Category: socialware-online

I have been thinking about using some other tool to capture news content which dont quite make its way into slashdot and boingboing. Metafiter is one which Tom uses religiously but I dont quite like the old style of web product it is. Del.icio.us popular is good, but the content tends not to be totally news. Sometimes you get the links related to the news, instead of the actual news. This also applies to technorati popular.

So I'm testing out Digg which has a slashdot type model but allows for a much more longtail approch via friend aggregation as well as the whole digg nation. Talking of which, there is a show hosted by Kevin Rose and Alex Albright called Digg Nation where they take the top few stories from a week on digg and simply talk about it. Its kinda of slashdot review but with comments turned on. The show comes in audio and video podcast forms and I have to give huge kudos to Jon for making there feeds compatable with a RSS TV setup. Do check it out if you have a Azureus with RSS automation type setup. But back to Digg Nation for a moment. Currently Digg Nation only deals with the most digged/popular news. But theres nothing stopping someone doing a show about the most digged/popular content between certain groups of friends or a certain category. If they adopted a tagging option too, it would be endless.

For now digg is getting the thumbs up for me. I know I would love to see digg move away from the website. What I mean is tools and applications which mean I dont need to log into digg to do everything. At the moment you can get RSS feeds for pretty much anything in Digg, but it woud be great to see APIs for posting and digging news. Maybe directly in the RSS reader via a RESTful url like how a trackback ping works. So you can see what being digged and dig it by simply clicking a link at the bottom of the entry.

Comments [Comments]
Trackbacks [0]

Honestly I dont really go for these social networks for the sake of social networking. Linked in, Friendster, etc have been some I've looked at but have not been that interested in. The one I invested the most time in is Microsoft Wallop, which is still an invite only beta. At the moment its simply reading my RSS feed from this blog and using it as a base for my own wallop blog. Weirdly enough, it also counts anything I link to as a resource of my own which it then stores. Anyhow, i've recently started using O'Reilly's connection beta which is another social network but focused around developers, net type people and designers. The ability to get FOAF data out seems really simple and I have to admit it seems quite simple and not this super elaborate affair of wallop or orkut.

I do however have to wonder once again, how long I'll stay with this one? All the social networks I use all the time tend to be more like social tools. Del.icio.us, Flickr, Audioscrobbler and even recently Yahoo's new killer/social tools, seem to take a different view of social networking. There much more intergrated into daily life that a place to hang out. If O'Reilly's connections was to open things up a litle more, so for example using other services for job listings, etc. Then maybe it would stand a better chance at staying relveant. I do have a feeling that they will base it around there books and conferences in the future but will that be enough? I mean is it a career development site? no is my gut answer but what room is there for connections in face of the other tools?

I was kinda of upset that it didnt import my current FOAF profile. I mean I would have uploaded if they liked, but I would expect it to look at my website url and read the linked FOAF file without a bother. But oh no, I need to add all the data again via multiple html forms, which really sucks. Isnt this the point of FOAF in the first place? I have no idea if there's the ability to see inside the network from outside? For example here is my profile view and the FOAF data which is created from it. Please note, although I said I was based in the UK, it still puts US in my FOAF data. Maybe this is a bug in there processing? It must be because for some reason it hasnt even got my surname – forrester! I'm also not happy about the fact it uses many redirects to make up the FOAF data. http://connection.oreilly.com/users/profile.public.php?user_id=1671/ is not my homepage sorry.

There is no way i can use this as my FOAF profile till its sorted out and reliable. Also note everyone in FOAF profile has to be registered with Connections for it to add a entry to my FOAF data. This is deeply inward looking and this is why I think Yahoo 360 will succed where others failed. I mean think about it for a quick second. Buy a social network/tool like flickr with millions of users and billions of user generated data then buy yourself on to peoples desktops through Konfabulator. The possiblities are pretty endless, now I can see where the 360 nature comes from.

