This year, I've used a few of the Web 2.0 tools to enliven my courses. This year, these techniques were used on two modules, DSA and IAD. Since I also teach web technologies, it seems appropriate that students gain some experience of these tools as part of the courses.
This is a summary of the value my students and I have found in some of them.
Background
The standard VLE at UWE is BlackBoard (UWEOnline). There are a few difficulties with the way it is administered here, particularly around the restricted access to material. Another problem is the lack of ability to link to modules or documents to UWEOnline, limiting the possibilities for integration.. Nonetheless, this the vehicle students expect to be used. Of course in this subject area much of the material is based on server languages like PHP, MySQL, and XQuery and thus resides on our faculty server.
Blogs
Blogs were used in two ways - as a means of communicating from teachers to students, with student commenting (DSA and IAD), and as a component of each students work.
Tutor blogs
The blogs are all hosted on Google's Blogger. During the year this improved in ease of use an in the ability to tag any item. Thus all items on lectures, or coursework 1 or PHP can be tagged and then browsed as a group.
Over the year there were over 50 posts to each blog. The tutors on the modules all had write access. Activity on the blogs was monitored by Google Analytics and showed the expected increase in traffic around lecture time and towards coursework hand-ins. Since the blogs were public,there was a background of hits from all over the world, and some items, such as one on the Periodic Table of Visualisation methods were linked to and gained considerable traffic. Comments on both blogs were few, from students and the public, but there were some.
I also occasionally post to my personal blog (here)and it was sometimes unclear where an item should be posted. I also tried to run one for the new 1 -year programme in Internet Application Development but this did not take off .
Lessons:
Students prefer a consistent interface to learning materials, although they also valued the narrative structure of the blog. However there are difficulties in using it to organise information which must be easily findable, even with the tagging facility. Thus, with some reluctance, I have decided to revert to putting lecture notes on BlackBoard rather than my own web site. BlackBoard will also contain the workplan. I plan to use my own blog to add my own commentary and additional material, with items tagged for consumption by one or other module even though the blog will only be able to refer to items in BlackBoard by name.
Student blogs
Students on IAD were required to keep individual blogs as a record of their reading over they year. This was not successful, partly as a result of concern for the damage to reputation that poor quality web appearances could have on future prospects. In this regard, it may be preferable to use the private blogs in BlackBoard for this purpose.
Wikis
I have used wiki's for some time, and one was used for web2.0 technologies. It is doubtful that this was beneficial, with so much excellent material on wikipedia. However for details of implementations and tools which are specific to CEMS, there is a need for a local wiki. This was established as the CEMS wiki and is being populated, though much more could be done before it becomes the first palce of enquiry about the use of a language, tool or technique in CEMS.
I also set up a student wiki in CEMS for use by students in organising and presenting their research into individual topics. This was quite successful, both as an opportunity to gain experience in wikis and as a means of organising their material. However, the student entries were more in the nature of online articles, with a low level of linkage. Students were required to hand-in a copy of their wiki material (since this is needed for external examiners) and this perhaps led to the lack of a linked structure. A more collaborative, group-based approach might be tried to get a larger network but there are obvious problems with attributing authorship.
Google docs
On one module,the tutors kept a register of attendance. We opted to use the recently available GoogleSpreadsheets for this, partly to explore the value of collaborative documents. Technically, this was very successful, and a big improvement on passing Excel spreadsheets around. However (there seem to be rather a lot of 'however's in this account :-( ) the spreadsheet feels rather cramped and lacks the ability to hold row and column headings in place when scrolling. (freeze panes)
RSS
An advantage of blogs is that they provide an RSS feed which can then be aggregated with other feeds. (unlike BlackBoard which does not) An attempt was made to use this feed to populate the announcements in Blackboard which was only partly successful, creating an additional point of failure. Take-up of RSS is still slow but the new myUWE portal supports RSS feeds so it will be suggested that the blog feed can be added by the individual student.
Podcasts
A couple of lectures were recorded on audio and posted to the blog. I also used Evoca to record short messages.
Whether these initiatives were valued is difficult to say. I think the audio alone is of limited value and integrated slideshow, video and audio is the ideal. Creating such resources for small to medium-sized classes, in such as rapidly changing field as the web, may not be economically viable.
Conclusion
These new technologies have great promise, but the fit with a corporate culture an institutional E-Learning policy creates tensions. This year's experiences are certain food for thought.
Thursday, June 14, 2007
Saturday, June 09, 2007
XQuery functions
Over the last few weeks, I've been writing some general XQuery functions for our own system and in collaboration with Dan McCreary. There are XQuery collections to which these could be contributed, notably Priscilla Warmsley's //www.xqueryfunctions.com/xq/
but some of my functions depend on the eXist function libraries, and I would also like to prove the functions work and show examples of use in unit tests.
I'm currently working on some system tools to browse the code, develop full coverage tests and help refactor the code for the FOLD application. Part of this suite is an XQuery unit test tool. As an experiment, I've combined the two and put up a simple demo on our public eXist demo server.
