A Consuming Experience

Sunday, 11 December 2005

Feeds: how to enable autodiscovery for both full and partial feeds

In my previous post on how to offer your readers more than one type of feed to subscribe to as they choose (full, excerpts or headlines only), I said I'd explain for those who don't know how to implement autodiscovery for their feeds, and I also mentioned using just one of your multiple feeds for autodiscovery.

For those who don't know, this post will cover how to insert code in your blog template to enable autodiscovery - which is a way to enable your visitors' feed readers etc. to automatically pick up your feed URL from your blog URL (if you use Feedburner the feed URL will be quite different from your Blogspot or other blog provider's feed URL). And it's actually quite easy to enable autodiscovery for all your different types of feeds.

Most blogging platforms will automatically include in your template the autodiscovery code for the standard feed that's associated with your blog. But that just won't be good enough if you've:

customised your template heavily (e.g. by deleting the <$BlogMetaData$> tag in Blogger, which can cause all sorts of problems)
burned a Feedburner feed (see my previous post on why you'd want to, and how to do it), or
want to enable autodiscovery for several feeds, where you've implemented the option for your readers to subscribe for longer or shorter feeds.

You'll have to edit your template yourself, in that case. But it's pretty straightforward. Here's how:

1. Note down your feed URL (or URLs, if like me you've decided to offer multiple feeds to your subscribers). If you're using Feedburner make sure you've enabled SmartFeed (see my previous post).

2. In your blog template, insert the following code between the <head> and </head> tag (but before the <$BlogMetaData$> tag if you're on Blogger - it seems to work even if you leave that tag in, but it may confuse subscribers as the standard Blogger feed URL will be displayed there too, even though it'll be last on the list, so consider deleting that tag - but watch out for the associated problems if you do). Then save the changes and republish your blog.

Of course you should first change http://feeds.feedburner.com/ConsumingExperienceFull, http://feeds.feedburner.com/ConsumingExperience and http://feeds.feedburner.com/ConsumingExperienceHeadlines to the URLs of your full, excerpts and headlines feed respectively (if you only offer one or two types of feed just delete the line for the feed you don't need), and also change "A Consuming Experience" after the "title=" to the name of your own blog, in each case, in your own template.

The "title" bit (e.g. title="A Consuming Experience (full feed)") is the description for the feed which your visitors will see in their feed reader, so the "title" in each line of the code is slightly different to match the different type of feed. Change "A Consuming Experience (full feed)" etc further if you like, but obviously it's most helpful to your readers if you make the description, well, descriptive - not "My feed" etc).

Here's the code:

<link rel="alternate" type="application/rss+xml" title="A Consuming Experience (full feed)" href="http://feeds.feedburner.com/ConsumingExperienceFull" />
<link rel="alternate" type="application/rss+xml" title="A Consuming Experience (excerpts feed)" href="http://feeds.feedburner.com/ConsumingExperience" />
<link rel="alternate" type="application/rss+xml" title="A Consuming Experience (headlines only feed)" href="http://feeds.feedburner.com/ConsumingExperienceHeadlines" />

and this is what it looks like when someone clicks the Live Bookmark icon at the end of the address bar in Firefox 1.5:

3. Some gotchas and hints:

include the code for the different feed options one after the other as above, with no other code like script tags in between them (some feed readers may not pick up the later URLs if there is other code between)
not all feed readers will show the "title" of the feed (e.g. Feed Demon doesn't, it just shows the feed URLs), so it's best to make the URLs of your burned feeds descriptive to help your readers (like with "Full" at the end of the URL in the case of my full feed)
put the line for the feed you want to be your "default" feed at the top, for readers and other things which don't recognise different options and just pick up the first one in your template (I'd suggest putting the line for the full feed first because the full feed is best for the search engines etc, see my previous post). It'll also help if you're on Blogger and decided to leave in the <$BlogMetaData$> tag (you can even change the "title=" for the first line to add something like "Pick this feed if in doubt", if you wish! Though it will obviously lengthen the title).

[Added later] Note: I don't know enough yet about how the search engines etc. which pick up and use your feed from your blog work. I believe and hope they'll just pick up the first feed URL in your template/page, so I've added all three options to my template - but there's no guarantee they won't choke instead if they find more than one.

So if it's more important for you that the search engines pick up your full feed (I've outlined the importance of that in my previous post) than that feed readers show more than one autodiscovery option, it may be safest to just include ONE autodiscovery line, for your full feed. (Obviously for those search engines where you can manually submit your feed URL that's not a problem, just submit the full feed URL - it's the ones which autodiscover your feed that I'm not sure about.)

If anyone knows for sure how the search engines work on this front, I'd be interested to hear from you.

Technorati Tags: feed, feeds, newsfeed, newsfeeds, autodiscovery, feed autodiscovery, RSS, Atom, full feeds, RSS feeds, partial feeds, Feedburner, Blogger, Blogger.com, Blogspot, template hacks, Javascript, hacks, blog, blogs, blogging, blogging tools, tips, tricks, Magical Sheep, Improbulus, A Consuming Experience, Consuming Experience

Technorati: how to check when Technorati last indexed your blog

The Technorati blog have announced that Technorati have updated their ping page, as Stowe Boyd has noted.