Comments [Comments]
Trackbacks [0]

Just a quick note that I'm attepting to keep a small link blog of things I read, listen to or watch which gives me a little inspiration. The blog is simply called inspiration and will be more of a challenge in my process of blogging that anything else. Ideally I would just mark something as good enough for the link blog that loading up a blogging app and blogging it. I'm sure this is very possible with something like Automator for OSX 10.4 but on windows without knowing VB? Maybe not…

Comments [Comments]
Trackbacks [0]

I recently sent the guys at blogmatrix a email asking whats going to happen to Jaeger in the light of Sparks! I use Jaeger everyday because it pretty much does what I really need a RSS reader to do. It stays out of the way of my browsing and allows me to glance over many feeds (almost 250 now) very quickly while working away. Its not perfect however, there are bugs like the sync bug which ignores your proxy server settings. So I cant sync my work machine with my laptop and home machine. There are also lots of little things I would love to add if I was a python hacker, like more search engines, send via email (generates a email with the link, title and small description already written), post to blog or del.icio.us and output xml/rss. I'm almost able to hack this together myself except its a XML or XHTML only affair. There seems to be no way to have both. The templates are luckly velocity templates but there seems to be no way to specify other templates Jaeger should server up. On that same vein, I'm not able to get Jaeger to serve up HTML outside of the local machine. Not sure if this a constraint, bug or something I'm doing wrong. Anyhow, here's the email, I'll be interested in what others feel.

Hi David and others,

Quite a while ago I wrote in regards to Jaeger support and what's going to happen to it? I just wanted to ask again, with all the support and push for Sparks, I'm wondering if Jaeger would ever make it to a version 2? I understand the development and time constraints attached to having 2 products Jaeger and sparks.

Dont get me wrong, I like Sparks! But its well over kill for myself, there are also some things which have been lost in the transformation.

* Single Panel (works with my standard browser)
* Offline/bulk reading seems better in Jaeger
* Synchronization! (This is a lifesaver when working with 3 Jaegers clients, I even got my pocket pc rss client to read the stored OPML every week now!)
* Universal Search Engine (I'm actually hacking around with Python to add support for A9 and others)

If Sparks had all these I would upgrade, but it strikes me that Jaeger and Sparks are different products and thats a good thing. So in light of this and the development effort, I wonder if you had ever considered,

* Open sourcing Jaeger or making it more open at least (if I was a python programmer I would love to re-engineer Jaeger, right now my XHTML+CSS templates are custom to my preferences which is the best I can currently do. But I had thought about hacking Jaeger into outputting XML which I could then transform into a Konfabulator widget, screensaver or something else)
* Releasing a version without the licence nag? (Now sparks is free, shouldnt Jaeger be too? I believe you were charging $30 per client for Jaeger a while ago?)
* Make a donate Jaeger version (I would be happy to donate to Jaeger but not pay $30 for each version of Jaeger I install on my many machines)

Anyway, I hope you understand this isn't an attack of Blogmatrix, I personally think Jaeger is great and still beats some of the mature rss readers like blogbridge, feeddemon, etc. But I'm left pondering what future it has everytime I get the licence nag. Plus there are bugs in Jaeger which seem to be fixed in Sparks but I'm unsure there ever make it to Jaeger ever

I recieved a email from David Janes from Blogmatrix after a couple of hours.

Hi Ian,

The source code for Jaeger is available on sourceforge, if you'd like to
play with it. I'd be more than willing to offer help as time allows.
FYI: Our llicense agreement says it can't be rereleased without the nag
unfortunately.

So it looks like Jaeger will be supported but its going to take lots of time and resources from the blogmatrix guys, till then I'm hoping some open souce developers pick it up and do some interesting things with the code. Yep a little lazyweb but hey theres nothing wrong with that.

Comments [Comments]
Trackbacks [0]

(If the picture above never gave you a big enough clue)

Well I have not been blogging as much recently, thats one thing.

