Zmarties

A view from the palace

2007-06-17T13:48:00.001+01:00

Taken, processed, and uploaded as part of Hack Day London

Improving CustomEyes

2007-03-26T22:51:00.000+01:00

GlassesDirect offer a feature they call CustomEyes which allows you to "try on" the glasses they offer online. It works by you uploading a photo of your face, and then you can overlay an image of the glasses on top of that.

Unfortunately they don't really use the power of the computer in doing this. CustomEyes is a silly little Flash application, whereas this could all have been done in far more accessible HTML. You currently get to see a small image of your face, and then have to select the frame you want to see by name from a drop down. The Flash controls do allow you to rotate the glasses - but who wants to see glasses that are rotated off of the horizontal anyway? In addition, they also allow the glasses to be resized, which is necessary since you they don't know in advance what sized face image you have uploaded - but is doubly necessary because they don't even make all the glasses images the same size - some are twice the size of others!

I'd suggest that the application could be much improved by doing the following:

do away with the Flash, and make this a straightforward dynamic HTML page
select from an illustrated list of frames, rather than simply by text name
for each frame that you select, add a new copy of the uploaded image to the page, so that you can actually fill the page with images that you can compare side by side, or which you can print out and compare that way
either automatically recognise where the eyes are in the image when it is uploaded, or ask to be told once, and thereafter make sure that all the images of the frames are scaled to be the correct size to fit.

What are Zooomr up to this week?

2007-03-07T23:17:00.000Z

Just setting things up so that I'm all ready to try out the new Zooomr version 3 when it's ready (which Thomas Hawk is saying is very soon now).

moon Hosted on Zooomr

YouTube blog

2007-02-14T21:03:00.000Z

YouTube has millions of users.

It also has a blog at http://youtube.com/rss/global/our_blog.rss

According to Bloglines, there are just 9 subscribers to that blog via Bloglines!

I wonder, is it that YouTube users are not very blog savvy, that the YouTube blog is not very interesting, or are YouTube blog readers not doing it via Bloglines?

(By comparison, the official Google Blog reports 43,697 subscribers via Bloglines).

Ask maintains a blacklist of IP addresses

2007-01-28T23:09:00.000Z

Someone at Ask.com should realize that IP addresses are dynamically given out by ISPs to their clients (potentially) each time they connect - so blacklisting "clients" by IP addresses is not a very smart thing to be doing.

I recently found that the IP address I was now allocated was somehow on Ask's blacklist, and that instead of getting Ask's search engine, all I got was a page saying:

Your client does not have permission to access this site.

Please refer to the Ask.com terms of service page.(Ask.com and Syndication).

If you feel that you have received this response in error, please send an email to unauthorized@ask.com. Before sending this email, please refer to our terms of service page, accessible at the url provided above.

Please copy and paste the information below into the body of the email.
fff9480d-2fff9480dfff9480d

I'm not sure if emailing that address is an automated system to remove the IP address from the blacklist - you can probably spot that the text they ask to be included in the body of the email is simply the hex representation of the client's IP address, (repeated 3 times for some reason). (I've obscured the actual address in the example above, but kept the same format).

Having sent the email, I've certainly not gained instant access to Ask again - I rather think that it will be quicker to get a new IP address from the ISP's DHCP server than to sort out Ask's broken blacklist. If Ask wants to use a blacklist system then they should probably expire IP addresses off the blacklist say a few hours after whatever behaviour it is that triggers the inclusion on the list was last detected. That way, zombie machines that continue to do something bad via Ask will remain blacklisted and hence blocked, but legitimate users who inherit an IP address that was previously used by a zombie are not permanently banned.

Findory rides into the sunset

2007-01-14T23:48:00.000Z

Searching is for when you know roughly what you are looking for.

An alternative way of finding information is for it to be recommended to you. One of the leaders in recommendation systems is Amazon.

Findory tries to apply some of the ideas Greg Linden learnt when building Amazon's recommendation systems (and what he's learnt since!), but unfortunately Greg has found it difficult to continue to grow Findory - and has just announced that "Findory rides into the sunset". Roughly speaking, this means that development of the site has slowed to a crawl, and that Greg is to concentrate on other things - which for the moment he explains means health and family.

I'm sorry to hear this - Greg's was one of the best blogs in the search area, backed by his ongoing research interests and development of Findory, and whilst I'm sure his interest in personalizing information will continue, without the need to attend to it on a daily basis, I suspect we will be hearing less from Greg in the future. I wish him well, and sorry he couldn't find a way to continue expanding Findory.

Keyhole website still up and active

2006-10-05T22:20:00.000+01:00

When Google bought Neven Vision recently they were extremely quick in tearing down all the pages on the websites that Neven Vision had.

I was therefore rather surprised to find that although

http://www.keyhole.com

redirects to Google Earth, if you just add index.html to the url thus

http://www.keyhole.com/index.html

then the old Keyhole site is still there, and apparently fully functioning.

OCR problems in Google Books

2006-09-29T07:24:00.000+01:00

The accuracy of OCR is vastly increased when the words you read exist in a dictionary. In that case the OCR, when trying to decide between more than one possible interpretation, can use the dictionary to help determine which is the more likely.

As the Google Books project announces its extension to scan European books in Madrid it will have to adjust the dictionary it uses - it's no good scanning Spanish books using the same dictionary as American English books.

However, there are some "English" documents that Google has scanned already where the OCR process has gone very wrong. Consider this page of old printed English, with the "long s" symbol, which looks like a modern "f" character. Looks like the OCR was not told that this dated from 1796, so to look out for long "s" - hence it has identified lots of "fuch" and "fale" rather than "such" and "sale" on the page. A simple dictionary check would have helped here - but only if the process expects "f" and "s" to be confused.

Picasa Web Albums adds email digests

2006-08-16T19:39:00.000+01:00

There is a new feature just appeared on the Picasa Web Albums settings page. Between the existing Public Gallery URL, and the Content Controls, there is now a section that allows you to set up an Email Digest.

It looks like Google are defaulting this to a weekly digest, to get the news out to people, and then giving them an option in the email to turn it off. The 4 choices given are no digest, or at a frequency of daily, weekly, or monthly. The description alongside explains:

What is an Email Digest?

When people you have marked as favorites create new albums, upload more photos, or comment on your photos, we send you a summary of these activities to your email address at the interval you specify.

The digest email itself is an HTML formatted email as follows:

The links in the email for changing settings are generic - they just take you to the main Picasa Web page, where you can then log in if needed. This means that you can forward the email to someone else without security issues.

The images do of course link through to the album that has changed. The thumbnails shown are embedded in the email, not just links to the Picasa Web site, so they can be seen when offline, or when external images are turned off (as is the case with most email programs these days).

I didn't have any new comments for this example to notify me about, but the source of the email indicates that they would follow after the list of changed albums.

(This should have appeared on my Documenting Picasa blog, but I'm having a few publishing issues at the moment).

Update:It did eventually make it to the Documenting Picasa Blog.

Most popular tags on flickr

2006-08-13T13:32:00.000+01:00

Whilst thinking about Searching Picasa Web Albums I did a bit of investigation into the most popular tags on flickr.

flickr provide a tag cloud, but that is just a visualization of the data, and I couldn't see that they provide the data in a direct form anywhere on the site. However, with the aid of the flickr API I was able to take all the words in the tag cloud, and feed them to the API, to get a count of how many matching tags there were for each word. In addition, I also fed those same words to the "free text" search API call, which looks for the word in more than just the tag field - considereing comments and descriptions for example.

The table below shows 3 columns - the tag I searched for, the number of tags of that word found, and the number of photos found using a free text search. The table is ordered by popularity of tags.

wedding	1697455	2269471
party	1475681	2350563
family	1160057	1557739
travel	1113576	1275557
friends	1093997	1649955
japan	1074281	1167494
vacation	1020423	1162147
london	927172	1100791
beach	864728	1493985
california	837012	960129
birthday	829670	1371859
trip	817324	1786838
nyc	774884	879161
summer	732387	1122946
nature	728059	816467
italy	686976	776812
france	669406	759215
me	664111	2886782
paris	662024	786996
art	643563	1104574
flowers	627830	1439335
sanfrancisco	618604	620523
europe	610237	708284
china	604334	707039
flower	596891	1439327
newyork	572049	575557
water	565749	973599
people	552582	975146
music	551493	740282
cameraphone	544369	634883
australia	543923	611789
christmas	537883	834827
usa	531563	619628
sky	527376	756538
germany	526931	595724
new	526549	2723772
canada	512067	612492
night	510466	1210035
cat	504840	852699
holiday	501819	703551
park	498178	1820673
bw	495568	514618
dog	486692	859611
food	478691	684164
snow	478068	643189
baby	463882	837637
sunset	462862	752411
city	446866	1294211
chicago	445639	529140
spain	441158	491290
taiwan	437633	461265
july	435057	1033708
blue	431367	862404
tokyo	427181	505507
england	425920	510575
mexico	406752	535804
winter	403827	579849
portrait	396759	644632
green	394571	671410
red	393719	848593
fun	392680	688356
india	386409	446473
architecture	383526	423993
garden	382303	981077
macro	380663	428872
spring	373441	759260
thailand	368331	411644
uk	362555	468979
seattle	357614	421000
festival	357137	658357
concert	350224	471579
canon	349327	580369
house	347404	1236984
berlin	343694	403534
hawaii	340744	384573
street	339503	1036943
lake	324783	937873
zoo	324442	460997
florida	322534	387322
june	321243	729667
may	316693	903231
white	313703	867135
vancouver	312657	383569
kids	312551	635702
tree	312542	1019630
clouds	307407	479446
toronto	304648	359194
barcelona	293035	341533
geotagged	292631	294655
home	291227	778635
sea	288469	546365
day	287287	2136618
texas	284851	360478
scotland	284615	309559
car	281730	788266
light	281436	909673
halloween	280515	362042
camping	279238	654077
church	273001	535265
animals	270757	542946
trees	270098	1019628
washington	266921	470046
river	266128	668903
nikon	265291	393117
april	262835	581807
boston	261935	330032
girl	258663	857847
ireland	258647	305876
graffiti	257081	283560
amsterdam	256794	307304
rock	256142	795001
landscape	254799	343825
blackandwhite	252987	253364
cats	252971	852700
newyorkcity	250213	250430
san	247739	1271943
rome	246027	286082
roadtrip	245326	255484
urban	244511	321849
honeymoon	241904	271611
ocean	240941	359337
dc	235053	363788
newzealand	232776	233573
march	231491	532229
black	231180	708269
museum	230643	562979
york	230193	868773
hiking	226067	398781
island	224110	844505
mountains	223030	759360
yellow	221515	347290
sydney	221447	278644
sun	218685	503464
hongkong	216193	222298
show	216155	850535
graduation	214681	306454
color	213967	579521
film	210509	351853
mountain	210046	759360
animal	208416	542946
losangeles	207932	208309
school	205891	550486
moblog	205613	210954
photo	205047	2308835
dogs	203610	859613

Documenting Picasa

2006-08-07T08:03:00.000+01:00

I've mentioned Picasa, the photo software from Google, a number of times on this blog.

Of all the Google services, this is perhaps the worst documented (with no API, and a tendency from Google of making the old forum / groups support postings unavailable), so I'm looking to do something about that. I've created a website called Documenting Picasa, together with the Documenting Picasa blog, which both look to cover the downloadable Picasa client software, and also Picasa Web Albums.

Refreshed betas for Picasa and Google Earth

2006-07-10T08:26:00.000+01:00

The beta versions of Google Earth 4 and Picasa 2.5 have been refreshed.

Picasa moves from build 31.99 to 32.01, though its unclear what fixes have gone into this.

Google Earth moves from build 1565 to 1657 on the PC, (1658 for Mac, 1660 for Linux), and the changes are listed on the release notes that is installed with it, and available via the Help menu. Google Earth Blog has a post summarizing the changes.

Distorted images on Geoportail

2006-07-08T09:29:00.000+01:00

Now that the French mapping site Geoportail is finally available, I've had a chance to have a quick look round.

