Tag Archives: links

Microblog auto-link differences

One reason I like microblogging on my own web site is that I can control the links and simple formatting. I’ve noticed lately that Twitter can’t consistently auto-link even certain domain names, for example.

This difference is illustrated well in a post I made this morning, which included timetable.fm and micro.blog. Twitter auto-linked the .blog but not the .fm. The cross-post to App.net auto-linked the .fm but not the .blog.

Here are the 3 versions:

Link differences screenshot

In the final screenshot — the original from my own site, from which the others were pushed out automatically — you can see how I’ve specifically linked the domain and phrases I wanted to. It’s a minor thing, but it just looks better when the author has a little control over the formatting. (And while I don’t use it here, my own short posts can contain text in bold or italics via Markdown, too.)

Inline links in microblog posts

When I was first trying to figure out how my microblog posts should look, I was thinking more like tweets and less like HTML. Eventually I settled on HTML for publishing and display, with Markdown for writing.

Here’s what a microblog post looks like in the timeline for my new web app:


You can compare that to how it looks when cross-posted to Twitter. It’s not exactly a fair comparison since the tweet was truncated, but it’s still incredible to me how much better these posts look if you allow inline links and some more characters.

Twitter and the cost of links

Federico Viticci covers the news that Twitter will expand from 140 characters to 10,000, nicknaming the feature Twitter Notes. His nickname is appropriate given this latest transformation to become more like Facebook, since Facebook’s Medium-like capability for long posts is also called Facebook Notes.

The tweets and blog commentary on this have really missed a key aspect and cause for concern, though. Many posts – including even my own first attempt – have focused on whether Twitter Notes would water down Twitter’s unique strength. They then conclude that it’s better to include a long-form text feature rather than the compromise hack of screenshot text and tweetstorms. Federico sums up this endorsement with the following:

“Unlike other recent additions to the service, I want to believe that third-party developers will be able to support the feature in their clients (Jack seems to suggest as much) and that the iPad won’t be left behind again. I may be disappointed when the day comes, but if done right (see Matthew’s points here) and as long as Twitter Notes are intended as attachments for regular tweets with real text, I don’t see why I would be against them.”

Here’s why this matters, and it gets back to my post last week about the hyperlink. Closed platforms want to trap all activity, not send it out. The danger in Twitter Notes isn’t that they will replace textshots, it’s that they will replace external blogs.

For all of Twitter’s problems, at least right now most of the good writing we see on Twitter is actually linked out to external blogs (and yes, increasingly Medium posts). To shift that to be stored more on Twitter itself would be a setback for the open web. It would slowly train a new generation of timeline surfers to prefer Twitter-hosted content instead of blogs.

I wrote the above in draft form, and then later saw Ben Thompson’s daily update about the Twitter news. His take is the first I had seen that directly covered the issues of linking, even suggesting that no one really clicks on links anymore. But while he’s worried about Twitter from a business standpoint, I’m more worried about the attack on the web.

Ben also mentioned the clever trick Jack Dorsey used in writing his response as a textshot. Daniel Jalkut pointed out the same thing in the latest Core Intuition. Jack could have posted it to a blog, or to Medium, but he deliberately picked the worst way to work around Twitter’s current 140-character limit, to underscore his point.

Now, Will Oremus writes for Slate about the potential new Twitter walled garden:

“What’s really changing here, then, is not the length of the tweet. It’s where that link at the bottom takes you when you click on it—or, rather, where it doesn’t take you. Instead of funneling traffic to blogs, news sites, and other sites around the Web, the ‘read more’ button will keep you playing in Twitter’s own garden.”

I know we can’t rewind the clock to the heyday of the blogosphere. But we can still do more. More to encourage bloggers, more to spread awareness about how the web is supposed to work, and more to value open APIs. I think it starts with 2 things:

  • Build tools for independent microblogging, to make blogging just as easy as tweeting. I’m trying to do this.
  • Make the web faster, so the cost of clicking on a link goes down. Google’s helping this with AMP.

I was encouraged when I saw that Known had added support for AMP. They have their doubts about AMP, but at least they were quick to try it. From the Known blog:

“We’ve shipped support for AMP because we see potential here, and recognize that something should be done to improve the experience of loading independently-published content on the web. But attempting to bake certain businesses into a web standard is a malformed idea that is doomed to fail. If this is not corrected in future versions of the specification, we will withdraw support.”

Maybe AMP ends up being too ad-friendly to become a good standard. I don’t know. But if so, we’ll move to the next idea, because the web has to be faster. Slow pages are like a disease for links.

Anyone with a blog should be concerned about what could happen with Twitter’s 10,000-character push. We won’t feel the effects right away, but years from now it will matter. We should do more not just to promote blogs and writing on the open web, but to also make it easier for Twitter alternatives to exist through independent microblogging.

