Thanks! That's good to know - I've already built out authentication in the Puppeteer-based sync tool I'm working on, but haven't yet got further, and it'll help a lot once I do to know to account for this.
Granted, that means still a third step to orchestrate - auth ending in cookie capture, then album list fetch and reconciliation, and finally (on the other end of some queue) album downloading. Thank goodness they didn't make it too simple, I guess - I'd have hated for their annoyingly inconsistent login captcha to have been the last arbitrary challenge I had to overcome.
(I can't fairly give them too hard a time, I suppose - they're doing an incredible amount with a tiny team and in that context even having pulled their public API is reasonable, however much I wish they hadn't. Just that it'd be nice if they had the scope for dealing with use cases that aren't the adversarial mirroring they've clearly optimized against...)
Granted, that means still a third step to orchestrate - auth ending in cookie capture, then album list fetch and reconciliation, and finally (on the other end of some queue) album downloading. Thank goodness they didn't make it too simple, I guess - I'd have hated for their annoyingly inconsistent login captcha to have been the last arbitrary challenge I had to overcome.
(I can't fairly give them too hard a time, I suppose - they're doing an incredible amount with a tiny team and in that context even having pulled their public API is reasonable, however much I wish they hadn't. Just that it'd be nice if they had the scope for dealing with use cases that aren't the adversarial mirroring they've clearly optimized against...)