As you zoom in, the images show one version of the data (I'm assuming a browser scaled version of the low res data), then replace it with the higher res data. In many places that I looked at, the two images do not line up - leading to a strange image shift on switching between the two.

I wasn't sure what was causing this, but wondered if it was something to do with the images being warped over a 3d terrain model. However, the net effect is simply one of image distortion - if it is a terrain effect, then their model has bad data.

The example image below shows the runway at St Martin in the Caribbean - where as you might expect, a runway is straight, flat, (and in this case at sea level, which should mean no terrain height problems). It's a spectacular runway at the best of times, with landing jumbo jets passing closely overhead of sunbathers on the beach at the end of the runway, but I wouldn't fancy landing on it if it's as uneven as the distorted images on Geoportail suggest!

(I havn't found how you bookmark a particular view using GeoPortail - is it possible?)

St MartinHosted on Zooomr

Overplot mashes quotes heard in New York

2006-06-26T20:17:00.000+01:00

Mihai Parparita, a Google employee, has put together a Google Maps mashup he calls overplot. It mashes together the quotes from Overheard in New York, via the Google Reader feed of the data, with some geocoding via the Google Maps API, so that the quotes can be seen in their appropriate positions on the New York street map.

The technical details include having to use an overlay instead of individual pushpins, since there are so many clustered results to show, and the use of the canvas object to draw areas. Although there is a fallback to ExplorerCanvas to support IE, the code makes no allowances for earlier Mozilla browsers that didn't support the canvas object, which is a shame, since it looks as if this was a late change to improve performance - presumably the earlier code could have been left in place (with reduced performance) to cover a fuller range of browsers.

Geoportail - impressions of the homepage

2006-06-25T09:20:00.000+01:00

France's new mapping service at GeoPortail continues to be suffering from huge demand, and hence to be generally unavailable. I did however once succeed in getting up the home page, a screenshot of which is shown here, and there's enough to be seen there to allow for some comments.

The site is all in French! OK, so thats the political reality of this being set up by French government dictact, but if the objective is to encourage interest in France and French culture, then making it more accessible to those outside the country, speaking different languages, would seem to be a valuable feature. (The site is also (un)available at geoportail.com, but this just redirects to geoportail.fr)
There are 3 stages outlined

Summer 2006 (now) - aerial images and scanned maps
Autumn 2006 - 3D viewer, and initial data sets
2007 - a rich set of public information available, plus WebServices

The page is hardcoded for a screen size of 1024x768 - if you have a smaller screen then there's a lot of scrolling needed, and a larger one wont give anymore useful info. This is not the way to handle such graphically rich applications as mapping - they really do need to use all available screen space to show as large a map as possible. The layout is achieved via tables, rather than via CSS.
Coverage entends to

FRANCE
CORSE
REUNION
MARTINIQUE
GUADELOUPE
ST MARTIN
ST BARTHELEMY
GUYANE
ST PIERRE ET MIQUELON
NOUVELLE CALEDONIE
MAYOTTE
ILES KERGUELEN
ILES CROZET
WALLIS
FUTUNA

The initial map shows the whole globe - with no zooming or other controls on it. If you move the mouse over the map, certain hidden areas are clickable - merely indicated by the mouse pointer changing to a hand pointer. (These areas are of course those listed above).
There's a built in feature for bookmarking favourite locations. The current top 3 are

Le Terminal Trans-Manche (62)
Le Mont-Saint-Michel (50)
Le Château de Chambord (41)

The geoportail site is for the "visualisation"; there is a separate site for the "geocatalogue" (and this site, since it has nothing but a holding page until Sept 2006, is available). The visualisation does not include any search box, to direct the map to a particular location.
In a commendable effort to be openly available to as many users as possible, the site is

compatible with IE6 and greater, Mozilla 1.7 and greater, Firefox 1.0 and greater, Safari 2.0 and greater, and OSs Windows 2000, XP, Linux (Redhat, Suse, Debian), Mac OS X

Much of the site is using php to serve up the pages, even for what would appear to be static information - perhaps that's part of the problem as to why the servers cant cope with the demand.

French support of Google Earth

2006-06-23T19:11:00.000+01:00

With the (possible) launch today of http://www.geoportail.fr, which has been rather hyped as the French challenge to Google Maps/Google Earth, I was doing a search for current French use of mapping, and came across the Gers Chamber of Commerce and Industry geomatics technology centre.

There they have some very well crafted KML files, which show information about the interior region of France, to the west of Toulouse. The files on offer include both sets of placemarks, and also presentation of statistics such as employment and population figures. Neat features about the KML used include snippet control, the addition of a fixed logo, and a copyright string down the left of the view, though unfortunately the statistics key now appears just where the (semi-transparent) controls of Google Earth 4 are.

It looks as if the data files were prepared with the help of GE-Data, a French company which specializes in producing data files for Google Earth.

... and as for geoportail itself, as I write this, the website is unavailable, so all we have to go on are some screenshots and early user reports.

GNT, a French tech blog, has coverage of the inaccessibilty of geoportail, IGN (which produced it), and of GNT itself, which happened to also rank highly in Google for the search "geoportail". The screenshot included in this report is quite unlike the others - I assume it is from an earlier prototype of the site.

Opera 9 ships, but still fails in its support of xml

2006-06-22T22:52:00.000+01:00

The Opera browser has often been praised for its standards support, and with the release of the latest version, Opera 9, I had hoped that it may have added good support for xml.

Although web browsers are primarily used to browse HTML formatted documents, they can also be used to view a number of other types of documents. Foremost amongst these are xml documents, a close relative of HTML documents, but in many ways far more powerful - that x stands for Extensible. Viewing a raw xml file is rarely the best way to do it, so xml files are generally viewed with the aid of a stylesheet - written in xsl (Extensible Stylesheet Language), and the the process of transforming the xml into a viewable form is known as XSLT.

The Opera 9 spec document states that

XML support

Opera can parse and display XML documents. Opera can be both a validating and non-validating processor.

Documents with Content-type "text/xml", "application/xml" or with a subtype ending on "+xml" will be treated as an XML document. If a Content-type is not available, the ".xml" file extension will also make the document be treated as XML. Opera does not use US-ASCII as the default character set for text/xml, but otherwise follows RFC3023. We recommend using application/xml instead of text/xml or use explicit character set declaration.

XSLT, XPath, and XSL-FO

Opera has near-complete support of XSLT 1.0 and XPath 1.0

Opera does not support XSL Formatting Objects.

It sounds from that as if the support of xml is pretty extensive, but in practice there's aparently a lot of flexibility in those words. The "near-complete support of XSLT 1.0" seems to mean that they don't support the vitally important document() function - which rather takes the teeth out of XSLT. The document() function reads xml from a named document URL, making the XSLT process powerful, allowing it to pull information in from a number of xml documents. Consider a page of product information that pulls in current prices from a specific price document, and stock levels from another document.

(The previous version, Opera 8, had no support of XSLT, so partial support might sometimes be considered an improvement. However, in many cases its a step backwards - in Opera 8, since the XSLT instructions associated with a xml file are ignored, at least you get to see the raw xml data. In Opera 9, it starts to process the file, but fails on an unsupported base feature, so nothing but an error message is displayed).

Without fully supporting XSLT 1.0, Opera languishes a long way behind the other browsers in supporting xml - IE 6 has great support for XSLT, as does the Mozilla / Firefox family (if a little slow at times).

Update: A fascinating interview on Slashdot, with Håkon Wium Lie of Opera. He notes that:

From that perspective, it makes sense to leave ### half-implemented. You can claim support (and many journalists will believe you), and you also ensure that no-one can use the unimplemented (or worse: buggily implemented) features of the standard. The only way to change the equation is to remind ### how embarrassing it is to offer a sub-standard browser. And to use better browsers.

How true this is - the sentence could so easily apply be completed with XSLT and Opera filling in the gaps. Actually, the published quote was about CSS2 and Microsoft!

Håkon Wium Lie was the father of CSS, and I think reading the whole article gives an insight into the thinking at Opera - they view CSS as all important, and are happy to let XML and XSLT take a back seat. This also comes through across in another reply, which notes

... CSS is an intrinsic component of AJAX. The "AJAX" name sounds great, but allow me to propose a few alternate spellings that I find more accurate:

AJACX: Asynchronous JavaScript, CSS and XMLHttpRequest
ADJACS: Asynchronous DOM, JavaScript and CSS
ADHJACS: Asynchronous DOM, HTML, JavaScript and CSS
AJAHCS: Asynchronous JavaScript, HTML and CSS
AJACS: Asynchronous JavaScript, HTML and CSS

Bloglines upgrades their Atom parser

2006-06-21T08:21:00.000+01:00

Although Bloglines has been able to process Atom feeds for a while, it didn't do a very good job of it - in particular it would lose whitespace around hyperlinks, which made reading atom feeds a bit of a pain.

However, in "Duplicates; real and imagined" it mentions that it is introducing a new Atom parser, and the Atom protocol makes it much easier to detect duplicates, especially when articles are syndicated into different feeds.

It looks like this parser is available now - so now I can read Tim Bray's Ongoing in its full text form, rather than in just the summary form that he also made available in RSS format. (Looking at the stats from Bloglines, I see that Tim has 49 Bloglines subscribers for his Atom feed, but 3465 via his RSS feed - expect that pattern to change now that the much better Atom feed can be displayed properly).

An API for Picasa?

2006-06-20T23:44:00.000+01:00

I've mentioned before that Picasa is one of the few Google offerings without a (published) API. However, although unpublished, there are a number of integration points with Picasa that form a sort of unofficial API.

Specifically, Picasa offers the following

upload to a Web Album (new in the Picasa 2.5 beta version)
upload to Joga (in the customized version of Picasa available at Joga.com)
upload to Blogger

This uploading interface is undocumented, but has been reverse engineered by tabblo, who use it to offer a way to upload to the tabblo photo service direct from Picasa. They offer few clues as to how they did it, beyond this comment on a flickr forum, to the effect that

"we had to spend a couple of days reverse-engineering a complicated desktop application/server-based protocol, then more time debugging it, and as of late, more time bringing it up to date for the 2.5 release of Picasa."

Picasa has some other potentially useful integration points

the oem.xml file is used by the Joga customized variant to supply the customization of the title bar
Picasa 2.5 uses kml files to interact with Google Earth, with the communication in the reverse direction acomplished by Picasa acting as a web server
Picasa also shows a webserver interface when you use the Ctrl-L shortcut, which gives a close approximation to the Picasa UI within a web browser
Picasa Web Albums offer a "download to Picasa" link that works by using a url of the form "picasa://downloadfeed/?url=http%3A%2F%2Fpicasaweb.google.com%2Flh%2FrssAlbum" so Picasa must register a url handler for the "picasa://" protocol.

Whilst the above are all undocumented, there is documentation provided for another integration point. If you look in the web directory under your Picasa installation, there is a documentation directory, and within that is a index.html file which documents the Picasa Web Templating System. This describes the templating system used when doing an export as web page from Picasa. In addition, you can also export as xml, which provides another fairly easy way to extract data from the Picasa system, for use in other programs.

Keyboard shortcuts in Picasa

2006-06-15T21:50:00.000+01:00

Picasa is a great tool from Google for organizing your photos.

However, its documentation is a bit sparse, and hard to find (and from a developers point of view it's one of the few Google offerings without an API).

Certainly underdocumented are the many keyboard shortcuts that the program offers, which both serve to make using the program easier and faster, and in some cases offer facilities that are not available with the mouse, or via the menus.

My favourites are:

CTRL + ALT - displays the image the mouse is over at full screen size
CTRL L - brings up a web browser view of the whole Picasa interface

Google Book Search - Shakespeare

2006-06-13T23:49:00.000+01:00

Google Book Search has put together a page specifically on Shakespeare, which ties in with their sponsorship of some theatre perfomances in New York's Central Park this summer.

