I have released my Multi-Language Translation Script (as seen on http://www.mobi-blog.eu) as Open Source. Please feel free to use or adapt it for your own site(s) - you can visit the project homepage at:

http://code.google.com/p/multi-language-translation/

This script will automatically translate any web page or site that contains text written in more than one language, into a single language requested by the user. This can be particularly useful for sites involving a lot of user-generated content - such as blogs, forums, and other community websites. It harnesses the Google AJAX Language API to perform translation and language detection; has the ability to translate between more than 50 languages; and can be deployed easily on virtually any modern website.

The script contains a number of useful and interesting features; performing translations dynamically on screen after a page loads or a language is changed. Users can view the progress of a current translation; and their choice of language will be remembered whenever they visit any page on the site. The source language (and optionally the source text) of specified regions can also be displayed to the user. Any translation errors or timeouts are handled effectively in the unlikely event they occur...

All feedback would be extremely welcome - including comments, enquiries, suggestions and problem reports. This will help the script to evolve as further development takes place over time.

A Bit of History:

This project emerged as a result of work carried out by OdinLab on the Mobi-Blog website during the early summer of 2009. To improve participation, we decided that students should be allowed to post entries in their own native language. This did however present problems with people who did not speak the same language as us or other users - and we agreed that it would be impractical to manually translate the text into various languages ourselves.

Initially, we used the Google 'Translate My Page' Gadget to embed automatic translation functionality into the site. This did initially work very well when the site content was written primarily in English, but there were significant problems when multi-lingual content was added. Indeed, it got to the point where the gadget could not achieve something as rudimentary as translating English text into English!

I was therefore asked to investigate alternatives; and decided to use the Google AJAX Language API - mainly because it did not suffer from these limitations. I created a fully working prototype within around a week; working out how to overcome various technical challenges, as well as creating a user interface and additional functionality. Once we actually had something online to play with, we suddenly realised that the script could in fact be useful for many other sites - as a lot of websites involve users of different nationalities submitting content. Unfortunately, at the time the script was tailored very specifically towards the Mobi-Blog site.

Therefore, a few months later, I started work on making the script far more flexible, so that it could be used on any website. I also made some substantial improvements to the way it worked. After a week or two this was complete, so I then released it as Open Source... And this project was born!

I hope you like the script, and that it is useful to you!

Whilst trying to sort an issue with user icons not uploading, I noticed Elgg 1.1 out-of-the-box generates some rather silly errors.  So I thought I would fix a couple of them to save filling up my log file while I am getting the This Is Me server working rather better than it did by default.

So, first one I dealt with is a simple problem in mod/messages/start.php which reports an "invalid argument supplied for foreach()" - easy solution, don't try iterating on a collection which is empty!

       function count_unread_messages() {

           //get the users inbox messages
                    $num_messages = get_entities_from_metadata("toId", $_SESSION['user']->getGUID(), "object", "messages", 0, 10, 0, "", 0, false);

                    //set a counter
                    $counter = 0;

                    //loop through the inbox and count the number unread
                    if ($num_messages) {       //PNP check it exists first, ppl!

                     foreach($num_messages as $num){

                      if($num->readYet == 0)
                        $counter++;
                    }
           }

           return $counter;

        }
 

At Pat's request, I'm posting some brief instructions about adding additional languages to Moodle (this issue came up whilst working on the Muvenation project's moodle site).

1. Go to the Moodle site and log in as an admin.

2. On the Site Administration panel, click on the 'Languages' link and select 'Language Packs'.

3. Select the required language from the listbox on the right, and click on the 'Install selected language pack' button.

4. Zlib must be enabled on the server for the language pack to install properly - otherwise an error message ('abort pclzip.lib.php : Missing zlib extensions') appears at this point. The instructions below explain how to enable zlib.

5. connect via SSH (using a tool such as PuTTY) to the server upon which the moodle site is installed. Log in as a user with root level access privileges.

6. When logged in, enter the command 'sudo /sbin/yast -i php5-zlib' (without the quotes).

7. Allow the installation of zlib to complete (this may take a while).

8. You can check if zlib is installed properly with the command  'rpm -qa | grep -i php' (zlib should appear on the list of modules).

9. It MAY be necessary to edit php.ini to contain the additional line 'extension=zlib.so' and change the 'zlib.output_compression' setting to 'on'. This wasn't needed for the muvenation site though (even though these options aren't set in the file), so try testing without making these changes first...

10. Restart apache (command: '/usr/sbin/apache2ctl graceful'). Return to the Moodle site, and try installing the language pack(s) again.

CEUR-WS provides a service for publishing your conference or workshop proceedings - ftp to their site, and you retain ownership.  You can even delete it again later if you want.

Now, there is no guarantee of quality for the material there - in that there is no guarantee that it is peer reviewed (not that that is much a guarantee of anything, frankly) - but everything I have read so far is really rather good.

The problem, and this is a big problem, is that there are about 2000 papers there that I want to read... anyone got any suggestions on how to do this?

I like Tortoise.  However, whilst setting my new machine up, I installed the latest build (last weekend) and it couldn't authenticate- it just sat there.  When I had the opportunity to try and fix things, today, I installed the latest version and whilst it can let me browse and export, I cannot checkout from my repository because instead of treating svn+ssh:// as the protocol it appends a local path to it, and then unsurprisingly says it can't find it.

