Times have been wonderful in learning about the industry, both media/music industry and IT industry.

As such, we have moved on. And who knows, maybe we will restart this project when opportunities permit.

Thanks to all who have been supporting Project MuSMo’s concept, values and effort.

Using ifenslave, you can bond two ethernets together. When two network devices are bonded, both of them will take on the master’s MAC address (or you can predefine it in /etc/network/interfaces). Now, this is not interesting when your server reboots for some reason. The secondary NIC will not be able to be bonded again because it won’t be called eth1 when rebooted but rather aliased with a different name due to how the udev rules are generated.

In Debian Lenny, /etc/udev/rules.d/75-persistent-net-generator.rules will generate the necessary ruleset, depending on your network hardware configuration, to /etc/udev/rules.d/70-persistent-net.rules. By default, the following exists in 70-persistent-net.rules…

SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:19:db:f7:00:b7", ATTR{type}=="1", KERNEL=="eth*", NAME="eth0"
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:19:db:f6:7a:d5", ATTR{type}=="1", KERNEL=="eth*", NAME="eth1"

75-persistent-net-generator.rules will generate the above rule to add a network device based on its MAC address. This is definitely not interesting for us.

We can start writing udev rules after finding out the info we need using udevinfo -a -p /sys/class/net/[device name]. Instead of using MAC address to determine our NIC alias, we’ll use the driver, vendor and device id to be 100% sure. With those info, we can create the following udev rules…

SUBSYSTEM=="pci", ACTION=="add", DRIVERS=="r8169", ATTR{vendor}=="0x10ec", ATTR{device}="0x8168", NAME=="eth0"
SUBSYSTEM=="pci", ACTION=="add", DRIVERS=="r8169", ATTR{vendor}=="0x10ec", ATTR{device}="0x8167", NAME=="eth1"

We can be pretty certain that both these NICs will be bonded regardless of the MAC changes.

Am sure we can also change the way 75-persistent-net-generator.rules generates the ruleset. But it’s pretty pointless to do it when you already know which devices you have.

This method would be a good direct solution.

If you’re still using TWiki, best to migrate to Foswiki ASAP since all core developers have moved, and the unfix security exploits is not interesting in TWiki. :/

So, I’ve made a simple tweak that allows me to search all the existing topics in the web as I type. Result? I’m able to search for a parent topic by its name. Not exactly perfect, but it fits well so far.

There are 3 parts to achieve this.

1. Create SEARCH term to list all the topics in the web

%SEARCH{ ".*" type="regex" web="%WEB%" scope="topic" nonoise="on" nototal="on" multiple="off" separator=", " format="'$topic'" }%

2. Integrate the above with JQueryPlugin
We’ll be using jquery.autocomplete for the completion of the input field.

<script>
  $(function(){
    $("#topicparent").autocomplete(
      [ %SEARCH{ ".*" type="regex" web="%WEB%" scope="topic" nonoise="on" nototal="on" multiple="off" separator=", " format="'$topic'" }% ]
    );
  });
</script>

3. Modify edit.natedit.tmpl
We need to modify /<path_to_foswiki>/templates/edit.natedit.tmpl to include all the above. So it should look like this…

  <div class="twikiFormStep foswikiFormStep twikiFormLast">
    <h3>%MAKETEXT{"Parent topic"}%:</h3>
%JQSCRIPT{"jquery.autocomplete"}%
%JQSTYLE{"jquery.autocomplete"}%
<script>
    $(function(){
        $("#topicparent").autocomplete(
            [ %SEARCH{ ".*" type="regex" web="%WEB%" scope="topic" nonoise="on" nototal="on" multiple="off" separator=", " format="'$topic'" }% ]
        );
    });
</script>
    <input class="twikiInputField" id="topicparent" type="text" name="topicparent" size="40" value="%TOPICPARENT%" />
  </div>

View this document for more information!

Congrats to the release manager, developers, testers and translators in making this a success! It’s definitely going in a better direction that TWiki, IMHO. Quite a significant amount of the core is refactored, new features are added (especially a more stable TinyMCE), and overall a thriving community that looks at users rather than its commercial implications are just some advantages that Foswiki have been working on.

Now I just need to find time to complete the migration from TWiki to Foswiki by following this guide.

With a combine of 10 years of insights & experience in both digital music and the ICT industry, our consideration of the market, changes and viability related to this project continue to show positive prospects. Successful digital entertainment company acquisitions, changing of digital media laws, wider user adoption on Internet-enabled mobile devices, and the ever expanding media content around the globe are indications of the purpose and prospective success of our Project MuSMo. Additionally, with the continue growth from telecommunication industry, ISPs and mobile device manufacturers, we can be sure that user media entertaintment within the cloud computing has arrived. The Project MuSMo was conceptualise in 2006.

