Tuesday, July 10, 2007

Portrait of a Taxi Driver

Meet Mr B T Ndlovu, who drives the taxi with registration number VRC 586 GP. He decided to overtake me on the left side, across the yellow line, and then cut in front of me. When I saw how close he was driving I hooted, but he was so busy showing me his middle finger that he drove into the front of my car before I could brake, scraping the front left of my car.
Notice his white paint on my tyre, and the dent caused by his bumper.
His vehicle has minor damage, which probably won't be fixed, even though the owner of the vehicle, who wouldn't give me his name, wants me to pay for the damage. I explained to him that overtaking on the left and then cutting in front of me was not my fault, and he'll have to pay for my repairs as well.
There was plenty of space for the taxi to overtake on the left, even though it is illegal, because I was allowing space for people wanting to turn left onto the highway. I was going straight, and so was he, by the looks of things.
All his passengers got out and took another vehicle, so he made a U-turn across 4 lanes into oncoming traffic to collect another load. The police case number from the Linden police station is AR 169/07/2007, and now I have to try to track down the owner and claim damages. What a mission! I contacted OUTsurance, who have been most helpful so far.
Update: the contact number supplied by the driver is incorrect. I wonder whether this person actually exists, or whether the vehicle is correctly registered? Time will tell.

ATM Crime in SA


Footage of ATM being blow up, vandalised, and demolished by a front-end loader. Also footage of people having their cards stolen or taken at gunpoint. Brought to you by The Times and FNB. Eish!

Harry Potter in South Africa

I love to visit the week of Madam and Eve page.

Monday, July 09, 2007

Sunday, July 08, 2007

The Great Defrag Shootout XVIII: Outlook Accelerator 2.0

Time for a slight diversion from complete hard drive defragmentation: Outlook Accelerator only defragments the data files for Microsoft Outlook. I originally thought it optimised the contents of the Outlook PST file, but I was mistaken.
Because the software is an add-in to Outlook, it "knows" where the data files are. Outlook data files are not easy to find, and if you get a lot of messages the file grows regularly, so it is easy for it to become fragmented. Outlook Accelerator claims to defragment these files on a regular basis. There are various options available to determine how often they should be defragmented.
It's not clear to me whether DefragMentor Software is still operational, as the web site does not appear to have been updated since June 2004. Their e-commerce facility works, but my attempts to contact technical support have failed. I paid the $19.95 required to test the product, but when I encountered a problem and contacted them I received no reply.
The problem I encountered was that the program doesn't appear to do anything, and does not report when the defrag is complete, or how far it has progressed. I also found that if I quit Outlook and then tried to re-load it, nothing would happen. I had to terminate the "OUTLOOK.EXE" process in Task Manager before I could start Outlook again. I have been unable to arrange a refund and have uninstalled the product. I think it's overpriced at $19.95, even if it worked correctly. This was definitely a bad buy.

The Great Defrag Shootout: Part I | II | III | IV | V | VI | VII | VIII | IX | X | XI | XII | XIII | XIV | XV | XVI | XVII | XVIII | XIX | XX | XXI | XXII | XXIII | XXIV | XXV | XXVI | XXVII | XXVIII | XXIX| winner | all | why

Thursday, July 05, 2007

eNaTIS response to hackers is really sad

Yesterday the eNaTIS web site was hacked twice. This morning a defiant message has been posted on the eNaTIS web site. Unfortunately the message is not in the least bit reassuring, and is clearly on the defensive. It reads:
eNaTIS "Hackers" on wrong track
Some media hype has suggested that the eNaTIS system was hacked recently[most of the reports specified the web site]. This was apparently due to someone leaving a comment on a page of a section of the eNaTIS public web site (this site)[So that's how they did it]. The suggestion that eNaTIS was hacked is actually laughable. The eNaTIS public web site is in no way connected at all to the eNaTIS system. [Not yet, but on-line transactions will be possible "later this year"] This choice was a deliberate design choice.
The eNaTIS system and database is still secure and cannot be accessed via this web site.
The truth is that the eNaTIS web site is running on a public hosting area on a public hosting service. The hosting service is not inside the eNaTIS data centre at all. There is also no connection of any kind between this web site and the eNaTIS system.[Not yet] The Department of Transport deliberately decided to host the web site on a completely different server than the eNaTIS system servers to ensure that any hacking attempts would be fruitless.
Any attempt to hack this web site (www.enatis.com) is totally fruitless in respect of the eNaTIS system. The eNaTIS system can only be accessed by work stations that are authorised to access the system and all communication with the eNaTIS system is encrypted. In addition, a pre-defined user name and password is needed to connect to the eNaTIS system. [I wonder how often these passwords are changed?] An eNaTIS user will only be given access to the system after signing a confidentiality agreement regulating the security of passwords. The South African public can rest assured that the eNaTIS system is not open to the public and hackers of the web site will not get one millimeter closer to the eNaTIS database by doing this.
Now for the hard truth about eNaTIS: the web site was hacked twice because the people responsible for the web site are incompetent.
It's reassuring that access to the eNaTIS database cannot be gained via the web site yet, but the facility for on-line transactions has been promised "later this year" according to the same web site.
Hopefully this security wake-up call will be correctly interpreted as such, but judging by the tone of the statement posted they have completely missed the point: the security of the web site is as important as the security of the rest of the system. Judging by the reports in Beeld, the rest of the system is not secure either.

