Bend in the Weather

MrPointy, This blog post is just for you! 


Your wicked Java way,
Ensures dodgy apps are hear to stay.
We all know Java sucks,
Yet you insist we try our luck.

Process freezes and applications hung,
Dont worry, Java has just begun!
We know you're a Java nut,
But don't think we'll fall in the same rut.

Write once, run anywhere,
Sucky applications we just can't bare.
Eclipse and NetBeans; your preferred IDE,
Give me a simple text editor; and forget IcedTea!

Java certainly has it's place,
Use it in Windows with it's poison lace.
System programming requires C,
Leave Java out of it, let us be!

Show me one decent Java app,
That memory it doesn't sap.
Just one app that will not crash,
That doesn't make my system thrash.

Python, C, and Perl is what to use,
It's bound to avoid the blues,
Use them next time you need a script,
Not that silly Java dialect!

There is no more to state,
Java just isn't that great!
Avoid coding in Java at all cost,
So your reputation is not lost!

Many claim Java is here to stay,
Though we know it's had it's day.
It's time to leave the plight,
And come to see the light!


NB: For those confused, I work with MrPointy who is an avid Java fan and coder.  We all take pleasure in poking him in relation to his one-eyed views on Java.
       If you are wondering on his choice of alias, read up on Buffy the Vampire Slayer, and her favourite wooden stake.

It appears VMware has known about a critical bug in their desktop virtualisation software for some time, but have yet to release an update.

This includes all their desktop products; but not their server products (ESX, Server).

Core Security has released Proof of Concept code after they claim to have been trying to get VMWare to acknowledge the problem for around 4 months.  Their press release gives more background information.

VMware has now at least acknowledged the problem. 

To date, the fix is disable shared folders where guests are not trusted.

To be honest, I'd be more concerned if it effected the Server and ESX products (likely to have many more untrusted guests), though having said that, you can expect malware/trojan/virus writers will add this to their bag of tricks to try.

I would recommend disabling shared folders and relying on network file transfers in the interim (ie: sftp, scp, ftp to the host).

Strangely enough, this isn't some quirky artist title.

<rant>
I am indeed being flooded by postcards from Telstra. 

What makes matters worse is that they are for BigPond's NextG product.

Let's make it clear:

1. I don't use them for my Internet connectivity.
2. I don't use them for my mobile phones.
3. I don't eat green eggs and ham.
4. I only have 2 land-lines (that I have to -- to get ADSL services).

Riddle me this... why am I subjected to their gorilla marketing campaigns?

What's worse, this form of 'marketing' appears like a legit postcard that one of your friends would send.  (Though there are tell-tale signs for those who look closely enough.

I suspect this campaign is hitting many people and will arrive in batches.  I haven't seen others comment on it, so I suspect I'm in the first 'batch'.  (Lucky me!)

For those wanting to know what dodgy marketing from Telstra looks like... here's the front/back of each of the postcards.  Feel free to click on them for an enlarged view.

I recently blogged that an inexpensive tablet PC could provide the magic silver bullet when it comes to a notebook suitable for children in primary and secondary education.

Blow me over, and tickle me pink! 

Seems that a UK firm Elonex have come up with such a beast.  Called 'ONE' the tablet PC features some stunning ideas:

• About half the price of the Eee PC.  (shipping for £99 whilst the EeePC is approximately £200).  Expect it to be around $200-$250 by the time it hits Aussie shores.
• Ships with a 7" screen.
• Runs Linux. 
• The keyboard can detach completely, and the 'screen' can be used as a tablet instead as a traditional notebook.
• The system has 1GB of storage internally, and the option of using a 'digital wrist band' that can store upto 8GB of data.  The wristband can then be synced with any computer via a USB cable.
  
• A bluetooth enabled version of the notebook is also available.  It also comes with 2GB of storage and ships for £20 more.
  

Details beyond this are hard to come by, with ZDNet stating that any hard specs are not being released until the 28th February presentation at The Education Show in Birmingham, England.  Several websites are slowly picking up on the story; including The Times Online and  Engadget,

It will be an interesting wait (the end of next week) to see what this little beauty packs hardware wise and the choice of Linux distribution it ships.

In any event, it's bound to throw further cats amongst pigeons as hardware vendors attempt to capture this 'cheap educational computing' market.  Lets hope the hype being generated in the UK at the moment is worth it!

