Life is grand

Corank made me giddy an hour or so ago. It is Digg but I get to pick the 14 year olds I associate with. Digg filtered through a contacts network.

Then the reality started to seep in and the giddiness turned to sadness. Corank is close but my cigar remains unlit, the fat lady is still in her chamber. The first problem is that when you bookmark something the tag field is limited to a few characters. I enter a lot of tags when I bookmark a URL, it helps me find things later. I can’t work as expected with Corank.

The other problem is the filtering is all or nothing. Only items bookmarked by people I specifically add as sources make it onto my front-page. I’d prefer if Corank had a Digg-like front-page and that the “sources” I subscribe to only influence the front-page. That way I can keep a finger on the pulse of the internet but make sure that people I think are smart get to influence that pulse. They must not define the pulse. As Corank stands I’d have to add 50 million people to my list making sure I exclude the 14 year olds. At the least Corank should flip it around.

So it seems the whole Snap Preview outcry is having some effect. TechCrunch has modified how Snap Preview works on their site. Before, any external link would pop-up a little window previewing the site on hover. Now they have moved that feature by inserting an icon to the right of any external links and showing the pop-up on hover of that icon.

Is it an improvement? Yes. Is peace and tranquility restored to the interwebs? No. You can’t take a well intended but bad idea, shift it ten pixels to the right and declare the problem fixed. The idea was bad in the first place; admit it, kill the idea and move on. Now when I read paragraphs on TechCrunch the flow of words is broken. Typography is important. Readability is important. Preview links are not. The icon is meant to be subtle in contrast (it is off-white) but it still takes up 16 pixels of width which throws off gaps between words which is a big no-no in typography and readability.

Ah well. You can still disable the damned thing, until you clear your cookies and have to go disable it again.

FeedDemon’s solution to a tough problem is elegant and simple. The problem is how does one find popular items within the feeds one cares about. Other systems are applying all sorts of voting systems, tagging pools, text analysis and “machine learning” which is all well and good but often doesn’t work quite the way one expects.

FeedDemon says “count the number of referrals and list from highest to lowest.” Simple. That it is applied just to the feeds you are subscribed to makes it so much more powerful. You have already gone to the trouble of selecting the feeds you care about so why throw that out in favour of some complex “blogosphere wide” analysis engine that gets it wrong a good deal of the time.

evhead makes some good points about how pageviews are going the way of hits in the grand game of website analytics:

Pageviews were never a great measure of popularity…

So what’s a better measurement? Good question. Like many good questions, the answer is “it depends.” If you’re talking about what’s important to pay attention to on your own site, you have to determine what your primary success criteria are and measure that as best you can. For some sites, that could be subscribers, or paying users, or revenue, or widgets deployed, or files uploaded, or what have you. It may even be pageviews.

At Blogger, we determined that our most critical metric was number of posts. An increase in posts meant that people were not just creating blogs, but updating them, and more posts would drive more readership, which would drive more users, which would drive more posts.

Essentially it depends on your site and your intentions. A search site has very different measures of success to a site such as MySpace or Flickr or a web-app like Megite.

evhead goes on to say that someone should tackle the tough job of providing a single metric culled from various data. I doubt one metric can be found and doubt it is even worth it. The measure of success between different vertical markets is too great for one metric to be anything more than vainglory. It is largely worthless knowing that MySpace is on the march to beating Yahoo! in pageviews/traffic/something. Two different sites.

Looking above you can see my latest Google Analytics report. It is pretty and has some interesting data points but the interest is the same as a funny video on YouTube. Interesting but useless.

Google Anlytics isn’t useless. For many sites it is a power tool but in my case and the case of many personal bloggers it is pretty pictures and nothing more. My measure of success is connections with readers. Readers who email me, readers who comment and readers who mention my posts in theirs. Not pageviews or sources, not search keywords or length of time spent.

A metric per vertical market would be useful, for advertisers and investors and even us users, but one, grand, over-engineered metric isn’t worth our time.

A friend has a really good idea. Allow users to do a full-text search across the sites they have bookmarked. Wether that be your Firefox/IE bookmarks or your del.icio.us/Blinklist links. With systems such as Rollyo this would not be hard to do.

