OpenID is generating a reasonable amount of buzz recently. OpenID is a vendor-independent approach on bringing the infamous concept of single sign-on to the web. in a nutshell, OpenID allows users to use a single, URI-based token to access numerous OpenID-enabled websites. this post on Read/WriteWeb features a screencast explaining the concept and a comparison with the single sign-on efforts by Yahoo, Google & Microsoft.
Niall Kennedy writes about a new del.icio.us API allowing to query the top tags & number of bookmarks of any given URL. while the API hasn’t been released officially yet, Yahoo! developer network already features a screencast introducing some of its capabilities. currently, the API seems to consist of a single URL which returns data in JSON-format when queried with the md5′ed hash of the requested URL. f.e., a check for del.icio.us metadata of play.fm returns the following result-set:
[{
"hash":"f636b74442302d5a3e1c7d646f530df0",
"top_tags":
{"mp3":6,
"house":6,
"techno":14,
"archive":4,
"stream":9,
"radio":36,
"streaming":6,
"audio":13,
"music":45,
"electronic":12,
"webradio":13},
"url":"http://www.play.fm/",
"total_posts":71
}]
go to JoostdeValk.nl for some simple php wrapper-code, or subscribe to the official del.icio.us-blog to stay updated on the matter.
update: del.icio.us now also offers a widget exposing the same data for easy integration.
Google just released an API for their Spreadsheets product as part of the larger Google Data APIs (”GData”), which also allow 3rd parties to access Calendar, Base, Blogger and Code Search. the API supports read/write-access to existing documents, though - according to Google Blogoscoped - it’s not possible to create new spreadsheets (yet).
Aaron Hopkins (Google) published a great, in-depth article on optimizing page load times, especially considering AJAX-apps with their inflationary number of http-requests. Hopkins gets dirty analyzing http-implementations of current popular browsers, eventually deriving a set of useful optimiziations no web-developer should miss. recommended!
Coghead, a web-environment & -framework for rapid develoment of data- and process-driven online-applications, launched into closed beta this week - and received a lot of good press. Coghead is aimed at information workers who have developed desktop/LAN-applications using Microsoft Access or Visual Basic in the past (there’s plenty of them!). while competitor Dabbe DB is solely focused on data-collection and -storage, Coghead seems to go further by offering form editors and even a tool for modelling business processes. sounds very interesting - expect more details as soon as I’m in the beta!
(Coghead Form Editor)
(Coghead Process Modeler)
looks like Google ain’t going to slow down these days: Code Search - Google’s effort on indexing source-code on the web - was launched earlier today. Code Search offers developers powerful ways (read: regular expressions) to search for code-snippets in any particular language, under any specific license. a quick test-drive unveiled that Code Search suffers from similar problems like established competitors Krugle and Koders: string-based search-results often mistake underlying semantics. when searching for an id3-implementation in PHP, results often lead to external function-calls, but not the actual source implementing id3. still, indexing source-code will definitely help making open source knowledge more accessible to average developers, not able/willing to spend hours/days to dissect code. looking forward on Danny Sullivan’s commentary on Code Search…
update: here’s a collection of [queries for]code that was not really supposed to be public 
Mahemoff’s Software-As-She’s-Developed-blog gives a brief overview on eight methods for creating dynamic graphics within the browser. besides more obvious choices like SVG/VML (Vector Markup Language, Microsoft’s equivalet of SVG) or Flash, the post also introduces lesser known options like the Canvas-element (think of a fully scriptable -element - currently supported by Firefox, Safari & Opera) or CSS-based graphics libraries. don’t get too excited though, as most methods are not supported by all browser-platforms yet.
Ajaxian released the results of their 2006 ajax-survey, asking more than 860 developers about their coding-habits. according to the survey, the leading ajax-framework clearly is Prototype, with about 43% of the developers using it to date, followed closely by the visual-effects library Script.aculo.us (33%) which is built on top of Prototype. Dojo, DWR and Moo.fx close in between 11 and 19%. personally I’m using Script.aculo.us for the eyecandy and some convenience-features, but prefer to do ajax-calls directly via xmlhttprequest - interestingly, so do 25% of the polls’ respondents.
ajax backend-coding is predominated by PHP (49%) and Java (37%), with some smaller chunks for .NET (16%) and Ruby-on-Rail (14%). check out the detailed results…
the recommendable All about Linux-blog features a detailed walk through of many features introduced with the new version 7.0 of texteditor Vim (Vi-improved). highlights include spellchecking, ‘omni’-code completion for many predefined languages, tabbed user-interface and undo-branches. the latter is a life-saving time-machine: typing ‘:earlier 5m‘ rewinds the editor-buffer to its state 5 minutes ago - even if the file has been saved meanwhile. similarily, ‘:later 20s‘ fastforwards 20 seconds. Vim is available for unix, Windows and Mac OS-X.
I just discovered O’Reilly Code Search (Beta), a search engine indexing all the sample code of more than 700 books published by O’Reilly (that’s about 2.6mio lines-of-code :)). snippets are annoted and linked with similar code from other books - a great resource for those owning an O’Reilly-collection worth several metres of shelfspace.