A sidebar on the page suggests "Take a Literary Field Trip - Download Google Earth to visit the Globe Theater and other Shakespearean landmarks right from your desktop.", but unfortunately it doesn't provide a KML file link on the page to actually provide this tour.

There is also no actual search box on the page, so although you can browse through the various Shakespearean plays by title, you cant do a simple search for a word, or half remembered piece of dialog - surely some mistake from a search engine!

Major mapping upgrades from Google

2006-06-13T23:13:00.000+01:00

Google's been releasing a whole load of new mapping features all at once:

Version 4 (beta) of Google Earth, which is now available for Windows, Mac, and Linux. This has an improved UI that give lots more space to the photo data, and allows textured 3D models
A new version of the KML file format, that supports the new 3D model changes, improves network links, adds regions for different details at different zoom levels, adds radio buttons
A new version of SketchUp, now available for Windows and Mac, that produces textured 3D models
Geocoding in the Google Maps API - available for US, Canada, France, Germany, Italy, Spain and Japan, with data returned as xml, kml, or JSON
Support for KML in Google Maps

Its going to take a few days for people to get their heads round all these new features, but its certainly going to make for interesting opportunities. There are of course plenty of things that can still be improved, for example

why no UK geocoding suport? - Google claim that "More countries will be added as Google Maps launches in new countries", but the UK was the first country after the USA to get comprehensive Google Maps coverage
Supporting KML for Google Maps as well as Google Earth is great, but it looks as if the Maps version is severely limited - for example loading my Countries of the World KML file gives corrupted balloons, just 67 data points, and a message that "Parts ... could not be displayed because it is too large"

Sony SLR digital camera

2006-06-13T08:22:00.000+01:00

Following on from Sony's acquisition of the photographic assets of Konica Minolta, they have now launched their first SLR digital camera, called the Sony Alpha DSLR-A100.

Its an obvious development of the Konica Minolta 5D, using the Minolta lens mount, and so compatible with many existing lenses. The highlights of the camera include

10 megapixel CCD
camera integrated anti shake
anti dust mechanism
40 segment honeycomb metering
2.5" 230,000 pixel LCD monitor
Lithium ion battery with a claimed capacity of 750 shots

Oh, to get my hands on one to play with. In the meantime, there are hands on reviews (of preproduction models) at

Livesearch from AllTheWeb

2006-05-10T22:07:00.000+01:00

We havn't heard much from AllTheWeb recently, (now part of the Yahoo stable), but today they released a new search interface they call livesearch.

This takes the Google Suggest technique of suggesting words as you type in the search box, and goes one step further by actually doing the suggested search for you to the right of the screen, automatically. This is obviously a resource intensive operation which leads to lots of unnecessary searches being done, so you can see why Yahoo are trying the experiment on an almost forgotten engine, rather than on their main search properties.

The words (or phrases) that come up after typing just one letter are an interesting mix, and it's informative to compare them to the words that Google Suggest uses for the same letter. Some letters agree, but others are widely different.

And what better way to illustrate the chosen result other than by comparing them via another of today's other search engine launches, Google Trends?

In all cases the livesearch result is given first:

a) american idol vs amazon
b) bank of america vs bbc
c) craigslist vs currency convertor
d) both agree on dictionary
e) both agree on ebay
f) facebook vs firefox
g) google vs gmail
h) both agree on hotmail
i) internet explorer vs ikea
j) jessica alba vs jokes
k) kelley blue book vs kelly blue book
l) both agree on lyrics
m) myspace vs mapquest
n) natalee holloway vs news
o) both agree on orbitz
p) photobucket vs paris hilton
q) qvc vs quotes
r) runescape vs ryanair
s) southwest airlines vs spybot
t) both agree on target
u) unique baby names vs ups
v) verizon vs valentines day
w) both agree on weather
x) xanga vs xbox
y) both agree on yahoo
z) both agree on zip codes

I think the K choice is particularly interesting, where they both suggest the same thing, but Google plumps for the much more popular incorrect spelling!

Google Trends

2006-05-10T21:28:00.001+01:00

Released today in Google Labs, Google Trends allows you to get graphs of search and news volume over time for keywords you provide. (Its thus a great complement to the Google zeitgeist, which shows whats hot over a short period of time).

If you provide multiple keywords, separated by commas, then the graphs are overlayed, so you can compare them. The graphs also have little lettered marker flags, which correspond to particular news stories, so you can sometimes see the reason for a particular peak in the searching. The graphs carry no scale, so you can't see absolute numbers, just trends over time, or in comparison to another keyword.

There are a few interesting examples provided

Yankees v Red Sox - with a big peak when they were head to head in the World Series (and the winners search volume was twice the loosers)
major US holidays, peaking as you might expect

and a few of my own examples

da vinci with a huge peak last year (when Leonardo's birthday was a featured Google doodle), and recently increasing news volume obviously due to the film of The Da Vinci Code.
olympics with expected peaks for the Greek summer games, the announcement of the 2012 hosts (London), and the winter games
linux, vista, xp

Free personal version of SketchUp from Google

2006-04-27T08:44:00.000+01:00

Having recently bought the company and 3d design product SketchUp, Google have now taken the next step of releasing a free personal version of the SketchUp software. Currently available for Windows XP, with a Mac version "coming soon", the program allows for simple drag and drop design of 3d models - which amongst other uses can then be displayed in Google Earth. The pro version remains available for commercial use, with lots more features.

Google are also introducing 3D Warehouse, designed as a repository for 3d models created in the program. The models can be viewed in Google Earth via a network link, so you can see geolocated models as you browse the world, rather than having to explicitly download them. Google has pre-populated the warehouse with a number of models which range from complete complex buildings, such as the Taj Mahal, through to individual design elements such as traffic lights and furniture.

Photo words

2006-04-24T21:15:00.000+01:00

The oneletter and onedigit groups at flickr provide the raw data for a rather quirky photo font, where you can put together words formed out of photos of individual characters.

The task is of course made a lot easier by additional software, such as the online spell with flickr which provides an easy interactive way to select the photos to use. (Simply type in the words you want to form, then click on individual letters to change to an alternative image).

Ex Googlers and Ex Amazonians

2006-03-27T23:01:00.000+01:00

Xooglers started out about 5 months ago as a great resource by ex-Google employees, with some very informative posts about the early days at Google.

Posts have slowed down recently, and the recent article on high paying adwords ("mesothelioma lawyers", "tax attorney", and "mortgages" all get mentioned) from certain angles looks just like an attempt to attract high priced click income to the site. However as pointed out in the comments, these keywords dont carry nearly as high a price as made out - the prices quoted are for results on Google search pages, not on third party sites.

(In fact, the original story seems to lead back to CyberWire who published a long list, though its also been picked up by a good few heavyweights, such as John Batelle).

If stories of the early days at what have now become Internet giants interest you, can I also recommend Geeking with Greg, by Greg Linden, who often touches on his early days at Amazon, when not giving in depth insights into personalized search in general, and Findory in particular.

Riya photo service can't count

2006-03-27T20:54:00.000+01:00

Riya, the online photo service whose widely advertised feature is its face recognition, finally opened its doors to everyone this week.

In just a few days they raced to over a million photos uploaded, though for all I know, they may be several times that by now.

I noticed, that they are obviously indexing (some of) the EXIF data to be found in most digital photos, and was curious as to what makes of camera were most popular.

So, I did a quick set of searches, which gave me the following results

canon - 25741
nikon - 21442
minolta - 19704
kodak - 25543
pentax - 18633
fujifilm - 20008
olympus - 21259

Those totals are suspiciously close too each other, so I frankly do not believe them. I therefore expanded my search some more, to look at other popular search words - namely some numbers, in which I would expect the number 1 to be massively more popular than any other number. The results of this were as follows:

1 - 22899
2 - 23429
3 - 22329
4 - 23200
5 - 22226
6 - 22590
7 - 23133
8 - 22937
9 - 23876
10 - 23060
100 - 21954
2005 - 21079
2006 - 18662

Again, all the totals are from a very small range, which suggests that none of these results can be trusted to be accurate. We've grown used to search engines giving us "estimated" numbers, especially when their indexes are up in the billions, but when you are hovering just over a million, yet always get about 20,000 as the answer to any query, then it looks as if someone can't count!

As requested, to help the team at Riya find posts on problems with their service, I'm hereby noting that riyasux.

Konica Minolta's Garage Sale

2006-03-25T09:04:00.000Z

Minolta were the first manufacturer to market an autofocus camera, variously produced with the Maxxum, Dynax, or Alpha brands depending on geographical location, and over the years have had many innovative cameras. However sales always placed them a long way behind Nikon and Canon, and their inexplicable decision to hold off producing a modern digital SLR for so long must have led to many photographers deserting their platform for one of their competitors.

(For the curious, they did indeed have an early digital SLR, the RD175 in around 1995, but for that to have been your only digital camera capable of accepting your lenses for a full 10 years in this fast moving market must have been a death knoll. Their second digital SLR was produced for their APS based Vectis lenses, of which there was a very limited range, and so had few attractions).

When it did eventually come out, the Konica Minolta 7d had a few tricks up its sleeve - body integrated anti-shake being its unique selling point. It's a great camera, but came too late to save the company, even with its cash injection from merging with Konica. This has been followed by a cheaper entry level difgita SLR camera, the Konica Minolta 5d, which uses more plastics in its construction to bring costs down, but is otherwise mostly feature comparable with the 7d.

With just these two cameras in its digital range, and much of the entry level (equals huge sales potential) already gone to Canon, Minolta have found that photography does not make economic sense for them any more, and in Janauary 2006 announced their withdrawal from the photographic market. The assets are being sold to Sony, who will be producing new cameras to fit the existing Minolta lenses.

In the UK, Morgan Computers, a well known outlet for end of line equipment at good prices, has secured what they describe as a "massive factory clearout" - they are offering both digital and film camera's for sale. I suspect stocks are limited, especially for the prime items.

Google Reader does JSON and Atom 1.0

2006-03-23T21:48:00.000Z

The new share feature in Google Reader has been noted in a few places, not least of which is on the main Google Reader Blog.

A brief but informative post at Persistent Info by Mihai who worked on the feature, gives a few key technical insights into the implementations. Key amongst these are the fact that

the "put a clip on your site" feature is producing the data in JSON format - so although it is easiest to display it as styled by Google, it's also possible to use the data in a radically different form
the feeds produced by Google Reader are in the Atom 1.0 format

Blogger has long used an obsolete Atom draft format, so its about time that they moved to the released spec. JSON is an interesting technology, and Google is in this instance behind Yahoo, who have offered JSON output via many of their APIs for a while now.

Update: It's also worth taking a look at Niall Kennedy's post where he reverse engineered the Google Reader API.

Update 2: Also very worthwhile seeing Tim Bray's comments on the Atom data they are outputting.

Google making information inaccessible

2006-02-27T09:04:00.000Z

Google state in their Code of Conduct

Our goal is to build products that organize the world's information and make it accessible to our users.

However I'm noticing that they are more and more making information (that was once accessible) inaccessible. I'm not talking about data that they merely link to, but data that Google themselves control.

As a couple of examples:

The data displayed in Google Earth and Google Local is being changed as and when Google gets hold of different aerial images. For some people having the newest images would be the best data to have available, but for others having the highest resolution data even if older is better. Google recently updated their view of Heathrow Airport, replacing a composite image taken at a number of different times with one taken all at the same time. However, the replacement image is several years older than the image they used to show - as witnessed by the fact that it shows 3 Concorde aircraft still in service - whereas the previous image showed the single retired museum piece Concorde.
Google's image organizer Picasa used to have a very useful and well supported user forum, which was full of informative posts built up over the years. This has now been superceeded with a "Google Group", and the forum has been deleted, so all that information is lost.

Group Shot for making photo montages of groups of people

2006-01-25T23:23:00.000Z

Microsoft Research have released an experimental program called Group Shot that, in their words:

creates a composite image from a series of photos. The photos must be of the same scene, taken from the same point of view within a short period of time. The application uses regions that you indicate to build a composite image automatically.

