I’ve moved to Chrome on the Mac

September 19th, 2011 § Comments Off § permalink

Although Chrome’s the best browser for Windows, I’ve favored Safari over Chrome on the Mac until a couple of weeks ago. Safari looks better and and is easier to use, in my opinion. But Chrome now works better.

I thought I was the only person experiencing the spinning beach ball when leaving lots of tabs and web apps open. I switched to Chrome on the Mac a couple of weeks ago, and there’s no question it performs far better for this purpose.

Tim Bray [via Gruber] notes the same solution:

In recent releases, Safari has been re-architected, with some of the work farmed out to a thing called “WebProcess”. This doesn’t seem to be working out that well. Specifically, I note that:

[...]

To be fair, part of the problem is the proliferation of Web apps that are extremely heavyweight in terms of the amount of JavaScript code sloshing around in the background. A notable example is Google+. Even sprucing up this blog’s typography has involved quite a bit of industrial-strength JavaScript. But that just seems to be what a modern Web browser has to suck it up and deal with.

Have we reached peak Patch?

September 2nd, 2011 § Comments Off § permalink

Patch traffic and page growth appears to have flattened out. I’ve been following the rise of Patch traffic on Quantcast for about a year, and have been impressed with their steady linear growth. Until now.

Even accounting for summer doldrums, this looks like they’ve peaked.

Patch 110902

Meanwhile, Huffington Post, a more mature AOL news property, grew at pretty healthy rate over of the summer, even ignoring the fact that Quantcast’s chart doesn’t start at zero.

Huffingtonpost 110902

I have mixed feelings about Patch. It is better than nothing — and nothing is what many communities were getting before Patch. But it doesn’t exactly break my heart to see this corporate competitor to hundreds of locally-owned community news sites hit its peak.

Not all local retailers reside in big boxes

June 30th, 2011 § Comments Off § permalink

This week, Amazon.com dumped its California affiliates (including me) because it was unwilling to collect sales tax in the state.

I’d say it’s time for Amazon to grow up and be a good citizen, but that’s years overdue. Danny Sullivan has an excellent analysis of the issues, both economic and ethical.

There’s no question that Amazon is using what amounts to a loophole to gain unfair advantage over local retailers (not all of whom reside in big boxes).

I’ve been an Amazon affiliate for years, but never got around to posting the links on my community news site, mainly because I live in a small town, and I support my local retailers. I also buy a ton of stuff from Amazon, but I didn’t want to actively support the gutting of my downtown. To say nothing of the gutting of social services, state parks, schools, and quality of life resulting from the state’s tax drought.

I have no sympathy for all those affiliates chasing “passive” income. We’re seeing plenty of pro-Amazon of comments from those creepy libertarian nihilists. I do have a lot of sympathy for those affiliates who work hard to provide their readers with real value and are trying to make a go of it online. But they should be looking to Amazon for relief, not Sacramento.

RIM CEO’s BBC interview evokes SNL skit

April 14th, 2011 § Comments Off § permalink

“SEOMimic” attacks comment whitelists

March 2nd, 2011 § Comments Off § permalink

I just saw a new attack on my member whitelist, in what may be a new wave in comment spam designed to attack sites that use member whitelists to control posting.

On Coastsider, my community news site, all commenters must verified before their comments are released and their accounts are taken off pre-moderation. This is relatively easy for me, since my service area is small and I ask new registrants for the name of the community they live in. That’s more work to fake than most comment spammers have time for. They don’t even try, usually using nonexistent places, such as “farm road” and “western trench”.

I verify most new members if they post a reasonable comment on a entry. Today I had a registration from a new user who left a comment that could have only been left by an informed local:

Tough to imagine the county of the present doing anything right in this part of Fitzgerald, where they can’t even locate the California Coastal Trail where it has already served well on an ad hoc basis for the last decade and a half.

No wonder: He lifted that comment from one of the informed locals who had already left a comment on that story.

I wouldn’t have figured that out if “Tom Smith” hadn’t also left a comment in Chinese on another story, and used the email address “seomimic@yahoo.com” and the real — but unlocal — community of “UK”.

I doubt SEOMimic is going to be as careless the next time it shows up on my site — or yours.

I get plenty of comment spammers on Coastsider, but this is the first attack of this nature I’ve seen. My guess is I’ll be seeing more — and more sophisticated — SEOMimic attacks in the future.

WordPress succeeded by going where the money wasn’t

February 10th, 2011 § Comments Off § permalink

How did WordPress win over Movable Type? Majordomo has a good review of Six Apart’s strategic weaknesses and errors, and how WordPress/Automattic exploited each of them.

