All posts by silvia

Annodex the solution for ethnographic researchers

Digital Media, Open Source

A few years ago when I was still at CSIRO, I was contacted by Linda Barwick from PARADISEC to research into the use of Annodex for linguists. The main problem was that ethnographic researchers are publishing research outcomes on paper or even HTML, which are essentially discussions about small sections of field recordings of exotic languages – however, they had no means to do citations of these sections through hyperlinks or any other simple interactive means. In the time of online media, that should be a trivial task, right? But it wasn’t. Annodex and the timed URIs provided the right basis for a solution.

Fast forward lots of months of work in the EthnoER project and you get a solution for ethnographic researchers which is unique and completely based on open formats and open source software. Check out Linda’s blog entry of today!

Congratulations to everybody who has put all that effort into the project – Nick Thieberger, Linda Barwick, Shane Stephens, Stuart Hungerford, Jonathan McCabe, and all the others whom I forgot. EthnoER and Annodex might have changed the way in which linguistic research online can be published – not a small feat at all!

Editing the Skeleton and CMML standards

code, Digital Media, Open Source

In the last few weeks, I’ve created an Internet-Draft (I-D – a draft specification of an IETF RFC) for the Ogg Skeleton meta track, and updated the CMML I-D to include a new element called “caption” (CMML DTD). All of this is work that should have been done a long time ago, but I only got the motivation for it through the WHATWG work on HTML5 which will take Ogg Theora and Ogg Vorbis as baseline codecs. Since liboggplay is the key open source library that implements this baseline codec support, and liboggplay supports Annodex, it seems plausible that Annodex (which adds essentially Skeleton + CMML) will be available in Web browsers of the future. So, now is the time to fix up the few open issues that remain and cast the specifications into readable I-Ds.

If you haven’t seen the great functionality that will be available with liboggplay, you should check out the liboggplay javascript API. I’ve seen Shane make a demo web page through which you can toy with the javascript API, but haven’t got the link available right now.

Usability testing for Web2.0 sites

vquence

Let’s say: you run a Web2.0 site. And you know it’s not perfect (no website is perfect – people just make do with what’s there). And you are actually determined to improve it. Who do you ask?

No: I am not asking you to send me a quote – go away! I don’t want to spend thousands of dollars on a professional agency to undertake usability testing and get some cool new designs for the interface and processes. Chances are I find out later that this apparently “perfect” solution makes processes too complex or uses the wrong colours for our audience or … well, it’s wrong again in some aspect. Our Web 2.0 site is targeted at teenagers and I predict that no agency out there can give me the right feedback – simply because they are corporate.

The solution is very simple indeed – and it will make for a few very happy teenagers – happy to be able to have input into a new online site and to tell their friends about it. All you need to do is ask a few intelligent and outspoken teenagers for a bit of time and invite them to your office or home for a few hours to do some semi-formal usability testing. You shall be amazed about the results!

I’d like to say special thanks to Charlotte, Will, Peter and Alice for taking the time on Friday for 3 hours to do a thorough usability testing exercise with us at Vquence. It was a lot of fun – and I dare say: for everyone involved. Here is the schedule that we followed.

We started with an introductory 30 min where everybody, including the testers, introduced themselves and we started talking about what we do online, what we love and what we hate – including the testers. This is important because you break down the barriers and create trust.

Then we had a questionnaire for them to fill in (the middle part of the schedule). It took them through the functionality of the site in a logical manner, asking for feedback on the design, the processes, the functionality and the usefulness of the site.

Then we got together again to “vent” all of that frustration. And … boy, did we get feedback! It was totally awesome – the positive feedback actually made us value our previous work a lot higher – and the negative feedback was great because it tells us what to fix or improve. They were quite direct in their criticism and had some great ideas for improvements. Since they know their world and what makes them and their friends click, they are really the only ones who can tell you what is great and what sucks.

Once we fix all the things that we’ve been told, we should probably do another round of testing – with some new teenage blood. But for now, I am extremely happy with the feedback we got – and will need to look for further investment to get all of this great feedback rolled out! 😉

Thanks guys!!

WordPress bug? (was: Usability testing for Web2.0 sites)

Open Source

Or .. why does password-protection suck in WordPress.

I wrote this entry and wanted to get feedback from the people that it was about, before publicly posting it. So I thought – wordpress has this great feature of password-protecting posts. Let me use that! But then the problems started: the post was actually added to my RSS feed and sucked in by a few planets and people started complaining that they were not able to read it. Well, it’s great to get feedback from the community that my posts are actually being read. But it sucks that wordpress handles password-protected post in such a bad manner. Is it something I did wrong or is that indeed a bug in wordpress? Leave me a comment!

LCA Multimedia Miniconf

FOMS, LCA, Open Source

The organisers of LCA have found another slot for a miniconf and ours is it! Yay!! We shall have an audio/video miniconf at LCA! This is particularly important since we will bring to Australia a large number of key open media application developers for FOMS. These guys will also be able to provide deep insight and understanding during talks provided to the more general LCA audience. Expect some awesome media talks at LCA!!

