applied procrastination

hdiutil and OS X data recovery

2012-01-20T10:17:00.000-08:00

Your hard drive fails, so the first thing you do is grab dd and pull a byte-by-byte copy of the drive's contents:

   dd if=/dev/bad_drive of=bad_drive.img conv=noerror,sync

Problem is, if the drive was from a Mac and you want to use a recovery tool like Data Rescue or Drive Warrior, you'll probably need to attach your drive image as a loopback device, and double-clicking the file to mount it doesn't work when the filesystem is fragged.

Enter hdiutil:

   hdiutil attach -nomount bad_drive.img

Resuming truncated 'cp's

2011-12-21T14:51:00.000-08:00

I'm sure this isn't too hard to do w/ dd, but let's say you've got some slow network disk mounted under /mnt/crappy_SMB_share and were 100 gigs into a 101G file when the Windows server went down, and now you want to finish off that last gig.

  cd /mnt/crappy_SMB_share
  curl -C - -O file:///original/file/location

from here.

AirPlay: streaming from iTunes and your iPhone to a Linux box

2011-09-27T21:52:00.000-07:00

So I've got a Debian machine as a WiFi access-point / file-server, and I decided to plug some speakers in and stream music from my iPhone.

I tried Erica Sadun's AirPlayer, a Python script that, as it turns out, implements version 2 of the AirPlay protocol -- the one that supports video and images. What it DOESN'T support is sound, at least not without the appropriate crypto key.

Which led me to Shairport, vessel of the secret sound-support sauce. Shairport only implements AirPlay v1, which can't handle images of any kind. Which means that the video portion of YouTube can be streamed from my iPhone to AirPlayer, and the audio portion to Shairport, but I have to choose one. Huzaah.

This worked well enough for iTunes, but iOS, ie, my iPhone, didn't work until I disabled IPv6 on my Linux box.

Yet more Crescent City

2011-06-03T21:24:00.000-07:00

Those giant game-of-jacks pieces are called dolos.

Hello from Crescent City, CA

2011-06-02T19:53:00.001-07:00

Another in a series of brief tests of Wolfram Alpha

2011-05-04T15:27:00.000-07:00

If you've spent the last couple of years behind a few inches of lead, protecting you from the nerd-radiation of the Wolfram|alpha launch,well, you should read up on it. It claims to be, not a search engine, but something smarter, a compendium of systematic data (temperature records, physical constants, business data, you name it), and the algorithms necessary to process that info.

I've been disappointed thus far. I tried another reasonable-seeming query today: "kinetic energy of a ring with mass 1 kg and radius 1m rotating at 100 rpm"

(If you're wondering, I was hanging out by a centrifuge, which just begged the question, ya know?)

And what did Alpha have to say? "and (English word)"

It recognized the word "and" as English. That's it.

The answer didn't show up until the 8th hit in google's results, but there it was,

E = 0.5 * I * ω
I = m r²

etc etc. Granted, I still have to do the math, but at least it's something.

The lively faces of poverty

2011-01-28T09:26:00.000-08:00

At the risk of sounding like Tom Friedman, the most striking thing about Saigon was the bustling, crowded buzz of the place -- the way every square inch of public space was somehow occupied by a person, streets completely filled with Vespas and every sidewalk a public market.

The average Vietnamese lives on about $3,100 per annum, less than half what her counterpart in China makes, so we're talking about a very poor country. But that poverty doesn't manifest itself as obvious misery--not where we visited, at least. Instead, you see this:

This Hanoi street is lined by key-cutting carts, as could be found on every other street corner in Saigon. These are industrious, hard-working people, but in the US, copying keys must occupy 30 minutes out of the day of a Home Depot employee; the rest of the time, he's off stocking shelves while the machine rests in the corner. American labor is simply too expensive to waste waiting for customers; instead, each worker has a whole stock of capital goods (the key-cutting machine, a forklift, etc) at his command, keeping him occupied each and every minute.

In Saigon, you can find any food or service you need, right there on the sidewalk, because in Saigon, you can make a living cutting keys for one blocks-worth of people. It's actually kinda nice if you don't mind crowds ...

