Affiliate Me With Google Toolbar's AutoLink

The relesae of the new Google Toolbar v3 in Beta has caused a widespread level of debate around the new Autolink feature that allows users to click a Google toolbar button and convert ISBN numbers (among other things) into links (Google Toolbar's AutoLink & The Need For Opt-Out gives an insight into some of the reasons why people think this is a bad idea...).

In the Beta, the ISBN links are to Amazon, although it's likely that users will be able to tune this in the final release. I can see why many webmasters will feel uncomfortable with this ability for Google to effectively rewrite pages, and this forms one of the major arguments in favour of allowing webmasters to opt-out. But then, Firefox extension coders revel in developing such tools which can, in many cases, provide a useful service.

At the risk of upsetting everybody, Google could always allow either webmasters to embed their preferred book reseller and affiliate code in page headers, so that IF the Google Addlink is used, the original site at least has some influence where the link goes, unless ovwerridden buy the user, who should also have the same freedom. This suggests a default hierarchy where user prefernces over-ride site preferneces, which in turn over-ride Google's defaults.

It would also be nice if as well as pointing to booksellers, the toolbar could be configured to take users to their local library... I'm thinking of a service like Jon Udell's LibraryLookup tool, of course...

Not Quite Useable - Txting London's Blue Plaques

The seemingly unmaintained HandHeld History website - a joint venture between English Heritage and tech aware historiy freaks, apparently - tells you how to go about using your mobile to find out just a little bit more about the lives of the famous people who's homes, birthplaces, or occasional overnight resting places are immortalised by blue plaques.

If you txta message to 82222 starting with hhh and followed by one of the codes listed on the Blue Plaques SMS codes page, you'll get an informative txt back about the person named on the plaque. The codes are generated from the initial and suname indicated on the plaque, mostly - but there are exceptions, which is where this service falls down I think...

How much better if there wre a unique code on all the plaques?

The slightly premium rate Blue Plaques story codes service (which gives you an audio description) uses an 8 digit code made up of the year of birth and the year of death detailed on the plaque.

Given the number of plaques around the country, I can quite imagine that these numbers are not unique - but this numerical code is more reliable than the txt key, I think, if it is to be generated by your average punter (assuming all the plaques carry at east one date, and you know the rule to repeat the date if only one is shown). In cases of duplication, you might be informed of the two or more possibilities relating to Plaques sharing a key and asked to reply with your choice of which Plaque you'd like to know more about (and the hhh people may find that punters then go on to request info about the alternatives that shared bith and death dates - for an extra paid for message, of course!)

Micro-Info Radio: Traffic Info (and music downloads?)

Catching up with myself again, as the notion of micro-information services as short self-describing, auto-discovery and/or standards based short messaging protocols develops, here's a bit of background information on RDS Radio Data System.

(If you've never really got you head round the PTY and TA buttons on your car radio, the BBC produce this useful RDS cribsheet).

Digital radio also supports digital radio text services, of course, which can similarly provide information about current tracks, or artist information.

Of course, in my integrated radio/mobile phone/iPod in-car entertainment system, I'd have a single button that made use of the artist info and track identifier to pop a download of the song currently playing onto my iPod, and all for the price of a premium rate data call....

Traffic and travel information has traditionally been one of the drivers, though, of RDS uptake, and this area of information provision is only likely to become more widespread as location awareness becomes cheaper to support. This quitew elderly article on the provision of traffic and travel information services rehearses some of the arguments that have presumably informed the development of current TTI services.

It may (or may not....) be worth noting that TTI services seem to be quite an active area of international standards development (TC204 Intelligent transport systems).

I guess while I'm on the subject of traffic information, it's probably worth putting in a pointer to Transport Protocol Experts Group (TPEG), a protocol that has been developed to enable the broadcast of Traffic and Travel information by DAB (Digital Audio Broadcast) radio broadcast. A slightly more accessible overview of the standard is available here, but you'd still have to be really keen to look through it!

Location Aware Txt Alert Bookcrossing?

BookCrossing is oh so simple an idea that can be used to pass on a book (possibly) to someone who really wants to read it. Leave your book in a particular location, register it on the Bookcrossing site, and posssibly, possibly it will go to a good home.

