Konfabulator is one of those things which I have been meaning to check out for quite sometime. Since seeing the Dashboard idea in OSX 10.4, I have thought it was a pretty good idea. For example, the amount of times I'm sitting at my laptop and I want to quickly do a complex sum with the built in calculator. Well I have to click start menu, find accesssorys, find calculator. Yes I know I could assign a keyboard short cut but come on. So using widgets in Konfabulator, I am able to click a hot key and get my calculator in without doing any navigating or finding. But thats only the start of things… Theres this really nice widget (UK Train Timetable) which grabs the latest train times from the official Train timetable website and displays any delay's or cancellations.

So from own understanding, Konfabulator is using the Mozilla javascript engine on top of the operating system which allows widgets to be built in the same way as you create applications on the web. This is very interesting when you consider the real push behind dynamic web applications like google maps. Ajax or as I prefer remote scripting creates a inituative user interface to underlined webservices. With widgets in konfabulator its easy to imagine taking applications right into the operating system. I understand this is exactly what the Avalon/XAML idea in Longhorn is meant to do. But you know what, its available now and being built on Standard Javascript/DOM is good idea. I have been meaning to improve my Javascript/DOM skills for quite some time now, this is the perfect reason why. I wonder how easy it is to connect to the web and the file system? Does it use xmlhttp or something else? There are a;ready plans in my head to talk to cocoon and other web applications. How hard would it be to read a xvid file on my filesystem and search for the informaton for it on IMDB.com? I'm sure the interaction with the user interface and filesystem is limited but not that limited surely? I mean for example there is a nice wireless strength indicator which can be downloaded. For it work, it must plug into the operating system and read that information in some way? There are also many system monitors which to me indicates that access to operating system information must be easy or at least flexable.

There are performance issue with layering a Javascript engine over an operating system, sure. On my 1.3ghz Tablet PC Konfabulator realisticly takes about 2 secs to switch on my button switch. I have about 8-10 widgets running. While on my desktop 1.8ghz Athlon box it takes nothing more than half a second at very most. Bar Avalon and somewhat Dashboard, what else could be used instead of the javascript engine? Java? Python? what else? See this is the thing, Javascript drops the entry level down real low, you do not need to learn a while programming language to create widgets. This is good! Yes performance is one of the trade-offs but come on a widget is meant to be small and simple. For example there is a nice little search box which I have as a widget, it simply converts my string entry and passes it on to the locally installed web browser. Makes sense. It wouldnt be too difficult to send the request to google search webservice and get results back in the widget as a small list. Cool, but why bother? Unless your building a browser as a widget, just pass the query to the browser at the start.

Not trying to do a lazy web, but heres some thoughts for future widgets.
Del.icio.us adder – What ever webpage you happen to be on can be added to del.icio.us through a widget instead of the browser. There is already a firefox extention which I use which does the same but it would be cool to see a widget too.
Upload to Flickr – Uploading to flickr is simple and there are 2 ways to do so right now. Download a application or do it via a webform. What about a widget which allows you to drag folders of files to be uploaded. Once its finished it opens a webpage where you add the metadata. Or you could do it from the same application too I guess (I prefer not to).
Blogging it – I like wbloggar and other blogging clients but come on, its all heavy weight for posting to a blog. My blog software already has a javascript bookmarket for posting, so why not a widget too? Actually it would work quite well because maybe it can accept drag and dropping urls and files too?
Tell me a route – I use transport for London's Journal planner all the time. It would be so cool if I could just put where I am and where I'm going and it could send the request to TFL. Then chart a route or send the response to the locally installed browser. This would also work with Google maps.

Following on from my last post about RSS and Azureus, I'm thinking once I finally get XML or something structured out of Azureus and its completed or in progress queue. I should build a widget which shows me the latest TV shows downloaded or in progress. Yeah I know Azureus has a little download bar but its too abstract and through cocoon I could get updates for in the same way even when I'm not on my local network but at work or even roaming. Pretty powerful, you have to agree? Yes all I'm doing is reformatting what comes out of cocoon but hey, it can be alot more with a lot more time and coding. Instead of just a list of whats downloading and whats done with percentages. How about a tv lookup so you get a image and nice small percentage instead? Simple and effective.

So generally, yeah I do not think Konfabulator is the best thing since sliced bread, but I do rate it. I will have a better feel for how I rate it once I get into writing widgets of my own. I may be so wrong and widgets could be based on a twisted version of javascript and requires super human knowledge to write. Maybe connecting to the web is difficult and reading machine or operating system information almost impossible. But on first judgement its good and effective. I'm happy with it for now. There was another program which I downloaded but didnt end up testing called Samurize. Its open source which is good because konfabulator is payware but quite cheap at only 30 dollars.

Comments [Comments]
Trackbacks [0]

I stopped listening to Adam Curry's Daily Source code quite a while ago. Tell a lie, I do still download the podcasts, but Blogmatrix's Sparks usually does delete the files before I get around to listening to them. At first it was interesting, well produced and a great chance to get a feel for what was going in the podcast world. However podcasting has moved on, theres a lot more choice and there is no need to know whats going on as such. (Its a bit like a blog about the blogosphere, however I do listen to blogosphere radio now and then). Anyhow around the same time as listening to the Daily Source code, I was listening to Dave Slusher's Evil Genius Chronicles.

