Saturday, December 06, 2008

Why are news readers so buggy?

Well, I know that the answer is: because they try to use too much JavaScript and AJAX and all the minor differences between how to do stuff and reference objects in the browsers make them very prone to being buggy. Also, sometimes it's not only how buggy it is, but what you have to do to get around some flawed design decisions, which makes the experience even worse. But let's get to some examples, shall we?

Google Reader

I was very happy with Google Reader until one day that I think Google was having some problems with their servers and opening any links was very painful. That's when I realized that it was painful because all links are actually links to their tracking page that redirect you to the site you want to get (and, by the way, Gmail does the same thing, but let's not talk about Google's evil ways today) and their tracking service was apparently having a slow day. I don't really mind people tracking what I'm doing, because I know that in the end it actually will help other people find things. What I mind is that it was interrupting my reading experience. So I stopped using Google Reader.

But today I went back to see the changes that they made and started just clearing all the lists marking everything as read when I saw an interesting bug: when I read all posts on one feed that was inside a folder the count in the feed went to zero but the folder was still showing that I had unread posts. I wished I took a screenshot of it, but after some time there seems to be some "refresh" task that cleaned it up.

Bloglines Beta

I decided on using the Bloglines Beta because I was too used to the Google Reader model in which things are marked as read only when you scroll to them. The classic Bloglines marks everything as read automatically as you select the feeds to read.

More than just bugs, Bloglines Beta had a very simple annoyance: it wouldn't refresh the feed you are reading. You had to navigate out and back into the feed. That took a lot of time, considering that I read my feeds once a day or once every two days and I usually have on the order of 500 entries to read. When I get to part of them, there are new waiting.

A very well known bug of Bloglines though, and what made me move away from it, was the fact that it just didn't refresh some of my feeds. The blog itself had new messages, Google Reader would show them, but Bloglines wouldn't. It's something that has been observed by other people, but it was time to move on...

NewsGator

My latest trial is NewsGator. It's a pretty simple system that does not try all the fancy AJAX things that the previous two had: it does not try to see what you are reading to auto mark them as read. You have to go to the end of a page and click on "mark all posts on this page as read". It took me some time to get used to it, but in the end it made it ok to set waypoints on reading entries instead of one at a time. It was way lighter-weight, so it made me happy.

But not everything is great in this world here. NewsGator seems to have a very odd bug on the counts of unread emails on the summary on the left. Sometimes I mark things as read, but it does not get in sync with the list on the left. At least in this case there is a button "Refresh Feed List" that fixes it. The only problem with it is that it takes you to a "feed discovery" page, instead of keeping you in the context of where you were. Quite annoying.

The conclusion is that there seems to be no online RSS reader that doesn't annoy me in one way or another. I'll keep using NewsGator for a little longer, but let's see what happens.

Oh, all tests above were done on my Mac using Firefox 3. I haven't really tried other browsers.

2 comments:

Unknown said...

I use Google Reader and have had occasional bugginess with it as well. It stems from its use of AJAX, which I find occasionally gets out of sync with what the server thinks. For example: AJAX will decrement some count by three, when the server thinks it is only 2.

Have you tried any desktop based ones? I don't know what's available for a Mac, but it might help some of the issues you are having...

Michel said...

I have tried desktop ones before. The problem is that there is a 5-10% of the time that I want to read my feeds from a different computer and that's what took me away from them.

Also, in general, performance to features-wise, online versions always seem to be better. Which something quite interesting about current online technologies. Things like Search seem to be quite resource intensive and work better in the server side, or, as some people would call it, "in the cloud".

But I have to agree that so far the only problems I've seen with desktop readers was rendering bugs for HTML and odd bugs that would crash the application. No counting errors.