Last minute update:  It appears that it is a re-badged  Fontastic A-View.  Specs are:

Hardware

CPU: Aday5F-300 MHz X86

Memory: 128/256 MB DDR2 SDRAM

4 MB Flash ROM

Storage: 1 GB/2GB NAND FLASH

Display: 800x480 7" TFT LCD

Network interface: Ethernet 10/100 Mbps

Wi-Fi 802.11b/g (optional)

Bluetooth (optional)

Webcam: 2.0M pixels (optional)

USB sockets: USB 2.0/1.1 port X 2

Audio jack: 3.5mm in/out

Battery: 3 cells, 2200mAH/cell

Weight: 950g(with keyboard)

Dimension(WxLxH): 230 mm x 146 mm x 33 mm(with keyboard)


Software

Operating System: LINOS 2.6.21

Applications: Picture Viewer(bmp, jpg, gif)

Picture Express

Music Player(mp3, wav)

Video Player(mp1,mp4,shockwave)

Web Browser

Internet Messenger(Pidgin)

WI-FI AP(optional)

Bluetooth(optional)

It also appears that the machine in question pops up under a series of different names.

Looking at it, it appears to be a 'glorified' digital photo frame.  In fact, one company even makes that comparison.


Whilst it's commendable - a 300Mhz laptop just isn't going to cut it.  It's bound to be sluggish and likely that 256MB of RAM will just run the applications if utilising modern software for an Office Suite (like Open Office).

Ideally -- my vision of a usable tablet PC that meets the mark would include:

• A screen around 9-12".
• Battery life of 4-6 hours.
• A CPU in 800Mhz - 1.2Ghz range
• 512MB of RAM
• 2-8GB of SSD or a 30GB HDD
• 10/100MB Ethernet port, and supporting 802.11b/g wireless.

Whilst this will likely break the $200 mark, I'm sure it's something that under $500 could be produced.

It seems the EeePC will also ship in Q2 8" and 9" display models (recently announced at CES2008).
The 2nd generation EeePC's could indeed fall into this "suitable" category.  If so it will firm the current dominance that Asus has in the field.

Well, that's enough ramblings by me on notebooks for education for the next few days.

Early this morning I blogged in relation to the OLPC and my own thoughts on solutions for Labors Digital Education Revolution policy.

Russell Coker picked up my thoughts and wrote an interesting article regarding Laptop vs Book Weight and made some good points.

To continue the discussion in the Blogsphere (man I hate that phrase!), I thought I would answer some of his questions, and clarify some of my initial comments.

What OS does the  ClassMate PC ship with?
The Intel Classmate PC by default ships with Microsoft Windows XP.  Close inspection of the images on their website suggests so.  The Interactive Manual also clearly demonstrates a Windows based OS.  As such, the default shipping system does not meet my own requirements -- without the removal of pre-installed operating system and software.

What age bracket does the ClassMate PC suit?
I did state in my article to mention that "The Classmate PC looks like more of a rival to the XO laptop more than anything." and whilst not clearly stating it; I see it more focused on primary school children then actual high school students.  As such it fails the Digital Education Revolution policy (Years 9-12).
As Russell correctly points out it is 'designed for primary students (ages 5-14)'.  It is also important to note that Intel have also have left the OLPC project.  They are now squarely focusing on the promoting the Classmate over the XO laptop.

What computer candidates for high school children then?
I pointed to the fact of the Everex Cloudbook and Asus EeePC as two possible candidates.  In addition it appears that others are about to hit the market in the next few months including a model from MSI as confirmed.

I do agree with Russell that whilst the EeePC is probably the front runner (with a SSD over a traditional HDD), though the screen size (7") is far too small.  I would add that the keyboard would also be annoying for any length of time.  WithUpdate: potentially 6 hours a day working on this beast, it's likely to be light enough for students to throw across a class-room in frustration.

I am also of the belief that Tablet PCs could also be reworked into a small form factor and low price (sub $500-$600AUD).   It  allows students to either elect to use a keyboard or a pen-based solution if preferred.  In addition, drawing diagrams on a tablet is much easier. TabletPC screens generally are larger as well, making it suitable for viewing for long periods.

 There is some partial Linux support for Tablet computing, but traditional Linux applications are likely to require linking to the correct libraries to provide support for native 'writing'.  It seems Toshiba are also thinking along the same lines about a Tablet PC rework.  TabletKiosk shows it can be done, and use OpenSuSE as an option on their range of Tablets.

Russell's main point that the he disagrees with my statement "A traditional notebook is far too heavy for high-school kids to lug around".   In the adventurous mode that is Russell; he went digging through his old school books, and a set of scales to prove that current high-school text books are indeed heavier.

I could have saved him the pain...  they are indeed heavier. 

However, I'll stress again that my personal belief is that a traditional notebook is far too heavy for high-school kids to lug around.

