I have been a faithful user of Feed on Feeds for about a year. I thank Steve Minutillo for all of his hard work and dedication and I was really pleased with the experience while it lasted.
However, with the increased load on my server, the 250 odd feeds that I tracked every hour and my goal to save all feed information, FoF was just killing my server for hours at a time. An hour was not enough to parse through the list of feeds and F0F would start a new update process while the old one was running, thus effectively DoSing itself. Time well spent with a great product but now I need to move on.
I have tried other aggregators and have had various levels of success. I just love FeedDemon but I miss the ability to view feeds online and I simply cannot lug my laptop around everywhere (and it does not work on Windows CE). I have tried Bloglines (which is what I have resorted to using at this time) but there were some features in FoF that I sorely miss such as the ability to look at a digest of all updated feeds in a coherent and succint format and then “mark read till here” from FoF. The interface to Bloglines is zippy but too many doodads ruin the experience. Newsgator is just weird. They tout their product more than anything else and it seems like it is aimed more towards the novice than the power user. NewzCrawler is another client, no dice.
The Wikipedia page on News Aggregators is very full of information but those are just too many to try. I have thought of digging into the FoF code and cleaning up some of the hangups and writing a python or perl collection module for it, but that would not happen in an evening. For now I am stuck with Bloglines until something better comes along or I either write something better or fix FoF to my “itch and scratch” liking.
What do you use for aggregation? Any products that really stand out beside the ones I have mentioned above? Is anyone thinking of writing something like FoF that is lightweight, fast and can keep up with volume?
[EDIT] I am back with FoF. It looks like there were some seriously borked feeds (including Feedster which has more than 200 errors from one of their feeds) and it had corrupted the mysql tables. I tried to repair/optimize but opted to export my feeds (thank God for backups!) and then re-import them back into a fresh FoF install. Now, I am updating 288 feeds in about 5 minutes and reading the rogue feeds in Bloglines 🙂 Let them deal with the validation errors, they have more resources than I do.
Not sure what cron you’re using to call the FoF update thing, but there are crons out there smart enough to not kick off a process again if it’s not yet done executing. fcron being the one I use. I’ve made a small handful of mods to FoF myself, adding a full-text search to be able to find stuff I remember reading on a feed but don’t remember which one, as well as a couple other things. I can create a patch for you against the base FoF if you’re interested — drop me an email if you are.
Have you tried gregarius? I use it and so far love it. Functionality is not as extensive as FoF though (I’m a former FoF user), but I keep using it.
SharpReader.
When I was starting with FoF, the two that people mentioned to me the most were SimpleAggregator and Temboz.
I don’t suppose you’d like the suggestion that you fix whatever’s wrong with your server? 😉
I’m subscribed to 369 feeds, on cheap-ish shared hosting, and an update run seems to take six or seven minutes. Where are you bottlenecking?
Hmmm….it MIGHT just be my server and I should have checked on that first. The performance continued to degrade with time, so I never looked into it.
I would love to continue using FoF!
You may also look into replacing the built in FoF RSS engine. MagpieRSS can be deadly slow in comparison to something like LastRSS – switching engines could alleviate a lot of pain as well – but I don’t know how easy that would be.
Additionally, I think an update for FoF is due out any time now with a slew of improvements.
Temboz is what I am using. I hacked it a little bit, but mostly to have it run with PostgreSQL instead of SQLite – but there were no technical reasons, just my personal preferrence of PostgreSQL and the nice effect of being able to use the database from multiple running programs.
I’m with Phil, Mark; I average around 288 feeds aggregated, and I don’t have problems often. When I do, I sometimes can a feed. [I hear Mark Pilgrim crying softly.]
FWIW, Steve is working on getting v0.5 out to beta. [So much for incremental code work, I guess …]
I hadn’t heard of FoF before, thanks for mentioning it. I have tried just about every desktop aggregator I have found, and not liked any of them perfectly. I tried building my own web-based aggregator by using MT plugins and frames, and that wasn’t too good either. I’m currently using Bloglines but I’m not thrilled with it…and yet all the other web-based aggregators aren’t quite what I want, either.