The function is defined in the prolog which is prepended to each test before execution. The result can be either text or a node which is compared with the expected result as strings or with deep-equal respectively.
The test scripts themselves, viewed as XML look a bit of a mess but the source shows the formatting. CDATA sections are now preserved in stored files in eXist which is convenient when round-tripping XQuery code containing XML, but is controversial since CDATA is only intended to encode XML characters on input but is not part of the infomodel.
Links
but some of my functions depend on the eXist function libraries, and I would also like to prove the functions work and show examples of use in unit tests.
I'm currently working on some system tools to browse the code, develop full coverage tests and help refactor the code for the FOLD application. Part of this suite is an XQuery unit test tool. As an experiment, I've combined the two and put up a simple demo on our public eXist demo server.
The function is defined in the prolog which is prepended to each test before execution. The result can be either text or a node which is compared with the expected result as strings or with deep-equal respectively.
The test scripts themselves, viewed as XML look a bit of a mess but the source shows the formatting. CDATA sections are now preserved in stored files in eXist which is convenient when round-tripping XQuery code containing XML, but is controversial since CDATA is only intended to encode XML characters on input but is not part of the infomodel.
Links
- CommonCode
- The test script for a function Dan wrote
- The test script for a BandMap structure
Sunday, March 25, 2007
SPA2007 - day 0
I tinker endlessly with the eXist workshop material, still some solutions to do and the new server isn't having anything to do with the Java Webstart client. Hopefully it will get fixed before Wednesday afternoon. I'll also have to figure how to create multiple folders with the same initial contents - should be able to use a backup but not sure how to install in a different place.
Monday, March 19, 2007
FizzBuzz
Here's a quick XQuery solution to the FizzBuzz problem posed in David Patterson's blog.
I've taken the liberty of splitting the hyphenated string into two attributes.
I've taken the liberty of splitting the hyphenated string into two attributes.
let $config :=
<fizzbuzz>
<range min="1" max="100"/>
<test>
<mod value="3" test="0">Fizz</mod>
<mod value="5" test="0">Buzz</mod>
</test>
</fizzbuzz>
for $i in ($config/range/@min to $config/range/@max)
let $s := for $mod in $config/test/mod
return
if ($i mod $mod/@value = $mod/@test)
then string($mod)
else ()
return
if (exists($s))
then string-join($s,' ')
else $i
Thursday, March 15, 2007
Timelines
Here are some ways in which timelines are being used in the web:
Links to other people load their timelines (although some are incomplete and not working).
Multiple event streams can be displayed in multiple bands - e.g. here are the lives of two family members
It would be great to get a subset of world history events from TimeSearch to include as a separate band.
To do:
- TimeSearch
- GoogleEarth timeline
- HyperHistory
- SIMILE timeline from MIT
- Provides a Javascript API to create a complex panning timeline.
- Events are uploaded in XML format which defines for each event
- start title end? isDuration? latestStart? earliestEnd? image? link? and body
- Cant locate schema for the event stream
- Event date format is not xs:date
- body must be a string i.e. with < not <
- Set the mime type to application/xml
- example XML timeline
Links to other people load their timelines (although some are incomplete and not working).
Multiple event streams can be displayed in multiple bands - e.g. here are the lives of two family members
It would be great to get a subset of world history events from TimeSearch to include as a separate band.
To do:
- Convert xs:date to Timeline format
- Handle photos with missing dates
- Focus timeline when no date of birth
Wednesday, March 14, 2007
My Family History
Bamber Gascoigne's TimeSearch got me thinking about my family history project again.
The Family History project started with the idea of putting some family photos on the web, together with some meta data about the photo. Most had a list of subjects but only very occassionally a date. I had the idea that if the birth dates of the subjects were added, and the age of just one subject could be guessed, the system could infer the date of the photograph, and hence the age of all subjects. With birth and death data included, a timeline could be extracted for a person, including birth, deaths and marriages, photographs and even world events.
The resultant eXist/XQuery/XSLT prototype is here.
What would be nice would be be able to combine such personal histories with events in world history, or to compare the timelines for famous people with that of a family member. A simplistic approach would be to deep link into history sites and this is what the prototype has done, creating links like
These generated links are not as clever as those in TimeSearch, which I would guess are hand edited.
This is simple to accomplish, but its rather one-side - my site can link to another, but it can't mashup the data with my own. A mashup requires an API and a published format for exported content. I suppose we might look at hcalendar as one possible format but it would need extending it to support tags. RSS is another possibility.
The Family History project started with the idea of putting some family photos on the web, together with some meta data about the photo. Most had a list of subjects but only very occassionally a date. I had the idea that if the birth dates of the subjects were added, and the age of just one subject could be guessed, the system could infer the date of the photograph, and hence the age of all subjects. With birth and death data included, a timeline could be extracted for a person, including birth, deaths and marriages, photographs and even world events.
The resultant eXist/XQuery/XSLT prototype is here.