Let me clarify this point.

1. Text books (those awful paper form) are currently required.  Adding another 1-1.5kgs (weight of a notebook) to a schools bag in unnecessary.
2. Parents are already concerned with the weight of school bags already, adding extra kilograms to the bag won't go well with them and is likely to be met with resistance.
3. The Victorian government has already published details on concerns of the weight of school bags.  Some interesting statistics come out of a study undertaken by the Australian Spine journal:
• The weight of the average backpack is heavier, proportionally, than the legal load-bearing limit for adults.
• 79.1 per cent of children say their backpacks feel heavy.
• 65.7 per cent report feeling fatigued by the weight of their backpack.
• 46.1 per cent report back pain caused by their backpack.
The site  goes on to state:
• The backpack should weigh less than 10 per cent of your child’s body weight – for example, a child of 40kg should carry less than 4kg in their backpack. Ideally, the child in this example should only carry around 2–3kg of books.
4. I agree with Russell's idea of using digital text books and ensuring the children only carry the notebook.  However I do see situations where this will be hard to enforce:
   • Book publishers are unlikely to be swayed into a digital age for distribution, as much as other publishers in the entertainment industry (movies and music spring to mind)  have been dragged kicking and screaming.
   • Whilst Creative Commons licensing would be the nirvana; I could only see this working if a 'redemption' process works.   ie: Each student registers with the school the texts they are using (based on the enrollment of a course), this could be then billed to the students family and the revenue collected is given to the publisher).
   • Another process would be to have Book publishers provide their text books via USB Keys.  A student places an order in a central location, and a USB Key is shipped to them; with all their classes loaded on.  This form is however likely to require some form of DRM; which many (including myself) will be opposed to.
     
   • The most secure approach would be centralization of text-books.  A school intranet or a government based education server that requires authentication is utilized that records the student accessing which book.  A semester bill is then generated.  Internet access is almost mandatory for each child, and if not available at home the local library or school premises could be used to access the text.  This is probably my preferred method.
     
5. Using the argument, 'it was fine for me when I went through school' some 20 years ago, doesn't hold any water.   When I went to school, corporal punishment was still in place!  Things change over time, new research and practices also come to light.  Little work had been undertaken (around 1988) in relation to lifting loads of children at that time.  Much has changed.   Remember, Linux wasn't even an itch in Linus' pants 20 years ago.

I think the discussion and response from Russell is healthy.  I appreciate Russell's input and he does raise some of the questions that will need to be addressed.  Hopefully more people feel prompted to express their opinions and insights.  The FOSS community is wide and diverse; many of these hurdles can be addressed and a superior and less expensive solution is likely to be found.   Whilst traditional system vendors will only look at the opportunity to increase revenue, those in the FOSS community are likely to focus on the end-user experience.

The current focus by the Federal government also helps narrow down the desired solution for high-school students.  It is a perfect time to introduce our next generation of adults on the virtues of OSS.  These students will take that experience into the workplace of the future.   This will assist in a wider adoption of OSS solutions not only in the workplace but in homes around the nation.

Placing computers in the hands of our students and replacing legacy paper-based text books will address two areas in one swoop.  That surely is a Digital Education Revolution and one that both doctors, parents and students will be grateful to take a load off their backs!

Update: Steve Walsh correctly identifies that the ClassMate PC does indeed ship with a Linux variant. I admit, I also missed that myself! As such the ClassMate PC is also suitable for an alternative to the XO Laptop.

Recently I penned about the OLPC and Labour's Digital Education Revolution policy.

After some consideration the following day I drew the conclusion that the OLPC XO is better suited to primary school children than that of secondary children.

Jeff and Pia Waugh both contacted me with exactly the same sentiments. 

In fact, Pia informed me that work was underway in relation to trailling the OPLCs in Australian primary schools.  A website has been setup (content still coming!) and a mailing list is available.

Those interested in being involved should join the mailing list.

Those in high school some of the latest sub notebooks are more appropriate such as:

• Asus Eee PC
• Intel Classmate PC
• Everex CloudBook

With the focus from the Federal government on secondary schools for computers in Year 9-12... these little machines might be perfect for high-school kiddies.   The CloudBook or the Eee PC are probably the choice machines at present.  The Classmate PC looks like more of a rival to the XO laptop more than anything.

A traditional notebook is far too heavy for high-school kids to lug around, and a sub notebook is light (normally under 1kg) and could be even loaded up with school text books to save on back strain.

I'm hoping that the government advisors undertaking the tender process look at such items and just don't become 'deer eyed' in front of the glare of glitzy sales-pitches from traditional laptop makers.