Its aim therefore is to let you take a bunch of photos of the same group of people, and then to produce a composite one where the participants weren't blinking or looking away from the camera etc.

Its certainly a specialist photographic requirement, but one that many people occasionally hit.

LinkCloud

2006-01-19T20:22:00.000Z

Weblogistan are providing LinkClouds, which seem like a bit of fun, hence I thought I'd place one here.

Initially I'll populate it with some of my own favorite products and services, but it will be interesting to see how it may change in time.

Google music search

2006-01-13T08:46:00.000Z

Google have a largely ignored specialist search for music available, which was announced last month.

The system is not specifically labelled as beta, but is fairly experimental, and has severe display problems in my Mozilla Firefox browser, but looks OK in IE.

Searches bring up matches in separate categories for Artists, Songs, and Albums, with results illustrated by 90 pixel square images served up from Froogle. Clicking through on the artist brings up all their albums, on an album all the tracks, and on a track details including a snippet of the lyrics plus links to other sites with the lyrics.

Where appropriate, there are links to actually buy the album or song, either as a CD or as a download, and the artist pages also include links to do searches on other Google properties, including the web, images, news, and groups.

Google Earth for Windows and Mac

2006-01-12T21:23:00.000Z

Google Earth for Windows has supposedly graduated from beta to full release status as stated on the main Google Blog. However, despite this claim, the current version available for download is the same build number 3.0.0762 from November last year.

Maybe they thought that this build was "good enough", which is a shame since it still proclaims it's beta status quite prominantly:

When you sign up, if there are problems signing in it states "Google Earth is in beta, and we're still building out our ability to take on new users. We're making good progress, and expect to be able to accept new users shortly, so we recommend try accessing again in xx hours."
The extra program GPSBabel which Google distributes with Google Earth, and which is used to convert output from GPS receivers into data that can be displayed in Google Earth is at version 1.2.1_beta. This version is seriously out of date (and in my experience so bugged as to be practically useless) - I'd recommend making sure you get hold of the latest version (either 1.2.7, or a beta of 1.2.8) as soon as you can.

The same announcement also notes that Google Earth is also available for the Mac OS X 10.4.

Globe Glider marries Google Earth and Google Maps

2006-01-05T23:28:00.000Z

Globe Glider, as announced on the Google Earth BBS, is a web application that runs inside the browser window of Google Earth, linking the aerial photo views of Google Earth, with the pre-drawn map views of Google Maps, along with other relevant information.

The application interacts with Google Earth via its unpublished scripting interface, and as such requires a small registry tweak before the browser window can be used.

Searching on another plane with dohop

2006-01-05T08:15:00.000Z

dohop is a fairly recent entry into the specialized flight search arena, but with its December 2005 upgrade to include 660+ budget and full fare airlines, it really has become an incredibly useful one.

The features I like best about it include:

as you type a destination, it provides a dropdown of matching locations
it recognizes that for many cities such as London, Paris, New York you have a choice of airport, and that you probably want to search all of them at once. (This feature could do with some tuning - it brings together all 5 major London airports, but only 2 New York ones; it also does not include the budget airlines secondary destinations, which are often in neighbouring cities, but still convenient for many travellers).

The initial search just brings up possible itineries based on timetables - if you then ask for details on a particular flight, it does more searching and adds a price for that flight, and probably all others with the same airline. This two stage process does mean you get a view of the flights very quickly, but it would perhaps be better if they could cache the price info, or else fetch this in the background so that its available without the second action being necessary. Prices are shown in the currency of your choice. Sometimes sponsored results come up - which moves them to the top of the list, even if they are not necessarily the best matches.

Once you have a list of flights, you can filter them by time, airports involved (including for transits), airlines, and number of stops.

The site makes good use of AJAX to provide a snappy experience, but unfortunately there are a few technical problems with the implementation. As you navigate around the pages, the url address displayed is locked to the homepage, since you are really only navigating an inner frame, yet its easy to get to a state where that outer frame is inaccessible, so all the AJAX stops working. In fact the site does not degrade gracefully if Javascript is unavailable - it simply stops working at all.

url(x) for shorter urls

2006-01-04T08:02:00.000Z

url(x) is a new "shorter url" service, and as such joins a long list of alternatives, of which TinyURL is probably the best known.

urlx however has a few tricks up its sleeve:

urls optionally can contain (part of) the original domain name, which does make them longer, but serves as a very useful reminder mechanism
the main form for creating them is AJAX enabled, which means the new url is ready for you as soon as you have finished typing it in
there is an API, so that other web pages can create shorterned urls (REST input; text, JSOF, or XML output)
the stats pages uses some neat sparkline graphs, mostly just for fun

The AJAX feature seems to be a bit too eager, at least with my hesitant typing - it came up with shorterned urls for lots of partial urls that I hadn't finished entering yet.

Commercial packages using Autostitch panoramic software

2006-01-03T23:54:00.000Z

As I've covered before, Autostitch is a program and algorithm that is capable of producing some quite amazing stitched panorama images with little effort.

The free demo version continues to be available direct from the author, but the system has also been licenced and included in two newly released commercial packages:

Autopano Pro www.autopano.net
Serif PanoramaPlus2 www.serif.com

It's also interesting to note that Industrial Light & Magic (ILM), (a Lucasfilm company) has licenced Autostitch software to produce panoramas for film production.

PAQ7 compression released

2005-12-29T20:51:00.000Z

PAQ7 is a command line file compressor and archiver, that achieves far better compression than the standard ZIP based archivers (and in fact better than almost anything else available).

The better compression is achieved by trading memory usage and speed for compression, and the program uses a neural network to combine a number of predictor models that predict what the next bit in the file will be. Of particular interest are the predictor models that are specialized to work with tiff and jpeg files - in my tests the program achieved around about 20% compression of jpegs in the archive. Technically what it does it to partially decode the jpeg stream, back to the image coefficients, and simply apply better lossless compression to that level of data.

PAQ7, as its name suggests, is one in a long line of experimental leading edge compression programs - and this version is roughly 3 times as fast as its predecessor. The program's source code is available under the GPL, and builds for many platforms, including Windows and Liinux.

Also recently released is a beta of the closed source archiver and compressor WinRK version 3. which similarly achieves very high compression ratios whilst taking a long time to compress files. WinRK is a GUI based program, only available for Windows. It also has a special jpeg recompression algorithm.

Currently neither of these jpeg recompression modes achieves quite as high a compression ratio as Stuffit achieves, which I talked about earlier in the year.

Yahoo and Google widget APIs go head to head

2005-12-14T08:52:00.000Z

Yahoo! and Google have both just released new API's for building widgets. The Google offering is new, and allows you to add the widgets to your personalized Google homepage, whereas the Yahoo! offering is at version 3 - but you need to have downloaded the Yahoo! Widget engine to run the widgets (hence they are limited to Windows and OSX).

Both types of widgets are written in Javascript, so it is fairly easy to adapt from one API to the other. At the moment the Yahoo! widget gallery has 1600 entries, whereas the brand new Google widget directory has just 5 staff written samples.

New Virtual Earth SDK

2005-12-09T08:59:00.000Z

I've not had much time yet to investigate Virtual Earth v2, now branded as the instantly forgetable local.live.com, but I did have time to notice that all existing mashups are broken. This seems like a very silly thing to do if Microsoft actually want people to develop to their API - who will bother developing to something that will be broken at whim by Microsoft?

I also downloaded the help for the new API. This is distributed as a self extracting zip file. Why? In an age where people are understandably wary of exe files, why does Microsoft decide to distribute a help file as something that requires an extra step before it can actually be used? The help file has internal compression anyway, so the zipping up into the exe increases the size of the file, not decreases it. It's also a Windows only format (both as a .exe, and as the .chm it extracts to), yet by nature of being a web service, the mapping control is cross platform.

Update: Not only are many existing mashups using the previously published (v1) API broken, but as Microsoft's official developer site ViaVirtualEarth notes, the sample code given in the v2 SDK doesn't work either. In fact the only way Via Virtual Earth can get it's new sample to work is link to entirely different URLs for the Javascript and stylesheet files.

London and the UK revealed in very high detail

2005-12-03T12:15:00.000Z

The Google Earth dataset has been updated to add very high resolution (one pixel per 6 inch) aerial photographs of London in particular, but also a large number of other significant parts of the UK.

The BBS announcement lists:

London 2001,
Bedfordshire 2002,
Berkeshire-2003,
Birmingham-2001,
Bournemouth-2000,
Bournemouth-2003,
Bristol-2003,
Broadlands-2003,
Cambridge-City-2002,
Cambridgeshire-2003,
Corby-2003,
Coventry-2001,
Dudley-2001,
Exmoor-2003,
Farnborough-2004,
Gloucester-2002,
Gosport-2001,
Isle-of-Wight-2004,
Kingston-upon-Hull-2003,
Lancashire-2000,
Luton-BC-and-Airport-2003,
Manchester-2000,
Mansfield-2003,
Medway-2003,
Merseyside-2000,
Milton-Keynes-2000,
North-Lincs-2002,
Peterborough-Hampton-2002,
Poole-2000,
Portsmouth-2004,
Reigate_and_Banstead-2003,
Runnymede-and-Spelthorne-2002,
Sandwell-2001,
Sevenoaks-2001,
Solihull-2001,
South-Cambridgeshire-2003,
South-Oxfordshire-2003,
South-Yorkshire-2002,
South_Ayrshire-1999,
Southampton-2004,
Southend-2003,
Stansted-2004,
Suffolk-Coast-2000,
Surrey-1999,
Swansea-2003,
Tyne-Wear-Gateshead-2001,
Tyne-Wear-Newcastle-2001,
Tyne-Wear-North-Tyneside-2001,
Tyne-Wear-South-Tyneside-2001,
Tyne-Wear-Sunderland-2001,
West-Yorkshire-2002,
Wolverhampton-2001,
Wycombe-2003,
Wyre-Forest-2004

- Scotland (Aberdeen 2001, Craigmore 2003, Dundee 2001, Glasgow 2002,
Renfrewshire 2002)

- Northern Ireland (Northern Belfast 2001, Northern Carrickfergus
2001, North Downs 2001)

- Wales (Blaenau Gwent 2001, Bridgend 2003, Caerphilly 2001, Cardiff
2001, Merthyr Tydfil-2001, Newport 2001, Rhondda Cynon Taff 2001,
Torfaen 2001, Vale of Glamorgan 2001)

This imagery is provided by The GeoInformation Group, who also power the Cities Revealed website, where you can buy printouts of the data.

There are also smaller updates to other international data:

Updated/improved Digital Globe .7m data
---------------------------------------
- Alexandria, Egypt
- Asuncion, Paraguay
- Bamako, Mali
- Bangalore, India
- Cairo, Egypt
- Hyderbad, India
- Kano, Nigeria
- La Paz, Bolivia
- Mexico City, Mexico
- Montevideo, Uruguay
- Omaha, Nebraska
- Provo, Utah
- Pune, India
- Quito, Ecuador
- Sydney, Australia

NYGIS (updated imagery)
- Albany, NY
- Columbia, NY

USGS
- Las Vegas, Nevada
- Springfield, Massachusetts
- Mt. Helena, Montana
- Nashville, TN

Other
- fixed mask for Galapagos Islands
- Digital Globe imagery of Pakistan earthquake area
- Nanaimo, Canada
- NW Arkansas

Google Desktop plugin to search Google Earth files

2005-12-02T23:37:00.000Z

Google have released a plugin for Google Desktop which allows it to index and hence search Google Earth data files.

This was announced via a Keyhole bbs entry. After installing the plugin, any KML or KMZ files created or updated will be indexed - though as the readme states, existing files of these type will not be automatically crawled (though you can force them to be found by adding their location to the directories to be explicitly crawled).

It's author is Davis Eustis who appears to be a new hire - at least he hasn't made any previous posts to the Google Earth BBS.

A high definition web experience

2005-11-17T20:56:00.000Z

If you read a few of my posts, you'll notice that I'm not a fan of sites that use Flash, where particular examples which I've commented on include Flickr (now thankfully gone the AJAX route), Yahoo maps and most recently Google Analytics.