FOMS 2008 support by Mozilla Foundation

Digital Media, FOMS, LCA, Open Source

It is awesome to see FOMS – the Open Media Software developer workshop we ran for the first time this year – turning into a major audio and video developer event for Linux. FOMS 2008 will be in Mel8ourne in January and will focus on audio on Linux (in particular libsydneyaudio) and on native Firefox support for Ogg Theora (in particular liboggplay). Because of the latter, FOMS has attracted sponsorship by the Mozilla Foundation. This sponsorship is very welcome since most of the relevant developers come from overseas and are not part of large organisations that could afford to pay the expense. Check out the current list of participants on the site – it will be another milestone event for open media! And … thanks Mozilla Foundation!

Good Manners for Adobe Flash on Linux

Digital Media

Today, I had to deal with some badly behaving Flash content. The debugging process involved some extensive use of google and brought out a few interesting facts, which I thought would be good to be put together into a coherent story.

Webpages that embed Flash together with other layers of Web page content generally suck – in particular on Linux. It’s not the Web design that sucks – in fact: being able to use “depth” in a Web page is actually really nice for fitting more content on a page without making it too crowded (or hammering those flash ads at us – sigh).

Indeed, it is the technology that sucks because Flash on Linux doesn’t do what it’s supposed to do. Flash misbehaves on Linux by being bullish and always trying to stay dominant in the foreground over the top of DHTML content, even though being told not to. Don’t be prejudiced though: it’s not Adobe’s fault – or at least “not only”.

Before we try to teach Flash good manners, let’s first understand the problem.

When you embed an Adobe Flash 9 swf file in a Web page, Adobe provides a parameter called “wmode” (windowless mode), through which Flash is being told to be either “transparent” or “opaque”, either of which will prevent a Flash from playing in the topmost layer and allow you to adjust the layering of the movie within other layers of the HTML document. Layering in HTML is done by using the z-index.

Now, using wmode and “transparent” teaches Flash good manners on most operating systems. However, when the Flash ends up on a Linux computer, it’s old dominant personality returns and it turns evil again.

Yes, it is a bug. However, it is quite an interesting and long-standing bug – and longer-standing in Mozilla than other Linux Web Browsers, too.

Here is the story: On the 12th April 2002, Braden registered Bug 7189 with Mozilla about window-less plugin support on X. It was soon established that in principle window-less support is possible on Linux and the bug was re-focused on Flash. It turned out that multiple levels of fixes were necessary to get this working: at first, X needed window-less plugin support – a patch to fix this was provided by Dec 2004 by Pete Collins. Next thing necessary was support by Adobe to have an active “wmode” parameter, which was provided by end 2006. In April 2007 finally there was a first patch for Mozilla. A later patch was finally applied to trunk on July,2nd 2007. W00t!!

So, while we are all waiting for this patch to make it into Firefox and Linux distributions, to finally end up on our Desktop and make Flash behave, there is indeed a way for keen Web developers around this problem. And I can’t believe I’m actually going to say this: iFrames were the answer to our problem in my July’s blog post – and iFrames are again the answer here. Evil, detested, but oh so useful iFrames. Who would have thought…

A long story of logins

code, vquence

The new vquence website, that lets you create vquences yourself, has been out for more than a week now. But we were unhappy to talk about it because the login code was screwed up. It all happened in the last minute: a crash between the beauty of design and the reality of Web browsers. Who won? Well …

Between Julian and I, we had decided to do a cool sliding effect for user login and signup (try it out on the site – it’s hard to explain in words). Very Web 2.0 – very shiny. It all worked well and we quite happily ran the code for a long time.

Then came the time that we wanted to give outside users access and that our more security-conscious people suggested to exchange user passwords over https rather than in plaintext. Fair enough – no worries – quick fix to change the protocol, right? Bah – all wrong – because now you have http and https elements on the same Web page needing to talk to each other. Big sandboxing issue in Firefox – and worse in IE.

Anyway – I leave the details of this problem – and the ultimate solution for it – to somebody better suited to explaining it. I will only say that it has cost us days of pain and suffering and bug #93 will forever be remembered.

But – we fixed it – and now you can happily create video playlists on www.vquence.com and share/socialise them. I have one on my facebook page. 🙂

Vquence gets its first real competitor

vquence

Today, I heard about SlapVid. SlapVid is a demonstrator built on a p2p system built by 4 CMU graduates and supported by Y Combinator. The SlapVid flash player has similar functionality to the Vquence player in that it concatenates slices of videos from a collection. And – like we did for the Vquence alpha – SlapVid’s website presents the top videos from youtube in their demonstrator. Seems a bit of a copycat. 😉

At the moment, you cannot generate SlapStrips yet, but they’re progressing toward it. So are we – and fast – expect to author vquences by the end of this week!

I for one welcome the arrival of our first competitor (even though SlapVid may not be called a complete competitor, because they largely rely on p2p to provide their service, while we rely on our own infrastructure mainly, which will allow much better control over the user experience). A competitor in the market place validates your ideas and business model. It makes the discussion a lot easier than if you have to explain the concept over and over again. Also, if somebody else is doing it, we’re obviously doing something right. Let’s just try and stay ahead of them. 😉