WordPress.com’ RSS feeds suck

The primary feed reader that I use is the open source RSSowl. I also use NewsGator (Web and Outlook edition, though I’ve pretty much abandoned Outlook except for testing), AmphetaDesk, and Yahoo!RSS.

The http://scobleizer.wordpress.com/feed/ stopped working on October 29 and began working again today. I always got full text before the 29th and again today.

RSSwowl does give the error as to why a feed isn’t validating. Though, I’m afraid that I didn’t note the error message at the time.

It does sometimes happen. I would guess that very few users or sites hand generate the RSS, RDF or Atom XML for their feeds. The author publishes the post, and some script generates the feed automagically for them. Even SF Gate wasn’t validating for a few hours this morning. BBC screws up about once a week. No doubt better script debugging is the answer.

Robert, glad to have you back in my RSSowl.

The primary feed reader that I use is the open source RSSowl. I also use NewsGator (Web and Outlook edition, though I’ve pretty much abandoned Outlook except for testing), AmphetaDesk, and Yahoo!RSS.

The http://scobleizer.wordpress.com/feed/ stopped working on October 29 and began working again today. I always got full text before the 29th and again today.

RSSwowl does give the error as to why a feed isn’t validating. Though, I’m afraid that I didn’t note the error message at the time.

It does sometimes happen. I would guess that very few users or sites hand generate the RSS, RDF or Atom XML for their feeds. The author publishes the post, and some script generates the feed automagically for them. Even SF Gate wasn’t validating for a few hours this morning. BBC screws up about once a week. No doubt better script debugging is the answer.

Robert, glad to have you back in my RSSowl.

On my blog I’ve left the terms RSS and Atom behind – I just kall them (news/web) feeds. For a visitor it’s only really important that there are abilities to subscribe to the content and not what format they come wrapped in.

For the technically inclined it’s easy to see that I serve all the feeds as Atom 1.0. I believe it’s important to have that transparency – we don’t go around and call web pages html-pages or even php-pages or asp.net-pages, do we?

Oh, and on the matter of services vs running your own. Running your own can take work. We do it. Well over a year ago we started testing various open source blogware, including WordPress, b2evolution and plog. Our web sites were on a shared server at XO at the time. In addition to bandwidth and storage, XO uses a “resource” calculation to see what amount of CPU and RAM your using in the shared environment. Our domain hosting costs shot up – and we were posting infrequently and only one friend read us. After a few months, we settled on b2evolution [we needed the multi-blog, multi-author capabilities PLUS the community was much, much more responsive and helpful than WordPress], and we got a dedicated server from ServerBeach.

We use so little of the resources of the server, that we’re running a hosting service: open source software for small businesses, including b2evolution. Hey, it’s cheaper than what XO’s additional “resource units” were costing us, but it’s still a lot of money per month for the machine.

At any rate, my point is that unless you’re willing to dedicate a lot of your time to maintaining the underlying software, a service, be it Typepad, LiveJournal, WordPress.com, ours, or any other can be fine. Though, I do have to admit that I shudder whenever I want to pick up someone’s feed, and my only option is feedburner. Something just doesn’t feel right there. Also, a service should allow a user to have their own domain [or subdomain like blog(s).domainName.TLD], as well as an exit path that preserves their look and all archived posts if they should wish to move to another service or self-hoting.

Oh, and on the other point in these comments, I haven’t seen a feed reader yet that doesn’t support RSSv0.92, RSSv1 (a.k.a RDF), RSSv2 and Atom 0.3. Few support Atom 1.x yet.

I do agree with Marc Canter: the user is best served when the services support as many standards as possible.

Oh, and on the matter of services vs running your own. Running your own can take work. We do it. Well over a year ago we started testing various open source blogware, including WordPress, b2evolution and plog. Our web sites were on a shared server at XO at the time. In addition to bandwidth and storage, XO uses a “resource” calculation to see what amount of CPU and RAM your using in the shared environment. Our domain hosting costs shot up – and we were posting infrequently and only one friend read us. After a few months, we settled on b2evolution [we needed the multi-blog, multi-author capabilities PLUS the community was much, much more responsive and helpful than WordPress], and we got a dedicated server from ServerBeach.

We use so little of the resources of the server, that we’re running a hosting service: open source software for small businesses, including b2evolution. Hey, it’s cheaper than what XO’s additional “resource units” were costing us, but it’s still a lot of money per month for the machine.