Downgrading to version 1.4.8 solves this - in case you need to know.

The only bit about this which really leaves me uncomfortable is that the changelog says that the authentication problem was solved in version 1.5.1 - unfortunately the official release is 1.5.0 and the latest nightly build is also still only 1.5.0.  Not sure whether there is a time machine involved or just a typo combined with a mistake - neither of the 1.5.0's work in the way the changelog suggests 1.5.1 should (which is fair, I guess, in a way...)

We had a power outage last night.  It wasn't supposed to affect our room, but it did anyway.  My machine this morning decided it no longer liked Firefox, and over the first three hours of the morning it was having trouble accessing various networks, presumably because some routers were slowly being re-configured somewhere.

This made trying to get mobi-blog configured for users to self register an interesting task, finding the arcanely hidden options between re-configuring proxy settings to wend my way through that minute's network settings (ok, it probably wasn't really changing on a per-minute basis)

By the time that was more or less sorted, and I had come to discover that the wordpress installs are presumably using 'mail' rather than any smtp settings and are thus apparently unconfigurable from within the interface, I suddenly got 6 lots of emails to tell me my registrations had been successful.  Which was good, as I was running out of email addresses.  WP doesn't seem to want to let go of my email addresses even now I have deleted the accounts, which is rude.  Not letting me cancel an impending request was fairly rude too.

Judged a couple of Loebner prize entries, and had discussions with Will re emotions and their place in intelligent systems.  Also discussed potential for a national PhD workshop on consciousness and cognition briefly.  Might be involved in a strand on consciousness at a workshop in Barcelona next year too.

Got verification that I am going to Leeds just before ALT-C in Sept to take part in a 24 hour workshop on technology and the future of HE in the 21st Century.  Rather pleased about this one - only 25 attendees.

Uploaded graphics for Mobi-blog.  Dealt with several PhD conference issues (though am starting to think that so few people will submit, we may as well cancel it).

Managed to get MeAggregator to compile and ran the java based stuff on the Mac.  This is actually the only thing I have done today that was part of my plan - which was to be documenting MeAg this week.  So far, 0 documentation, but then in some ways keeping Euro partners happy is probably more important.  Oh, and thanks Karsten for taking time out to help me make it work on the Mac :-)

Also been trying to figure out what to write for my paper for the PhD conference.  Deadline at the end of the week, and I am not even sure what topic I want to present on.  I think I will actually do a comparison between VLE and VRE requirements, as a theoretical study, grounded in the similarity of learning and research - research being the generation of knowledge which must become public to be of value and learning being the generation of knowledge internal to the learner (and which can become public) - both of them need essentially the same tools and both of them seem to benefit from collaboration.

So, not as succesful a day as I might have intended - but a bit of customer service goes a long way, I guess.

Both socially and workwise.

Socially I was at a Christening on Sunday (yay Max!), an 18th on Wednesday (Yay Jenna!) and will be at an engagement party tonight (yay Edwin and Harriet!) - makes me worry whether I will get a wedding and a funeral next week...

Obviously there has been the Unconference, and there was the news about Eduserv funding, I went to a great seminar yesterday about possible mechanisms supporting intelligence, and spent a chunk of time doing SSE Conference organisation.  I also got round to joining the BCS and have been investigating the way they deal with competencies/topic maps/open learner models (call them what you will) which is work, training, professional development and PhD related all in one go.  It will take a while to actually go through and map out what I do, and how well, and even longer to try to do the same thing for my goals.

I have spent most of today getting the Mac to behave itself, and copying files across from the old Windows box.  I think I am just about there, but I still have to install some programming environments (apart from the fact it seems to have come with Ruby and subversion installed, which is nice :-))

Also ran into Nick from Business school today.  Had been to a seminar by LS about using ontologies to describe learner competencies.

Ah yes, also started reading Groundswell (Charlene Li and Josh Bernoff) this week (after skimming bits of it last week) which is a good book.

Not quite sure what I am planning to write for my SSE Conference paper.  I keep having ideas about which strand of research to explore, but havven't yet reached a decision.  And I have the new book chapter to start getting in order with the same deadline.  Now, I wonder if anyone can guess what I am likely to end up doing...

Room booked, food ordered, joining instructions sent out, projector booked; still need to get flip charts, pens, post-its etc organised. Looking to be a brilliant day :-)

I have wanted to write some advice on how to choose the tools which go to make up your Personal Learning Environment (PLE) for some time.  A PLE is the set of 'technologies' you use to help you learn, organise your learning and generally do anything related to learning.  From this perspective it includes things like Blackboard (though you don't get much of a choice on this one), RedGloo, Google Scholar, Del.icio.us, Clipmarks, Flickr, YouTube, email, IM etc.

But it isn't always obvious what constitutes a useful addition to your arsenal of tools, so I wanted to write something to help people decide what they thought was useful on a cognitive basis rather than on the basis of "Fred is using it, so I started, and now my data is there and I can't be bothered to change..." or similar.

I am trying to decide whether to try and put it together in RedGloo, or use ITNGs wiki.  I am leaning towards the wiki - that will make it nice and easy for other people to contribute/correct my (probably biassed) ramblings.

I have booked myself a place on

Conference: Emerge Community online event: Exploring User 2.0: the shape of future users, 23-25 June 2008