In my opinion, Six Apart’s biggest error was its choice of focusing on the enterprise market just as blogging was poised to go mainstream. This was an understandable mistake. They had investors to satisfy (not necessarily a bad thing) and there was no clear path to profitability for a software platform in the mass blogging market.

This lead directly to the Six Apart’s disastrous change in their license, which created WordPress’s moment of opportunity.

I was a happy Movable Type user at the time, but I was moving in other directions. I adopted Expression Engine for my next site, because MT didn’t have the membership management tools I needed. By the time I was ready to refocus my efforts on my MT site (MediaSavvy) in 2009, it was actually easier (by which I mean “possible”) to move it from an old version to MT to the latest WordPress than it was to upgrade to the latest version of Movable Type. Things had changed a lot in four years and the game was over.

The next time you justify a strategic decision by saying “That’s where money is”, you should remember that Willie Sutton was not a brilliant strategist.

Huffington Post finds its sweet spot

February 8th, 2011 § Comments Off § permalink

Written in reply to a friend who asked for my thoughts on AOL/HuffPo.

My first thought on hearing this was “I hope they got cash, because I wouldn’t want to be a long-term investor in AOL.” Of course, they did.

I think HuffPo is a boring collection of gas-bag blogs. I was stunned by the price and startled when I looked them up on quantcast: They’re huge and growing fast, but page views per reader are declining.

So, long term, there are two kinds of content: the AOL/Yahoo/Demand Media/Rupert’s Daily dross, and the good stuff. The first type comes with a (problematic) business model. The second type has never been much of a business, but we continue to make it. Everything in the middle (me-too content not produced at scale) has no reason to exist.

Speaking as a competitor to Patch, I hope they continue using their cash to buy stuff.

Netflix uses an API to get ready for new platforms

January 23rd, 2011 § Comments Off § permalink

Netflix is using an API, not to make their streaming services available to developers — which would cause all kinds of problems with their suppliers — and not for a handful of special projects.

Netflix has made their API a pillar of their multi-device distribution strategy. The other pillars are, suprisingly, HTML5 and Webkit (the open source Web browser engine). Instead of creating custom applications for each new device they want to be on, they’re porting Webkit to that device, if it’s not already there, and then building their custom application in HTML5.

The result, as I’m sure you know, is that Netflix is everywhere. Their customers can watch their movies on about 200 different devices. This video (from Mashery’s Business of API’s conference) changed the way I think about the impact of API’s on publisher strategy. I recommend watching starting at about 9 minutes into the video.

But their breadth of distribution is only one headline. The other headline is that they can rapidly change their appearance on all these devices quickly, with much less programming and a single platform: HTML + CSS + Javascript.

If Netflix can solve a distribution problem like this using standard Web technologies, publishers should reconsider non-standard solutions targeted at single platforms, such as iPads or iPhones. Right now, Apple is clearly dominant in non-PC distribution platforms, but we can expect rapid innovation from both hardware and software (Android, Windows, HP’s Palm, RIM, and more) manufacturers. Market fragmentation could become the norm in the near to mid term. And these new devices will be even better platforms for reading than for video.

Highly customized, device-specific “electronic magazines” and “digital newspapers” demo well and dominate the blog-driven news cycle, but they’re hardly the flexible platforms we’ll need in the next five years.

When mulitplatform distribution becomes the norm, we know that Netflix will be ready.

How many publishers will be ready?

Updating Facebook and Twitter from your blog

January 23rd, 2011 § Comments Off § permalink

It’s hardly a social strategy, but posting your site’s headlines to your Twitter feed and Facebook page is an important first step for a news site.

I’m pretty sure that most WordPress and Drupal users can do this with a plugin. Things aren’t so simple for users of Expression Engine or some other content management system. You may also want to use a third-party service if you’re reaching the limits of your current system. I’ve wrestled with this problem for year or two. I thought I’d share my experience and save you some time.

My first impulse, and probably yours as well, was to try get my Facebook page to pull in my feed. There’s a way to do it, but I defy you to discover it yourself, even though it’s accessible right there on the page. Once you know the non-obvious directions, it’s easy to do. But this is a very slow way to do an update. Facebook polls RSS feeds at long and seemingly random intervals. It can take hours for a new post to show up on your Facebook page. These days, this isn’t acceptable for J. Random Blogger, let alone a news site. And, of course, that doesn’t get your feed onto Twitter.