Microblog links

Brent Simmons points to my post on microblogs and asks:

“Is the web we lost gone forever? Was it a brief golden age before the rise of Facebook and Twitter and The Algorithms of Engagement?”

But he quickly follows with an alternate view: that it’s a blip and we’ll get back on track. And that’s what I believe.

Instead of accepting a common opinion that Twitter is slowly replacing RSS readers, we should flip that around. What kind of changes could be made to RSS readers to embrace microblogging and make Twitter itself less important? Because once we do that, we get back control of our own short-form content and at the same time encourage open tools that will survive independent of whatever happens with Twitter and Facebook in the future.

I received some other great feedback about defining what it means to be a microblog post. One question that I didn’t address is links. Noah Read writes:

“It has consistently annoyed me that Twitter and App.net’s links count against my character count. It seems to run counter to what I love about microblogging, carefully chosen words communicating a succinct idea. I often have a pretty good tweet composed and then I paste in the link to a site or image and have to rework the whole thing.”

And David Ely says that a microblog post…

“Contains a single thought, a link with short commentary, or a photo with a caption.”

Whereas a full blog post would often contain multiple links. Certainly a lot of what is posted to Twitter and Facebook is just a single link with short commentary.

I also noticed recently that Dave Winer’s Radio3 includes links in the text when tweeting, but in the RSS feed the text and the link are split out. The URL goes in the RSS item’s link tag. While this is easy enough to support in tools, it’s surprising if you consider the link part of the content, not metadata. (I also expect inline HTML links to become even more common.)

Wii Transfer takes over internet

Okay, not really. But this has been a crazy and surprising week for my “little” application, Wii Transfer. Putting 8 hours each day into “VitalSource”:http://www.vitalsource.com/ (I have a post coming about that tomorrow, by the way) and then juggling home responsibilities, putting out various other fires, and sitting down to work on Wii Transfer until 3am is just not healthy.

Luckily I slept great last night and took a 3-hour nap today. So time to blog again. :-)

Over a week ago I released Wii Transfer 2.0 and made a big mistake, and since I’ve been programming for the Mac for over a dozen years now, I really should know better. It was buggy. And not just a few minor cosmetic problems, but at least two serious crashers. I simply had not tested enough. It’s difficult (sometimes impossible) to regain a user’s trust after their first experience with an application is a bad one, so I got to work that weekend fixing problems and releasing beta builds to customers to get a few extra eyes on the software.

Then Monday came, and all hell broke loose.

Links from “Daring Fireball”:http://daringfireball.net/linked/2007/january#mon-22-wii_transfer, “Ranchero”:http://www.ranchero.com/?comments=1&postid=1523, and “The Unofficial Apple Weblog”:http://www.tuaw.com/2007/01/22/wii-transfer-2-0/ were followed by “Jostiq”:http://joystiq.com/2007/01/25/wii-transfer-for-mac-os-turns-your-wii-into-a-media-center-but/, “Infendo”:http://infendo.com/2007/01/wii-media-center-software.html, “4 Color Rebellion”:http://4colorrebellion.com/archives/2007/01/22/wii-transfer-for-mac-reaches-20/, and a bunch of others. Ironically one of the only gaming sites I read that never linked to Wii Transfer was the only one I had actually sent an announcement to (“GoNintendo”:http://www.gonintendo.com/). Traffic and sales were way up (“here’s a Mint screenshot”:http://www.manton.org/images/2007/riverfold_mint.png from one day last week).

But meanwhile, the application was just not that stable. I started rewriting most of the web server inside Wii Transfer and fixing lots of issues with iTunes and iPhoto libraries stored on external drives. Then I made my second mistake: I added a feature (album cover artwork!). Obviously, adding a feature in the middle of bug fixes just delays the original fixes and introduces new problems.

I also quickly realized how many things could go wrong with how music and picture sharing works. It relies on the Nintendo Wii and your Mac being on the same local network. Because Wii Transfer pings a bookmark server to register your IP address, you also have to make sure the app picks the right IP if your Mac is on both ethernet and wireless networks. Worse, many people have the Mac OS X built-in firewall enabled, so users are required to manually open up port 9000.

At one point on Tuesday when sales were coming in, every time I received a PayPal notification email I literally groaned. “Stop buying this software until I can make it work reliably,” I would say to the computer. The thing that got me through was that all customers who sent in support email were extremely helpful and patient. The other good news is that with version 2.1.1, it’s looking pretty solid, and the next update should wrap up any remaining fringe issues.

To everyone who gave Wii Transfer a try, thanks! I think you’ll like what comes next.