Robert Scoble just linked to the Flash Troll Generator, of which he says:

"Hate Flash? You'll like this site. Done by Oliver Steele. Personally I don't agree with Oliver."

This seems to suggest that Robert thinks this site is anti-Flash, whereas by trotting out the regular trolls, and labelling them as such, it seems to me that it's precisely the opposite.

On the other hand, Robert then notes, "It's time for a higher definition web", with the implication through proximimity that Flash is the way to a higher definition web (especially when you read his next post, which is about Laszlo mail, which is an email client written in Flash).

I'm fully in agreement with Robert in wanting a higher definition web, but Flash doesn't seem to me to be stepping in that direction.

A large part of higher definition to me means that I can zoom in at will to see the details I want. Examples of this in my regular browser, running regular HTML and AJAX based web sites are:

the literal ability to zoom in, by adjusting the text size as offered by most browsers
the more specific ability to zoom images, as offered by the Image Zoom extension in Mozilla or Firefox
the ability to make my browser window larger, to show more data (and yes I realize that not all HTML designers are capable of the fluid design thinking that makes this a reality)
the ability to target in on the text on the page I want simply by starting to type (the search as you type feature of Mozilla and Firefox)
treating any text visible on the page as a hyperlink by selecting it and selecting the "web search for selection" offered from the Mozilla or Firefox context menu
autolinking, an example of which is the Google toolbar which autolinks ISBN numbers to external references

All of these features work on all web pages - that's a high definition web experience to me.

There are plenty of other examples I could use which are currently more specialized, but which all look forward to an even higher definition web. These include:

being able to extract structured data from any web page, as shown by Piggy Bank
"on page" mashups, as performed by many a Greasemonkey script
tools that clip and blog the info on the page, whether to online sevices such as del.icio.us, or to local repositories such as Onfolio.

In contrast to all of these, tends to Flash provide me with a low definition experience. The data is trapped in what is often a fixed size box on the page. With it's gaming and presentation heritage, the artwork is often literally low definition - eyecatching no doubt, but not particularly useful. Simple web conventions and behaviours are broken - whether its the back button, chosing to open new links in new windows or tabs, bookmarking specific information, saving the page to disk etc.

I've no doubt that it's possible to program high featured applications in Flash, and Lazlomail is up there with the best of them, but for me high features on their own (rich client if you prefer that terminology), does not on its own make for a high definition experience, when I lose so much in getting there.

It's similarly possible to program AJAX badly, but when done well the result is also a rich client experience, without throwing out all the existing browser experience that makes for a high definition web.

Robert ends his note on Lazlomail with a comment that Lazlo are building a platform - and that Flash is just the first target of that platform, with AJAX yet to come. That gives me hope - we may reach a very high definition web when we can mix the existing high definition web browsing experience with the undoubted innovation that some rich client experimentation is showing us now.

Why does the Google Analytics UI suck so much?

2005-11-16T21:08:00.000Z

Google Analytics launched this week, and has certainly had its fair share of teething troubles in the first couple of days - with registration suspended for periods as the servers stuggled to cope with the demand.

However, now that I've managed to get it set up, I'm very unimpressed with the UI, which is poor in a number of respects.

First off, it requires Flash, which is an unnecessary overhead (and having taken the time to download and install Flash in my browser because of it, I find my general browsing experience is now much worse - distracting ads now appear on websites that previously were quite clean, and I've suffered an attack of popups that have been missing from my browsing for years now)
The meat of the data is within a seemingly fixed size rectangle, yet there's so much detail there that I need it to be resizable so that I can make it as large as my screen will allow
The maps are the worst that Google has ever produced - yet we know they have great mapping technology, so why dont they use it? Using the right mouse button there is a zoom function offered, so you can drill down into an area of the map in more detail, but if you do, then all the text tooltips also get bigger, and since they are limited to the map rectangle, they are thus truncated and unreadable.

The animation of elements is pretty annoying (for example bar graphs where the bars grow from zero up to their final length), and is wasting time - my need for web stats needs the data to be there for me to view instantly, not taking its time to appear as it currently does.

Having said all that, the rest of the experience outside of the Flash displayed content is actually quite good. There's a good amount of AJAX driven dynamism to the site, and the data is laid out reasonably (with the possible exception of a wide strip of unused white space down the right of the page - could this be reserved for advertising I wonder?)

I'm also glad to see that there is the facility to export the data in alternative formats of tab separated data, xml, and as Excel data. However, this always seems to be restricted to the current view - I can't see a way to export all the data at once, which would be very useful.

Another disappointment is the poor support for printing - there is a "print" button, but all this does is hide a bit of the navigation - the layout of the remaining items is not adjusted at all to take advantage of the capabilites of the printed page. Compare this to Google Maps, which makes sure that printed driving instructions look good on paper, with a quite different appearance to that which is appropriate for on screen display.

Update: Found another problem with the map - when you click on a result on the map, it shows you a set of details in a popup - but the popup always appears below and to the right of the point. So, when you have a load of points in Australia to look at, the popups are truncated since they can't go outside the Flash borders, so you can't see the data.

Official Google Maps API blog

2005-11-16T15:38:00.000Z

Bret Taylor, Product Manager of Google Maps and Google Local, has announced the opening of an official blog to support the Google Maps API.

The first post is uninspired, but it does promise:

An (approximately) weekly changelist listing the updates we are pushing to the API that week
Official announcements and updates
Workarounds for common API problems
Techniques for using the API and AJAX in general, e.g, techniques to avoid memory leaks in IE
API sites that have caught our attention

This should certainly be an easier communication mechanism to follow than the existing Google Group.

Grid coordinates in the Netherlands

2005-11-09T19:37:00.000Z

Whilst considering producing a mashup with a database containing locations in the Netherlands, I was dealing with coordinates in the Dutch nation grid system known as RD coordinates (Rijksdriehoeksmeting).

For one off conversions there is an online converter which will convert to latitude and longitude.

As is very often the case, the wikipedia article on the Netherlands is a very good general reference.. It explains

There is a west-east coordinate between 0 and 280 km, and a south-north coordinate between 300 and 620 km. The reference point is the Onze-Lieve-Vrouwentoren (Our Lady's tower) in Amersfoort, with RD coordinates (155.000, 463.000) and geographic coordinates ca. 52°9′ N 5°23′ E

Lots of releases today

2005-11-03T22:27:00.000Z

The past 24 hours has seen a flurry of new releases:

Google Desktop 2 has upgraded, and has supposedly graduated from being a beta. There is a new mapping plug-in in the side bar, but this in particular seems very beta quality to me. (It's very hard to use - the sidebar does not accept drag and drop, so you must type in where you want to go, and there is no "go" or "Search" button, so you can't simply paste with the mouse, since you must then touch the keyboard to use the enter key).
Ask Jeeves Desktop Search has been upgraded - see the blog for full details. One of the features they particularly claim is increased stability.
Yahoo have introduced a new mapping beta. This goes against the grain, and requires Flash, which I think is a backwards step, and also falls in to the same trap as Virtual Earth did of believing that America is the only location that counts.
Yahoo also slightly redeem themselves by offering a new mapping API, which does offer an AJAX based embedded map, and also a REST based geocoding service (again unfortunately only for America).

Virtual Earth mashups competition results

2005-11-01T08:30:00.000Z

Mashups using Microsoft's Virtual Earth have been slow to arrive - probably because although the API may be easier to use, the underlying mapping experience is much poorer (and noone really wants to tie their application yet to what was obviously a very rushed beta).

In an effort to kick start things, Microsoft's arms-length developer site organized a competition, offering a $1000 first prize, for mapping applications developed using their API.

The results are now available, with the top prize going to

MapStats which shows where website visitors come from. This rather neatly works around the severe limitation of Virtual Earth in that it only covers the USA in any level of detail, by not needing to show much detail! Since most websites get visitors from around the world, a world level view does just fine.

Honourable mentions are also made to

CREMaps, which would appear to be a property locator (but I could not get it to show me any results). It also has what I assumed was the (otherwise missing) map scroll control (containing arrows pointing in the 8 compass directions), but which turns out to be simply a branding logo.
Zoom to Hotels which presumably allows you browse hotels, but which failed to show me any map (getting a Javascript error instead).

It's apparent there is a sharp drop off in quality with these second placed entries. Perhaps they work better in IE rather than the Mozilla browser which I have installed.

(I feel that the competition may have shot itself in the foot - as well as the first prize, there were other lesser prizes offered to the "first 25 entries to arrive" - which rather makes a quickly coded but poorly produced entry a worthwhile proposition).

Bloglines interface changes

2005-10-13T07:37:00.000+01:00

Bloglines have made a number of interface changes, perhaps prompted by last week's introduction of Google Reader.

The ability to step through posts using the keyboard (j - next article k - previous article s - next sub f - next folder A - read all r - refresh left pane)
The "unread" count associated with each folder is now a two part number - the number of new messages, plus the number of "pinned" messages

The keyboard hotkeys are certainly one feature Google pushes hard, but not one I actually use. In many areas I'm a keyboard junkie - using a command prompt in preference to flashier but ultimately less useful visual tools most of the time, but for me, feed reading is an entirely mouse driven experience. This is where Bloglines, with its plain HTML rather than AJAX interface wins massively for me - the mouse wheel lets me scan through lots of posts very quickly since Bloglines presents the data in a long pane, with a standard scroll bar.

I pin a lot of messages, so the change to the way of presenting the count I find very useful. I'd like to take it a stage further, and have a way to just show me real new messages, without the pinned messages.

Update: The new features are now mentioned on the Bloglines About page.

Bigger maps at Google Local

2005-10-12T13:59:00.000+01:00

Google Maps (aka Google Local) have rev'd their javascript again, to version 25.

Not sure what the changes all mean, but I have noticed that if you type in a placename into the search box, such as "New York", you now get a full width map - the results panel down the left is removed and given over to showing more map.

Hurrah - perhaps they can also do something about reducing the height of the banner across the top of the page, so that we can also get to see more map vertically.

Google Earth and large data files

2005-10-07T19:12:00.000+01:00

As people discover the power of Google Earth, they are producing larger and larger collections of placemarks.

Some fairly large collections include:

my own Countries and territories of the World with ~260 entries
UNESCO world heritage sites with ~790 entries
World Club football teams with ~1800 entries

However, these are all eclipsed by two collections (presumably produced from the same raw data) of about 36500 cities in France: Collection 1, Collection 2.

Beware though - Google Earth shows why it it still in beta - it's extremely unstable, and when given sizable files is very prone to crash. Note that a crash will lose any placemarks you have accumulated since you last started Google Earth - even if they are in the "permanent" section of the placemarks (ie. not just in the Temporary Places section).

So Google Earth team, please get your act in order, and do two things:

write placemarks to disk as soon as you can, so they are safe if the program crashes (ie dont just wait until the user exits the program cleanly before you write anything to disk)
add some proper validation so that it is not possible to crash the program with either malformed KML, or simply large collections of placemarks.

I find that the safe limit on the number of placemarks in a file is somewhere in the low hundreds - I've not had a problem with my countries file (and noone else has reported problems with it to me), but stepping up beyond 500+ placemarks more often than not causes a crash.

iPod iWay providing maps for iPod and nano

2005-10-07T08:29:00.000+01:00

iPod iWay are offering a service that packages Yahoo Maps driving directions for display on the iPod or iPod nano.

The proceedure is to enter your source and destination addresses, which you can then preview to check that the directions are OK. When satisfied, you then select the "get directions" button, which packages them up into a zip file which you down load, and unzip into your iTunes folder, from where it is transferred to the iPod. The directions are a series of GIF images, comprising the text directions converted to images, along with the turn images that Yahoo already provide.