What would be nice would be be able to combine such personal histories with events in world history, or to compare the timelines for famous people with that of a family member. A simplistic approach would be to deep link into history sites and this is what the prototype has done, creating links like
These generated links are not as clever as those in TimeSearch, which I would guess are hand edited.
This is simple to accomplish, but its rather one-side - my site can link to another, but it can't mashup the data with my own. A mashup requires an API and a published format for exported content. I suppose we might look at hcalendar as one possible format but it would need extending it to support tags. RSS is another possibility.
TimeSearch by Bamber Gascoigne
The item on the BBC's Start the Week' by Bamber Gascoigne on the history site he has created called TimeSearch sparked my interest TimeSearch is a history search portal, using event stubs to provide links onwards to a wide range of sites - from general news sites, Wikipedia and Google images to specialist on-line resources. Events can be selected using two hierarchical category systems, location and them as well as year. I confess I had some trouble deselecting categories once selected.
This is the latest form of delivery of the extensive material Bamber has created and put on line in:
This is the latest form of delivery of the extensive material Bamber has created and put on line in:
- History World
- An online encyclopedia of history with great little quizes.
- OCEAN index
Tuesday, February 13, 2007
SPA 2007
I have been working on my session for this year's SPA for a while now. Seems a timely session to run according to Elliotte Rusty Harold's predictions for XML in 2007 (which mentions eXist).
Mostly I have been vacillating about which case study to use - I have so many part-finished projects. At first I revived the Whisky project, but after a lot of work, this felt like a bit of a slog - it was good as a modeling exercise but didn't reveal the power of XML and XQuery to my satisfaction. I then switched to the Bus Timetable project, which is much more realistic and one which I want to solve to replace the existing PHP/MySQL application with its hand-crafted timetables. However the data here is very complex, and the data files I have so far are incomplete. Nevertheless, this is the case study I settled on, although I did a short detour into the DVD hire example and of course there is always StudentsOnline, the main site we are developing with this technology.
Timetables come from TravelLine South-West in the form of TransXChange files. These first need processing to a more amenable structure. The actual data model is highly interlinked which makes for some XPath fun. Conversion is required to calculate absolute times (TXC has time differences between stops) and to convert from OS Easting and Northing to Lat/Long . With the simplified structures, I can present the data in various ways. HTML pages show the stops and the departures times but the neat idea is to generate kml files to overlay on Google Earth. Using a NetworkLink, I can display changing times on the map. The XPath date and time functions are a great help.
Other bits of functionality include an Ajaxian incremental Bus stop locator and membership registration rretainig favourite stops and services.
Here are some useful links for this project:
Mostly I have been vacillating about which case study to use - I have so many part-finished projects. At first I revived the Whisky project, but after a lot of work, this felt like a bit of a slog - it was good as a modeling exercise but didn't reveal the power of XML and XQuery to my satisfaction. I then switched to the Bus Timetable project, which is much more realistic and one which I want to solve to replace the existing PHP/MySQL application with its hand-crafted timetables. However the data here is very complex, and the data files I have so far are incomplete. Nevertheless, this is the case study I settled on, although I did a short detour into the DVD hire example and of course there is always StudentsOnline, the main site we are developing with this technology.
Timetables come from TravelLine South-West in the form of TransXChange files. These first need processing to a more amenable structure. The actual data model is highly interlinked which makes for some XPath fun. Conversion is required to calculate absolute times (TXC has time differences between stops) and to convert from OS Easting and Northing to Lat/Long . With the simplified structures, I can present the data in various ways. HTML pages show the stops and the departures times but the neat idea is to generate kml files to overlay on Google Earth. Using a NetworkLink, I can display changing times on the map. The XPath date and time functions are a great help.
Other bits of functionality include an Ajaxian incremental Bus stop locator and membership registration rretainig favourite stops and services.
Here are some useful links for this project:
- eXist database home
- an XQuery tutorial from Howard Katz
- A basic Timetable implementation (pending)
- TransXChange documentation
- TravelLine South West
- Google Earth
- kml description and reference
Monday, January 29, 2007
Computed synethesia
The Cloud appreciation society lead me to the Cloud Harp project. I'm listening to a recording made in 2004 - very ethereal.
It re-generates my interest in artificial synaesthesia generally and perhaps a return to my Bristol Harbour project, using image stream and image processing to pick out the frequency and direction of wave patterns to modulate music, with self-contained stations set up around the harbour.
Information arts often has artificial synaesthesia at its heart, converting information in one medium to information in another, with the aim of thus enhancing our appreciation and understanding of the source domain and creating beauty and interest in the target domain.
It re-generates my interest in artificial synaesthesia generally and perhaps a return to my Bristol Harbour project, using image stream and image processing to pick out the frequency and direction of wave patterns to modulate music, with self-contained stations set up around the harbour.
Information arts often has artificial synaesthesia at its heart, converting information in one medium to information in another, with the aim of thus enhancing our appreciation and understanding of the source domain and creating beauty and interest in the target domain.