At any rate, my point is that unless you’re willing to dedicate a lot of your time to maintaining the underlying software, a service, be it Typepad, LiveJournal, WordPress.com, ours, or any other can be fine. Though, I do have to admit that I shudder whenever I want to pick up someone’s feed, and my only option is feedburner. Something just doesn’t feel right there. Also, a service should allow a user to have their own domain [or subdomain like blog(s).domainName.TLD], as well as an exit path that preserves their look and all archived posts if they should wish to move to another service or self-hoting.

Oh, and on the other point in these comments, I haven’t seen a feed reader yet that doesn’t support RSSv0.92, RSSv1 (a.k.a RDF), RSSv2 and Atom 0.3. Few support Atom 1.x yet.

I do agree with Marc Canter: the user is best served when the services support as many standards as possible.

Yes, these sort of issues you raise have bothered me too. I have been a bit apathetic in the past towards these issues, but I do support your strong comments on this.

Availability of full-feeds is one of the most important issue to be addressed. After reading the comments I see there is a solution but I think it should be the default view in RSS readers if that’s what the author intended.

> Also, a service should allow a user to have their own
> domain [or subdomain like blog(s).domainName.TLD], as
> well as an exit path that preserves their look and
> all archived posts if they should wish to move to
> another service or self-hoting.

1) Feedburner lets you (for a small fee) use yourdomain.com/feed or whatever (while still taking advantage of their cool services).

2) Even on free Feedburner accounts, if you decide to ditch FB, they’ll kindly redirect accesses to your FB URL to whatever URL on your own server you’d like for 30 days (likely enough time for feedreaders to note the 301 redirect and react accordingly).

Note that I am not affiliated with FB in any way except as a happy user.

Oh, and Robert… it’s certainly possible that there’s SOME feed reader out there that can’t handle Atom 0.3, just like there are still some browsers out there that can’t handle proper CSS (and lots of other stuff), e.g., Netscape 4. I don’t know about you, but I gave up on those folks ages ago.

> Also, a service should allow a user to have their own
> domain [or subdomain like blog(s).domainName.TLD], as
> well as an exit path that preserves their look and
> all archived posts if they should wish to move to
> another service or self-hoting.

1) Feedburner lets you (for a small fee) use yourdomain.com/feed or whatever (while still taking advantage of their cool services).

2) Even on free Feedburner accounts, if you decide to ditch FB, they’ll kindly redirect accesses to your FB URL to whatever URL on your own server you’d like for 30 days (likely enough time for feedreaders to note the 301 redirect and react accordingly).

Note that I am not affiliated with FB in any way except as a happy user.

Oh, and Robert… it’s certainly possible that there’s SOME feed reader out there that can’t handle Atom 0.3, just like there are still some browsers out there that can’t handle proper CSS (and lots of other stuff), e.g., Netscape 4. I don’t know about you, but I gave up on those folks ages ago.

Garrrh! I can’t believe we’re still doing this. Robert, you’re absolutely right to complain and push wordpress into doing the right thing. Getting a non-validating RSS feed from a hosted service is unacceptable. And if the RSS feed doesn’t provide the information you want to be in it, complain about that as well. And “just use Feedburner” is not an answer. You shouldn’t *have to* use Feedburner. That should be a choice if you want to use their added value, not a requirement.

A few other thoughts:-
- We should be way past people handcoding RSS. It should be transparent and just work. If you need to read the XML to work out what’s going on, then you can also read specs and do research.

- There’s nothing funky about CDATA. It’s a part of the XML spec and has been for years. It’s one solution to the double encoded & and it does just work.

- Your RSS feed doesn’t have full text in the item.description and some html tags have been removed. Putting the full text into a content:encoded RSS2 extension tag is ok, but it assumes (wrongly) that most aggregators will be able to read it.

- Inventing new MIME types is a pain in the neck. (people who serve OPML and RDF-XML like this, please note). My browser has a perfectly good XML reader and it can handle stylesheets in XML. If I click on a link to an XML file I want to be able to read it. Not get some dialog forcing me to make a choice I may not understand.

- I’m not sure if you’ve started in MS word or used an MS system to cut and paste or what but you’ve got high order versions of ‘, “” and a few other characters in there. Character sets are a PITA, and while UTF-8 handles it, the feed may not end up in something that does. I do wish that when the content is basically aasci-7, all the systems involved would use aasci-7 for characters that exist in it. Microsoft please note, your smart quotes hurt developers because we frequently have to write code to convert them.