Running a Linux based OS also makes sense for teenagers, as it severely limits the ability for the machine to become infected by worms, trojans and viruses.  These traditionally target Windows operating systems, and become docile on a Linux based machine.  Whilst I would still recommend an AV scanner on the machines, the impact and damage these nasties can take is severally limited; thus reducing the overhead in support costs keeping them clean.

Plenty of food for thought...

I found it refreshing to see at the last Federal election that the Australian Federal Labor Party had policy on providing each and every secondary student in Years 9-12 with their own computer.

The Digital Education Revolution was a major piece of government policy that saw the party win in a landslide in 2007.

The policy goes on to state:

• $1bn over 4 years to be spent.
• The National Secondary School Computer Fund will provide upto $1m for schools to provide new and upgraded ICT equipment.
• $100m to be spent on high speed FTTP provisioning to schools.
  
• Training of school teachers in the effective use of ICT to assist student learning
• Development of web-portals for parents to participate in their child's education.

This will mean a complete transformation of many schools, and bringing them in line with 21^st century teaching practices.

Whilst this all does sound appealing, in it's present form I can see Microsoft and their solution providers clammering on the software/operating system space to bleed the government dry of the funds.

I think this is a perfect location where the OLPC could see the amount invested into such a scheme.  The use of XO Laptops are perfectly suited for the classroom, and even more so in remote rural areas.

Take into consideration the requirements for most students:

• Ability to write documents  (Word processing, spreadsheets, presentations, etc.)  The traditional 'Office' application
• Ability to communication with fellow student (Collaboration works)
• Ability to send Email
• Ability to browse the Web
• Ability to use digital content (Audio/Video/Graphics)
• Ability to undertake course material in a digital age

These are areas where the XO Laptop shines.  It's not a general computer, but one specifically built for students in mind.

The $1m AUD could therefore be used to ensure ALL children in high-school (Years 7 - 12) are provided with a laptop, and thus ensure to double the reach of the initiative.

OSS software in the school would be seen as something that could greatly enhance to dollar spend and provide a rich and full learning experience for the student, assist the teacher in providing the appropriate tools.

If the average laptop is say $1200AUD, it is likely that they can replace one such laptop with 4 OLPCs.  If a traditional laptop requires say $300 worth of software (Operating System, Office, Utilities, etc), this further increases the value of the OLPC as the software required is built into the laptop, thus avoiding any additional expense.

So comparing functionality and cost -- it's likely 1 traditional laptop could be replaced with 5 OLPCs.

The OLPC is rugged and made for demanding students, add to it that it's easily repairable, ensuring ongoing costs are minimized.

I would expect that money saved with laptop rollouts could be used to further enhance the OSS stack and provide a full OSS stack in the school.

• OpenLDAP authentication servers
  
• Samba file servers
• OSS Webmail solution (something like Zimbra that allows students/parents/teachers the ability to access it outside school hours).
• Linux based firewalls protecting the school infrastructure.
• Web-servers based on LAMP.
• The use of an ECM like Alfresco for ease of content management.
• Collaboration servers such as Jabber.
  

The XO Laptop was on display and handed out to many of the conference attendees at the recent LCA08.   Many people marveled at it's simplicity and pure beauty.   (I'm sure it's something Intel and AMD will pour over for months to come!)

If people from Linux Australia are interested in putting together a submission to the Federal government on such a proposal I'm happy to donate my time.

I see investing in the future education of our children paramount to moving Australia into the digital age, and the use of OSS ensures they get the best start possible! 

I feel this is one area the OSS community can make a massive contribution to computing in Australia that is likely to touch nearly every Australia family.  Whilst deploying OSS solutions will also ensure it's familiarity is likely to rise to unseen levels in this country.  Adopting OSS principles at an early age with students is likely to ensure that the next generation of Australians are very OSS savvy and will continue the great work their previous generation have commenced.

If you're interested, please contact me.

Alfresco is one of the darlings of the OSS world.

For those unfamiliar with Alfresco they are makers of an ECMS.
Q: "What's an ECMS?" I hear you ask.
A: ECMS: Enterprise Content Management System. 

Q: How does this differ from a traditional CMS -- such as CMSimple, Drupal, Joomla!, etc?
A: Whilst traditional CMS most people know about is actually a WCMS (Web Content Management System).  An ECMS however handles more than just web-content.  It will handle all types of content.  It also has the ability to undertake items like:

• Document Management (Revision Control/Release Management/Document Review)
  
• Workflow/Business Process Management
• Record Management
  
• Fine grained Roles/Groups per department
• Desktop and Office application integration.
• Common dashboard for each department.
  