To "ping" a search engine means to ask them to revisit your blog or webpage to re-index it so it can be searched for on their site. You'd usually do this because your content has changed and normally on receiving a ping they will send a spider or bot to "crawl" your pages.

To me the most useful change is this: if you've signed up and logged in to Technorati and have claimed your blog on Technorati, when you go to their ping page you will see that against the name of each claimed blog it gives the number of days or hours ago that Technorati last indexed or crawled your blog, like this:

This is obviously a good way to check that Technorati have indeed been indexing your blog regularly. (But remember you won't see that info unless you've both signed in AND claimed your blog.)

You'll see that there's also now an "Update ping" button to manually ping Technorati for each of your claimed blogs - so they've made it easier to do a manual ping from their ping page, or alternatively indeed you can do that from the Blogs tab (visible once you're logged in, which summarises info about your claimed blogs) as the button is there too:

However, you really don't need to ping Technorati manually for many blogging platforms. It's much easier to set things up so that Technorati is automatically pinged whenever you publish a new post, and Technorati have provided step by step instructions on how to configure this for major platforms like:

In Blogger, at least, this should be automatically set up for you anyway so you shouldn't need to do anything, but if you want to you could check your Blogger ping configuration just in case, and change it if necessary.

My wishlist item? It's great that Technorati now provide this info, but it would be even better if Technorati included the "Updated X days/hours ago" information on the Blogs page of logged in members, rather than us having to visit the ping page to see it. That would be the logical place for it, to me. (The ping page link is also not the easiest to find as it's oddly enough hidden at the bottom of Technorati's pages in the same line as the copyright and other info, rather than with their main links).

And while it's good that they are regularly indexing, I wish they would fix the problems with their tag pages (or maybe tags indexing or tags database), which clearly people are still experiencing - I've found myself that my post on how to offer different lengths of feed to your subscribers isn't showing up on their tag pages though it's clearly on their index. I don't know if it's because I included code in that post, but some guidance as to what can break their system would be helpful so we know what to avoid.

Technorati Tags: Technorati, ping, ping Technorati, search, searching, search engines, indexing, blog, blogs, blogging, blogging tools, blog tools, Improbulus, A Consuming Experience, Consuming Experience

London Underground tube map by travel time

For those who've not encountered it yet, Oskar Karlin's Tube map measured by travel time is very interesting and useful - an ingenious and well-executed idea. You can download it in PDF - there are different versions, that's the best one in my view.

It's obvious when you see it, though harder to describe in words. He's overlaid onto a Tube map "contour lines" based, not on the traditional height above sea level, but on "distance" in terms of travel time (in 5 minute intervals - 5 minutes, 10 minutes, 15 minutes etc) from Elephant & Castle Tube station. Very helpful for figuring out rough travel times to different stations. Obviously more useful if you start from the Elephant than elsewhere! It was part of a student project and the Elephant was where his college was based, hence that starting point.

As he says in his blog post, Mr Karlin has capitalised on the key fact that today time is the scarcest commodity for many of us, and that when it comes to travelling people generally think in terms of travel time, not distance. I certainly do: I won't take on a weekly commitment if the venue is more than an hour's travel away for me, irrespective of physical distance (in London, for me at least, an hour each way is about the average "distance" in terms of time for most of my regular commitments). But if I have to go somewhere only once a month, I might be OK with up to an hour and a half of travel time, etc.

Some caveats. The map was drawn around the end of 2003 so it may be a bit out of date. As mentioned, it only gives the times from the Elephant. Mr Karlin also seems to have estimated the times himself, it's not clear exactly how (but if he did the travelling it must have taken him forever). Plus, for reasons he's explained (logical for him maybe, but sadly less so for users of the map!), he hasn't followed the traditional colours in his map e.g. red for Central Line. He has put up a version with the usual colours but unfortunately that doesn't include the "contour lines" which are what makes it useful. Perhaps if there is more demand he'll be kind enough to produce one...

I'd also join in with those who would like:

a map with the Tube lines matching the original colours e.g. yellow for Circle Line
incorporation of official travel times (actual ones, not optimistic London Underground estimates!), with realtime updating of the database for live up to the minute travel times (what with near constant signal failures and other delays on various lines etc) - or, as Kosso has suggested in a comment, harness the power of people, the Net and aggregation ("social timemarking", perhaps, to coin a phrase): " I can see a way where I could use OPML with a suggested ‘time’ attribute to store and display this data on their maps. It wouldn’t be too difficult to provide an interface for people to add to this data, aggregating different travellers’ experiences and the time it took to get from A to B."
an interactive version, based on a database, where you can (say) enter your chosen start and end stations, or better still (as Bill commented) generate a map centered on your chosen station on request
ideally something which factors in interchange time.

I hope that London Underground and other travel networks take this concept on board and design and provide maps, ideally interactive ones, like this.