And finally, does anyone know who is reponsible for making & a reserved character in XML? Because shooting is too kind. I want to poke them with pointy sticks for eternity! And the same goes for the person who chose instead of / in Microsoft systems for folder path delimiters.

Yep, noticed your feed in Bloglines was down for a few days. Had to click on the main head to get through to your blog.

BTW I’ve had the same problem with RSS2 on my standard WordPress blog (via Fantastico). I set it to full text, but nothing happened for a few days, then a partial feed arrived in tha aggregators. So it’s a generic problem, not just WordPress.com.

Yep, noticed your feed in Bloglines was down for a few days. Had to click on the main head to get through to your blog.

BTW I’ve had the same problem with RSS2 on my standard WordPress blog (via Fantastico). I set it to full text, but nothing happened for a few days, then a partial feed arrived in tha aggregators. So it’s a generic problem, not just WordPress.com.

I’ve had discussion with Matt via Dave Winer about the WP RSS feeds, the implementation regarding enclosures and podcasting usability is terrible – it turns all links to files as enclosures – not just the first or one…great functionality if you need it; if not you can’t turn it off or delete the extra enclosures…in a Douglas Adams way the ‘delete’ button just serves to make you feel better about your life, but doesn’t actually delete. So I followed Ben over at Tracks up the Tree and intentionally ‘broke’ that function…not great, and I’m not the first to do or say this.

I seem to remmeber I had to do some hacking to get full posts in RSS2 – initially I kept getting the summaries, NOT the post. There are two different WP functions, one for summaries and the other for posts, never the twain shall meet.

It’s annoying, and a shame because WP is a great blogging platform…but for podcasting and for out-the-box it’s got a long way to go…atm I can’t really tell non-techy types to go WP when I know techy people have issues – WP’s loss is Blogger’s gain, sadly.

eek Monday morning and I’m a bit slow..

Actually just checked my feed and I have the same issue – truncated post…I remember now I tried the full post WP function and got nada…it’s very odd. Let us know if you find a solution…

It’s odd because aggregated sites from my feed, such as Podcastalley et al show the full post? Curiouser and curiouser…

eek Monday morning and I’m a bit slow..

Actually just checked my feed and I have the same issue – truncated post…I remember now I tried the full post WP function and got nada…it’s very odd. Let us know if you find a solution…

It’s odd because aggregated sites from my feed, such as Podcastalley et al show the full post? Curiouser and curiouser…

I read your feed through SharpReader, and I’ve had no problems with it whatsoever…

I don’t know that sucks is correct, how about inconsistent? I’ve been online for the past seven hours and my reader just picked up this post along with 19 others. Sometimes they’re timely, sometimes they’re not.

I don’t know that sucks is correct, how about inconsistent? I’ve been online for the past seven hours and my reader just picked up this post along with 19 others. Sometimes they’re timely, sometimes they’re not.

Robert: Hosted services have to pick reasonable defaults for all customers. The fact that the default doesn’t suit your specific needs doesn’t mean it “sucks”.

You know this, of course. So I’m left wondering why you’d waste time with virtiol instead of simply asking how to tweak the default.

“I just tried my Atom feed and I couldn’t even get it to load up in the browser. That certainly is not good usability…”

Agreed. That means your aggregator is broken.

When you request a properly-served Atom or RSS file, your desktop aggregator should launch and help you subscribe to it. Unless you use a web-based aggregator, in which case you should be clicking your autodiscovery bookmark instead of playing “hunt the feed”.

“I never got complaints” isn’t a perfect measure of how good something is: for a while, when I was using Bloglines, I set my subscriptions to show “Summaries if Available” so I could quickly skim through, opening interesting posts in a browser tab. Since your feed didn’t have separate summaries and full content, I had to scroll past your whole posts when they were on something that didn’t interest me, but since I knew you were using a tool that wouldn’t produce more flexible RSS, I didn’t complain.

“I never got complaints” isn’t a perfect measure of how good something is: for a while, when I was using Bloglines, I set my subscriptions to show “Summaries if Available” so I could quickly skim through, opening interesting posts in a browser tab. Since your feed didn’t have separate summaries and full content, I had to scroll past your whole posts when they were on something that didn’t interest me, but since I knew you were using a tool that wouldn’t produce more flexible RSS, I didn’t complain.

Works fine in Newsgator for outlook. Always has.

always has worked in Feed Demon right from day 1.

always has worked in Feed Demon right from day 1.

Feed has always worked great in RSS Bandit.