The service could be improved in a number of ways:

this is one example where frames makes sense - have the preview in a frame, so that you can still see the "get directions" button, without having to juggle multiple browser windows or tabs
There is no overall view of the route provided, either as a map, or as text image - so information that Yahoo provides is missing from the iPod version - I'd suggest adding these, so that you get the overall distance and time estimate, plus the full starting and ending addresses
images are all very well (and its a neat idea to convert the turn directions text into an image alongside each turn map), but also including all the text data as a single text file as well would not add much to the download size, and would prove useful

However, the biggest drawback is outside of iWays' control - Yahoo Maps only offers directions for USA and Canada!

The demise of Google Maps

2005-10-06T23:21:00.000+01:00

Google Maps is no longer - it has been absorbed into Google Local, and the combined product, under the banner of Google Local is now out of beta.

This transformation took place in a number of stages yesterday - I first noticed the map had swapped sides, but then there was a further round of changes - the javascript has been revised again, to version 23.

I've taken the time to redo my CSS file which modifies the mapping UI. The main change is to make the map occupy the whole of the browser window, and to use transparency for the other controls that get placed in front of the map.

The revised CSS file is now available - as before I recommend using the Mozilla/Firefox developers toolbar to apply the CSS file to the page.

Google maps new layout

2005-10-05T10:02:00.000+01:00

Google Maps has a new layout, where the most obvious change is that the map has moved from the left to the right of the page.

Unfortunately this breaks my CSS redesign of the page - I'll let you know when I have had a chance to fix it to work with the new version.

The javascript indicates it is up to version 22, in just 8 months (the original maps went live on 8th Feb this year), so the rate of change of the service is very fast. I wonder why they didn't take the opportunity of this redesign to add the one useful feature that MSN Virtual Earth has that is missing from Google Maps - the ability to change zoom levels with the scroll wheel.

Update: 6/10/05 I've updated the CSS file now. It's now apparent that the early changes I spotted were part of a much larger change - where Google Maps has now been absorbed into Google Local.

Concorde Compendium

2005-09-04T21:21:00.000+01:00

To follow up my Countries of the World data file for Google Earth, I went for something entirely different. Having seen that Concorde pops up as a regular "find" on the bbs, I produced from scratch a file that locates all 20 Concordes, linking to a page about each specific aircraft, and also where possible to the museum that now houses it.

A good number of the aircraft cannot actually be seen in Google Earth - because they are inside the museum buildings where they can be protected from the elements.

The Concorde Compendium is primarily a Google Earth datafile available via the Google Earth forums, but it also comes in a web page version.

Countries of the World in Google Earth

2005-08-25T23:09:00.000+01:00

Having used Google Earth for a while, I decided to come up with my own data layer for it. The result is a data file covering all the countries (and a few dependent territories) throughout the world.

For each country, it shows the flag of the country, and provides links to

* Books about the country (via Amazon)
* facts from the CIA World Factbook
* Google search for the country, plus images and news
* Lonely Planet info about the country
* the country's entry in the Wikipedia

The file is proving quite popular on the official Google Earth forums at

Countries and territories of the world

You can also access the kml file from
http://www.zmarties.com/earth/about/countries.htm

That page will also take you to a web page based version of the same information, with additional links to maps in Virtual Earth, Google Maps, MultiMap, and Yahoo! maps.

The smallest possible zip files

2005-08-17T22:31:00.000+01:00

Whilst there are many file formats that offer better compression, the ZIP file format is still very widely used, and in fact experiencing a surge in popularity again, since it is the basis for a number of other file formats, including

jar files used in Java
kmz (compressed kml) files used by Google Earth
Microsoft's upcoming xml and zip based Office document format
OpenOffice's document format

With this in mind, I did a quick and largely unscientific test of a number of Windows based command line zip utilities, to see which one produced the smallest files. In all cases, I set the parameters to give what I believed was the most compression. I was compressing just one file at a time (since that was what my particular scenario needed).

The results were:

pkzip command line version 4.0 (parameters -level=9) produced a compressed file which I considered my baseline file size
info-zip version 2.31 (parameters -J -X -j -9) produced a file that was about 0.5% smaller than the baseline until the compressed file size was about 3K, at which point the file size tended to be about 1% larger than the baseline file size
7-Zip version 4.23 (parameters -tzip -mx9 -mpass=4 -mfb=255) produced a file that matched the best of the previous two for very small compressed file sizes

So the overall winner was 7-Zip - across my selection of 42 files of varing sizes, it produced a total compressed size of 104K, compared to 109K for each of the other 2 programs.

Innovative uses for the iPod

2005-08-17T14:22:00.000+01:00

There are a lot of innovative uses being found for the iPod, simply using it as a client that millions of consumers carry around with them:

http://www.ipodsubwaymaps.com/ has subway maps from around the world, formatted to fit on the iPod screen
http://www.ipodlinux.org/Doom Doom has been ported to run on the iPod
http://www.makezine.com/blog/archive/2005/06/make_ebooks_for_1.html Producing ebooks for reading on the iPod
http://www.zapptek.com/ A number of applications, including presentations, email, RSS
http://macs.about.com/od/ipod/a/ipod_email.htm A round up of the applications that do offer email on the iPod
http://macs.about.com/od/ipod/a/free_ipod_utils.htm A round up of free iPod utilities

Google using their own mapping API

2005-08-11T11:39:00.000+01:00

Whilst its not surprising for Google to make use of their mapping capabilities elsewhere on their own site, I was slightly surprised when I came across the Google Mini success stories map to note that it's actually using the published API (complete with registration key), rather than the private methods that most other maps from Google have used to date.

Google News adds RSS/Atom feeds

2005-08-09T22:57:00.000+01:00

Finally, Google is waking up to syndication technology.

Google News is offering RSS and Atom feeds for both its pre categorized news, and also for news search results.

Unfortunately News searches seem to have a 10 word limit, which means it's hard to remove noise from the results using lots of "-" terms, but at least they do respect the "&num=100" tagged on the end of the URL, so we can now get feeds with a respectable number of items in, not just limited to 10 results as other providers are.

The feeds are in RSS 2.0 or Atom 0.3 format, and give the generator as "NFE/0.8" (which is simply the webserver that all Google's news sites use). Feeds are available on the .com site, as well as UK, Canadian English, India, New Zealand and Australian news sites - basically all the English language sites.

This follows soon after Google added the ability to add any feed to is personalized home page.

Space Shuttle, and other maps

2005-07-27T08:05:00.000+01:00

More worthwhile mapping mashups and resources:

track the Space Shuttle and ISS
Global Coordinate presents all manner of worldwide data on Google Maps, and also offers a Landsat view as well
UK weather reports
zoom to a rectangular area
USAPhotoMaps - an offline program that makes the most of data it downloads from TerraServer

Improving the UI of mapping sites using CSS

2005-07-26T19:13:00.000+01:00

When run in Mozilla/Firefox, MSN Virtual Earth does not have quite the same visual appearance as it does in Internet Explorer - specifically the banner header holding the search controls at the top of the screen is solid in Mozilla, but is semi-transparent on IE.

I realized that this could be fixed by a simple change to the CSS. Whilst I was at it, I went further to see what else I could improve in the Virtual Earth UI, using just CSS. The changes I put in my CSS file for Virtual Earth are:

make the banner semi-transparent, as it is in IE
move the menu up to the top of the banner
remove the "about" menu item, which adds nothing one you have read it once
move the scale to the corner of the map, and give it a solid background, so that it can actually be read
shrink the size of the zoom control slightly, and make it semi-transparent to obscure less of the map
move the compass rose, so it's not so in the way, and make it semi-transparent (but less so when you actually mouse over it to use it)
fix the size of the "x" box in the corner of info windows
increase the font size for permalinks, to make them easier to read and click on
hide the what and where help text hints
Stop the search result titles SHOUTING

I also made a similar CSS file for Google Maps, which gives a similar set of improvements:

enlarge the map to take up the whole of the browser window
place the results window in front of the map, in a semi-transparent form, which becomes more solid when the mouse moves over it to use it
reduce the size of the search form, which is similarly semi-transparent
remove the logo and banner which were taking up too much space
remove the text labels that just duplicate icons

These CSS files can be used in Mozilla by using the "Add User Style Sheet..." option in the Web Developer toolbar.

Update: 5th Oct 2005: There is a new layout to Google Maps that means this CSS file no longer works. I'll have to see if I can find time to fix it for the new layout.

Update: 6/10/05 I've updated the CSS file now.

VirtualEarth from MSN

2005-07-24T17:12:00.000+01:00

MSN have launched Virtual Earth - their attempt to catch up with Google Maps.

Having tried it out for a few hours, I think they've generally missed. There are a few areas where they exceed the capabilities of Google Maps, but these are outweighed by those where they fall short.

The capability that is likely to get the most press is the fact that they show the USGS colour data at full resolution (ie at the resolution that Google Earth uses, not the reduced resolution that Google Maps currently displays). However, outside of the USGS colour areas (around 30+ US cities), they only show the black and white USGS data - which is a very poor substitute for the sometimes lower resolution, but coloured photos that Google uses. Google also has detailed coverage for many cities throughout the world, whereas Virtual Earth is limited to detailed data in the USA only.

A possible other reason to prefer VirtualEarth is it has a larger part of the screen given over to the map, since it does not reserve space alongside the map for search results.

The list of areas they fall short of Google Maps is longer:

This is an extreme US centric launch - there is no attempt to provide photo data yet for the rest of the world, beyond the 1 pixel per km NASA Blue Marble image, and the "road map" view does not show any road data outside of the USA (showing just limited place name data).
Keyboard navigation is harder with Virtual Earth - no use of pageUp/pageDown to move the map in big chunks (but I do like the fact that the scroll wheel can be used to zoom)
Virtual Earth uses large, opaque, and ugly controls for zooming and scrolling
Bringing up help places a pseudo popup window in front of the map - so you can't view help and the map at the same time
Unfortunately the Virtual earth view of the world is flat - you cannot scroll west from the USA, but must scroll east to get to anywhere else
The distance scale shows miles only - no metric measurements
It is not possible to scroll around the whole world at a particular zoom level - when the data runs out (generally when you leave the USA) the system simply locks you from scrolling anymore in that direction, rather than showing you the data at whatever resolution it may have

Virtual Earth comes with what they call an API - which simply means that they have documented the URL format used to bring up a page at a specific location and zoom level etc. Unfortunately the URLs use the pipe character "|" when constructing the URLs - I've always considered this an invalid character (though the RFC may be ambiguous in this respect), and I know of plenty of tools which will not cope with such URLs - wikis for example, which often use the pipe character to separate a URL from the label a link will use. It's worth noting that its possible to separate the latitude from the longitude in the URL using a comma instead of the pipe character, but they really need to remove the pipe character altogether.

Hybrid mode for Google Maps

2005-07-23T11:25:00.000+01:00

Google Maps has gained a hybrid mode, which overlays mapping data on top of the satellite images. To achieve this, they have had to rescale their photographic images, which were not to the same scale as the maps. This appears to have been done by resampling the existing images, so for example the Google copyright watermarks have been stretched along with the rest of the images.

In most cases the maps and the images line up well - though as always people take great pleasure in finding the exceptions. A discussion on Slashdot found the following examples:

just minor misalignment in Toronto, Canada
misalignment at Camden Town, London
the creation of imaginary roads (there is no E 42nd St), and E 43rd St does not exist beyond the crossroads - its a drainage ditch!
the satellite image is more recent than the map in Ottowa, Canada - the road has been repositioned south of the building
the map data is more recent than the satellite in Westminster, Co

Update to Yahoo Maps API to allow POSTING

2005-07-18T23:16:00.000+01:00

Whilst I'm unconvinced of the general usefulness of the Yahoo maps API (since the maps cannot be integrated or otherwise embedded into other pages), they have just made them potentially a bit more useful.

The original API took a URL that supplied the RSS feed describing the map points. This URL therefore had to be accessible to Yahoo's servers - thus ruling out use on intranet sites, and on many sites where the data could not easily be written to a file.

The changed API allows the data to be supplied directly via the API, in the form of an HTTP POST. Details in a new POST section on the API documentation page.

Google Maps showing a scale

2005-07-18T10:16:00.000+01:00

Google Maps have started to show a scale on their maps, in the position previously used by the Google logo.

