A Django site.
March 18, 2008
» IEEE workshop on Next-Generation Networking and Open Source Technologies

IEEE NGNOPS- First International workshop on NEXT GENERATION NETWORKS: OPEN PLATFORMS & SERVICES is being co-hosted with the IEEE NGMAST by the Integrated Communications Research Centre, University of Glamorgan in Cardiff between 16th -19th Spetember 2008.

Why this workshop?
Converged networks and services are becoming a fast reality to provide anytime, anywhere networked services. While the standards bodies and the processes are providing the necessary normalisation by way of standards, the implementation of standards compliant networked systems to provide for converged services require development and test deployments before they go into production networks. Open Source software packages in these areas will make a difference by their availability and therefore providing a means of test deploying the new services. However, the challenge is to identify what is available as well as what is expected to be available from the open source community in terms of such software packages. The proliferation of mobile computing provides scope for many innovative mobile applications. Emerging open mobile devices, and devices that come with open development kits would make available the development environment to a large developer community. Such resources, projects and/or communities need to be identified and/or facilitated. From a business perspective, there is a big opportunity to realise cost effective services by providing cost effective solutions. This is true of markets in South and South-East Asia. Can open source solutions make a difference here? Are they interoperable with other commercial solutions? Again, what is available and what will be available? Finally, are there any technical concerns in the open source developer community that are not addressed? Would this community require any additional facilitation? The NGN-OPS workshop intends to address these issues.

Objective of the workshop

The objective of the NGN-OPS workshop is to discuss the open source activities contributing to aspects Next Generation Networking (NGN), mobile devices, development environment and open source technologies enabling in creating an open ecosystem. We are inviting both theoretical, developmental and experimental papers as well as work-in-progress research projects. Topics of interest include, but are not limited to:

Open Source software development projects and community efforts relating to

o Next Generation Networking
o Mobile Devices
o Mobility Technologies
o Pervasive Networking
o Mobile Applications Development Frameworks and Tools
o Deployment of Community Networks using Open Source software and tools

Open Standards efforts relating to

o Next Generation Networking
o Mobile Applications and Web 2.0
o Network Management for NGNs

Open Source based

o NGN Services prototypes
o NGN Services test beds

Community based efforts in Wireless Networking

November 22, 2007
» J2me web emulator (show off your work)

"wouldn't it be nice to link to my j2me application where everyone can view it anywhere in the world without needing to explain any technical details to them??"

I am sure every J2me developer thought of this at a certain point and I can imagine some smart ones have had their own workaround ages ago.

Just yesterday, I found this free web emulator for J2ME application and seriously all you have to do as a developer is append your Jad link to the free web emulator link (more details here). That is it.

and your application is ready to work online :))

Now, send the link to everyone and give a live demo of your application and I can also see alot of potential in using this type of web emulators in Mobile application testing.

Just a note: the emulator only supports MIDP2.0 and CLDC 1.0

 Also, the camera is not supported at the moment. But, I am sure we are getting there. By the way, there might be others outthere. This is one is free though and open source for developers.

Demo the emulator: 

take the url to your jad file, simply prepend:


For example,




» Still crazy over mobile content

I spoke before about the growing popularity of mobile content authoring and adapting existing content to suit mobile browsers. What about research in this area?

The most popular way of web content adaptation is at the proxy level. This makes alot of sense , right? But, how the adaptation is done is what matters. Adaptation could be based on redundancy or it could be automated.

I will tell you a little secret, what triggered me to write this post was that i read in this book, that redudancy is something that we should plan for while preparing mobile learning content. So, I felt that I had to say nooooooooo in anyway possible.

So, following the proxy logic, imagine right before the content is sent to the mobile device, the software can choose from several pre-prepared formats of the content or it can just use a technique for automatic adaptation of the content (mostly an artificial intelligence technique: markov models, bayesian networks ..etc).

