DISQUS

Drew Kime · 11 months ago

If nothing else, maybe some competition -- or at least the threat of it -- will wake them up on support issues. I realize it's a free service, but I've been following a problem on the support boards for several months that gets absolutely zero response. (Email subscribers counts repeatedly disappearing from total feed count.)

If you check the Feedburner support forums, you see a *lot* of questions, and almost no feedback ... except when it's an obvious trivial user error, which they're pretty quick about solving.

gsmaverick · 11 months ago

I don't think the current state of pinging server is going to be good enough for indexing.

mikepk · 11 months ago

I may be nit-picking a little, but I'm not sure it's a feedburner clone service that's needed here. (I do understand your borader point though). I tried talking to steve about this on Twitter, but I think he got fed up with me and "hung up" on me :).

Feedburner came about to solve a different problem than real-time updates. They built a system with a few goals (I believe) to act as a generic feed translator due to the fractured nature (and frequent poor implementation) of feeds, to create a robust service capable of handling large volume traffic from feed requests, and to provide statistics on feed usage. Their design was never to provide real-time updates.

That's not to say they can't or shouldn't change things now, but it might be the non-realtime nature of the problem they were solving is baked into their system deeply. I also argued that Google may be succumbing to the "innovators dilemma". It's only a feeling, but I think the structures inside google are re-aligning to maximize and support only their *current* revenue generation efforts. As they continue to grow in size, the pressure to meet their wall street growth number is going to get massive. I agree this is the future, but just like the mini computers missed the PC business cycle because it was too future oriented and presented limited revenue potential (at the time), I think Google may be feeling the same kind of effects.

One of the things we built into Grazr was that as a feed processing system, it was *always* up to date in the widget. We built it with an eye towards real-time data, with caching in the system for performance but it always checks the sources for updates when running (using standard HTTP protocol caching semantics). Not the most efficient way to do things, but at the time it was the only guarantee of absolute feed freshness. Unfortunately no one, and I mean no one, cared about that. That was the environment FeedBurner was built in. Things are changing now (clearly) and real time will become more and more important.

evansolomon · 11 months ago

If you want a legit competitor to Feedburner, it's worth pointing out that advertising is probably at least as important, if not more, than RSS stats in attracting users. If what you actually want--and this is what it sounds like--is an open source app to compile server logs or something like that into a nice stats UI, that's a much different endeavor than a Feedburner competitor imo.

dave · 11 months ago

I think the ads are a relatively recent addition, but I could be mistaken,
because I think the idea of putting an ad in a feed makes as much sense as
putting an ad on an ad, which is to say none at all.

evansolomon · 11 months ago

I like the "ad on an ad" line, it's clever. However, I don't think it's accurate here. I can understand your comparison of a feed to an ad (if that's in fact what you were doing) but it's pretty obviously content that can be advertised against.

To keep the "ad on an ad" thing going, I bet a TON of companies would pay to crossbrand really successful commercial campaigns (like the Budweiser frogs from years back)--probably not as many would pay to crossbrand the HeadOn disaster commercials. Ads and content can merge, and to the extent that a feed is like an ad, I think that's a clear example of it.

Bill Fliier · 11 months ago

Dave, Pheedo serves large commercial publishers (NYT, CNET, PCWeek and more) with feed stats (and ads). Contact me if you are interested. Be happy to see if we can get your feeds updated quickly.

Best,
Bill Flitter
Founder
Bill @ Pheedo.com
888-495-8384

Jauder Ho · 11 months ago

If Feedburner adds SUP support, that would help.

Here's the request on Satisfaction. http://bit.ly/c1gU

David Recordon · 11 months ago

SUP is only useful if you are already fetching a lot of feeds and thus can discover their SUP IDs to look for in the updates stream. Thus, it's useful for FriendFeed as they already know about the majority of feeds that would be included in a SUP updates stream.

SUP is useful if you want a layer of indirection between the list of changes and URLs of feeds (it would be useful for advertising updates to non-public feeds) but as it stands still seems like more work than something like the Six Apart Update Stream (http://updates.sixapart.com/) which contains full content and can be easily transformed into a list of updated feed URLs with curl and grep. (n.b. I work for Six Apart)

Jauder Ho · 11 months ago

I was not trying to imply that this would be a catch all. There are quite a few sites pulling plenty of feeds these days and this would be useful in that use case.

I'll take a look at the Update Stream. Thanks for the link.

shokk · 11 months ago

Currently I'm using Mint with addons for RSS feed tracking, but everyone seems to prefer Feedburner. I seem to have something against paying for that service, so I would really welcome a free PHP based DIY solution to install on my own servers. I think it would really take off. I'm watching this closely...

Tumbleweed · 11 months ago

Dammit, beaten to the punch! Though I must admit, I wasn't planning on even attempting this until later this year. I was planning on using the domain 'feedsquisher.com' myself. Ah well.

vrypan · 11 months ago

Dave is right. Now that almost every web app has a couple of feeds, we need an open source "feed proxy" that will at least

1. will cache and throttle feeds
2. keep stats
3. Implement a ping server for the feeds it handles

Additionally,
4. do "smart things" with rss enclosures
5. do "smart things" depending on the client requesting the feed (browser, itunes, etc.)
6. RSS transformations

It looks like AppEngine is the ideal platoform to run this kind of app. It's (free) quota limits should be good enough even for blogs with high traffic.

smonev · 11 months ago

Feedburner solves 2 major problems:

1. It gives you stats
2. It helps you distribute your feed fast by putting your feed contents on their fast servers, so if you have 20000 subscribers polling every 30 minutes your own server wont die

It has some other features as well, but they are irrelevant:
1. Putting ads on feeds,
2. Ading social buttons bellow posts
3. Etc.

Stats are known to be easily faked and not quite correct.
Steve Gillmor has shown that the service fails to fix the other problem you have as a feed publiser: fast distribution.

So what does feedburner gives you? Just ads