Traditionally an ECMS will replace a SMB departmental file-shares. (Sorry Tridge!) It's much easier to manage and ensures all documentation and business process knowledge is located in one central repository.  (This is a good thing!)

Anyway, I digress; enough of the lesson and on to the Open Source Barameter!

I attended LCA08 last week and had a fantastic time.

There was a range of talks from the brilliant, the ho-hum and the insane.
The lightning talk on Oracle VM fell into the last category.

The method which Oracle guys attempt to push 'their solutions' down our throat makes me sick in the stomach.
Whilst the speaker (Shane Owenby) was quite easy to the ear, the whole talk smacked of Oracle corporatization gone bad.

Whilst what Oracle does in the Linux space isn't seen as breaching GPL it does rub many the wrong way.

• 'Unbreakable Linux' is merely a respin of RHEL sources (think CentOS), the then offered with a hefty support license (though less then RedHat's own support contracts).  So basically it's nothing more than customer stealing by Oracle without the heavy lifting of rolling their own distro.  Whilst I'm not privy to the inner workings of Oracle -- RedHat corporate relations, I'm sure this must have put a strain on the relationship.
• Oracle VM - which was being discussed at LCA08 had some interesting ideas.
• It basically runs a Xen-based kernel under Unbreakable Linux in Dom0.
• Guests can either be Linux (RHEL/Unbreakable based) or Windows based running under DomU.
• A range of Oracle products are certified to run under 'Oracle VM'.  These include the likes like Oracle Database 10g, 11g, Oracle Application server, Oracle Enterprise Manager, Oracle Siebel CRM, etc.

Whilst virtualizing the application stack makes sense, and allows for rapid deployment and hot-swap, the powers to be have decided that 'lock-in' would be one of the key features of running Oracle.

One of the craziest statements I heard during the talk was that Oracle will only support their products running under a VM if it's within Oracle VM.

If find this just crazy, considering that Oracle VM is purely a respun RHEL, and running a Dom0 with the database itself within a DomU.

Just tell me the difference of Oracle VM, to a customer running RHEL with Xen that has Oracle running in DomU?

The idea that support is only offered if running 'their stack' is beyond me, and likely to turn many customers off utilising Oracle under a virtual environment.  Many have invested heavily in their infrastructure (whether that be VMWare based or Xen based under another Linux distro).    Why on earth Oracle is forcing people to go down their path just smells of corporate greed.

Personally, if it's that hard.. I think it's high time many organizations look heavily into porting their application to PostgreSQL.

Oracle and their approach to Linux smells like a corporate wolf dressed up in a cheap lamb costume.  Look closely and it just doesn't look right.

If Oracle is serious about providing Enterprise support in the Linux space, they need to get the basics right.

• Start allowing administrators to install your products using native Linux packages (*.deb and *.rpm).  The current approach is a nightmare and quite unmanageable from an Enterprise level.  You provide system installers for Windows (Making it easy to install/uninstall via Add/Remove) -- why not the same level of support under Linux?  
  Use of mounting the CD, and running an 'install shell script' is not an Enterprise solution.
• Fix up the installer.  There is no need that the installer needs to run under X.   Databases are installed on SERVERS.  Servers DO NOT run graphical desktops.   It is criminal that you force people to install X libraries (and a bunch of other junk) just to install your product.  
  Less is more, and very much so on a server.   The need and demand from your installer for X libraries opens up more potential security holes.   Ask any administrator with half a clue and he'll tell you that a native system package that can be installed by the command line would be their preferred method.

These complaints aren't new, and have been chorused by Linux administrators for years now, yet nothing is done.

If you expect these same techs who you are championing Oracle VM to support it in their organizations, then at least look at getting these two major hurdles addressed.  There is no technical reason they need to remain, besides sloppy packaging.

Addressing these concerns is likely to ensure the DBA vs Admin hostilities will at least move away from deployment issues, and can then move to items like disks filling. (our usual turf wars!)

Well I turn the big 3-5 today.   I'm starting to feel old... well not young anymore!

I've already been spoilt...  Pauline has given me my pressie and I just love it.

I have so many geeky 'toys' that I really wanted something non-geeky for my b'day this year.   Another geeky toy just means I'll be spending more time on the computer... and I do enough of that already!

We saw some artists painting several months ago and in passing I mentioned to Pauline 'I'd love to do that as a hobby!'

You know the drill, away from the computer, out in the country-side taking in some sun and fresh air.

Well it looks like it will become a reality now! Pauline bought me a French Easel Art Set.