO: apache configs
$> cd schema
$> mysql -u root -p < SETUP.md
$> cat *schema | mysql -u root -p flickr
Now set up the application config file:
$> cp www/include/config.php.example www/include/config.php
TO DO: updating the config file, see also: https://github.com/straup/flamework-tools/blob/master/bin/make-project.sh
Backing up photos
After setting up everything above, and setting your API key callback to "http://YOURDOMAINNAME.com/auth/", visit /account/backups/. This will create your backup user account and then from here you can run the various backup scripts inside of the bin/ directory.
write files to S3 (see also: flamework-aws)
make sure video files are actually being fetched properly
API hooks (see also: flamework-api)
always fetch path_alias for contacts (call flickr.people.getInfo)
sets, galleries, groups
uploads (and then re-uploading to Flickr (see above inre: API hooks))
dates and timezones... sad face (also something is currrently very squirrel-y with the way dates are being indexed in parallel-flickr-solr but I haven't been able to figure out what/why yet (20111125))
cron jobs for backups
consider moving all the backup jobs (fetching data for individual photos) in to a proper queuing system - this should probably be a feature flag so that the whole thing can still be run in "stupid" mode and not spiral in to astronaut territory.
context-specific URLs (e.g. in-faves or in-WOEID)
search, see also: parallel-flickr-solr
duplicate key errors fetching faves
better layout, tested in more than just Firefox
See also: TODO.txt
- password reminders are disabled, only because I don't have a mail server set up