This scale does not always appear on the maps, but does seem to be there whenever you go to a map via a "link to this page" URL. The scale is provided with the tooltip "Scale at the center of the map" and gives a horizontal scale. As you scroll the map, the scale changes dynamically as the area in the center of the map changes, which dramatically illustrates the fact that the scale is not constant over the whole of the visible map. No attempt is made to provide a scale in the vertical direction, which is not normally the same as the horizontal one.

The scale is shown for both the maps view, and the satellite view, and gives both imperial and metric units (with the unit changing as appropriate to keen the numbers managable).

I've not been able to determine if there is a way for applications using the Google Maps API to get to show this scale.

Update: Its not been documented yet, but the API now supports the Scale control via a command such as:

map.addControl(new GScaleControl());

Nuclear blasts on Google Maps

2005-07-15T08:13:00.000+01:00

Eric Meyer has produced an interesting, if slightly bugged application of Google Maps with his HYDESim maps which are an indicator of structural damage to buildings from a nuclear blast.

The maps show rings of overpressure caused by the blast - which is unfortunately where they go wrong. Google Maps do not have the same scale both vertically and horizontally, so these rings should really be ellipses. The error is not too bad in much of the mid-USA, but moving further North or South the errors mount up, and switching to satellite mode the difference can be even greater. For example at -96.24298095703125, 63.87451171875 on the satellite image the ratio between the vertical and horizontal scales is roughly 23:10.

Pixsy - the random image search engine

2005-07-14T08:21:00.000+01:00

Pixsy is a new image search engine that "searches blogs, mobile blogs ("moblogs"), social networks, and other sites that have unique content" in order to find images that most search engines miss.

That's all very well, but unfortunately viewing any of the results seems to be a lottery. Like most engines, it does not show all the results on one page, but offers Next and Previous links. The URL of a results page is currently http://www.pixsy.com/Search.aspx?next=true - note that it has a parameter of next, that steps you through the groups of results each time you get to that page. Thus the standard operation of "view an image then press back to get back to the results page" does not take you to the same results page, but instead advances you to a new set of results.

Even on a single page of results, the details links are always numbered from zero for the first visible image, (so the first image on the first page and that on the second page etc all have the same URL), so it seems to be random when you view the detailed image result as to whether you will get the image you thought you would, or rather some other image that shares the same URL on a different page of results.

Weather maps of the USA

2005-07-06T08:49:00.000+01:00

It may not last long at this URL (which is just a development one), but Weather Maps is presenting Weather Underground information on Google Maps.

The markers for each point artfully show all of temperature, wind strength and direction, and weather observation, and the popup info window adds humidity and a temperature chart.

There are also links to webcams so you can see the realtime weather, and data can be downloaded for display in Google Earth.

Update: More weather presented on maps at:

Roundup of more mapping highlights

2005-07-05T22:45:00.000+01:00

Google Maps

GMaps Pedometer is another implementation of measuring the length of a path you click on the map
Mobile GMaps is free software to display Google Maps imagery on Java J2ME-enabled mobile phones
Google Maps Transparencies overlays a map directly onto the centre of a larger satellite image - which does show that the two do not line up exactly
UPS tracking with Google Maps
Scavengeroogle - a scavenger hunt to track down images seen at Google Maps

Google Earth

Network Link - an explanation of the feed subscription and location based URL format used in Google Earth
GoogleGlobe serves up Google Earth placemarks for user entered locations
Chicago real estate in Google Earth format - its about a 500K file, updated daily, so they really should be using the compressed .kmz format which would cut the size to 100K
GeoData from the German Wikipedia - covering about 2600 locations worldwide that have been geotagged in the German version of Wikipedia

Official Google toolbar for Firefox on its way

2005-07-04T07:27:00.000+01:00

Fritz Schneider has contacted the Googlebar developers to let them know that Google are expecting to release an official Google Toolbar for Firefox this week.

It will work with Firefox 1.0+, on Win2k/XP, MacOS X 10.2, and Linux 8.0+. With the exception of the popup blocker (which Firefox already has), it will match the features of the latest IE toolbar. Thus we can expect spellcheck, autolink functionality, and probably a PageRank display.

For a while now, if you visited the Google toolbar page using Firefox, it suggested that you install the independently developed Googlebar thus:

Googlebar: We encourage Firefox, Mozilla, and Netscape® 7.x users to install the open-source Googlebar, developed and supported by Google fans.

Google Earth feeds and files

2005-07-02T10:45:00.000+01:00

Google Earth is the downloadable application from Google that uses the same satellite data as Google Maps, but which literally takes it into another dimension. The images can be scrolled as with Google Maps, but can also be viewed from an angle and rotated - and since the data is augmented with altitude data, some of these views can be most spectacular.

[The downloads page has been temporarily disabled, whilst Google increase capacity of their servers]. The free entry level version of the application runs on Windows, but the enterprise version runs on Linux.

Data for overlaying the images with points of interest is stored in KML files. Google have put together documentation for the KML file format so that other people can produce these files.

You can search via Google for kml files (and their compressed kmz equivalent), though beware that these will also bring up some non-google earth kml files.

search for kml files (a claimed 5430 files at the time of writing this post, but after removing duplicates only 86 remain)
search for kmz files (a claimed 73 results at the time of writing this post)

Specific feeds of interest include:

USGS Earthquakes
Latest 50 geotagged flickr photos, updated every 30 mins
Closest 50 geotagged flickr photos within 100km to the point you're looking at.
Closest 50 geotagged del.icio.us pages within 2000km to the point you're looking at.

More static files of interest include:

An overlay to update the image of a Chicago airport
Tourism in Monaco

script.aculo.us - visual effects, and drag-and-drop framework

2005-07-02T10:18:00.000+01:00

script.aculo.us is a javascript framework which provides (currently) a number of visual effects, and drag-and-drop support. It's available for use under an MIT style licence.

There's an example puzzle, to illustrate the drag-and-drop, that although very simple is quite entertaining.

More Google Maps API applications

2005-07-02T09:37:00.000+01:00

Google Maps API applications are coming thick and fast now. A few that caught my eye:

amateur radio position tracking - illustrating the use of locked together side by side map and satellite images
Boston subway map - artfully uses different coloured polylines to show the different subway lines
Puget Sound Dive Maps - nothing particularly unique about this, just a nice example of how individual pages can make use of the maps

Lengthfinder - Google Maps based app for measuring distances

2005-07-02T09:13:00.000+01:00

LengthFinder is a Google Maps based page that allows you to interactively plot a path on a map, and display the length of that route.

The developer is Australian, so the map (or actually satellite image) comes up centred on Sydney, and distances are given in Km only. There is also a conversion to walking/cycling time - I guess they walk slow in the Australian heat since the conversion assumes just 4Km per hour walking pace!

Another maps web service

2005-07-01T08:30:00.000+01:00

Plotted.com is another site which aims to simplify using the Google Maps API, by simply serving up a map as a web service.

It also supports geocoding of locations, and is designed to be placed in a iframe.

Greasemonkey script to add Google Map to any page

2005-07-01T07:47:00.000+01:00

Greasemap is a Greasemonkey script which adds a Google map to a page if it finds any geocodable addresses or other geotags on the page.

The map is placed in an iframe served up from Vinq.com - which gets around the restrictions that the Google maps key registration mechanism imposes.

Currently recognised locations are of the form

123 foobar street, anytown, MN 95125
<meta name="geo.position" content="41.328,-110.292">
<meta name="ICBM" content="41.328,-110.292">

The actual script is very short - it simply packages up any geotags on the page, together with any addresses recognised by a regular expression, and then passes it via a URL to the server which produces the iframe contents to be placed in an injected iframe tag.

All geocoding of addresses is done on the server. The server is thus a useful webservice in its own right. Take a look at:

http://www.vinq.com/greasemap.html?locs=801+spring+St,Seattle,WA

- it produces a twin map layout, showing a wide angle view, and a close-up.

Yahoo maps API

2005-06-30T21:23:00.000+01:00

As well as Google, Yahoo have also launched a mapping API. Whilst the Google API has it's fair share of problems, the Yahoo API is quite simple - it's such a shame that the service is such a poor one, so as to make the API a non starter as far as I'm concerned.

Jeffrey McManus (who works at Yahoo) has a 7 point list as to why he thinks the Yahoo API is better than Google's, which comes down to this:

The Yahoo map is hosted on Yahoo - so this means that you can't integrate it into your site - what good is that?
Yahoo allows unlimited maps to be served up - whereas Google have a limit of 50K per day per page - that seems very generous to me
{juvenile name calling}
Yahoo allows US street addresses as well as the Lat/Long that Google uses - but Google's maps are worldwide, whereas Yahoo is limited to the USA only
{repeat of juvenile humour}
Yahoo have lots of other API's for other types of data - agreed - I'm a big fan of their other API's (which pass back data you can then serve up presented as you want), but that only serves to show how poor the Yahoo mapping API is
Yahoo uses modified RSS to specify points to show on the map - agreed this is currently simpler than Google's arrangement

So by my reckoning, there are 2 strikes against the Yahoo API here - the maps can't be embedded, and are USA only, with only the simplicity factor in their favour. I really can't see many people wanting to use this API at all.

The cream of the Google Maps API

2005-06-30T20:08:00.000+01:00

Google Maps has finally gained an official API, (though being beta, it's almost as likely to change as the unofficial API everyone has been using up to now).

Use of the API requires a key, which is causing no end of problems, since keys are tied to partial URLs:

you cant get a single key that works across more than one of example.com, www.example.com, example.co.uk
you cant get a single key that works when your page is migrated from example.com/news.htm to example.com/archive/june.htm
if your site uses URLs of the form example.com/tag/page.htm where tag varies, then you need a separate key per tag.

Lets hope they fix that ASAP, since thats a huge flaw in the implementation.

There's a very lively discussion group for the API, with lots of good questions, answers, and examples being posted already. A few that caught my eye are:

discussion on how to add a scale bar to the maps
calculating the distance between two points
scaling items with the zoom level
adding a side panel to the maps

Example uses of the API include

MapMaker which allows you to interactively add a path overlay to a map
Paris Star Forts which illustrates adding a side panel

Other interesting items to come out of the discussions:

geocoder.us which can translate addresses to latitude/longitude

More Google Maps fun

2005-06-14T20:52:00.000+01:00

A few more Google Maps data presentations:

gCensus presenting US census data
zoto photo sharing placing geotagged photos on the maps
gTraffic.info providing UK-wide traffic info, plus traffic cams in central London
dynamite local providing access to a lot of UK datasets, including weather, flickr photos, travel info, geocaches

Smart commas, but stupid ampersands

2005-06-12T22:30:00.000+01:00

Getty I mages are indicating that their search now uses "smart commas". The text describing this feature notes that:

If you're using words in the search field for your keyword searches, commas now work like AND:
- Put "tom cruise, close-up" in the search field
- Your results will return images that contain close-ups of Tom Cruise

If you're using image numbers in the search field for your keyword searches, commas now work like OR:
- Put "1945157, 1951923" in the search field
- Your results will return both image # 1945157 and image # 1951923

I approve of this - this seems a very sensible behaviour. However, the text also notes that:

You can also use an ampersand (&) in place of AND. But be sure to add a space between keywords and "&".

I don't understand this restriction at all - the examples show that you dont need to surround a comma by spaces, but you do if you use an ampersand. It must have been almost as hard work to write up that ampersands don't work as it would have been to fix it so that they do!

Why Google maps does not have a scale

2005-06-09T12:54:00.000+01:00

One of the missing features from Google maps is a scale. Whilst not impossible to provide, there is one very good reason why this is not provided - Google maps are not drawn to scale!

This Slashdot comment titled Google Maps are awfully distorted anyway explains that

Google Maps uses a fixed longitude/latitude distance ratio of ~0.772, while the true ratio depends on latitude (the ratio should be cos(latitude)). So Google Map is optimized for 39.5 of latitude (N or S), and the maps are increasingly distorted as you go toward the poles or the equator.