The future is already here – it's just not very evenly distributed. (An open letter to AT&T)

2010-12-30T23:50:00.000-08:00

Dear AT&T,

Attached you'll find a picture my wife and I took on a boat traveling through Vietnam's Mekong Delta. Our abilities with our new camera are admittedly lacking, so it's hard to make out the orange life vests lining the boat's ceiling and green jungle in the background, but then, this is my blog, not a court of law. The real crux of this photo is the 5 bars you see on my iPhone: I had good reception throughout the Delta -- much, much better than I get in America's second most density populated big city*.

AT&T, I've loved my iPhone these past 2 years, but long for a phone call that doesn't end mid-sentence, and will be thrilled to move to a network that can compete with the 3rd world the moment my contract expires.
-- kieran

* - These kinds of comparisons hinge on the definition of terms like "big city," of course, but scan through Wikipedia's density stats and look for a city you recognize that isn't NYC. And if you haven't tried AT&T in San Francisco, don't bother.

Konichiwa

2010-10-14T06:07:00.000-07:00

Hello from Osaka. God knows what's going on with that very uncomfortable rhino up there, but one thing I'll say for Japan: it's dense.

That shot out the window of a train is exurban, far-out Osaka. No sprawl in sight: dense, multi-story housing and ... woods.

OpenOffice makes me appreciate Firefox' search function.

2010-08-01T15:40:00.000-07:00

The search dialog that appears when you hit Command- [or Control-, depending upon OS] F is a bit of UI that's improved greatly but subtly in Firefox: instead of a separate dialog appearing, an embedded interface shows up at the bottom of the page.

This means the dialog location is consistent and predictable, but more importantly, it means that I can always search by hitting Command-F. Under OpenOffice, Command-F toggles between opening and closing the search dialog; if I was searching for something earlier, I've now CLOSED the dialog rather than opening it. Very annoying.

Quick link: shameless self-promotion edition

2010-07-22T10:18:00.000-07:00

In the interest of raising the site's Google Pagerank, and as a plug for my wife, who did the site design, here's a link to my sister-in-law's optometry site, Lake Vision Optometry, located in the verdant, exotic lands of Riverside, CA.

The Eee, Ubuntu, and Songbird

2010-07-21T16:46:00.000-07:00

After years reading about super-cheap Netbooks, I finally picked one up for a friend, though ASUS' Eee PC 1201t skates pretty close to the line separating Netbooks from Notebooks with its 1.6ghz AMD CPU, 12" LED LCD screen and 250G drive. All for $380!

Don't try to get an OS on the beast, though. That'll be tough. After determining that MeeGo doesn't support the Eee's Radeon GPU and finding EasyPeasy kinda clunky, I decided to put Ubuntu 10.04 on it. Unfortunately, Canonical has mostly dropped support for USB installers, and this little machine has no DVD drive, leaving me to pull off bits of cheap Taiwanese plastic until I hit hard-disk.

Now that Ubuntu's loaded, however, I couldn't be happier. All the hardware -- Wifi card, webcam, everything -- was supported without any tinkering, and Ubuntu looks great.

I do miss iTunes, though. Songbird (installed via Skyzim's deb) is a gorgeous replacement, so long as you don't want to listen to any w4a files from the iTunes store -- which I do. This thread explains the fix, though it's tough for the command-line-phobic. After installing the various gstreamer libraries, you'll want to open your favorite editor and add one line to /usr/bin/songbird:

export SB_GST_NO_SYSTEM=1

Quick links: succinct non-English words

2010-07-01T09:42:00.000-07:00

Care of reddit, a site devoted to non-English words that don't have a good counterpart. Shadenfreude hasn't appear yet (perhaps it's officially English now?), but drachenfutter has:

Literally translated as dragon’s food.

Drachenfutter are gifts that a husband brings to his wife after pissing her off. It is usually an attempt to avoid her wrath by giving her chocolate, flowers or other small gifts.

On leaking memory like a sieve

2010-06-11T16:47:00.000-07:00

I tend to leave browser tabs open for weeks on end, and if it's not entirely clear from the gratuitous screen-grab above, those 19 Google Chrome processes each consume anywhere from 16 to 108M of memory, for a grand total of 577M.