Recently I have been messing with Apache Cocoon 2.1.7 and been trying to get it to play nicely in a shared JVM environment. It looks like Cocoons security manager can not be turned off, so this may mean the end of the road for cocoon in this case. I've also been trying to sort out some of my ideas and projects. I see a personal wiki as an ideal way to do this but want to deploy one which might be able to do more content management type things later in the day. The two I'm looking at our xwiki and daisy. I was very impressed by the papers and presentations they both gave at Xtech and feel I could use them for myself and bigger.
Daisy is servelet based and basicly uses cocoon to do a lot of its transforming which means I can easily hack it into doing anything I like. Xwiki on the other hand seems to be a typical LAMP (linux, apache, mysql and perl) project and requires me to install all this other stuff which I dont normally use. Yes it is odd, but I find Resin 3 quick enough to server static files for me not to need apache on the front end. The Mysql server option is needed for both daisy and xwiki so it looks like I need to install that no matter what now. I'm hoping if all goes well I will end up moving either wiki to my live webspace in america.

So how far have I got with Daisy? Well I ran through all the instructions and have everything run except Starting the Daisy Repository Server causes a nasty error on my laptop. I type daisy-repository-server c:daisydata and get

C:daisy-1.3repository-serverbin>daisy-repository-server c:daisydata
—- exception report ———————————————————-

Exception: org.apache.avalon.merlin.KernelException
Message: Kernel startup failure.
—- cause ———————————————————————

Exception: org.apache.avalon.merlin.KernelException
Message: Cannot deploy application.
—- cause ———————————————————————

Exception: org.apache.avalon.activation.LifecycleException
Message: Unable to create a new component instance in appliance [/daisy/jmsclien
t/jmsclient] due to a component deployment failure.
—- cause ———————————————————————

Exception: javax.naming.NamingException
Message: Failed to connect to JNDI provider
—- cause ———————————————————————

Exception: java.net.ConnectException
Message: Connection refused: connect
—- stack trace —————————————————————

java.net.ConnectException: Connection refused: connect
java.net.PlainSocketImpl.socketConnect(Native Method)
java.net.PlainSocketImpl.doConnect(Unknown Source)
java.net.PlainSocketImpl.connectToAddress(Unknown Source)
java.net.PlainSocketImpl.connect(Unknown Source)
java.net.SocksSocketImpl.connect(Unknown Source)
java.net.Socket.connect(Unknown Source)
java.net.Socket.connect(Unknown Source)
java.net.Socket.Unknown Source)
java.net.Socket.Unknown Source)
org.exolab.core.ipc.TcpService.TcpService.java:122)
org.exolab.core.ipc.Client.Client.java:91)
org.exolab.jms.jndi.mipc.IpcJndiInitialContextFactory.openConnection(IpcJndiInit
ialContextFactory.java:210)
org.exolab.jms.jndi.mipc.IpcJndiInitialContextFactory.getInitialContext(IpcJndiI
nitialContextFactory.java:140)
org.exolab.jms.jndi.InitialContextFactory.getInitialContext(InitialContextFactor
y.java:109)
javax.naming.spi.NamingManager.getInitialContext(Unknown Source)
javax.naming.InitialContext.getDefaultInitCtx(Unknown Source)
javax.naming.InitialContext.init(Unknown Source)
javax.naming.InitialContext.Unknown Source)
org.outerj.daisy.jms.impl.JmsClientImpl.getContext(JmsClientImpl.java:217)
org.outerj.daisy.jms.impl.JmsClientImpl.initializeJmsTopicConnection(JmsClientIm
pl.java:101)
org.outerj.daisy.jms.impl.JmsClientImpl.initialize(JmsClientImpl.java:93)
org.apache.avalon.framework.container.ContainerUtil.initialize(ContainerUtil.jav
a:244)
org.apache.avalon.activation.impl.DefaultComponentFactory.incarnationefaultCom
ponentFactory.java:402)
org.apache.avalon.activation.impl.DefaultComponentFactory.incarnateefaultCompo
nentFactory.java:130)
org.apache.avalon.activation.impl.SingletonLifestyleManager.refreshReference(Sin
gletonLifestyleManager.java:144)
org.apache.avalon.activation.impl.SingletonLifestyleManager.commission(Singleton
LifestyleManager.java:60)
org.apache.avalon.activation.impl.DefaultAppliance.commissionefaultAppliance.j
ava:94)
org.apache.avalon.activation.impl.DefaultRuntime.commissionefaultRuntime.java:
86)
org.apache.avalon.composition.model.impl.DefaultSystemContext.commissionefault
SystemContext.java:362)
org.apache.avalon.composition.model.impl.DefaultDeploymentModel.commissionefau
ltDeploymentModel.java:122)
org.apache.avalon.composition.model.impl.Commissioner.run(Commissioner.java:185)

java.lang.Thread.run(Unknown Source)
——————————————————————————–

If anyone knows what could be the problem I would be greatful. Otherwise I may try it on my server and see if its just my unconfigured laptop which could be the problem.

