Tag Archives: davewiner

The algorithm has ruined Facebook

Dave Winer writes today about how because of the way the Facebook news feed works, sometimes you never seem to hear from friends again because they’re demoted by the algorithm. Your friends are posting, but you never see what they’re saying. Also:

For other people you are a missing person. You being the person who dutifully informs all your Facebook friends of what’s going on in your life. You, the friend they never seem to think of. No surprise they’re not thinking of you. The Algorithm decided you don’t count.

If you want to see this in action, visit Facebook in a web browser and see what it shows you. Don’t scroll or click anything, just wait a few seconds and hit reload. Then hit reload again. And again. Each time you’re presented with a completely different view of what’s important. It’s unusable.

Dave Winer on title-less posts

Dave Winer posted today about NetNewsWire needing better support for title-less feed items:

These items have no titles for artistic reasons. The author did not put them there. You, as a software developer, are not entitled to add them (haha that’s a pun).

I agree with Dave on this. Titles are clearly optional in the RSS 2.0 spec. The fix for the “Untitled” text that some feed readers use isn’t for authors to add titles where they aren’t needed, it’s for the UI in feed readers to improve so that they gracefully handle title-less posts.

(And this is not to pick on NetNewsWire. I’ve seen other apps and feed syncing services with the same assumption about titles.)

When I wrote about defining a microblog post, blank or missing titles was one of the fundamental points. If we want to have blogging software that’s as easy to use as a modern social network, titles can’t be required.

I’m hopeful that as feed readers adopt JSON Feed, developers will dust off their older code for feeds and make improvements for title-less RSS items as well. This is why we highlighted microblogging as a use case in the JSON Feed spec.

Medium.com updates

Ev Williams announced a batch of new Medium features recently:

“There’s always another level. Another level of polish and power in our product. Another level of breadth to our content. Another level of dialogue and discussion. And another level of progress. Today, we are announcing a slew of updates to bring Medium to the next level and in the process make it more powerful, more fun, more democratic, and more essential.”

Those updates include new mobile apps, @-mention support, a publishing API, and editor improvements. There’s also a new logo. (I know they put a lot of thought into this, and it’s a strong idea, but to me the logo’s design is so clever it’s actually kind of distracting. A little more subtlety in how they’re using depth could improve future iterations.)

Daniel Jalkut blogs about what’s included (and what’s left out) in Medium’s new API:

“One of the most unique aspects to Medium’s API is the provision for specifying a canonical URL and license on a post being submitted to the service. The canonical URL refers to another web location that should be considered the original, or most authoritative version of a post, while the license designates whether the post’s copyright terms stipulate a post is sharable as public domain or under a particular Creative Commons license. These attributes together indicate that Medium expects and encourages users of the API to contribute content that is not intended to be exclusive to Medium.”

While I generally think the trend to centralized writing platforms is bad for the web, I’m happy to see these changes from Medium, especially the API and expanding custom domain support. Medium has grown very slowly and carefully. I expect we’ll see quicker iteration on these new features now that they’re officially out.

In the process of experimenting with Medium posting, Dave Winer shared his take on post title support:

“It seems they have arrived at what I think is the correct answer: posts can have titles or not, and the content system has to be prepared for either case. That’s where this blog was in 1999, before other blogging tools and Google Reader pushed the world toward requiring titles. And then Twitter came along not having titles at all, and the intersection between all the kinds of blog-consuming environments became almost empty.”

I’m very interested in this because microblogging shouldn’t include titles. While Medium is mostly traditional essays, clearly comments don’t need titles, and Medium’s quick-posting UI encourages short posts. I hope this approach will get more RSS readers to gracefully handle title-less posts.

Little Outliner

Yesterday Dave Winer and Kyle Shank launched Little Outliner, an impressive JavaScript outliner that uses HTML5 local storage. It’s also completely hosted on S3:

“Thanks to the W3C and to Werner Vogels (for persisting in getting the ability to access the root of a domain from an S3 bucket). As a result, we get unlimited scaling with zero investment. Consider this an endorsement for both innovations.”

I used Frontier a lot back in the earlier days of the web, so I’m always looking out for what Dave does next. It’ll be fun to see what they build on top of this.

Deprecation mentality

Today, Twitter starts “shutting down basic authentication”:http://countdowntooauth.com for the Twitter API. One of my favorite Twitter clients, Birdfeed, will be allowed fewer and fewer requests until finally at the end of the month it stops working. Likewise for Birdhouse and Twitterrific 2. And the same for my “Wii Codes”:http://wiitransfer.com/codes/ site, until I have a chance to update it.

“Dave Winer wrote a fairly negative essay”:http://www.scripting.com/stories/2010/04/26/theToxicCoralReef.html a few months ago on this so-called OAuthcalypse:

“When Twitter breaks all the apps in the OAuthcalypse, they will break all of mine, and I have no intention of fixing them. I don’t expect anyone to care. But what you should think about is how many of the Twitter apps that you do care about will break and how many of them will say the hell with it? And how many of them will be around for the next time Twitter breaks everything, because that’s certainly coming unless Twitter develops some kind of philosophy about itself as a developer platform.”

I didn’t want to agree with him at first — I’m a big fan of nearly everything Twitter does — but it’s a fair question to ask whether backwards compatibility is getting the attention it deserves. Software moves fast, but this kind of thing hurts users, not just developers.

In the desktop world, OS APIs are unlikely to change so severely, and if they do you always have the option to run an older version of the OS or app indefinitely. For web services, though, you can’t keep an older copy of the internet around. Web apps are forced upgrades.

I’m not sure there’s a solution to any of this. It’s just part of tech progress, like moving data from old floppy disks to CDs to hard drives to the cloud. But it’s a bummer when apps get left behind as APIs are obsoleted. Over-aggressive deprecation was common in the Rails world, and “I was not a fan”:http://www.manton.org/2009/01/rails_4_years_later.html.

So, here’s to the future, Twitter. Keep new API changes versioned and maintain the old stuff. If this OAuth switch is a one-time cost, developers can focus on what makes their apps unique instead of always playing catch-up.

Still digging

Wow, Dave Winer is applying for a new job:

“As I’ve been talking with people about this, it’s been hard for them to separate me from UserLand, but that’s what I want to do. I’m going to get a new job with a new title, and it’s going to be quite different from being the CEO of a commercial software company. It’s time to set my life in a new direction.”

Good luck, Dave. Still digging! :-)