And here I thought Firefox was bad about memory management.

Filed under "I wish I'd thought of that"

2010-06-09T11:40:00.000-07:00

From "Comparing genomes to computer operating systems in terms of the topology and evolution of their regulatory control networks"

Faster than a speeding mushroom

2010-06-01T10:49:00.000-07:00

This growth occurred over the course of two days. Moving clockwise from the top left, pictures were taken at: 7pm on Wednesday, 8 am Thursday morning, 11pm Thursday night, and finally 7pm Friday night.

Accidental Mycology

2010-05-26T14:49:00.000-07:00

This houseplant has been looking unhealthy for a while, but I didn't know it was this bad:

I'm lame enough to post update pictures, so my non-existant readers can look forward to watching these mushrooms grow in near real-time. (PS: Bren should get credit for this one; she noticed the growth and took the picture)

Quick link: Knoppix on a USB jump-drive

2010-05-13T08:56:00.000-07:00

Everyone's favorite bootable Linux CD can be written to a USB simply by running the flash-knoppix command:

http://www.knopper.net/knoppix/knoppix62-en.html

It worked on an off-brand USB stick, but my SanDisk Cruzer refuses to play along. SanDisk implements their encryption via some weird juju that makes their USB drives very uncooperative; I wouldn't recommend them.

Banksy?

2010-05-06T16:27:00.000-07:00

Bren with a potential Banksy.

(And it seems that blogger strips EXIF tags from uploaded files. No fun!)

Python debugging quickie

2010-05-05T10:12:00.000-07:00

I've gotten into the habit of using pdb to interactively debug python code. I'll plug import pdb; pdb.set_trace() into the middle of an uncooperative function in order to examine machine state as the code runs. (Note that I don't need to full capabilities of a debugger here; I'm not stepping through the code line-by-line.)

iPython is a much better choice, as this posting lays out:

import IPython.Shell; IPython.Shell.IPShellEmbed(argv=[])()

That's all you need to drop out of your function and into an iPython shell. When you're done twisting knobs, ctrl-D will exit iPython and return to your function.

Inchoate musing: what does it mean to own a device?

2010-05-03T10:31:00.000-07:00

There's talk that the DoJ is looking into antitrust action against Apple for their recent decree that iPhone / iPad developers only use Apple-approved tools, ie, C/C++/Objective-C under Xcode.

I don't really know how antitrust law works, but it's not clear to me that Apple's move violates it. Microsoft was only prosecuted for bundling IE because they already had a monopoly on desktop OS's; it would be hard to argue that Apple has a smart-phone monopoly.

But I still think Apple's wrong, and government intervention is probably right.

In my mind, this falls under the broader problem of technology companies prescribing some narrow range of approved uses for the products we buy. A pernicious example is the license shipping with every camera that records video as MPEG or h264, which forbids making any money off any video you record with a consumer-grade camera, because that would make your work "professional." So if you've ever uploaded a video to YouTube, you probably violated your license.

I think I'd like to see something like a Gadget Owner's Bill of Rights: if I buy it, I can use it as I see fit. The vendor can eschew support for altered devices, of course, but I can't be legally prevented from modifying the device, and third parties can't be prevented from helping me. (And let's pass laws to keep the 'Net neutral, while we're at it!)

Only moderately tangential update: Reddit links to another great example of the kind of legal machinations that enable device manufacturers to claim excessive control over what you do w/ your device: a 92 page Terms of Service agreement.

More Python weirdness

2010-03-17T14:33:00.000-07:00

You can almost use Python as a functional language, but ... you run into trouble. The following code, discovered by Nithin Reddy, will raise an exception:

funs = [lambda i: i + j for j in [1,2,3,4,5]] del j funs[0](1)

The issue is that lambda, which executes in the context of the lambda call rather than the lambda creation. Using a proper def won't work either -- the following code fails in precisely the same way:funs = [] for j in [1,2,3,4,5]: def f(a): return a + j funs.append(f) del j funs[0](1)

The fix is to curry your function:

def partial(f, *args, **kwargs): def bound_f(*args2, **kwargs2): x = kwargs.copy() x.update(kwargs2) return f(*(args + args2), **x) return bound_f import operator funs = [partial(operator.add,j) for j in [1,2,3,4,5]] del j funs[0](1)

I'm not quite sure what to make of this. Is function context bound on return?