Sunday, January 28, 2007
Cloud Appreciation Society
One of my Christmas presents was 'The Cloudspotter's Guide' by Gavin Pretor-Pinney and published in 2006 by Hodder & Stoughton on behalf of the cloud appreciation society.
This site includes an excellent example of a subject-specific photo site, with many excellent photographs of cloud formations, optical effects and associated phenomena. The site was interesting in the context of the DSA module because it combines the problem of creating a photo site with the problems of tagging and categorising information, so I thought I'd write a review of it here, as perhaps as an example of what might have been included in theFlickr review. Next year I think I will base the coursework on a selection of sites like this rather than a monster site like Flickr. I welcome suggestions.
Navigation
The home page is a page featuring the last photo to be loaded.
Search by category and by partial string match within the title string are provided on this page. These lead to two different album pages - a paged layout in the case of the category, a list with thumbnails and titles in front of the page from which the search was launched in the case of the title search. Its not clear why there should be two different approaches. Usefully, the category selector shows the number of photographs in that category.
Direct links to the categories to which the current image has been assigned are provided and a featured category (Cloud Lookalikes) is at the top.
Each page includes a selection of other photographs. The criteria appears to be just the adjacent photo accession numbers and is not based on any criteria of similarity, but nevertheless it does encourage serendipitous browsing. Curiously the aspect ratio of these photos is determined by that of the main photo, leading to mal-shaped images.
Clicking on the photo rather surprisingly links to the next photo in sequence. There is also a 'Previous Page' link which is functionally a back button - it is not quite clear why this is included since this is a standard browser button. Perhaps Next and Previous links (in the accession sequence) would provide a better navigational mechanism.
A search (or link) to a members photos would be useful.
Photo Data
Images are jpegs at medium and thumbnail resolution.
Meta data include the name of the copyright owner (or is it just the member?), a descriptive title, data and time (but of upload I would guess, not of the photograph itself). The title can contain links e.g. to related sites.
Each photograph is related to several categories.
It is a pity that the photographs do not appear to be geo-coded, because it would be nice to mash these up with Google Earth. There is often a description of the place which could be translated through a geo-coding service. Since some photos are taken from planes, altitude data is needed too.
Tagging and classification
Each photograph is classified into one or several of around 40 categories and hence linked to other photos in the same category through that category. The category system itself is interesting as a information construct. On the surface, it would seem that it would benefit from some hierarchy. e.g.
* All photos
** Cloud Type
*** Cumulus
**** Kelvin-Helmholtz wave cloud
***
** Cloud features
*** Contrails
***
** Time-of-Day
*** Sunrise
*** Sunset
** Optical Effects
*** Rainbow
*** Halo
***
and probably others.
It also seems that some category meta-data is embedded in the name - the numbers on Cloud types is a reference to the Chapters in the book. Clearly a general description of the category itself could also be part of the overall data structure. The book provides a hierarchical classification of cloud types which would be useful to include. There are also occasional synonyms embedded in the name -eg Mamma (also know as mammatus) which could be treated uniformly as category meta-data.
The task of classifying each photo into these categories is not open to the public or members but is seen as an expert task. However there are places, such as in the category of 'Clouds that look like things' where folksomony would seem to be appropriate. Photographs also sometimes contain objects e.g. balloons, planes, boats for which arbitrary tags seem suitable.
Upload
Rather surprisingly, photo upload, even by members, is not supported, and photos have to be emailed to the webmaster with accompanying meta data for upload and categorisation.
Rating and comments
The public can easily rate photos and add comments. The comments are generally expressions of appreciation of no interest to anyone but the photographer. There is no information on the basis of the rating - indeed all photos I've seen are rated 4.It is not clear what value this adds or indeed what is being rated - the photograph or the subject.
Compatibility and Accessibility
Images have alt tags but these are the copyright names, not the description of the image which would better aid the reader.
There are keywords in the page meta data but these are not photo specific.
XHTML compliance - main photo page shows 1 error and 17 warnings. Typically these are problems with table and div tag nesting and problems with & . The error is a non-HTML tag.
The display of some characters in the title is broken (on Firefox and IE) e.g this
API and Feeds
The site provides RSS2.0 and Atom feeds of the latest photo additions and a Feedburner link.
Technology
The scripting language is PHP, images are held as files, not held in the database. Not sure what database is used (probably MySQL?). CSS is used for styling. Small amount of Javascript e.g. for GIF rollover on the rating buttons and call to update the rating. One function is defined (flip) which is not used.
The program architecture uses a single index.php script with a parameter to determine what page type to return, rather than multiple scripts. Interesting design issue here.
Data Model
This site includes an excellent example of a subject-specific photo site, with many excellent photographs of cloud formations, optical effects and associated phenomena. The site was interesting in the context of the DSA module because it combines the problem of creating a photo site with the problems of tagging and categorising information, so I thought I'd write a review of it here, as perhaps as an example of what might have been included in theFlickr review. Next year I think I will base the coursework on a selection of sites like this rather than a monster site like Flickr. I welcome suggestions.
