Turns out it doesn’t, but that’s easily remedied with this Termsrv.dll patcher.

Note that (I think) you have to manually open port 3389 in the firewall, assuming it’s on, for this to work.

I managed to pair the headset but it sounded like crap. Turns out you need to jack up the minimum bitpool setting for audio devices in OS X because the default negotiated with many headsets is wrong or bogus or something:

I tried using the Bluetooth Explorer as detailed here, but I couldn’t get the settings to persist – if I closed and reopened the Special Options window, the settings were back to 0.

With more Googling I found out how to set it using the defaults command in a Terminal:

$ defaults write com.apple.BluetoothAudioAgent “Apple Bitpool Min (editable)” 50

Both ’40′ and ’50′ work much better than nothing. It seems like ’40′ actually sounds better than ’50′. Not sure why that would be though maybe forcing the minimum value too high causes the audio codec to forcibly clip in order to conserve buffer space on the device or something? Yes I’m waving my hands wildly…

Start->Run->secpol.msc

Find “Network security: Minimum session security for NTLM SSP based (including secure RPC) clients”, untick all the boxes.

Find “Network security: LAN Manager authentication level”, change to “Send LM & NTLM — use NTLMv2 session security if negotiated”.

I fucking hate CIFS/SMB/Samba/whatever-the-fuck. And I hate computer security. What a bunch of bullshit.

1. Automatically guided media purchases
   The basic idea would be to have a service which automatically spends money out of a dedicated/auto-debited account with a preset limit. The funds are spent on a selection of movies, music, TV shows etc. You could control the ratio of how you spend your dollars and/or how you want your content composed; e.g. $20/mth on stuff my friends liked, $20/mth on what Pitchfork thinks is hot shit, and $20/mth on Amazon recommendations. This makes the most sense to me when combined with actual ownership a la iTunes, but I suppose the selection is really the point so it could be just as useful as a set of filters on top of say Rhapsody. Actually, I think you could do this just for Kindle books, like guided Kindle purchases. I wonder if Amazon has thought of this already.
   
2. Automatic lyric quoting via Last.fm/Scrobbling + some lyrics service
   I don’t know if there are any good free lyrics services out there; I assume there’s some decent freetard one. So the idea here would be to just have a Facebook app that automatically updates your status with quotes from songs you’re listening to in realtime. Basically automating what I do with my Facebook status by hand. Yes, I am that lazy.

I find it intriguing – and somewhat unsettling, since perhaps it indicates a personal rather than a communal shortcoming – that when I try to categorize a few big trends, I can’t definitively state or really even recall where some of them showed up first:

• Web Access – did more people first start browsing regularly at home, or in their offices?
• E-mail – Ok I’m fairly certain that this was much more prevalent in the workplace before making it into the home.
• Mobile messaging and Push E-mail – Were BlackBerry devices already prevalent in the enterprise before the SMS explosion? I would have guessed the former, but according to Wikipedia the average mobile user was sending 35 messages a month in 2000, while the first BlackBerry smartphone was released only in 2002 (though their original device was released in 1997. The NPT lawsuit debacle originated in 2000, which gels with my reollection, in that I’d heard of BlackBerry before graduating college, so I believe the pre-smartphone devices were already quite popular. Wikipedia does confirm that BlackBerry was the first major push e-mail service in the U.S. and it seems quite reasonable to use the BlackBerry as the barometer for the advent of this era in the corporate sector.
• Online video – I’m very uninformed about the use of digital video in business; I feel like it’s increasingly commonplace but I have no idea how mainstream it is – and I’m guessing that my work environment is quite atypical and therefore not a useful indicator. Nonetheless, I’m reasonably certain that movie piracy was the leading-edge of the trend, and that by and large, unidirectional streaming video is still largely a consumer-driven thing. Video conferencing meanwhile seems to be more evenly distributed, though it feels like it’s still a year or two away from being truly widespread.

Ok, so the reason I’m rambling about this is because it occurs to me that even assuming that the centralized SaaS-form of VDI makes any kind of sense as a way of doing computing, it’s still a toss-up as to whether it takes off at home or in the workplace first. Will VDI become mainstream before this Midori Cloud Computing platform from Microsoft shows up on a Dell machine?

I suppose an alternative outcome might be the emergence of a stable partition between consumer and business computing models, reflecting a fundamentally distinct ordering of priorities and acceptable tradeoffs between user experience, security, reliability and flexibility in these two domains – but I wouldn’t bet on it.

http://www.pitchforkmedia.com/rss/best_new_music

When I get bored or motivated enough I’m planning on writing a little RSS transformer that transforms this feed into one that sources Mininova. Actually I think that’s exactly what Yahoo! Pipes is supposed to be for…

Update: Holy fuck, that was easy. I’m impressed. This link works in uTorrent.

The Sprint 8130 is really sweet: GPS, EVDO, Bluetooth DUN, camera with video, external SD card slot, and standard 1/8″ headphone jack.

Anyway, the first thing I do with a BlackBerry (I can say that now that I’m on my third one) is to unlock the vendor themes via the standard “delete Vendor.xml and edit BlackBerry.alx” trick described on the BlackBerry forums. This gets me access to the sweet Vodafone theme that everybody likes.