The automatic adaptation option is certainly more difficult but imagine:  no copies, no reversioning, no inconsistencies, no problems :) 

so again no redudancy and more intelligence please :) 

This type of research mostly falls under the human computer interaction research. And what I really like about using AI in adapting content for mobile use is that you can take in consideration user's preferences and history of use (more). You can really learn more about your users before your talk to them, what a nice customer service, right ;) 

I am using Bayesian networks to achieve this type of content adaptation. If it relates to your work in anyway or if you feel like discussing, drop me a comment.

For more, See:





» VOIP-mania


I don't know how  a simple search for a USB VOip phone got me into reading about SIP adapters and voip routers.

 The Usb voip phones are simply user friendly handsets that could make your chatting over voip life easier especially, for users who are not that familiar with internet chatting (like my parents). But of course, you have to be connected to the internet through a PC to use these USB phones.

But, no that is not enough? The temptation is to forget about the PC at all. We need to use these voip phones regardless of the existence of PCs. So, you will find WLans phones and you can just download the voip software or client you need to chat. Also, there are  Dual Mode phones which act as both Dect cordless phones and voip phones. The dual mode phones are more independent and of course much more expensive.

I have to say, I was very tempted to buy one of these dual mode ones but the trick question is which one to buy?  They are first generation devices, too new to actually decide what is good or what is bad. We need some real user experience and feedback, please tell us if you have used them before.

It was like a closed circle search. I am back where I started as I don't know what to buy :) 

Sip appears in the picture again. Many voip software on the phones supports sip or even through the sip adapter i mentioned earlier.

  The sip adapter allows you to use either phone line or voip and according to the features, it switches automatically. I need some time for this image to sink in. But, it seems that it is going to be a great hit. Rafal , my colleague is telling me it is scam though and such device doesn't exist. So be careful before you buy it.  His point is that this sip adapter is said to be certified by skype but skype doesn't even support SIP. So, that is why he is assuming that it could be an unreal thing. SO BE CAREFUL BEFORE YOU BUY.


» A word from the wise guy

I just couldn't stop myself from sharing something I learned today. With a phone that doesn't support serial port connection or network access over bluetooh, I can still connect to the internet over bluetooth and using the Pc's internet connection YAYYYYYY

I have spoken before about how you can do this kind of connection using your default bluetooth software only in case

  • the network access is a service that supported on your device and with your adapter. ( I love using this on HP ipaq)
  • Or through serial port service in case your phone is s symbian based smartphone and you have an mRouter running on your PC that will support this virtual connection.

But, it was the first time to do this on a Nokia N70 that doesn't support serial port, network access through bluetooth and doesn't come with an mRouter by default. Here is a detailed manual with snapshots to show you how to do this for your Nokia N70 and download  an application on your phone (only symbian based) gnubox and mRouter as well.  

 It will take a bit of time to set it up but it is really worth it. Especially if you are a developer and you don't want to spend money on using GPRS to test.


You need to know that when you do this workaround connection then it won't be really used by the built in applications on your phone. Meaning that, this connection can be used by OPERA browser for example but it can't be used by the main web browser of the phone. Same for mail and messaging applications. You will find many free third party applications to download to replace these applications if you are interested. You will find links to these apps here.

Luckily, Midlets can detect this type of connection and use it as if it is a normal GPRS connection but of course faster :)

» Watch your back


When working with J2ME, you better watch your back. You will have to design the navigation in your J2ME application according to how many views you need to  show your user. You could have a limited number of views for example to add/Edit/View or you might have unlimited views that will depend on the content coming from the server side.

With the limited views, the navigation is of course easier but in both cases, you need to watch your back. You need to keep a trail of all the views viewed by your user to allow him to go back. I noticed that a common problem among J2ME developers is that they lose track of their past views. The navigation is limited on a mobile device and we wouldn't want to make it even more complicated for our users.

One technique I have discovered is that you can keep a "Displayable" object of the view you currently have in your application in a Vector. It is like keeping a snapshot of the screen that can be displayed later on.