Navigation
The home page is a page featuring the last photo to be loaded.
Search by category and by partial string match within the title string are provided on this page. These lead to two different album pages - a paged layout in the case of the category, a list with thumbnails and titles in front of the page from which the search was launched in the case of the title search. Its not clear why there should be two different approaches. Usefully, the category selector shows the number of photographs in that category.
Direct links to the categories to which the current image has been assigned are provided and a featured category (Cloud Lookalikes) is at the top.
Each page includes a selection of other photographs. The criteria appears to be just the adjacent photo accession numbers and is not based on any criteria of similarity, but nevertheless it does encourage serendipitous browsing. Curiously the aspect ratio of these photos is determined by that of the main photo, leading to mal-shaped images.
Clicking on the photo rather surprisingly links to the next photo in sequence. There is also a 'Previous Page' link which is functionally a back button - it is not quite clear why this is included since this is a standard browser button. Perhaps Next and Previous links (in the accession sequence) would provide a better navigational mechanism.
A search (or link) to a members photos would be useful.
Photo Data
Images are jpegs at medium and thumbnail resolution.
Meta data include the name of the copyright owner (or is it just the member?), a descriptive title, data and time (but of upload I would guess, not of the photograph itself). The title can contain links e.g. to related sites.
Each photograph is related to several categories.
It is a pity that the photographs do not appear to be geo-coded, because it would be nice to mash these up with Google Earth. There is often a description of the place which could be translated through a geo-coding service. Since some photos are taken from planes, altitude data is needed too.
Tagging and classification
Each photograph is classified into one or several of around 40 categories and hence linked to other photos in the same category through that category. The category system itself is interesting as a information construct. On the surface, it would seem that it would benefit from some hierarchy. e.g.
* All photos
** Cloud Type
*** Cumulus
**** Kelvin-Helmholtz wave cloud
***
** Cloud features
*** Contrails
***
** Time-of-Day
*** Sunrise
*** Sunset
** Optical Effects
*** Rainbow
*** Halo
***
and probably others.
It also seems that some category meta-data is embedded in the name - the numbers on Cloud types is a reference to the Chapters in the book. Clearly a general description of the category itself could also be part of the overall data structure. The book provides a hierarchical classification of cloud types which would be useful to include. There are also occasional synonyms embedded in the name -eg Mamma (also know as mammatus) which could be treated uniformly as category meta-data.
The task of classifying each photo into these categories is not open to the public or members but is seen as an expert task. However there are places, such as in the category of 'Clouds that look like things' where folksomony would seem to be appropriate. Photographs also sometimes contain objects e.g. balloons, planes, boats for which arbitrary tags seem suitable.
Upload
Rather surprisingly, photo upload, even by members, is not supported, and photos have to be emailed to the webmaster with accompanying meta data for upload and categorisation.
Rating and comments
The public can easily rate photos and add comments. The comments are generally expressions of appreciation of no interest to anyone but the photographer. There is no information on the basis of the rating - indeed all photos I've seen are rated 4.It is not clear what value this adds or indeed what is being rated - the photograph or the subject.
Compatibility and Accessibility
Images have alt tags but these are the copyright names, not the description of the image which would better aid the reader.
There are keywords in the page meta data but these are not photo specific.
XHTML compliance - main photo page shows 1 error and 17 warnings. Typically these are problems with table and div tag nesting and problems with & . The error is a non-HTML tag.
The display of some characters in the title is broken (on Firefox and IE) e.g this
API and Feeds
The site provides RSS2.0 and Atom feeds of the latest photo additions and a Feedburner link.
Technology
The scripting language is PHP, images are held as files, not held in the database. Not sure what database is used (probably MySQL?). CSS is used for styling. Small amount of Javascript e.g. for GIF rollover on the rating buttons and call to update the rating. One function is defined (flip) which is not used.
The program architecture uses a single index.php script with a parameter to determine what page type to return, rather than multiple scripts. Interesting design issue here.
Data Model
Saturday, January 20, 2007
Where I've been
Just discovered this little application which generates a GIF map of the world showing the countries you have visited.
So many places yet to see, and so much of places visited unseen like most of the States:
From Douwe Osinga, a Google engineer
So many places yet to see, and so much of places visited unseen like most of the States:
From Douwe Osinga, a Google engineer
Thursday, January 18, 2007
Voice snippets
I've been experimenting today with http://www.evoca.com/
A free account allows you to record and store up to 60 minutes of sound. This interface is simple to use. Sound can be captured from a number of sources but the easiest is directly from a microphone on the computer. Everything is done through the web interface. When the recording has been made, an HTML snippet is provided to paste where you need it, and this invokes a flash player.
I'm trying these as an addition to my blog for the DSA module
I suppose it will just pollute the airwaves, but maybe it offers another mode for information dissemination which may just help some students. For two lectures last year I recorded the whole lecture and mounted it as an MP3 file, but they very long and un-synchronised with the Powerpoint slides. I tried calling out slide numbers but it was reminiscent of the early soccer commentaries which used numbered squares (the origin I learnt today of the phrase 'back to square 1') .