Unfortunately, the unlock trick also resets BES activation status. I have already set/reset my status three times in the last month: once when I first got the 8830, next after I did the unlock trick for the 8830, and then again when I got the 8130. I was too ashamed to file yet another ticket when the new 8130 OS finally became available (access to the OS install package is a prerequisite for performing the unlock).

Fortunately, I figured out a way to install the Vodafone theme OTA without requiring an OS update or re-install of any kind:

1. Install the new OS package.
2. Navigate to %PROGRAMFILES%\Common Files\Research In Motion\Shared\Loader Files\<release>\Java.
3. Find the theme you want. They’re named according to type and vendor ID. The Vodafone vendor ID is 120, and the theme I wanted is named net_rim_theme_120_today_240x260_b.cod.
4. Copy the .cod file to a web server, and follow these instructions to create a .jad file to wrap the .cod for OTA installation.
5. Add the .htaccess file, then point your BlackBerry browser at the .jad URL.

So, without further adieu, here’s a link to install the Vodafone theme for the 8130 Pearl OTA:

[WARNING: Don't try this on anything other than an 8130; the .jad is not well formed and you might screw up your phone. I will not be held responsible for any damage you might do to your phone or anything else that happens in your life after you install this thing.]

BlackBerry 8130 (Pearl) Vodafone Today

Unfortunately, aterm has a couple of limitations. One is that it doesn’t handle XK_ISO_Left_Tab, which is incredibly annoying since it means you can’t back up in e.g. a bash or VIM completion list. I think this has something to do with changes in when and how XK_ISO_Left_Tab gets sent versus a simple shifted-Tab, and I think the whole thing is as usual the fault of motherfucking i18n which I hate. I sent a patch to fix this problem to the aterm maintainer but I never got a response. Perhaps I should try pushing it through Ubuntu instead. Anyway, I digress.

Another limitation, which is the subject of this post, is that aterm’s transparency is of the cheap hack form where it just reads the window background and blits it to the terminal window. After giving up on gnome-terminal for like the twentieth time this morning, I went back and hacked in slightly better transparency support using the _NET_WM_WINDOW_OPACITY hint. Support for this hint in the EWMH spec was proposed like three years ago, and for some reason it’s still not there, but it works in the window managers I care about (compiz, xfwm4, and metacity w/ xcompmgr).

So, my new hacked aterm can read a new commandline argument of the form -tp %, where % is the transparency value (greater than 0 and less than 100). It can also read the same value from a new resource, aterm*transparency. It takes this value, converts it into a fractional multiple of the opaque value (0xffffffff), and then sets the _NET_WM_WINDOW_OPACITY property accordingly (I just copied the code from xcompmgr).

I’m uploading an AMD64 binary that runs on Ubuntu Gutsy. If you’re running 32-bit, go fuck yourself. No seriously, perhaps if I get more motivated I will build a 32-bit version, and upload packages as well. In the meantime, I’ll upload the patch too – it’s quite simple and it’s pretty easy to see how you could do this for any program. I’m also including a gratuitous screenshot that probably contains confidential information that I shouldn’t be posting on the intarweb.

The main remaining limitation is that this makes the entire aterm window, frame, text, and all, transparent according to the specified percentage. The right thing to do is to use an ARGB visual when creating the background window, so that the text and frame remain opaque. This is what Keith Packard’s hacked transparent xterm does. I may give that a shot if I get bored/annoyed by this in the future.

• aterm (AMD64, Ubuntu Gutsy)
• Patch to handle aterm*transparent/-tp argument and set _NET_WM_WINDOW_OPACITY
  
• Screenshot with some transparent aterm windows

My ‘test’ was to run a loopback tbench run with three clients. Make sure your VM isn’t doing anything intensive, then issue:

$ tbench_srv &
$ tbench 3


This will run a tbench test with three clients connected to the server, for 10 minutes with 2 minutes of warmup time. The comparison between VMI and non-VMI kernels on my machine was quite favorable: I got almost a 2X speedup when running with the VMI-enabled kernel! The numbers were 55MB/s before, and 95MB/s after.

I suspect this is some kind of pathological worst-case for the non-VMI kernel, since most benchmarks aren’t nearly this favorable. But nonetheless, I was happy with the result and so I didn’t question it too much.

More about dbench and tbench, from the manpage:

Netbench is a terrible benchmark, but it’s an “industry standard” and it’s what is used in the press to rate windows fileservers like Samba and WindowsNT. Given the requirements of running netbench (60 and 150 Windows PCs all on switched fast ethernet and a really grunty server, and some way to nurse all those machines along so they will run a very fussy benchmark suite without crashing), these programs were written to open up netbench to the masses.

Both dbench and tbench read a load description file called client.txt that was derived from a network sniffer dump of a real netbench run. client.txt is about 4MB and describes the 90 thousand operations that a netbench client does in a typical netbench run. They parse client.txt and use it to produce the same load without having to buy a huge lab.

dbench produces only the filesystem load. It does all the same IO calls that the smbd server in Samba would produce when confronted with a netbench run. It does no networking calls.

tbench produces only the TCP and process load. It does the same socket calls that smbd would do under a netbench load. It does no filesystem calls. The idea behind tbench is to eliminate smbd from the netbench test, as though the smbd code could be made infinitely fast.