With the industry, Musmo Media aims to conquer the digital world market from sound experience, insights and working culture. At Musmo Media, we ensure that each team members are not only competent and responsible, but most importantly willing and able to learn and change to meet the demands of our customers.

For Project MuSMo, we are adopting Python-based Zope 3 application framework to develop a flexible, robust and secure platform. With Python, known as one of the easiest programming language to adopt and a wide range of usable libraries, we can be sure that rapid prototyping is made simple in this ever-changing market.

This Request for Collaboration hopes to engage companies and individuals in co-sourcing this aspiration. You will find adequate information in regards to our project culture, terms and conditions, technical requirements, and team member expectations to which we expect your duly response.

Feel free to contact KwangErn Liew for further enquiries about the project.

Free. Open. Future? by Mark Surman

It’s interesting to get a point of view from Mozilla on the future of open source, its values and principles on Study, Copy, Modify, Share. What really got me thinking is, how applicable is this in Asia? It’s a norm to study, copy, and modify. However, on sharing a piece of code, it’s not quite in the agenda when it’s commercially lucrative. Same goes for any software companies that rely on building applications as the revenue resource.

As Mark Surman pointed out, what people must understand is, SCMS drives innovations, standards and freedom. Though not all encompassing for the perfect FOSS as Mozilla like to portray, Google can be a good example in that they don’t share most of its platform’s recipe, but yet they drive for innovation, review and setting standards.

Debian by Bdale Garbee

This talk got me interested because am a strict Debian user, except that all my kernels are compiled from source. Bdale Garbee raised some very important insights on the birth of FOSS and Debian’s perception about it. Debian is all about freedom and very user-centric as per their Social Contract. Debian actually led the open source definition before anyone did.

Over a decade now, Debian is still growing strong with/without its derivatives. It’s pretty amazing.

What’s notable from this talk is, the value of values. Bdale Garbee highlighted that having strict values provide an anchor in rough times. From clear values, vision is focused and understood to create a strategy that benefits the organisation through its objectives. What’s equally important is how we define values for the common good.

OSI: Recent Activities and Future Directions by Michael Tiemann

Many OSI members introduced their initiatives and OSI’s role in the global efforts in promoting FOSS.

Open Source Observatory & Repository is a good source to understand the widespread use of FOSS in Europe. Besides efforts made by the Japanese government, OSI is pushing FOSS ideologies in India and Africa. What’s not interesting is to know that OSI doesn’t have a presence in China.

Let’s hope there are more initiatives in non-Western countries for FOSS.

OpenAMQ by Pieter Hintjens

OpenAMQ is a middleware for asynchronous application messaging. Pieter Hintjens introduced Zyre, which makes implementing AMQ easy via RestMS based on HTTP REST. Unfortunately they don’t have any examples or good materials to engage developers with. Nevertheless, being able to process 100-500 messages per second is pretty good. 0MQ‘s ability to do over 5 million messages per second is even more impressive! But then again, average web apps don’t need 0MQ. RestMS has lots of potential, will be keeping an eye on it.

FLOSSMetrics: providing data about FLOSS development by Jesus M. Gonzalez Barahona

FLOSSMetrics is a database of the development of FLOSS projects. I have actually thought this talk would provide a summary of different results on FLOSS projects, however, it didn’t. But anyone is able to download the results from Melquiades. The libresoft-tools used to harvest such data can be found at Morfeo.

Scala – A Scalable Language by Martin Odersky

Totally mind-blowing, since I understood almost nothing. Coming from a non-programming background, it was extremely difficult to chew on what Martin Odersky presented. Nevertheless, I left with 3 things, 1) scalability demands extensibility, 2) Twitter moved to Scala and 3) Lift is a Scala framework.

On web development, there have been scalability wars in terms of what is best for scalability. And what do you know? It doesn’t matter, because every language and framework can be extended. So what really matters for a project to be scalable? The most basic requirement, competent human resource through time.

Grid Computing with Debian, Globus and ARC by Mattias Ellert

Grid systems are always interesting for cloud computing to shine. Mattias Ellert introduced NorduGrid and Grid Packaging Toolkits as important tools to enable grid computing. What’s interesting to know is that NorduGrid comes with data storage management modules rather than just pure data processing.