Too many options

2010-03-10T10:18:00.000-08:00

The autocomplete module that comes with jQuery UI is not the same as the jQuery Autocomplete plugin. Totally different interface. Using the documentation for one while running the other will not make for an enjoyable afternoon.

jQueryUI comes with all manner of functionality beyond autocomplete, so that'd be my recommendation.

Write once, read many

2010-03-05T14:54:00.000-08:00

Nice: the PEDANT database contains the output of many common genome annotation tools against 3,000 complete genomes, notably including the human genome. I've been looking for the results of running, say, TMHMM against every human gene.

(For more, see the paper: PEDANT covers all complete RefSeq genomes.)

Tiny, neat Django trick

2010-02-15T10:42:00.000-08:00

In SQL, you can retrieve a list of records via the "IN" syntax:

SELECT * FROM mytable WHERE id IN (1,2,3,4,5,6);

Django supports near-identical syntax:

MyTableModel.objects.filter(id__in=[1,2,3,4,5,6])

nice!

Happiness

2010-02-13T13:37:00.001-08:00

Small bit of Scientology bashing: Is it possible to be happy?

Screen recordings under Snow Leopard (OS X 10.6)

2010-02-11T16:24:00.000-08:00

For all your software demo needs: Snow Leopard will create screen recordings out of the box.

Just run Quicktime, go to File -> "New Screen Recording."

(Thanks to Nithin Reddy for the heads-up)

Animal adjectives

2010-02-11T10:08:00.000-08:00

That I found this on a site called "PubQuizHelp" seems appropriate enough. You probably know that "simian" refers to all things monkey-related--hence the adjective's popularity during the Bush administration--and you may have heard that "ursine" is the word for all things bear ... but did you know that calling someone "lapine" likens them to a rabbit?

Very useful stuff.

jQuery 1.4

2010-01-29T09:16:00.000-08:00

The jQuery Javascript library, without which cross-browser Javascript development is unthinkable, has reached version 1.4.1.

Particularly useful is the ability to create new DOM elements. (See this brief discussion, though no anchor link means you'll have to search for "Quick Element Construction.")

What those docs overlook is the .text property, which enables you to create a proper element:

$("#foo").append( jQuery("<div>", { id: "foo", text: "I'm a new DIV element", click: function() { $(this).css("backgroundColor", "red"); }}));

or even create a nested elements via .html:
$("#foo").append( jQuery("<select>", {html: "<option value='1'>One</option>"}) );

The Great Pacific Garbage Patch, local edition

2010-01-24T12:07:00.001-08:00

This week's massive storm system seems to have washed up a small taste of the Great Pacific Garbage Patch:

We're doomed.

Multi-line strings

2010-01-21T17:26:00.000-08:00

Javascript doesn't have multi-line strings, but offers a variety of hacks ranging from concatination ("line 1\n" +"def") to the CDATA hack, but as always, your milage will vary depending upon your browser.

Python has dedicated multi-line string denominators, namely the triple-quote -- """ and '''. Quite workable, but not as elegant as simply allowing newlines in a string.

C and shell have the backslash:

char * s = "line 1\n\
line 2";

Haskell extends that to the double-backslash:

s = "line 1\n\ \line 2"

PHP, in a rare moment of elegance and usability, treats a newline in a string as a newline.

$s = "line 1 line 2";

Now why is that so hard?

Update: Jake is right: C-style backslashes work in Javascript (at least on Firefox).

stackoverflow.com, in which we circumvent Django's template system

2010-01-08T10:40:00.000-08:00

The whole separate-code-from-presentation imperative that inspired Django's template system ... let's say I have my doubts. From my experience, Django's particular approach to templates diverts much too much of my attention to circumventing its limitations.