For Xwiki I will install it directly on my server and see if thats any good.

Some of the other things I'm planning on doing at somepoint in the future includes.

Calendaring – Getting Sync4J up and working with Outlook, my 3g mobile and pocketpc. I assume this will stay on my local server. SchoolBell Calendar server (unless I can find a servlet version) looks very simple and seems perfect for calendaring between me and sarah. This might not be so needed with us both using Outllook again, plus sync4j should help in this area.

Email – Finally getting Zoe if its still being worked on up and running to store all my email so I can do away with the silly email problems I currently get when away from home.

Wiki – Explained above, either xwiki or daisy to store thoughts and do something which doesnt quite work in a blog.

XMLDB – I have done quite a bit of work with xml databases but would like to try it out for total replacement of flat filesystem files. Theres also quite a need for some serious search logic and I see Lucene being really useful in this area, which means I need to learn and understand what it can and cannot do. This could be quite challenging but thats a good thing.

Aggregation – Rome is a Java API for RSS and ATOM feeds. Cocoon can deal with these feeds no problem it seems to be lacking on caching ability? ROME seems to have caching sorted which means it shouldnt be too hard to use the ROME jar in Cocoon if needed.

IM – Not only am I really interested in learning jabber bots but also using IM as a transport network for content. Punjab seemed to be perfect for my needs as it was a XMPP to SOAP/XMLRPC/REST bridge. But it seems to need to be installed on the actual jabber server which is a pain because who would install it then open it up for all to use? I really do not want to run my own jabber server any time soon either. Theres are also some interesting projects opening in the area of Transporting ATOM notifications over XMPP. I also still need a decent jabber client for my Sanyo S750.

Wireless – I have yet to get deeply involved but I would like to help Bristol Wireless or some south London community wireless some time soon. I also signed myself up at warchalking London and should hopefully be delivering some nice blog posts from South London where there is open or cheap wireless access.

Webservices – I'm still generally collecting a load of RESTful APIs using del.icio.us idea being I will start using them to generate some quite unique services and applications. One I can let you in on is simpleflickr which will simply turn flickr into a image search engine. Yes I know flickr is quite simple already but I remember introducing my mum to flickr over the holidays. She found flickr quite confusing because there are so many aspects to every single part of flickr. So simplfying it so my mum can use it to see and print great pictures sounds like a nice easy project. Many afterwards I could use the APIs and routines which I've built up to generate something a little more wild and creative. Oh and I still need to do my attach images to bbc worldservice news stories for backstage.bbc.co.uk. Maybe I'll do it in time for backstage.bbc.co.uk Open Tech 2005

Other things holding me back from blogging recently. I am 100 percent legal to drive my Yamaha YP250 Majasty Super Scooter around the UK. I passed my A2 motorcycle test on the same day as the Geek dinner which was the 6th June. The A2 licence basicly means I can drive any automatic motorcycle up to 33 bhp for 2 years. After those two years (2007) I will be able to drive any automatic bike I like regardless of power or size. I have my eyes on a Yamaha Tmax (500cc) next or even a Honda silverwing (650cc), but till then my Majesty (250cc) is fantastic during this hot summer were currently having in London. Honestly I adore driving it around in everything except the rain. A lot of people ask why I went for automatic only? Well if you look at the range of touring bikes you will see Tmax's, X9's, Silverwings, Burgmans, etc all running between the ranges of 200cc and 700cc. There all automatic and work really well. So I see a great future for scooter driving. Maybe the first 1000cc scooter will launched soon and I bet it will still use a automatic engine. For me personally 500cc scooter will do nicely and I dont see the need for more.
I also just recently finally got Orange to swap my old 3g handset for a new firmware version which solves most of the bugs with bluetooth connectivity. It also comes with a few extras which the previous firmware never had but generally its the same handset and works no differently. Sanyo did launch there 2nd european phone which is the same as the S750 but with a different design. Yes you guessed what the model number is, yep Sanyo S750i.
And last of all, my main workstation computer went down recently and I have no idea why. All I did was put in a new fan/heatsink and a pioneer dvd burner. Neither should cause the machine to go crazy and not boot ever again. I tried swapping things in and out, but no matter what I do, I can never get the bios screen back. So I shifted everything over to my spare linux box ual Pentium3 500 xeons) and reinstalled windows xp. You'll be glad to hear I also put Ubutu linux and BeOS (5 max edition) on the machine too. Maybe in the future I will beable to put Apple OSX on there too…