The site supports alerts - so you can get an email if someone leaves a book in your area - but this is only useful if you have access to your email. It would be more convenient, perhapsm to get the alert to your mobile...

But who'd pay? There are plenty of sites on the web that allow you to buy some txt credit and then use it to send txts from the web, so if you trust Bookcrossing enough to allow it to send you txts via your account a solution presents itself. If you live in a popular Bookcrossing area, though, the txts may get expensive, so filters would be useful (e.g. to notify you of books on your booklist that have been left in your vicinity).

Two additional sorts of automation would also be possible

- using Amazon's similar items webservice, you can cast the net wider than your booklist, but still related to it;

- using location awareness (based on the location of your mobile) to set your current location preference on the Bookcrossing site.

A few projects there for a slow weekend, perhaps...

Where's That Mobile?

Although an asynchronous service, SMS has the potential to support one-way and two-way interactions based on the limited form of location awareness you potentially get for free every time you switch your mobile phone on.

Here's one brief review of Location and Presence in Mobile Data Services that sets the scene admirably.

One obvious question is - what sorts of location aware services are already out there making use of this information?

mapAmobile is one idea that raised quite a few privacy concerns when it was first released. Using the mapAmobile service, you can get your current location txted to your phone (send try mapamobile to 83118 (in the UK) and (allegedly) they'll send you details of your location by return txt. I'm not sure how many trials this is good for, though!). Once registered, you can look up the location of a phone registered with your account through a browser.

ChildLocate does something similar, and rather than going for the corposrate market, this service is targeted squarely at families.

Hmmm - seems like another blogger was been surfing similar things this time last year: One Stop Thought Shop: Little Brother is watching!. TagandScan looks weel worth a play here in the UK...it might not be too hard to build a simialr service from scratch using the location awareness service that comes with an iTAGG account.

Styling it Nicely...

Separating presentation from content is one of the bedrocks of current web standards. A great plus of this approach is that it becomes relatively easy to rip the look of another of site you particularly like the style of, and apply it to your own pages.

For a quick idea of what you can do to fixed content with a variety of style sheets, the css Zen Garden does, as its strap line suggests, demonstrate the Beauty in CSS Design.

Alternatively, the Blogger Templates blog shows just what you can do to a Blogger blog like this one with just a little bit of imagination...

If you're interested in customising pages yourself, it helps if you have a little knowledge of CSS... HTML Dog is as good as any place to start, with its clean design and useful blogs...

Generative Naming

If you can work out how a developer has structured the URI hierarchy of their site, you're well on the way to being able to URl hack it to find the stuff you want locate without having to do much browsing.

If a site is built on Generative Naming principles, and you have a model of those principles (or it is published for all to see) then as a user you can build all sorts of tools that exploit this principled naming convention.

There is a danger here, though, that the URIs are being used to transmit information they shouldn't. For example, the MIME-TYPE of a representation should be used to classify what sort of a representation it is (HTML document, pdf or png file, for example) rather than the document suffix. That is, we allow for the URI to be misleading - which is what generative, semantically useful naming confounds to a certain extent.

But as a shortcut, where the structure of generative URIs maps cleanly onto links and relationships between resources, we have a way of navigating from memory or rule. The Naming Isn't Navigating post considers this point, too.

You may note that both the above links point to pages on the RESTwiki... it's well worth a look, I think....

Firefox Search Toolbars

The number of Firefox toolbars available seems to be increasing day by day...

Steve Anthony's Universe offers several catalogue additions to the Firefox searchbox (such as the Canadian Institute for Scientific and Technical Information (CISTI) catalogue, and ScienceDirect). He also offers a CISTI Sidebar that builds on Peter Binley's Google Scholar/OpenURL extension.

The Science Library Pad: Firefox extensions blog has regular postings about new search tool additions to Firefox.

Shortening URLs

The DigBig - Long URLs shortened service lets you map long (often meaningless) URLs (such as deep links) to shorter equivalents that (unfortunately) are also meaningless...