A free account allows you to record and store up to 60 minutes of sound. This interface is simple to use. Sound can be captured from a number of sources but the easiest is directly from a microphone on the computer. Everything is done through the web interface. When the recording has been made, an HTML snippet is provided to paste where you need it, and this invokes a flash player.
I'm trying these as an addition to my blog for the DSA module
I suppose it will just pollute the airwaves, but maybe it offers another mode for information dissemination which may just help some students. For two lectures last year I recorded the whole lecture and mounted it as an MP3 file, but they very long and un-synchronised with the Powerpoint slides. I tried calling out slide numbers but it was reminiscent of the early soccer commentaries which used numbered squares (the origin I learnt today of the phrase 'back to square 1') .
Monday, January 15, 2007
Whisky - Laphroaig
This entry is a sample of data sources on a specific whisky - assembled here as a notepad for the workshop at SPA2007 and for the Data, Schemas and Applications module.
Laphroaig
Laphroaig Distillery,
Port Ellen,
Isle of Islay,
Argyll,
PA42 7DU Tel. 01496 302418
Home site
General
Types of information
Printed
Laphroaig
Laphroaig Distillery,
Port Ellen,
Isle of Islay,
Argyll,
PA42 7DU Tel. 01496 302418
Home site
General
- Scotland: Whiskies and Distilleries
- Wikipedia
- ScotchWhisky.net
- Whisky Magazine
- Scotland Whisky
- John Butler's site
Types of information
- Chronology
- Location
- Gooogle Map http://maps.google.com/maps?q=PA42%207DU
- Multimap http://www.multimap.com/map/browse.cgi?pc=PA427DU&scale=25000&cat=loc
- Wikipedia Geohack
- Pronunication - John Butler's site
- Taste http://www.whiskyclassified.com/laphroaig.html
- Tasting notes e.g. Mike Padlipsky
- Ownership - changing ownership - see the yearbook
- Bottlings http://www.laphroaig.com/whiskies/index.asp?expanded=our_whiskies
- Glossary http://www.whiskymag.com/words/
Printed
- Malt Whisky Yearbook 2006 p170
- Collins Gem Whisky p 171
Beyond Belief
I spent part of this weekend listening to the presentations at BeyondBelief2006 for which all sessions over the two and a half days are available as Flash video. A great resource to have provided and very well done. Ongoing discussion on the Edge.
Amongst the speakers, I was particularly taken with
Amongst the speakers, I was particularly taken with
- Carolyn Porco - the picture of Earth through the back-lit rings of Saturn is just wonderful.
- Harold Kroto - vega science trust
- Mahzarin Banaji - Implicit association tests
- Ann Druyan - Carl Sagan
Saturday, November 04, 2006
Double line spacing in Dissertations
It is standard practice to require dissertations to be double-spaced. In the age of word processors, one wonders why this practice persists. This must surely be a legacy from the days when dissertations were typewritten and students were allowed to make corrections in situ rather than re-print. Word processors have eliminated that problem. Double-spacing might also be suitable for drafts which require annotation for editing, but final copies of dissertations are not used in this way and modern techniques such as commenting do this job.
(Primo Levi's 'Chromium' in 'The Periodic Table' has several great examples of the way solutions to old probelms hang around even when the problem not longer exists.)
I have seen readability expressed as another reason for double-spacing. It's not easy to track down the evidence for this. Kruk and Muter(1984) report "single spacing produced reading that was 10.9% slower than that produced by double spacing" on a 'video' screen. Weller (2004) reports on a previous study that "It was discovered that single spacing of text requires more eye fixations per line and therefore fewer words are read per fixation, which increases reading time (Kohler, Duchnicky & Ferguson, 1981)." But Mills and Weldon (1987) report that this same study showed a 2% slowing of reading rate, hardly convicing evidence.
Other writers express the view that this convention is out-moded e.g. the author of the Tex manual : link
Princeton University now accepts single-spaced printed copies although ProQuest (an online dissertation repository) still requires double-spacing for the electronic copy ( and hence all printed copies therefrom). It is claimed that this improved on-line readability.
A PhD Student in the US has calculated that 20,000 reams of paper would be saved if ProQuest accepted single-spaced.
Double-spacing is often mandated in legal practice but even here the convention is challenged.
(Primo Levi's 'Chromium' in 'The Periodic Table' has several great examples of the way solutions to old probelms hang around even when the problem not longer exists.)
I have seen readability expressed as another reason for double-spacing. It's not easy to track down the evidence for this. Kruk and Muter(1984) report "single spacing produced reading that was 10.9% slower than that produced by double spacing" on a 'video' screen. Weller (2004) reports on a previous study that "It was discovered that single spacing of text requires more eye fixations per line and therefore fewer words are read per fixation, which increases reading time (Kohler, Duchnicky & Ferguson, 1981)." But Mills and Weldon (1987) report that this same study showed a 2% slowing of reading rate, hardly convicing evidence.