Comments [Comments]
Trackbacks [0]

Sorry being sitting on the information that Google maps has finally made its way across the ocean to the UK. Go check it out under http://maps.google.co.uk. All the features are there except the Satellite imaging.

Its quite interesting that the Train stations near me do not actually sit on the actual train tracks?

Comments [Comments]
Trackbacks [0]

Jon Udell has a nice and simple piece explaining the problems between delicious and delirious. From my own point of view I dont really care about the rip off-ness of the new service, end of the day there are many search engines and soical software apps which look simular but no one batters an eyelid. Honestly I would say its more of a compliment or a form of flattery. Anyhow the issue I do have is the open source-ness. I dont like my data not being mine, when I signed up to delicious I knew this but it was the only thing widely available at the time for bookmarking. Now I may change my bookmarks over to delirious because there is an alternative. I want my data to be under a creative commons licence which suits me and even though I dont have plans to build my own social bookmarking service, I'm more happy to know my data will not be locked into a service which could go bust. Yes you can get all your data out of delicious by using the API but thats not everything is it, I'm sure theres tons of linking information which I cant currently get my hands on. There's parellels between this debate and others like the attention.xml debate.

Comments [Comments]
Trackbacks [0]

I was using Flock for the last few years but moving to a shared servlet environment is making me reconsider using Flock. See this is the problem Flock has been worked on for the last 2 years and still has no ATOM support. It works well when its got a small amount of feeds but the fact you cant even change the duration of the feeds collected is troubling. After trying to install Flock on cubicgarden.com I got a reply from the excellent support from interadvantage.

I was excited to see that Flock runs on Tapestry, because I've never tried to get any Tapestry apps running on our servers. This is what I did:

1. I updated Apache's config to pass all /flock/* requests to Tomcat.
2. I then added security grants so any Webapp can read the net.sf.tapestry.* and flock.* system properties.
3. I then ran into a roadblock…

Flock is going to be problematic because of the ridiculous tendency of open source Java developers to assume their apps will only run on a dedicated Tomcat instance.

The problem is that it tries to write to /username/*. As you know, we put your home directory at /home/username. Further, the only way (I can see) to override this behavior is to set a system property, flock.home. Sure, I can set this, but it'll be set system-wide so if another user wants to run flock, you'll have to share a directory. That's obviously not acceptable. Why can't they just use a flock.properties file in the classpath? I've recommended this on their mailing list. Of course, I could set flock.home to /home/forrei/public_html/flock/WEB-INF/flock or something, thereby limiting the server to a single instance of Flock (yours). But that's not a good solution, so let's see whether the developers are willing to do anything for us on this one before we implement a kludge.

Yeah and I'm honestly he's right. So till some nasty little hack is decided on, I will keep on running it on my local server but I'm on the look out for something else to replace it. Any offers? or even any thoughts around this problem with flocks home directory? I actually remember having a simular problem a while ago when I was first setting up Flock. I didnt want it to stick all its data in the home directory in documents and settings, so I did the classpath trick to fix it. Luckly Blojsom is much better built and this blog is well on its way to moving to its new home under http://www.cubicgarden.com.

Comments [Comments]
Trackbacks [0]