The DigBig info page offers an example of a long URL into a multimap map for a particular postcode -

"This long URL (152 characters) ...
http://www.multimap.com/map/browse.cgi
?client=public&db=pc&addr1=&client=pub
lic&addr2=&advanced=&addr3=&pc=SW1A1AA
&quicksearch=SW1A+1AA&cidr_client=none
... is shortened to just 22 characters ...
http://digbig.com/3bbd"

How much more useful if I could also map that onto a sensible string, such as the following, for instance:
digbig.com//SW1A1AA
or possibly
digbig.com//postcodeMap/SW1A1AA
or
digbig.com//maps/westminster
or something a bit more intuitive....I personally don't think I can remember many more random 4 character strings!

The path would let me build up my own list of short (ish!) meaningful URLs, a bit like at ouseful.open.ac.uk/

New Ways of Communicating with Digital Library Users

Project Bluebird: notifications research at Talis is an ongoing umbrella project that is exploring how new technologies can be used to improve the ways in which academic and public lending libraries communicate with their users. There aren't many details on the Project Bluebird site yet (though there is this basic white paper on libraries and RSS), but it's probably worth keeping tabs on (now why don't they have an RSS feed about this project...?)

Here's something to be wary of on older sites, though - RSS in a library context is not necessarily to do with the sort of XML feed you may expect your blog to produce. It also stands for (or at least stood until recently for) Resources Sharing System, a "protocol-based automation system that managed borrowing and lending requests." It has since been renamed by its developers (if the site redirect is anything to go by) to
Universal Resource Sharing Application (URSA).

Anyway (and here's one I think I may have blogged before) for a quite wideranging peek at the way various libraries are using RSS feeds, have a look at RSS(sm): Rich Site Services. Blogs, news and what's new appear to be the mainstays.

And finally, I've seen online bank account consolidators before, but a library account consolidator is new to me. Still, the ELF Personal Email Library Reminder Service seems eminently sensible...

Give Web Services a REST...

A reasonable piece in IEEE Distributed Systems online: Critics Say Web Services Need a REST

Mobile Information Services

I've not been keeping tabs on what's been happening in the world of mobile information services (and in particular asynchronous, SMS services) for a bit, so let's see what's happenign at the old haunts...

Portalify seem to be slowly pushing out their Library applications, and their Coinlet mobile payment system.

Mobile ticketing seems a sensible application - I wonder if a crude derivative of this approach could be used to take payment for tea and biscuits and suchlike from charity stalls - send a message ("charitygift cup of tea") to a premium. 50p a shot number, and get a receipt back: "Thank you for donating 50p to /whatever charity/ for your cup of tea"! No more need for loose change - just post the quicktext number and keyword and let the txt take the payment. Show the receipt msg and collect your cuppa!

This looks like it may be worth exploring - text.it, which is apparently "the only official website dedicated to text messaging in the UK".

Now here's a thought (though possibly little more than that) - mGovernment: Mobile/Wireless Applications in Government. Using mobiles to co-ordinate activist activities (?!) is probably more likely: An Introduction to Activism on the Internet: Cell Phones

What Lies Beneath - What HTTP Offers REST

Advocates of the RESTful approach to webservices argue that HTTP provides all the verbs necessary for managing the transfer of representations via the web.

GET, POST, PUT and DELETE seem to map well on to the Create, Retrieve, Update, Delete (CRUD) primitives familiar in the world of databases.
If we take a view of the web as a medium across which we can access representations from some sort of universally addressable distributed database, then being able to map from GPPD to CRUD would seem like a useful place to start. Just how well these two do map will be left for a later posting.

If you had a look at the HTTP specification linked to from the W3C link above, you may still be none the wiser (in lay terms) about the differences between GET and POST in particular. (These are the two most widely used verbs, as you'll know if you've every created an HTML form). So, as far as the GET and POST methods go, what's the difference?

Txt Price Info

An old link - online book prices to your mobile phone by SMS - and a new one (to me, at least) - Car Prices by SMS.

The first one's free, the second one is pricey. Both work on a similar principle, though. Send a keyword and a unique identifier (ISBN in the case of txtbux, car registration number(!) and estimated mileage in the Car Prices case) and get a text message back with some useful price info - an Amazon book price on the one hand, car dealer list prices on the other.