The objects you have saved will be easily displayed using:

Display.setCurrent( DisplaybleObject ) ;

The vector you have created should be accessible globaly in  your main display midlet. Also, keep an index of where you are at a certain point and by decreasing or increasing this index, you can access the correct view requested by the user as he moves forward or backward. This is very helpful especially when you have an undefined and unlimited number of views in your application.  

Also, remember that you put a maximum to this vector, you have to be careful of how much memory you are using at runtime. Also, keep the limitations of the device you are using in mind.

» You test, you Learn

 Following up on my previous post about deploying your application over the air, I was assuming the use of User-agent header as my only source of recognizing the device and its capabilities.

Yet, 2 days ago, I had a device that only identified itself through the header "x-device-user-agent"

This small difference really affected the system 's behavior because a device with full capability was assigned a very limited profile because I couldn't find good information about it.


So, the user-agent looked like this { Mozilla/5.0 (X11; U; Linux i686; en-US; rv: Gecko/20060909 Firefox/ MG (Novarra-Vision/6.1) }

and the x-device-user-agent looks like this{ SEC-SGHE900/1.0 NetFront/3.2 Profile/MIDP-2.0 Configuration/CLDC-1.1 }

 of course, I had to improve my code and I had to contact the user to reinstall.

I believe it could be a provider specific thing because this is the first time I see this header even with the same device.

Do you have any idea what affect those headers??? PLease et us know. 

» Don't believe it, till you test it

 Wow, I haven't written a post for ages. I just needed to write this down somewhere to make sure I give this message for all the developers who maintain different versions of their midelts for different phones: Don't trust j2me over windows mobile.

Always test the new features because the jvm on PPC might not have the decency to tell you what is wrong, it will crash or adopt the silence approach.

I would say so far, treat your pocket pc version as if it is MIDP2.0 and CLDC1.0, just  to be on the safe side.

It doesn't matter if the device supports MIDP2.0 or JTWI, it still won't work on your fancy windows based PDA.

Let me be more specific in documenting my experience. The following capabilities won't work even on a JTWI PPC:

  • PIM functionalities of integrating with calendars
  • using the phone camera to take photos and videos 

Currently, I disabled these functionalities for the users who own a PPC device. The deal is that mobile windows now will come with a built-in "INTENT MIDLET MANAGER" to allow you to run java apps. But, it is only basic stuff and it doesn't work as good and as reliable as it does on phones. I don't know if it is the midlet manager fault (doubt it) or the operating system fault (more likely). I didn't try to download a different midlet manager since I don't expect my users to do this on their devices.  

If I find more incompatiable functionalities, I will document them here.

To be continued .... ;) 

» PTT - A Spill over from the previous post

Talking of PTT, I'll start as a pessimist . PTT although is like a voice IM is not as appealing as IM. The reason being, intuitively , one way communication with that delay in response works well in text mode, but in speech, you are not used to the delay in response ( thanks telephones which have gotten you programmed for no delay). Thus the walkie-talkie feature is not so much of a catch , unless you are bill conscious and PTT is offered for a lower price than normal voice calls. Furthermore sms has grown on users so much in these many years that PTT kind of simple application has got no attention. I strongly feel PTT will never be used in a big! way by normal users.

But, then all's not lost. I see potential and specialised applications for PTT taking off in a big way. Pust-to-Broadcast a voice message to a group of users as mentioned in the previous post is an innovative application. PTT in a carphone where a push directly connects you to your Car mechanic is an interesting auto-mobile application, PTT can be used in many military and medical emergency situations. And , what if call centres supported PTT , they would make sense on the bills, especially when you have to wait for long for the customer service agent gets back to you for each of your query. I don't know how much of this makes sense to you guys????

» IMS scenario

Just found this podcast from Ericsson education and it says in very simple words what is IMS (IP Multimedia subsystem). Something we should see working very soon.

What I like about the IMS scenario is the Push to talk use. I just never understand why features so simple are never used. Push to talk is a simple walkie talkie but when you speak about it, it is as if you are speaking in advanced chemistry or something. Anyway, if u like IMS, listen to this. If you don't like IMS, why don't you start liking it ;)

Just click on Podcasts below