Other writers express the view that this convention is out-moded e.g. the author of the Tex manual : link
Princeton University now accepts single-spaced printed copies although ProQuest (an online dissertation repository) still requires double-spacing for the electronic copy ( and hence all printed copies therefrom). It is claimed that this improved on-line readability.
A PhD Student in the US has calculated that 20,000 reams of paper would be saved if ProQuest accepted single-spaced.
Double-spacing is often mandated in legal practice but even here the convention is challenged.
Monday, October 23, 2006
Declarative Processes
I am getting back to the declarative process work I began a couple of years ago. Peter Marks and Ben Moseley's session at SPA2006 got me started again.
I now have a partial implementation of a DVD rental application written in a declarative process style using eXist and XQuery. The essence of the design is to retain all events and compute the current state by the evaluation of predicates over the event trace. The rules are hard-coded in eXist functions in this demo. Rules are mostly XPath expressions. For example
Much more work is needed to see how well this model will fit with other processes. Some applications include:
The Alloy language (Daniel Jackson, MIT)
Peter Marks and Ben Moseley - Functional Relational Programming, SPA 2007, FRP site
Work on rule-based language has had a rough ride - several projects in this area (e.g. BRML) are defunct. Some activity in R2ML and RuleML.
I now have a partial implementation of a DVD rental application written in a declarative process style using eXist and XQuery. The essence of the design is to retain all events and compute the current state by the evaluation of predicates over the event trace. The rules are hard-coded in eXist functions in this demo. Rules are mostly XPath expressions. For example
- To find the current address for a person $p given SetAddress events
- events($p)[SetAddress][last()]//address
- To find the address at $date
- events($p)[SetAddress][@date < $date][last()]//address
- To check if a member is suspended given Suspend and Reinstate events
- exists( events($p)[Suspend or Reinstate][last()][Suspend])
- the state of the system can be evaluated at any time in the past by filtering the event stream for events before the specified date
- if the rules are held as data, the state can be computed synoptically (i.e. according to rules applicable at the time) - but how should they be encoded - as executable XQuery expressions?
- changes in rules are supported, allowing re-interpretation of the past
- subjective rules are supported although only within the event language created
- 'what-if' analysis is supported, allowing questions to be asked about the impact of changes in the rules
- rules are robust to insertion of other events
- the event history is often required to understand what has happened and why so there is no addition storage requirement
- The computational cost is higher than simply fetching a simple value
- Writing declarative rules is not easy and will need a higher level language
- The approach requires a global model of the system, so its applicable to systems but not to distributed systems
- It's not yet clear what range of behaviour is expressable
Much more work is needed to see how well this model will fit with other processes. Some applications include:
- lightweight processes - e.g. academic processes such as setting and marking coursework
- descriptive models of real processes and their subsequent analysis for process improvement
The Alloy language (Daniel Jackson, MIT)
Peter Marks and Ben Moseley - Functional Relational Programming, SPA 2007, FRP site
Work on rule-based language has had a rough ride - several projects in this area (e.g. BRML) are defunct. Some activity in R2ML and RuleML.
Friday, August 04, 2006
America
Time to forget about University and the project. Tomorrow we're off to the USA for adventure and catching up with friends and relatives.
Here is a link to a GoogleEarth overlay containing the main places we are visiting:
Trip on Google Earth
Photos from the Minnesota trip are now up on Flickr: http://www.flickr.com/photos/paddlers9/
Here is a link to a GoogleEarth overlay containing the main places we are visiting:
Trip on Google Earth
Photos from the Minnesota trip are now up on Flickr: http://www.flickr.com/photos/paddlers9/
Sunday, June 04, 2006
eXist examples
I now have somewhere to host my example XML database examples.
http://www.cems.uwe.ac.uk/chriswallace/index.xql
This supercedes the previous hosts for my teaching applications on a friend's server and on the exist-db demo server.
There are a few there, but further development will have to wait till the end of the term. Few typos I've just spotted :- (
One problem I faced was to allow a visitor to view the Xquery files themselves. I initially saw this as a eXist problem, needing an eXist solution, and indeed Adam, one of the eXist developers, put in a feature to support this. However I slowly realised that if I wanted to put up a number of examples (there are about 9 in various stages of development) , I should develop a schema for an application index page, and individual XML configuration files instead of writing unique pages. I can then use XSLT to display the page. If I then wrote an Xquery to fetch a query and display it , I could use the application configuration XML to authorise access to the scripts.
This experience prompted or reminded me of a couple of computing adages:
In future, I want to enhance this configuration file to be a description of the application architecture and information flow, so I can also generate, via GraphViz, a clickable SVG image of the application.
That's the problem with eXist and XQuery: this technology greatly reduces the time needed to create an application, but it increases the range of ideas for applicatins which can be easily implemented even more!
http://www.cems.uwe.ac.uk/chriswallace/index.xql
This supercedes the previous hosts for my teaching applications on a friend's server and on the exist-db demo server.