Another SMS service I use regularly is the registration free SMS To email. Just start your txt msg with an email address, and it'll do the rest...

A Time Before the SOAP vs. REST - SOAP vs. XML-RPC

With the ongoing debate about the pros and cons of SOAP as compared to REST, the scholar in me got to thinking about the last round of debate between XML-RPC vs. SOAP. (Funny as it seems now, think the first time I came across REST was this article on Creating and Consuming Web Services With PHP - REST was mentioned as a final footnote...).

In particular, it seems that no-one's much interested in arguing around XML-RPC now, although a three-way comparison does occasionally turn up, like in this consideration of Consuming Web services in PHP.

This is possibly useful and/or interesting. What is it about XML-RPC that makes it no longer a topic of interest? Has REST taken over one of the lines of argument, or have some of the arguments in favour of XML-RPC lost out to SOAP? I seem to remember that the old Amazon webservice API gave examples for SOAP, XML-RPC and then later on for REST, but their new ECS API focusses on REST and SOAP (I think...)

SOAP'n'REST, cf. SRW'n'SRU

The webservice developer community has for some time been debating the relative strengths and weakness of SOAP based webservices compared to the REST approach (see the article Roots of the REST/SOAP Debate for one of many discussions on this topic, or alternatively this useful blog: SOAP vs. REST. This Second Generation Web Services article is also a useful starting point, along with the follow up article on REST and the Real World.

There also seems to be an almost parallel debating going on in the digital library world, (though lagging the SOAP vs. REST debate perhaps?), as described in Search and Retrieval in The European Library.

The two challengers represent different approaches to online search and retrival. In one corner, the SOAP contender is Search and Retrieve via the Web (SRW) and in the other the Search and Retrieve URL Service (SRU)

Developers are mindful of both approaches, though, as the OCLC Online Computer Library Center demonstrates...

Wot No URLs? When GET and POST intervene...

The increasing number of database driven websites means that in some cases a naive URL hack doesn't work... welcome to The Deep Web

The address may still be hackable of course, especially if you're looking at a library, say, where they may have opted into OpenURL, (and here's more than you need to know).

Where forms are used to generate search terms that at GET or POSTED to the server where it can make secretive database queries to its heart's content, you can always poke around the form's source code and work out what search terms you need. If you have access to server side rewrites and redirects, you can quite easily make a short, convenient alias to an otherwise bleurggghhhhy URl like this, perhaps?

How should I address my pages?

The previous posting about URL/address line hacking got me thinking about what protocols - informal or otherwise - influence the structure of websites and often as a result (server side rewrites aside) the most commonly used addressing schemes.

Here's what some of those folks at W3C think: Hypertext Style: Cool URIs don't change.

URL hacking

Sometimes its the only way...you know the info's there somehwere, but the onscreen site navigation is lacking...so what can you do but attack the address line directly?

If the site is organised in a sensible way, you can often get quite a long way just by making up a sensible path... take UK universiies for example. Sometimes the easiest way of finding a person (if you know their department) is to look up the department's homepage and stick people/ on the end. for example, the York computer science departmental website lives at http://www.cs.york.ac.uk/, so i reckon http://www.cs.york.ac.uk/people/ is a good bet for the staff listing...

So what does a quick Google trawl of url hacking turn-up? Hacking the URL: Do-It-Yourself Navigation Library hacks which naturally leads to the yet to flourish Library-hacks Wiki

This one I knew about from xml.com: Hacking the Library, which mentions a British Library experimental webservice, but doesn't yet give the location of it... If you're intereste, the address just happens to be http://herbie.bl.uk:9080/

OpenURL Firefox Extension

Peter Binkley's Google Scholar OpenURLs Firefox Extension added a tool that would allow users to retrieve the full text (wherever possible) of articles listed as search results in Google Scholar from their own academic library. The only problem was that the user themselves had to poke around around in the innards of the extension to set up a link to their own library's resolver....However, there is derived version of this extension now available that allows the user to set the resolver address throug a straightforawrd dialogue: Openly's OpenURL Referrer