Update:

In response to Jake's observation that "variables should always be set in the view, not the template," in general, I agree.

But the specific counter-example I'll supply is the very common title example. Let's say you'd like to create a base template like so:

<head><title>{% block title %}</title></head> <body> <h1>{% block title %}</h1> ...

And then you inherit from that template:

{% block title %}User Feedback{% endblock %}

This is a very simple, commonplace scenario, and it's simply unsupported. Django will error out on the second call to {% block title %}, and I think that's dumb.

Python's Zipfile library

2010-01-04T09:33:00.000-08:00

So you want to generate a big bundle of files all at once, and send 'em in one convenient package? Python's ZipFile library to the rescue! You can create each file, write it to a zip, then send the whole mess over the wire without ever touching the disk! (StringIO is pretty essential as well.)

Ahh, but wait: anyone unzipping your files will find each file marked unreadable/unwritable/unexecutable. And the Python ZipFile docs have zilch to say about it.

This turns out to be the requisite incantation, care of Stackoverflow:

info = zipfile.ZipInfo(name) info.external_attr = 0777 << 16 zip.writestr(info, bytes)
Obvious, no? Of COURSE you need to bitshift the permissions by 16! (File perms clearly occupy the first 16b of an int, which might even be the standard on Unix, but that doesn't excuse the Python docs' utter silence on the matter.)

One last thought on Wavelet image sharpening

2009-11-19T09:14:00.000-08:00

The GIMP's plugin documentation for the Wavelet sharpen plugin is interesting:

"The wavelet decomposition of an image results in multiple images with different frequency content. When amplifying the high frequency parts the recomposed image appears to be sharper than the original one. That way the frequency which should be amplified most can also be selected and a given unsharpness in the original image can be taken into account."

That matches with Tom Crimi's description of Wavelets as vaguely similar to the uniquitous FFT.

Sharpening photos like they do in CSI

2009-11-18T09:13:00.000-08:00

There's a recurring joke on nerd hang-outs like reddit about the "enhanced photo" plot device on dramas like CSI: a grainy, low-resolution image is zoomed and sharpened to bring out impossibly small details -- finding the reflection of the killer in a distant chrome hub-cap, that sort of thing. It's obviously stupid.

But after playing around with The GIMP's "Wavelet enhance" filter, I'm really impressed -- it really seems to bring out fuzzy details:

(My best effort at cleaning the image up with contrast & brightness on the left; the wavlet-enhanced version on the right.) Clearly, it can't restore missing detail, but it seems to do a good job of improving the contrast around edges / features.

Firefox can't print, and what to do about it.

2009-11-16T15:21:00.000-08:00

Say you've got a big, information-rich, scroll-heavy page, with annotations and doodads located a long scroll off the screen. But say it works: you can scroll around and find what you need; it's a solid app.

Firefox won't be able to print it -- not usefully, at least, b/c Firefox will print only to the right edge of the visible page, thereby limiting you to the width of your physical screen. Creating a new paper-type, some monstrously wide fictional ur-sheet, won't help things either. So what's a boy to do?

Enter CutyCapt, a WebKit-based command-line tool that renders to PNG. (Or PDF, or whatever file format your heart desires.) Better still, it will execute any Javascript on the page, so pages that are essentially blank until the JS runs will render just file.

You'll need Xvfb, the X virtual framebuffer, if you want to run it on a headless box, but Xvfb is a little miracle in itself.

Microsoft Outlook 2003: The Worst Program of All Time

2009-10-20T10:29:00.000-07:00

I'm forced to use Outlook 2003 by a clueless IT department -- forced, in fact, to sit an entire separate ThinkPad next to my iMac just for email, since a Mac can't possibly sit on the trusted network. It's a security threat, after all. Unlike Windows. Which is secure.

Having used GMail for years, I'm shocked at the sheer unfriendliness of Outlook:

Searching takes minutes, and only barely works.

Perhaps due to way Exchange is configured, Outlook won't auto-complete contacts by first name, only last, and even then it only auto-completes names from my personal contact list, not names in the Exchange directory.

