I recently put a Laptop for sale in a local community web page. Within about an hour I received the following mail:

From: bruce walker
subject: inquiry for laptop

dear seller,
i am interested i purchasing the item advertised by you online……….i would like to know the full working condition of the item.i am alright with the price and i would like to know your payment term.so get back to me as soon as possible because i really need the item very urgently.have a nice day.
business regard.
bruce walker

Strange, huh? I had posted details about the product and myself, but this mail quotes nothing, not my name, not even the brand of the laptop. Bruce wants to buy this “item”?

Well, I decided this must be a bot of some kind, and and judging by the spelling mistakes a bad one, so I didn’t reply. I did however do a search on Bruce’s email address and some quotes from the inquiry, but google had nothing for me. Am I really the first one to report this? If you have seen a similar message, let me know in the comments.

Background: About a year or two ago I replaced all Linux at home with Mac OS X. But since a few months I have a Lenovo T60p from work, which mainly runs Windows XP. It happened that the german c’t magazine recently came with a DVD of OpenSuse 10.3, and one day away from home (and my Macs) in some lonely hotel room I got bored and installed it. Everything worked just fine, but the laptop’s graphic card did not. It is an ATI, and I slightly remembered that they have a closed source driver for Linux, that needs to be installed somehow on top of any distribution that only ships software that is open source.

So my first step was to google for this topic, and I ended up on a page that explains how to install the ATI driver on Suse Linux. I followed Stefan Dirsch’s simple steps and ended up with the ATI driver in version 8.41.7 directly from the ATI Repository. Great. Only it still didn’t work.

Turns out that in the Change Log of the ati-fglrxG01-kmp-… package it is noted that the support for FireGL cards was “temporarily removed” one previous version ago. So what next? I found another page that describes how to install a previous version of the ati driver manually. It explicitly mentioned 8.40.4 but I was not sure for which graphic card that is. When I was browsing the ati pages, I could not find the Mobility VL5250 that I actually have. I selected something similar, and ended up with the version 8.35.5 download. That however is not helpful, as it support up to Xorg 7.1, but not the 7.2 that is in OpenSuse 10.3.

When I selected one of the Radeon cards, I finally found the right download of the ATI driver in version 8.40.4, and then I could follow Vichar Bhatt’s instructions to the end. The display now shows 1680 x 1050 with 16.7 Mio colors. Great. It wasn’t excatly straight forward, but I got it working. Thanks, y’all.

I retired my XP laptop some months ago, and still consider myself in the switching-to-Mac process. Not every application is required every day, so some things / topics to investigate just pop up after some time.

Today was such a day, when I had to add a new activity to my Logitech Harmony 525 remote control. Last time I programmed it was on Windows. Turns out that the Harmony application has all the settings web based, and a Mac version is available. It knew who I was when I connected the remote, I added the new activity (playing wii, which requires to switch to another input on the receiver) and finished wondering why not every application could switch so easily.

Well, some do. I keep using the google apps (mail, reader, spreadsheets) from both Macs and PCs. Great new world. Thanks y’all.

In this post I’m documenting my findings of a windows application running in Parallels accessing a USB2 external disk drive. I compared this setup to various other alternative disk locations. I first explain a little bit of the background and the 2 test that I did in each scenario, then show the results and discuss them.

Background: Since switching to the Mac platform I am using quite a lot new applications and appreciate every single one. However, there’s one Windows app that I still can’t let go, mainly because the alternatives that I looked at did not (yet) match its features. It is MediaMonkey, which for some nostalgic reasons is still my primary music tidy-up tool, before the tracks get anywhere near iTunes. Why I do this, I will explain in a different post later. Today let’s just be clear that MediaMonkey runs in Windows only.

My two speed tests are:

1. adding a 100kB album cover art to 39 tracks (177.5 MB), measuring from the OK on the Properties window until the status bar has no more “xx tracks to be tagged”. Note that none of those tracks had an album cover art before and I took a fresh copy of the old untagged tracks for each scenario.
2. modifying the same tracks from test 1. to have updated content in the artist and genre tags.

When adding the cover art in test 1. the whole file of the songs need to be re-written, since the tag is at the beginning of the mp3 file, hence it is quite a good test for file I/O. Modifying the existing id3 tag in test 2. is much faster, since only the tag gets updated – the rest of the mp3 file are left untouched.

All results are in seconds, less is better.

I used an external USB2 disk (LACIE 160GB) that came formated with the FAT32 file system. Hence I had the option of mounting it read+write in Mac OSX without installing NTFS support in any way. If your external drive is NTFS, you either need to get the NTFS write support going, or you can only use scenarios 1 and 8. A FAT32 drive can be used in Mac OS X, and then shared via CIFS to be used inside the Parallels VM – which comes with quite a speed improvement over USB2 inside Parallels – comparing scenarios 1 and 2.

Scenarios 3, 4 and 7 wrote to the internal disk of my Mac OS X host, scenarios 5 and 6 to another Mac OS X host in my local network. This second machine was accessed over Gigabit Ethernet in scenario 5, and “802.11n only (5GHz)” about 2m (6 and a half foot) away from the Airport Extreme AP.

In all 8 tests XP is running of the same disk – I have Parallels set up to use the Boot Camp partition. You see how fast writing to this Boot Camp partition is in scenarios 3 and 7. I was surprised that scenario 8 was quicker than 7. However, note that this Boot Camp partition is terribly full with less than 10% free space, while on the USB2 disk I still have plenty of space left.

Conclusion: USB2 speed in the Parallels virtual machine is not good enough for me. Boot Camp is much quicker, but requires reboots – a No-Go for me. I will use my USB2 disk mounted in OS X and shared into my Parallels as the best compromise between speed and comfort. YMMV – comments welcome!