There are a few there, but further development will have to wait till the end of the term. Few typos I've just spotted :- (
One problem I faced was to allow a visitor to view the Xquery files themselves. I initially saw this as a eXist problem, needing an eXist solution, and indeed Adam, one of the eXist developers, put in a feature to support this. However I slowly realised that if I wanted to put up a number of examples (there are about 9 in various stages of development) , I should develop a schema for an application index page, and individual XML configuration files instead of writing unique pages. I can then use XSLT to display the page. If I then wrote an Xquery to fetch a query and display it , I could use the application configuration XML to authorise access to the scripts.
This experience prompted or reminded me of a couple of computing adages:
- The only numbers of importance in computing are 1, 2 and many- with its meta counting variant: 1, 2, Schema
- Use of the configuration file to generate the web page and to authorise access to the script as an example of the Shanley principle - in which one component serves multiple purposes see Michael Jackson, Software Requirements and Specifications, Addison-Wesley 1995, p29-30].
In future, I want to enhance this configuration file to be a description of the application architecture and information flow, so I can also generate, via GraphViz, a clickable SVG image of the application.
That's the problem with eXist and XQuery: this technology greatly reduces the time needed to create an application, but it increases the range of ideas for applicatins which can be easily implemented even more!
Monday, March 13, 2006
RSS
Discovered bbc backstage this morning whilst researching for my lecture on Web services and REST. As an experiment, I thought I'd add a page of RSS-derived links to the FOLD. I shouldn't still be surprised by this, but it only took a few lines to add a primitive page.
I've put the demo up on the exist demo site. There is also a version which generates Voice+XML to make a real newsreader. It needs Opera 8 with the speech extension and there is only a selection of two American voices but it handles the task very well.
I've put the demo up on the exist demo site. There is also a version which generates Voice+XML to make a real newsreader. It needs Opera 8 with the speech extension and there is only a selection of two American voices but it handles the task very well.
Monday, March 06, 2006
Cellular automata and Processing
I've just re-discovered my Processing application which implements the generalised Conway Game of Life. The interface allows you to set any rule - for each of the counts of alive neighbours, whether a dead cell lives or a live cell dies. You can also run Conway's rule or generate a random rule.
Here it is
and here is my Processing page
What it doesnt have is any way to register interesting rules the viewer may discover. I discovered a nice one today: l2d=134&d2l=157 which shows constantly varying interesting behaviour - perhaps for a screen saver. Also l2d=125&d2l=3568&id=0.1 which is a slow converger to nearly all white with a few blinkers.
So two jobs to do: One to allow finds to be registered and commented on, the other to take automata specification from the URL in a form like that above. Both seem very Web2.0 - the first enabling participation, the second a unique url for every resource.
Apart from being a exercise in writing Processing, I use it in a lecture on Processes and Emergence.
'Processing' is a poor brand name (try searching for it!) [like eXist in that regard] but a wonderful little language for animations. Developed by two design students form MIT, Ben Fry and Casey Reas, it is open-source software with a really nice little IDE, Java-based (it generates a Java applet) with some great examples of both animations and information display. A much better introduction to programming than full-on Java I think and a candidate for a first language, partly because it places emphasis on algorithms and change, not structure and stasis. Its also much simpler than Flash Actionscript as well as non-propriatory.
Here are some of my favourite Processing examples
Zipcodes by Ben Fry - locates a US Zipcode
Flight Patterns by Aaron Goblin - visualising the flights of planes in the US on a single day
Dreamlines by Leonardo Solaas - a Flickr/Processing mash-up
Here it is
and here is my Processing page
What it doesnt have is any way to register interesting rules the viewer may discover. I discovered a nice one today: l2d=134&d2l=157 which shows constantly varying interesting behaviour - perhaps for a screen saver. Also l2d=125&d2l=3568&id=0.1 which is a slow converger to nearly all white with a few blinkers.
So two jobs to do: One to allow finds to be registered and commented on, the other to take automata specification from the URL in a form like that above. Both seem very Web2.0 - the first enabling participation, the second a unique url for every resource.
Apart from being a exercise in writing Processing, I use it in a lecture on Processes and Emergence.
'Processing' is a poor brand name (try searching for it!) [like eXist in that regard] but a wonderful little language for animations. Developed by two design students form MIT, Ben Fry and Casey Reas, it is open-source software with a really nice little IDE, Java-based (it generates a Java applet) with some great examples of both animations and information display. A much better introduction to programming than full-on Java I think and a candidate for a first language, partly because it places emphasis on algorithms and change, not structure and stasis. Its also much simpler than Flash Actionscript as well as non-propriatory.
Here are some of my favourite Processing examples
Zipcodes by Ben Fry - locates a US Zipcode
Flight Patterns by Aaron Goblin - visualising the flights of planes in the US on a single day
Dreamlines by Leonardo Solaas - a Flickr/Processing mash-up
Subscribe to:
Posts (Atom)