Threading. THREADING. I had threading in 1998 using Mutt, a command-line email client. GMail has had threading since day 1. An email and it's replies form a dialog, obviously, and Outlook's failure to group those emails into a thread makes it about as useful to me as my Nokia 3510's SMS functionality was.

On technolgical improvements in lighting

2009-07-31T09:03:00.000-07:00

Two out of three light bulbs burned out in our new lamp, so off to the store we went. We wound up with two different bulbs: an LED ($9), and a compact fluorescent ($7). This lamp has 1 incandescent bulb, 1 CFL, and 1 LED; would anyone care to guess which is which?

Wikipedia's entry on LEDs has an enlightening explanation of the temperature of such white LEDs; the phosphor that converts the LED's blue light into white light doesn't do a very good job of it, and much of the blue leaks out -- this graph captures the effect nicely.

PostgreSQL and capitalization

2009-07-24T11:22:00.000-07:00

The terminal-based PostgreSQL front-end psql, which ships standard, does not like capitalized table names.


ri=> \d KDplot_plotpoint
Did not find any relation named "KDplot_plotpoint"
ri=> \d "KDplot_plotpoint"
          Table "public.KDplot_plotpoint"
Column |   Type     | Modifier
...

Double-quotes are your friends.

On ZFS and terror

2009-07-21T10:16:00.000-07:00

I've had my main backup / nameless cruft repository on 3 x 1TB drives in a ZFS RAID-Z array for a while now, and finally found out what happens ZFS (or Linux ZFS-FUSE, at least) encounters bad blocks.

It loses it, completely.

I suspect this is largely a Linux problem: once a bad block is encountered, the Linux kernel (?) seems to go into an infinite-reread tailspin, flooding syslog with errors and generally making the SATA bus (and the machine at large) unusable. So it might not be ZFS' fault.

And a quick aside to the many posts claiming that "the OS should never see a bad block; the drive should silently remap the block to a spare and the OS will only be aware of it when you've run out of spares!": bullshit. I've had half a dozen drives turn up w/ bad blocks, and the SMART stats on each reported plenty of spares, and each drive was fine after a forced overwrite. (IE, I zero'ed out the drive.)

So I repeated my zero-out procedure on this drive, and the sequence of events gets fuzzy, but at some point I backed up /etc and at some later point I ran zpool export on the DEGRADED array. When the time came to put the newly-zero'ed drive back in the array, zpool wouldn't bring it back up: zpool import reported that "The pool can be imported despite missing or damaged devices," but zpool import tank ("tank" being the customary name for ZFS pools, I think) complained that it "cannot import 'tank': one or more devices is currently unavailable." zpool import -f tank yielded the same thing.

This when when I began to panic.

Despite having 2 fully functional drives, all my data would have been lost. But remember that backup of /etc, notably including /etc/zfs/zpool.cache? That was my salvation: restoring that brought me back to my pre-export, degraded state, which in turn let me replace the faulty drive.

My data is safe for now, but my confidence in ZFS is shaken. Why did exporting a degraded array make it un-importable? If I lose my zpool.cache, is all lost?

Developer preferences

2009-06-03T17:08:00.001-07:00

Ubuntu still ships with GHC 6.8, released 12 December 2007. Macports offers 6.10.3, which has been around for about 3 weeks.

I might be reading into this too much, but methinks there's something there.

On the prescience of Infinite Jest

2009-05-15T10:00:00.000-07:00

It's not a central plot point, but DFW's near-future Infinite Jest (1996) mentions Microsoft Pink, the operating system du jour.

Perhaps he has a fan in Redmond?

Are you sure you want to navigate away from this page?

2009-04-22T16:04:00.000-07:00

Scenario: you've been tasked with creating a simple web-based time-sheet application, so employees can log what percentage of their time is devoted to which project. Now, this hardly qualifies as "Bioinformatics," a domain claimed by both your job title and personal aspirations, but it's a small company and they pay your rent.

So you build the little beasty, and your cube-neighbor sits down to fill out a time-sheet. She spends 7 minutes selecting options and punching in percentages, and then realizes she wants to trim one field from 15% to 10%, clicks her mouse 4 pixels too far to the right putting the page, and not the text input, into focus, and hits Backspace -- equivalent to the back button.

