With the immanent termination of Google Reader it was an ever more urgent task to come up with an alternative to maintain, or even optimize, my work flow. Although announced way back in mid March, I took my fine time taking a decision. There was the faint chance, that Google may change its mind… more likely, though, was that the community would come up with an even better (and really free) alternative. The discussions were on, big time – at least amongst the nerds, that is.
I checked out and had hopes for Feedly but could not be bothered to use a service that would require me to install an extension potentially capable of recording my entire browsing activity. Also I want to be able to use the reader on any device (mine or not), store what I have read/stared, and search through it.
After some time it further more became clear, that I did not want again to run into the situation that a cloud provider would shut down a (pay-for or not) product, I so heavily rely on, so hosting it myself was ever more attractive. This criteria gave the final boost to the option of using Tiny Tiny RSS (TT-RSS) – setup is easy, and often described.
Google is kind enough to aid in migrating to other products by providing your subscriptions and Reader usage history for download in Google Takeout (that is for all Google products!). You download an OPML file (appears as a simple ZIP container), which contains, amongst others, a subscriptions.xml, which you can import (Preferences > Feeds > OPML) into TT-RSS to get your feeds integrated.
It was not initially obvious, but there is also a built-in Plugin to import Stared/Shared items from Google Reader, which can be used after activation (Preferences >Plugins > tick „googlereaderimport„), by uploading shared.json and starred.json, which you can extract from your OPML file.
I run it behind a proxy that provides SSL, which messes up the redirect (TT-RSS uses PHP’s REQUEST_URI to determine it) after login, but I don’t think blame for that can be assigned to TT-RSS. Did not take the time to look into that closer; not big enough a bummer.
Done. Works well so far, and does what it should – free after all.
- the folders cache, lock, and feed-icons have to be writable by the web server and have little to do with the software itself so architecturally it would have be cleaner to put all of them into a var folder (done that) – makes them easier to keep on update and manage their permissions
- admin user can not be removed/deactivated, so I gave it a ludicrous password
- a getting started wizard that notifies of the option to import Reader data would be handy
Number of cloud services I still use: Too many.