But why am I still listening? Well simply, Dave Slusher's podcasts have a much higher level of quality and narrowing that Adam's. I mean he knows whos listening and does not do this general radio style which I and others tend to hate. The Daily source code is a radio show as a podcast, its so general and does not take advantage of the nature of podcasts. Someone once said recently, Its NOT everybody (mimicing adam's voice). And in that statement, says it all. Dave Slusher plays music he loves and talks about subjects which interests him. Adam servers more like a radio dj reporting things which he has heard and been given. Yes he has a huge audience. Yes I do not like the music Dave plays, but screw it. Dave has a quality audience and the narrow band idea tied up.
Dave actually explorers this futher in this post and this podcast. And honestly I've been thinking about this whole area myself…

Its all about metric's, and Dave took the words out of my mouth.

The Podcast Alley fracas is mostly culture clash between the old methods and the new context. The more I think about this, the more I think the focus on the sheer size of listenership is taking the worst of the old situation and applying it to the new world. We don't need to think in channel-limited scarcity mode any more. It made sense when you could only have so many FM or AM channels max in any market, but it doesn't make sense when you have a nearly infinite variety of channels.

I dont really care whos number one on podcast alley, it makes no difference to who I listen to. But I do understand that old/dead media still does metrics by quantity not quality. This is echoed by Doug Kaye who is the owner and creator of IT Conversations. Who has a couple of times asked for listeners to vote on podcast alley, saying IT Conversations should be in the top 50 at least. While he and others (like myself) who listen may not care about what position its at, advertisers will be more interested if its closer to #1 at podcast alley. Its just the way they do metrics at this moment. The question is what can be done about it? Well there's hope from Doug Kaye. But in his answer, lies the actual issue…

I pitched the idea of a ratings system like Amazon, Netflix or IT Conversations, but as he pointed out, that doesn't work for his site. Chris can't just publish an 'average' rating for each podcast, even with some minimum number of votes required. Why? Because a podcast with five votes of “five stars” each, would then be rated higher than one with one thousand five-star votes and just one four-star vote. It's not a problem for IT Conversations and these other sites because 'ranking' isn't as important as the how-good-is-it rating for each item.

Why is the ranking system on IT Conversations, Amazon, IMDB, Netflix, etc not as important as the one on podcast alley? Is it because people realise that you can not compare one thing against another? That views are subjective and relative? What if the Daily source code is number one? Does that actually mean its better than IT Conversations? or vice-a-versa? What does being number one actually mean?

I blame the old/dead mediums for not growing up and moving on. THERE IS NO SCARCITY, anyone can podcast or write a blog, and the abundance of the internet through networking keeps the statement true. Its time to reconsider your metrics, because once again THERE IS NO SCARCITY and its no good trying to create a artifical scarcity. And the other point worth making…

The podcast infrastructure is very open to narrowcasting (I'd go as far as to say it is optimized for it). The popular podcasts in sheer volume of “units shifted” will always be the more general ones. However, a podcast that serves a small niche audience and serves it superbly well will always be lower in total downloads but could be very high in the axis of serving the needs of the listeners.

This was made very clear the other day when Doug Kaye asked listeners to send emails to people who could/would be interested in Underwriting with IT Conversations. With IT conversations narrowcasting to its target audience the Underwriting Campaign was a good success because of a quality audience. What more could a advertiser in the IT world want? Dave agrees…

People keep talking about how advertisers and sponsors want to see “big numbers.” I'm not so sure that is the best way. It is certainly not the only way. If a company has a product or service that is related to that niche interest, they might be getting a much better deal in sponsoring that podcast. The high affinity the listeners have for the show coupled with the focus of the interest may make it a great deal and a more efficient use of sponsor dollars that a general purpose show with a huge listenership.

There are no simple metrics to measure the relative affinity your audience has, or to determine the aggregate influence your listeners wield. In contrast, it is fairly easy to count concurent streams or determine download numbers so that will be what things are based on. This focus on volume, on popularity, on being the top in some ordered list – it all reflects vestigial thinking from the old way of doing things.

And in that lies the problem, its hard work. Its not something you can just count and be done with. I would go as far as say this is exactly what the long tail is all about. Of course large easy to count figures work well in the start of tail but as its spreads into the long tail you need to start thinking differently. Start thinking quality conversations with a your audience, not the old style everybody style broadcasts of yesterday. I know theres been some reaction to the long tail idea. One I heard recently was from Stephen Downes talk at northern voice where he asked, who really wants to live in the long tail?

So people talk, and people have talked a lot, about the long tail and they've said “Worship the long tail, mine the long tail, the long tail is where the action is.” And all of these people who are talking about the value and the virtue of the long tail have the unique pquality of not being part of it. I live in the long tail. And I can say from my own personal perspective that people who are in the long tail would probably rather not be part of it. They simply want to be read.

Stephen certainly has a point, but I don't believe its as simple as wanting to be read. For example, if I simply wanted to be read I could host cubicgarden.com on a dedicated server and spam all friends, family and there friends about it. Yes I would be read, but honestly knowing I'm read by people who are my peers and also my worst enemy's as such is much more interesting and also much more manageable. Imagine getting 100's of comments per entry? Is that better than recieving that one which points you in a direction you never considered before?I certainly think so and its the reason why I listen to Dave's podcast (with even the music i dont really like) over and beyond Adam's.

Comments [Comments]
Trackbacks [0]