There go her 7 tedious minutes of form-filling -- this app isn't sophisticated enough to autosave a draft in real-time, after all.

Enter the onbeforeunload function.

I've borrowed from this guy a bit, but here's the code:

var confirm_exit = true;

window.onbeforeunload = function() {
if( confirm_exit )
 return "";
return; };

$(document).ready( function() {
 $("#id_theform").submit( function() { confirm_exit = false; } );
} );

Originally, I'd planned to skip the ugly global variable and simply re-bind the onbeforeunload function, but apparently you can't rebind like that.

A Beginner's Guide to Hating iMovie '08

2009-04-10T13:07:00.000-07:00

Thank you, Apple. This is pretty much my first experience editing footage into a "movie," and with iMovie '08, you've made the process slow, painful, and constrained. What's more, I was lucky enough to get my computer weeks before the release of iMovie '09, so any of your stunning omissions fixed in the subsequent version will cost me $100. No thanks.

First off, titles. Adding chapter titles is a pretty common use for iMovie, I would think, but clips cut from one Project and pasted in another are stripped of their title. Thanks.

Oh, but wait: "chapter titles?" What's a chapter? iMovie isn't really meant to create movies for later burning onto a DVD -- who would ever want to distribute a DVD? So very 2003! The only way I've found to separate footage into chapters for iDVD is to put each chapter into its own Project. Annoying.

The perils of laziness

2009-04-09T11:07:00.000-07:00

I've begun cargo culting my way into Haskell, in the hopes that I'll eventually grasp such mathematical mysteries as anamorphisms. In the mean time, I'd like to write some code that actually works.

To this end, I've written a program to digest the C. elegans transcriptome in silico, then check the resulting fragments against data from a mass spectroscopy experiment (Multidimensional Protein Identification Technology --"MudPIT"--if you're curious). I use regular expressions to break up the protein sequences, store the fragments in a big Data.Map, and then test my MudPIT fragments against the Map.

I've greatly simplified my function to be a simple list-to-map conversion -- which would be stupid, since Data.Mat.fromList does this quite well -- but the real function involves plenty of off-topic complexity, so here's the svelt form:

  add_frags (i:l) m =
   let m' = M.insert i () m in
   add_frags l $! m'
 add_frags [] m = m

The important part is the $!, in bold. Given some function f and parameter x, f $! x is equivalent to x `seq` f x; the seq function, meanwhile, forces the evaluation of the first thunk. So f $! x applies f to x AFTER x HAS BEEN EVALUATED.

Without that forced evaluation, ghc will simply add more thunks to the stack until the stack is blown ... hours after the run has begun. You can pretty easily imagine this as a series of substitutions, which is exactly what we're talking about:

  import qualified Data.Map as M
 m = M.empty
 add_frags ["ACDEF","GHI","KLMNO","PQRST",...] m
  -> M.insert "ABCDEF" () (add_frags ["GHI","KLMNO","PQRST",...] m)
  -> M.insert "ABCDEF" () (M.insert "GHI" () (add_frags ["KLMNO","PQRST",...] m))

And off the stack grows, until pop.

RAID 5 on an Leopard Xserve

2009-03-30T11:45:00.000-07:00

This seems too trivial to merit a posting, but there are other posts that might lead one to believe that Leopard doesn't support RAID-5 on an Xserve, which it does, provided you've got a RAID card installed. When booting the installation DVD, just look for "RAID utility" under the "Utilities" menu. Just be sure to create the RAID before you pick an installation disk.

Update: Jashugan's right; I've updated appropriately. Though I still like the sound of "Xserver." One last thought: why ever does the RAID 5 volume build process take something like 2 hours / 100G? I understand the time required to rebuild parity after replacing a failed disk on a live array, but the initial build should be faster than that. Unless the OS is doing a bad block scan on the disks or somesuch, but ... why do that for a RAID and not under other circumstances?

Regarding encryption ...

2009-03-24T09:15:00.000-07:00

Not encryption in general, but one very specific instance of it: Apache SSL on my Ubuntu/8.10 box.