The example given shows some streets in Anchorage which are really at right angles, but which show up as a diamond mesh due to the stretching. (Ignore the follow up comments that claim the original post is wrong - they are pointing out north south aligned grids, where of course the right angles are maintained by the stretching anyway). If you want a further example, switch to satellite mode, and observe the round structure to the north west of the runway - which due to the stretching appears as an ellipse shape.

Translating UK locations to latitude and longitude

2005-06-08T21:08:00.000+01:00

The UK has probably the best national mapping service in the world, in the form of the Ordnance Survey. Their mapping data is used for almost all maps produced in the UK, and UK locations are normally specified using a National Grid Reference - which overlays a grid based on the metric system over the UK.

Google maps is different, even in its UK incarnation - its coordinate system is the universal latitude and longitude, which applies worldwide. The widespread use of GPS has also pushed latitude and longitude to the fore as an everyday system.

Thus there is an increasing need to convert from NGR data to lat/long data, and a few resources which can help are:

There is ready written code available to do the conversion in

UK postal addresses also have postcodes, which might similarly be converted to latitude / logitude locations. Resources to help with that are:

UK Postcodes

Piggy Bank brings the semantic web here and now

2005-05-25T23:23:00.000+01:00

Piggy Bank is the rather strangely titled Firefox extension that tries to turn the existing web into the semantic web - that is to say that it aims to allow extraction of structured information from the (on the surface) unstructured data presentation that makes up most webpages, and then to go on and use that structured data.

The extension, a substantial 4.8MB download, is the result of a joint project of w3c, MIT, and HP. Many other open source projects have also been utilized in producing this one.

It provides a mechanism to capture information from the web, to save it as structured data, to tag it, and to search through and combine that information in many ways. In particular, it integrates the Google Maps mashup mechanism, allowing the data to be seen in its geographical context. Data can also be shared, by publishing it, perhaps to a central "semantic bank".

The collection of data either uses some in built collectors (for existing structured data such as RSS feeds or RDF data), or relies on specially written screen scrapers which can extract structured data from particular web sites.

Unfortunately the use of screen scrapers is often fragile - depite being launched just a couple of
days ago, I could not get any of the sample screen scrapers to work - the web pages they target seem to have changed in such ways that the scrapers no longer work. Instructions are given for writing new screen scrapers - they are written in either XSLT or javascript, and their job is to take (generally) HTML, and extract the structured data from it into RDF form.

All in all a very interesting project, pulling together a lot of smart concepts - I will be following this closely.

Drag and drop Google results

2005-05-23T20:17:00.000+01:00

To showcase their new AJAX like toolkit, Backbase have put together some demo sites making use of their technology.

One such demo, BBGoogle, uses the Google API to get search results, that can then be manipulated via drag and drop - at the moment the order of the results can be changed, and you can drag results out into a bookmark area. You can also drag the three "windows" around within the page, and whilst the bookmark window grows as necessary to accomodate all your bookmarks, the search results window is of fixed size - a poor design, but I guess that wasn't what they were trying to show with the demo. Overall an interesting demo, but not all that useful as it is.

Unlike the Google drag and drop implementation used by Personalized Google which updates the layout as you drag without needing you to drop it, this implementation only does the action when you drop, which doesn't quite give enough feedback for my liking in this case. Its worth noting that Google is already on (at least) it's second attempt at producing a drag and drop personalization UI - there's also one used in personalization of the Google News results. The News version does not move the item until you release it, instead providing feedback by placing a coloured rectangle where the moved item will end up.

Flickr is now reading IPTC information from images

2005-05-16T07:41:00.000+01:00

Flickr has announced that it is reading IPTC information from images - in fact this was announced before the news about Flash, but only on the Flickr news feed, (which rather stupidly requires a login) rather than on the Flickr Blog (which anyone can read).

The discussion page on the feature documents the rules they are using as:

IPTC keywords, when available, are added as tags.
If there is no title specified (i.e., we were going to fall back on the image filename) and there is a IPTC *headline*, then we use that.
If there is no description provided with the upload, then we use the IPTC *caption* when available.
If there is no EXIF date, then use the IPTC *created date* when available.
Add the contents of *city*, *province state* and *country* as tags when available.

However, the developer then goes on to say

"we are IPTC neophytes and we probably looked at an old spec. We will review the spec you pointed out and review for changes to make. Do you have any partcular recommendations for mappings?"

so things could well change as the feature is worked upon.

Flickr ditches Flash for photo display

2005-05-14T09:24:00.000+01:00

Flickr is dropping Flash in favour of Javascript (DHTML, AJAX - call it what you will) for the simple act of displaying annnotated photo.

This is great news, fixing a major usability problem that plagued Flickr - obviously the level of complaints had got so loud they could not be ignored.

Update: Changed to clarify that is just the photo display function that is dropping Flash entirely. Flash will unfortunately continue to be used on other pages. There is detailed coverage of the change at Ajax Summit: Eric Costello of Flickr Presentation which gives the story direct from Eric Costello, a Flickr UI developer.

More Google Acquisitions

2005-05-14T08:47:00.000+01:00

I produced a list before that listed all the companies that Google now owns. It's time for an update to that. The extra entries are:

dodgeball.com - mobile phone community
Urchin - website analytics
zipdash - maps and traffic info for mobile devices
Where2 LLC - Australian developers who produced Google Maps

Update: Thanks to the anonymous comment that alerted me to the Where2 / Google Maps connection, and pointed out an interesting forthcoming keynote to be presented by the Google Maps developers.

Google Maps available for the UK

2005-04-19T23:30:00.000+01:00

Google Maps has expanded its coverage to the UK. Although it is available via the UK domain name, it still seems best to get at it via the main .com address, since that offers the "satellite" view that is missing from the UK domain entrance.

The maps used are unfortunately vector based maps - great in one respect in that they are easy to offer in multiple zoom levels, but they really do produce a poor quality map. The UK is very fortunate to have some of the best maps in the world, as produced by Ordnance Survey, and rival online mapping services at multimap and streetmap use that data.

The vector data also seems to be rather shot through with errors - I've found misnamed roads all over the place, and even when names are correctly named, the algorithms for automatically placing names on roads do a far worse job than individual human placement can achieve.

The search features now accept postcodes, but a national grid reference is not recognized, which is a shame since it's so easy to do, and that is how most UK locations have been accurately specified for decades.

The satellite data view is only available at very poor zoom level - again rival services are still well ahead of Google on this one.

The travel directions feature seems fairly competant, though I noticed that it tends to quote longer journey times than other offerings do - perhaps it is trying to apply US speed limits to UK roads.

Google Maps also seems to have acquired a prominent "Terms of Use" overlay link which appears on all maps - that's interesting to see , especially as Google have officially acknowledged the genius of Paul Rademacher Craigslist/Maps combo real estate listing, via endorsements on both the Official Google Blog and as the featured project at Google Code.

Autostitch for creating panoramic images

2005-04-14T21:05:00.000+01:00

Autostitch is a program that is capable of producing some quite amazing stitched panorama images with little effort.

The program is available as a fully working demo - which takes a directory of images, and produces an stitched panoramic image as output. "Advanced" controls allow quite a bit of flexibility in the process, but just letting everything default produces a good enough output with a significant "WOW" factor in most cases. The one thing many users will want to change is the output size, since the default produces a screen sized image, which is often smaller than just one input image given most of todays digital cameras.

There has been a good discussion about the program in the Panoguide forums. The researcher who produced this program also worked as an intern at Microsoft, and there is a similar feature built in to Microsoft's Digital Image Suite 10 based on some of his work.

Not AJAX, just CSS on its own

2005-04-14T20:49:00.000+01:00

AJAX is pretty much the flavour of the day for interactive websites, but that need for Javascript to be turned on in the browser is occasionally a problem.

Stu Nicholls, at the CSS Playground specializes in producing effects and demonstrations that use (or ab-use, dependent on your point of view) only CSS with no Javascript. He goes to a lot of effort to make his examples work in multiple browsers, yes even IE with its rather broken CSS support.

The photo galleries Mk I and Mk II are probably the most fully featured examples, but there are plenty of other great examples as well.

One thing I did notice as I toured the site (which practices what it preaches, in that the site as a whole uses many of the effects and ideas to good effect) is that CSS styled controls may look like standard controls, but at least in my browser (Mozilla on Windows) they can fall short on usability. In particular, there are a number of controls that show a scroll bar, but for which the mouse scroll wheel has no effect.

Why is this area obscured in Google Maps?

2005-04-11T08:41:00.000+01:00

I was looking for Southampton College, Long Island University, which I found in Google Maps fine in the map view. However, when I switched to the satellite view, the area has been deliberately obscured.

I wonder why this is - it's apparent that there has been quite a bit of heavy construction in the area when you compare the edges just outside the obscured area with the 1994 aerial image (black and white) which is available via TerraServer.

Blog entry spam by WebLogs Inc

2005-04-11T08:38:00.000+01:00

I read a number of feeds from blogs by WebLogs Inc.

This morning I find that they have engaged in a massive spamming operation of their own feeds. It appears that they have some new blogs to publicise, and instead of simply adding a new entry to their existing blogs (which I'd consider poor judgement since the new blogs have nothing to do with the subjects of the old ones), what they have done is orders of magnitude worse. They have actually added a new line to every existing blog entry - this means that suddenly my feed reader shows lots of apparently new blog entries for each of the blogs. In addition to the huge inconvenience this causes, wading through new entries that aren't, this also destroys valuable metadata in each of the posts, namely the last updated time. Previous to this spamming, I could see when a blog entry had been updated after it was published (perhaps to correct a fact that was wrong in the original post), whereas now all entries have an updated time of when the spamming was added.

This is certainly a very poorly judged action by WebLogs Inc - I wouldn't be surprised if it drives readers away from their blogs.

Lickr - Flickr, without the Flash

2005-04-10T22:13:00.000+01:00

I've been viewing Flickr for a while now, but have always found it hampered by its reliance on Flash for parts of its UI.

My solution up to now has been simple - I don't have Flash installed in my browser, so I get the fallback simple display of a jpg image, with all its advantages of ability to save, set as background image, zoom the image etc. The minor downside has been the loss of annotation functionality.

Now I have a further choice - via Firefox and GreaseMonkey, Lickr is a userscript which disables the Flash image feature, and replaces it with HTML driven code to do the same thing. This is great, and I hope that the Flickr team take the author up on his suggestion of using the code which he has very helpfully provided under a BSD licence.

Terraserver - the forgotten Microsoft service

2005-04-10T21:31:00.000+01:00

With all the interest that Google Maps is getting, especially with its "satellite" images features, Scoble notes that Microsoft has been offering the same aerial image data via its Terraserver site for a long time. In fact, the data via Microsoft is even more detailed, since it shows the USGS data at its full resolution (where objects 1 foot across can be seen), which is about a magnification of three compared with the full zoom at Google.

They also offer a feature like the Google Sightseeing blog, which they call Famous Places.

Google maps - more advanced usage

2005-04-10T21:21:00.000+01:00

The combination of Craigslist and Google Maps makes for a really impressive real estate listings service as a demo by Paul Rademacher.

Engadget has a good lesson in how to Make your own annotated multimedia Google map.

MyGmaps provides a service to make the whole process of using Google Maps with your own custom data very easy.

More on Google Maps "Satellite" images

2005-04-08T19:30:00.000+01:00

As I explore around more, its clear that most of the so called satellite images are anything but - the detailed images at high zoom levels are simply aerial photographs, probably taken from a plane.

There certainly are satellite images mixed in as well, for areas without detailed covereage, and my impression that the whole USA was covered is very wrong - there are many areas where (at particular zoom levels) the coverage just gives up - either reverting to grainy satellite images, or to the "no data" message.

Like many people, I've been doing a bit of aerial sightseeing, and can recommend:

Altamont Pass windfarms, California
The Boeing factory, Everett, Washington - the worlds largest building as far as I know
Mount St Helens
The Space Needle, Seattle
Cruise ships at Cape Canaveral, Florida

See also the Google Sightseeing blog for more interesting locations.