Wednesday, July 04, 2007

eNaTIS website hacked twice

At 3.30pm today the following story appeared on the News24 web site:
Cape Town - A part of the eNatis website appears to have been hacked, raising fresh fears over the security of the national system.
News24 received several tip-offs from the public on Wednesday that a section of the eNatis.com website had been breached by a hacker. When the "how do I" link on the eNatis.com website is clicked, the message "Sorry bro..anda terlambat. Patch By Tao. OK?!!." appears on a blank screen.
The words "!- Hacked by Tao -!" appear on the top of the screen.
No other parts of the website appear to have been affected by the hacker.
By 11pm the site had been hacked a second time. This time the front page was completely vandalised (as shown above).
Its ironic that the original page (see Google cached page) proudly states "New look for eNaTIS Website" and goes on to explain how they are using Joomla, an open source content management system. Obviously there are security holes on their server, which the first polite hacker pointed out.
I guess they had all gone home by 4pm and decided to fix it in the morning, leaving the system vulnerable to a more obvious hack. They clearly don't have a clue.

Sunday, July 01, 2007

The Great Defrag Shootout XVII: Raxco PerfectDisk 8.0.64

When I first started using Diskeeper Lite several years a go I wanted something that would work a bit better than Microsoft's built-in Windows Disk Defragmenter (WDD) and figured that if Microsoft had licensed their technology from Diskeeper (previous versions in DOS were from Norton Speed Disk) I might get an improved version from Diskeeper themselves. It was only after my drive got to less than 20% free disk space that I discovered the first major problem with Diskeeper, and eventually tried PerfectDisk.
It worked fine right out of the box, and I bought a copy of PD7, which I later upgraded to PD8. I can honestly recommend PerfectDisk to anyone who needs a simple defrag utility that just does its job as advertised without too much fuss. That's not to say it is "perfect": it isn't. It has some niggly quirks that I really wish the guys at Raxco would fix. But where it matters most, PD8 gets the job done, and it does it well, both on servers and workstations.
It has been difficult writing this review, because I can't give it a "first look", since I've been using the product for ages. I have put off the review several times, even though it passes almost every test I can throw at it. CNet gave it full marks too.
The first good point is that it installs without fuss, and the screen display is clear and useful. I quickly switched off the extra help on the right, because it explained concepts I already knew, but unlike most help screens, this one was actually quite useful, especially for first time users.
The second good point is that when you click "Analyse" it will examine your hard drive, building up a display of the files, and then calculate the best way of improving the file layout to speed up your drive. Sometimes this may mean a boot time defrag, other times it may recommend a free space defrag, depending on the status of your files. One of the PDF documents on the Raxco web site makes a big fuss about PD's free space defragmentation, and I must say I have found it helpful, especially when I am working with large files.
The third good point is that PD has a command line option, so you can create your own batch files for special defrag needs. If you don't like this idea you can also schedule defragmentation to happen during quiet times. On one of the servers the defrag runs at 1am, when no one is using the system. I scheduled a "Free Space" defrag and a "Smart Placement" defrag on alternate days, to ensure the best drive performance for this server.
The command line option doesn't allow you to change the "CPU Priority" so you can't have it set to "Normal" or "Idle" by default, and then change to "Turbo" when you are running the command line. I wrote to Raxco's technical support about this, but got no response.* I have yet to get any kind of technical response from them about anything. Writing to the sales people did get a reply, however. In the meantime I figured out that I could create small registry files with the required settings, and make the changes to the registry from the batch file.
One of the nice things about this program is that PD will monitor the disk usage and scale itself back if the drive is being used by another process or user. You have to enable this feature and it does mean that the defrag takes longer. I found that I could run the defrag and do less disk intensive tasks, like blogging or writing program code while the defrag is in progress. I don't do this normally, because I tend to do defragging when I don't need to use the PC. The defrag doesn't interfere with downloads in progress or eMule activity.