You need a third-party service to deliver your headlines as quickly and flexibly as your news cycle demands. Most of these services use the Pubsubhubbub protocol to know the moment you’ve updated your feed. I’m not going to go into a long explanation of how it works. I’m a little fuzzy on that to be honest. But, like the http protocol, you don’t need to know a lot to use it. Superfeedr will manage your push notifications for you. It won’t deliver your feeds to your social pages, but it will make possible for other services to do so more quickly.

Twitterfeed is designed will update social sites from your RSS feed. I found it to be reliable, nicely designed, flexible, and easy to use. It accepts Pubsubhubbub notifications. But its dashboard is a little buggy in my experience, and I’m not using it as much as I have in the past.

Ping.fm aims to be a switching yard for social updates. The good news is that it can you can update it immediately via email, allowing you to be more selective about which stories appear on your social feeds. The bad news is that (1) Ping.fm can’t take RSS feeds directly, making it harder to automate your workflow, and recommends Twitterfeed for RSS notifications. (2) I found the exact format of Ping.fm updates to my Facebook page to be unreliable — they changed over time. This resulted in a Facebook page that was a visual mess and occasionally just stupid-looking when Ping.fm (or Facebook) picked up the wrong image to display with a story. And (3) Ping also seems to have gone into a developmental hiatus since Seesmic bought them last March.

Dlvr.it feels like a more polished solution for posting your feeds. It accepts Pubsubhubbub notifications, is easy to use, has a flexible system for adding feeds and social destinations, provides excellent reports on items delivered and clicked, does a better job of formatting Facebook entries, and allows to to filter your posts based on their content. It can also deliver RSS output, which I used to buffer a friend’s RSS feed which my CMS started polling him too frequently.

All three services (Twitterfeed, Ping.fm, and Dlvr.it) look good and deal creatively with the problem of putting a simple, Web 2.0-style interface on creating complex many-to-many relationships. Ping.fm and Dlvr.it also offer API’s if you have the need and the skills to do something even more custom.

I recommend trying all three (using Superfeedr) However, I’ve found that Dlvr.it is the easiest, fastest, and most powerful way to update my social feeds from my news site.

Building an RSS aggregator with WordPress

January 21st, 2011 § Comments Off § permalink

I’ve been experimenting for years with how to create and structure community news feeds. Readers want to follow breaking news and relevant information as it becomes available — and not just on Facebook and Twitter. All this time, I’ve been looking for an easy and reliable way to merge a bunch of related RSS feeds into a single feed that would make this easier for my readers.

What seems like a simple problem turns out to be difficult to do well. The problem is that it’s an easy enough problem that everyone thinks they can solve it, so no one is willing to invest what it takes to create a quality implementation. I’ve tried a lot of methods short of writing my own aggregator, which have no interest in doing.

Speed is an issue for all aggregators, because they must recognize that a feed has been updated, read the feed, add it to the merged feed, and deliver the new feed to its subscribers. This can slow things down, because neither the aggregator nor the end user can poll too often. Also, many of these services do not scale very well and slow to a crawl as soon as they become popular.

A lot of sites aim to create merged feeds for you. About half those you’ll find on most lists via Google no longer exist. So how can you count on any of the others to be there in the long run, or to be fast and reliable enough to use in a production system?

Yahoo Pipes is ineffably cool, and a pretty good solution, but it’s slow, kind of obtuse, and unreliable. It doesn’t appear to be getting a lot of attention from Yahoo, which is a good sign you shouldn’t make it part of your infrastructure.

The Simplepie RSS parser has a nice-looking website and lots of links to interesting examples, but the project seems dead in the water. I was able to find a good example of an RSS aggregator that did most of what I was looking for, modified it to make it my own, and added it to my site. However, maintenance became a problem. First, the script slowed down my host. Shortly after that, it started to get aggressive in its reading of my friends’ sites. It was becoming clear that unless I wanted to go from being a publisher to a programmer, that rolling my own aggregator from Simplepie simply took too much work.

I finally realized I could solve this problem using WordPress, of all things. In a couple of hours, I was able to create a news aggregator using WordPress and a plugin called FeedWordPress.

My new aggregator can pull in headlines and summaries from multiple sites and deliver a combined feed. The aggregator itself isn’t intended to be a destination — right now I’m the only person who even knows where it is. It’s just a bit of infrastructure that works reliably and fast. I’m able to read the merged site’s feed with the magpie reader built into Expression Engine and deliver the results as a local news digest.

FeedWordPress has a terrible name and an ugly website, but it works really well and has plenty of options for managing the way you gather and display feeds. And it’s shareware. I sent in my donation.