Rollyo is a good idea but I never use it as the link between the sites I visit and value and the sites in my custom Rollyo search engine is not automatically updated. If Rollyo would monitor my Blinklist links adding sites as I blink them I’d have good reason to use Rollyo. Integration into Firefox(through the search textbox top right) or Blinklist would be ideal.

(Rollyo does have a “bookmark importer” tool but it isn’t automatic and only works with your Firefox bookmarks, not del.icio.us and co.)

Tags have, without a doubt, made my life better. My photos are better organised, my links are readily accesible, my posts categorised and I am labelling my emails like there is no tomorrow. I’m waiting for file systems to get in on the act and then life is complete. At least eLife is.

I want to see tags go a bit further though. I want their simplicty retained but their power expanded.

Many a time have I uploaded a photo to Flickr and wanted to have that photo automatically emailed to a group of friends based on a tag. Everyday I link URLs on del.icio.us and Blinklist that I want automatically sent to a group of fellow linkers. For example if I link “http://www.evoca.com” I would tag it with some standard metadata tags (e.g. “podcasting, audio, phonecasting, website, service”.) The next tag I would want to add is “group:podcasters”. The system would see the “group” ActionTag and send some kind of notification to the podcasters group.

Another ActionTag might be “email:bill@gates.com” which does exactly what it says on the tin (with some measures to prevent email abuse of course.) Maybe you want pingomatic pinged when you add a photo in which case “ping:pingomatic.com” is added as a tag. How about tying in text services “sms:+353868968944″ or API calls “api:basecamphq.com/api/tasks/add?t=evoca.com”. The last one is a bit complicated but having the capability would be very interesting.

I don’t want to complicate tags so in no way should an ActionTag aware system put any conditions on other tags. Listing tags for an item should either exclude or differentiate ActionTags in some way. ActionTags need to be simple to enter as well, maybe a colon to separate the key from the value (though that impacts a value of an URL.)

Mainly I want to automate my photos, links, emails and so on with simple commands.

We recently started an informal User Interface group here at the TSSG to create a UI and design resource along with generating an awareness of the importance of design. The group consists of a bunch of interested people from different projects spread throughout the TSSG building. Some I work with on a daily basis and others I haven’t actually met face-to-face yet.

On setting the group up I ran into a lack of collaboration tools. The mailing list (MailMan Listserv) and blog (WordPress) are adequately served but beyond that there is not much that fits.

Link sharing tools such as del.icio.us and BlinkList work fine for individuals but aren’t setup for groups. We could all sign-up for del.icio.us accounts and then tag relevant links with “tssguigroup” but I already have a link stream and don’t want to pollute it. Polluting your tag list isn’t a good idea either. I also want a system I can host internally and customise to our needs. Anybody got the del.icio.us source code to share? Didn’t think so.

A link directory would also be very useful. A directory with tags but also categories that allows us to build up a collection of useful links to resources. This is different to a stream of links as needed above.

Another requirement is sharing of images for peer review. For example if Edel wants to get some peer review of a new design then she should be able to post up a set of images, notify us all and we can then weigh in with our comments. Flickr is not setup for this. Once again I don’t want to pollute my Flickr stream with work images and there is the problem of keeping it all private. A private Flickr Group doesn’t cut it.

Then there are books. We all have design books on our desks which we are willing to lend out. How do we collect the books together, making the others aware of them and manage lending of them?

I am also looking for a good forum system which lets us post questions and discuss issues. What has happened to forum development in this day and age of blogs and wikis? There is still a need.

Speaking of blogs while WordPress works for our central blog I would love a system that allowed multiple blogs within a company. Setting up copies of WordPress for each person who wants a blog in the group is not ideal. Members should be able to just hit a “Create Blog” button and it gets added to a directory of blogs. Enterprise blogging! Tools?

I am looking forward to BlogBridge’s Library system as this would be great for getting group members up to speed with useful design feeds. For now though it will have to be a manually updated OPML file.

If you know of any tools that fit what I need then please speak up. I am not looking for a system that does it all, those generally don’t work.