The only time I got PD to interfere with anything was when I set up the "screen saver" defrag option, and then tried to run a normal defrag. The normal defrag kept getting cancelled by the "screen saver" option, which is a bit weird. You have to suspend the "screen saver" schedule, run the manual defrag, and then enable it again. What a mission.
On the subject of annoyances, there are a few more: the screen display doesn't update correctly in the area where the defrag has yet to be completed. The display shows the file number instead of the file name on NTFS volumes, and PD insists on placing part of a large file in front of a reserved space, and the rest of it afterwards, ensuring that the file is fragmented. See top image and then the one directly above. If you believe all the Diskeeper marketing hype, then this is perfectly OK, but I just find it a tad annoying.
These screen shots display the way PD arranges the files using "Smart Placement", where it puts the boot time files at the start of the drive, then the really old files (i.e. program files), followed by files that haven't changed in the last 60 days, and then the most recent files. It's probably not the most sophisticated file placement algorithm, but it works pretty effectively anyway.
I don't know what kind of magic PD does with directories, but I noticed an immediate improvement in the time to took to list a long directory after running PD, even though I had already used WDD and JkDefrag on the drive after reinstalling Windows from scratch. Whether it reorganises the contents of the directories or just places them in a better position on the drive, I don't know. What I can say is that the "everything faster" slogan for PerfectDisk is more than just marketing.
I have complained in other posts about PD being "slow", and by that I mean that it is pretty thorough at analysing and reorganising the drive, and my impatience runs faster than the defrag. But defragmentation is a disk-intensive task by definition, and the more thorough the defrag, the more data has to be moved. Vopt8 is pretty snappy, but it doesn't try to move lots of data around; this means that the drive is never completely reorganised either.
Here is an example of the "Analyse" results of my hard drive. Notice how there is a listing of the metadata size and fragmentation, the MFT size, and the free space size. PD wasn't able to fix the problem I had with a large, fragmented MFT, and it was only the unique freature of Paragon Total Defrag that sorted it out. I must admit that the problem was unusual, and that PD has improved the placement and fixed the fragmentation of the MFT on other drives I have tested.
I have tested PD on nearly full drives, as well as encrypted drives (TrueCrypt). It processes large, fragmented, compressed files without choking, and it can fix up my music files even when free space is low. From a technical point I can't fault either the way the defragging is done, or how the files are placed. It doesn't fall apart like DK does when free space is low, and it generally gets things sorted out in a single pass. Judging by the comments on the CNet Downloads site, my experience of PerfectDisk is pretty typical.
The only feature that is missing (apart from the annoyances mentioned) is the ability to identify which blocks belong to a particular file. You can defrag an individual file, but you have to guess which file is fragmeneted, which is weird. If PD added the file lookup features of O&O Defrag, it would be a killer app.
It's a pity the biggest competition to PD is a company that over-hypes and over-charges for its products. I got an email from a "Corporate Account Manager" at Diskeeper who claimed that it "has the 97.5% of the market in US" and they have sold 25 million licenses. He was unable to state the source of these numbers, so I assume it's just the usual marketing BS. I guess they need something to cover up their technical limitations, particularly the need for at least 20% free disk space. They certainly had me fooled for a while, and the user reviews at CNet are most revealing.
My conclusion: if you are willing to pay $39.99 for a defrag program, PerfectDisk 8 is the most complete product you can find.

*Update: someone at Raxco read this comment and was horrified. They are busy revamping their entire support system, and seem determined not to let anything "slip through the cracks" in future. I will post a separate blog article on how to turn "Turbo" mode on or off, following the advice I received.
Update 17 July 2007: Some readers have complained about PD8 being a CPU hog. I have not experienced this often, and generally the PD services use very little resources. The only time PD uses a lot of CPU time is at the closing stage of the "Analyse" process, which is to be expected. Another occasion was during a "single file defrag" of a large compressed file with many segments, during which time PD8 took over all the processing of the machine and showed no progress bar or anything else to indicate it was busy, even though it was.