The Science Blog outlines a method for teaching the computer various languages without any human involvement, i.e. in a data-driven manner.

[computers can learn languages] “…autonomously and without previous information infer the underlying rules of grammar. The rules can then be used to generate new and meaningful sentences. The method also works for such data as sheet music or protein sequences.”

Having quotes the above, I believe there is nothing truly novel to find in the study. The blog item does not have any mentioning of Markov models. That, in my blunt opinion, is proof that the author fails to grasp the real merits of the method. It also appears as if the author has limited knowledge in the field which is technically discussed. The ability to look at sequences of words up to a certain depth (as much as brute-force permits) presently produces nice textures in graphics and get good flow of coherent text (a paper-generating tool from MIT comes to mind).

The method can indeed perform more admirably than by making use of probabilistic models solely. It can extract grammatical content, but so can logical inference and theorem prover programs like Vampire (Voronkov and his students from Manchester University). As a matter fact, a few years ago, as part of a course on languages and semantics, we wrote programs in ML that translated English sentences into first-order logic. This technology has been out in the wild for many, many years.

Sounds like a royal phone

OR a while I have been reading about a big announcement, which is bound to be delivered by Apple’s founder and CEO, Steve Job. It now seems as if the iPod will be combined with a mobile phone, but let us wait and see.

Related local items:

• All iPod Owners Can Get Compensated
• Computers Become Hosts

Bill Gates arrested in his younger days (photo in public domain)

ICROSOFT have recently raved about their Get the facts campaign [rel="nofollow"ed]. This vigorous campaign aims to convince the public that Windows Web servers outperform Linux equivalents. The study upon which this is based is Microsoft-funded, who often wave big bucks at the face of academic research groups, as well as employ them. Microsoft-funded researchers have been caught red-handed in the past. They published fraudulent or questionable results when surveying security among Windows, Mac and Linux. Initially, little did they bother to mention who funded the research, but eventually, all conclusions were binned in a sector-wide embargo due to these biassed reports.

Sadly, for any impartial review, there must be funds allocated, which are often provided in the form of greedy commercial backings. Likewise, when some results are available, money can be used to run a propaganda. The extent of such propaganda is related to the size of the bank account — something that Linux, for example, is adverse to.

Quite recently, Open Source bodies refused to fall for industrial benchmark temptation and be victims to Microsoft’s marketing tricks that are constantly being pulled. The latest such manipulative situation is a scenario involving an offer from Microsoft to fund 50% of a research comparing enterprise Linux performance to that of Windows.

Microsoft’s open-source point man Martin Taylor has expressed interest to
jointly develop a study on deploying Windows versus Linux with the Open
Source Development Lab (OSDL).

…

“As far as working with Microsoft on a study, Microsoft could probably find
one negative line on Linux in a 100-page research report that it would
spend $10 million marketing while ignoring the other 99 pages,” he (OSDL
CEO) said in his e-mail to ZDNet Asia.

Quanta Plus – a Web development editor (click to enlarge)

wo of the most common arguments/wars in UseNet, as well as various forums, are over Linux distributions and favourite text editors, which are very fundamental tools, particularly in development.

There is no ideal text editor. Different editors suit different purposes and a different editor should be used depending the user’s level of skills. Below, for example, are various editors that I use for numerous distinct yet related tasks:

• NEdit – for editing C and tables (can make rectangular selections)
• KWrite – C++ due to colours, smart indentation, etc.
• KEdit – includes a well-integrated spellchecker
• Notepad – very light
• Wordpad- rich formatting
• BBedit (Mac) – much of the above, yet commercial (i.e. expensive)
• MATLAB editor – colour syntax highlighting, debugging capabilities (e.g. breakpoints) are integrated into the editors, consistency and cohesiveness with the rest of MATLAB
• Quanta – Web development in KDE – widgets for markup, code generation for tables, and more (see screenshot at the top)
• cPanel File Manager – editing on server-side using the browser

The bottom line is that there is no best editor (likewise there is no best Linux distribution). Perhaps this explains why there are endless arguments that end up nowhere. A developer would need to use different editors at different times. I can think of 5 distinct types of text editors that I use at the moment: MATLAB, C/C++, HTML and simple (plain) text. even E-mail and newsgroups clients can be considered editors just being themselves, not to mention textareas and forms in the Web browser.

cedemic institutes seek to provide a large number of workstations to their students. The more workstations, the more of computer-based education fits into the school’s curriculum. An observation can be made which says that schools buy the wrong computers. Licences for the software they run, despite the bulk discounts, can cost as much as the computer itself, if not exceed that cost. When accommodating a computer cluster and paying over $1000 per Windows-based machine, the number of available computers will inevitable drop. As Linspire.com shows, each student in Indiana will soon have his/her own Linux box at school. This is particularly important where kids do not have a computer at home. Will more schools follow suit? This certainly happens in Manchester as well, albeit rather slowly.

RedHat Linux

As hardware costs drop, it is gets harder to defend the purchasing of commercial software. The price of development does not not necessarily rise because of offshoring, but the availability of richer Open Source software goes up owing to the Internet, killing commercial equivalents in the process. This, in fact is one of the reasons for the success of Windows 95. Back in the days, plenty of software was bundled while getting new applications remained a time-consuming task in the absence of rapid access to the Net.

Three major risks are involved with the ownership of a handheld device. Below is a brief overview on fragility points to watch out for.

1. Vandalism

From Insecure Magazine comes a comprehensive document on Pocket PC security (PDF) which states:

These devices (PDA‘s) are easy to smuggle into a business and can be used to propagate an attack against network devices. Don’t make the mistake of assuming is a PDA is a simple data keeper. As the cliche’ goes… it is how you use it that matters.

2. Viruses

Palm viruses were created as “proof of concept”, but haven’t been found “in the wild” frequently, if ever. The Treos, however, might make the exception. Either way, AV software for the Palm seems unnecessary and you are advised to spend your money where it makes a greater difference and does not cripple your CPU. Data gets backed up during the frequent synchronisations in any case.

3. Privacy Invasion

Lastly, as more handheld devices incorporate Wi-Fi, it is worth mentioning how penetrable data packets actually are. To convince yourself that Wi-Fi is not secure at present, simply follow the links below:

• WEPCrack
• How To Crack WEP
• Cracking WPA

Cited by: PalmAddict

A ZDNet item that got Slashdotted moments ago reminds me why I back-up my data so frequently and frantically.

…A woman placed her laptop on top of her car while she got in. She forgot about the laptop, which slid off the back of her car, and she then reversed straight over it and reported hearing a ‘crunch’…

My seriousness when it comes to backup peaked a couple of months ago. As part of my ‘paranoia approach’ to backup, I have bi-weekly procedure which includes mirroring my entire hard-drive (~150,000 files), doing data integrity tests and a shallow sanity check using ‘diff‘ to ensure that no data was explicitly erased by accident.

Related items:

• Copy is Power – duplicating data excessively
• Cron-based Backup – backing up with pre-defined cycles
• One-Click Backups – automating the backup process
• Backup Fallacy – a common mistake
• Palm SD Card Backup – backing up PDA‘s using a tiny memory card