OWASP Testing Guide v3 and Secure Software Development by Matteo Meucci

Understanding OWASP is must for anyone serious about (developing) secure web applications. They have created 3 guidelines for ensuring secure web applications throughout the software development cycle. Nothing new I guess. But the following guidelines are richly informative!

1. Development Guide – important during the initial define and design stage
   
2. Code Review Guide – essential when developing/programming the application
   
3. Testing Guide – has tips and tricks to ensure the deployed application is maintained properly

Those interested, most books are free to download at lulu.com besides what’s available at OWASP wiki.

A notable example that Matteo Meucci mentioned is Cross-Site Request Forgery using img HTML element, which am sure is affected by many webapps these days.

Knowing the risks involved in your particular environment is essential in creating a user friendly service. Where risk equals to the likelihood of the event multiplied by the potential impact it can effect.

FreeIPA – Identity Management by Simo Sorce

Sponsored by RedHat, FreeIPA is an identity management mashup of LDAP, Kerberos, some additional modules, and custom Python scripts. Along with the German who sat next to me, we both found it way too complex and is adding security risks to a system that could be simpler.

I don’t think this system is a one-size-fits-all as Simo Sorce presented. Maybe it wasn’t his intention nor the intention of FreeIPA. But a simpler alternative to managing system administrators is some customised scripts on top of existing Linux tools, eg sudo and keychain.

Then again, if you’re a non-IT manager/director, FreeIPA could be the best fit with some questionable risks.

GSoC: A Behind the Scenes Look at Large Scale Community Management by Leslie Hawthorn

The finale of the event. Google just had to introduce Melange to the attendees. It’s basically an application that has been the core for organising and managing Google Summer of Code. It’s open source and released for the benefit of everyone else to organise programming events and yet keep track on what happened. Pretty nifty.

And who said Google is evil? They have been contributing back to the community…

Conclusion

Am glad the FOSDEM video are available online. Some events I didn’t attend and would be interesting to ‘attend’ them.

Also, I got to check out SUSE Studio, which is a great frontend for creating VM templates. Too bad it’s not distributed for other distros. Either way, it’s still possible to create VM templates manually but slightly more painful though can be rewarding.

Let’s hope next year would be a better year!

1. B.R.I.C.: There is a world of innovation beyond the usual suspects. BRIC, an acronym for Brazil, Russia, India and China will step up its drive in innovating brand new digital music services. Unlike most Western & developed nations, BRIC services are not sheltered under copyrights/IPR. In many cases, the BRIC nations have understood and embraced the “pirate culture”. This might surprise a few, but in countries traditionally known for Copycats/Clones, their services often include a few more tricks (to combat the supposed figure of 98% pirate nations). Is this the coming out year for BRIC services?
2. Business models: No one is totally absolved from the current financial tsunami, and with many Sugar Daddies (e.g. Vcs, PEs, Angels, IPOs etc) now out of the frameworks, it is time to go back to the basics – Building a sustainable profitable service business. This might sound cliché, but there are plenty Web 2.0 services with no visible or possible monetization model. Scale & amass users first, think about monetization later will not hold in today’s climate.
3. Innovation & Copyrights: These 2 are intimately related and will increasingly act as a double edged sword. This is also related to point 1., and why we see greater innovation to occur in the B.R.I.C., innovation protected by a loose IPR system and under the gazing eyes of executives in large copyright owners in the Western developed nations (thanks to the language barrier).
4. Experience: As a trend, we will all be continuing to push the boundary of digital experience. Learnings and exporting service experiences from the traditional brick & mortar sector helped established the digital service platform in the last 10 years. The time is ripe now to tackle concerns in the digital realm.

Critics of the music industries have accused many record labels and publishers of recognizing too late the value of digital.  The problem, however, is not only finding the right way to step into the digital arena, but in defining what music is nowadays.

1. Framework. Always work with an application framework. Choose the most suitable framework for your project. In other words, contextualise on how flexible, extensible and adaptable your project will be and how it relates to the framework you are working with. Do not neglect to take into account the original purpose of the framework, its security risks, documentation and resource availability.

2. Discussion. Development is not an individual effort, it is a team effort. And thus it requires on-going discussion of the project tasks on its requirements and specifications to ensure highest quality of clarity. If the team does not understand the task fully before acting, it will only lead to unnecessary reiterations.

3. Documentation. Without clear and detail documentations, it would be extremely difficult and time consuming to i) keep track on the project’s status, capabilities, components, features, functions, etc., ii) generate mature discussions on tasks, and iii) pin-point any potential and/or existing issues.

Our previous post on Outsourcing Software Development is intricately related to this post.