Ubuntu ships with two SSL modules available -- the canonical, OpenSSL-based Apache mod_ssl, and the newer mod_gnults (pronounced "noodles?"). I decided to try mod_gnutls for no good reason, and it seemed to work -- I could serve content via an encrypted connection! Except for the silent redirects. I'd point my browser to /, get redirected to my SSL login page as expected ... and then wind up on the login page over vanilla HTTP.

In situations like this, HttpFox is your friend. As are netcat and OpenSSL itself, which, if you'll pardon the tangent, has it's own netcat-like behavior indispensable for debugging encrypted network services:

   openssl s_client -connect myserver.com:443

Via openssl, here's the offending HTTP transaction:

GET /ri/account/login?next=/ri/ HTTP/1.1
Host: myserver.com

HTTP/1.1 301 Moved Permanently
Date: Tue, 24 Mar 2009 15:26:47 GMT
Server: Apache/2.2.9 (Ubuntu) DAV/2 mod_gnutls/0.5.1 PHP/5.2.6-2ubuntu4 with Suhosin-Patch mod_python/3.3.1 Python/2.5.2
Content-Type: text/html; charset=utf-8
Location: http://myserver.com/ri/account/login/?next=/ri/
Vary: Accept-Encoding
Content-Length: 0

I ask, over HTTPS, for the login page, and it redirects me back to the unencrypted version. And I never quite figured it out, either -- I just switched to mod_ssl and the problem went away.

I can't even say this is a GNUtls bug, but it's certainly some strange behavior.

python-ldap authentication and you

2009-03-23T11:56:00.000-07:00

Or me, at least.

Here's how I managed to make my python client code running on a Debian/Ubuntu box authenticate against a Windows Active Directory LDAP server:

Obtain the server's cert: openssl s_client -showcerts -connect myserver.com:636 > their_server.pem

Edit the resulting file to isolate just the cert you want. In my case, the file had two certs embedded in it, plus some identifying cruft. Preceding the first, desired cert was a block that looked like this: 0 s:/C=US/ST=MyState/L=MyCity/O=My Company Inc./OU=Information Technology/OU=For Intranet Use Only/CN=my.host.com, followed by -----BEGIN CERTIFICATE-----. Next came VeriSign's cert. Delete everything but the CERTIFICATE block for your server, including the BEGIN and END lines, and save to /etc/ssl/certs/myserver.pem -- you can save to any location, really.

And here's the code:

#!/usr/bin/env python

import ldap
#ldap.set_option(ldap.OPT_DEBUG_LEVEL,4095)
ldap.set_option(ldap.OPT_X_TLS_REQUIRE_CERT, ldap.OPT_X_TLS_NEVER)
ldap.set_option(ldap.OPT_X_TLS_CACERTFILE,"/etc/ssl/certs/myserver.pem")
l = ldap.initialize("ldaps://myserver.com:636")
l.set_option(ldap.OPT_PROTOCOL_VERSION, 3)
l.simple_bind_s("username@DOMAIN","password")

The trouble with DRM

2009-03-11T15:21:00.001-07:00

Cory Doctorow's brilliant denunciation of DRM, as delivered at a Google Author talk a couple years ago:

If you deliver to an attacker the cipher-text, the cipher, and
the key, and rely on the attacker not combining those except
under circumstances you dictate, you're living in a fool's paradise.

I was reminded of this by Roger Gregory, who directed me to a youtube video about cold-boot attacks.

Timing out an arbitrary Python function

2009-03-05T09:51:00.000-08:00

Update:

The problem with the signal.alarm approach described below is that the alarm WILL happen, thus interrupting a hung function in the failure mode we're trying to account for. But it will ALSO SIGNAL if that function returned as expected, thus throwing an exception at some random point in later code.

The solution is to disable the alarm, as described below:

Occasionally, child processes die. If you don't want your script to hang, you do something like this:


TIMEOUT = 10
process = subprocess.Popen(...)
(o, _, _) = select.select( [process.stdout], [], [], TIMEOUT )
if o:
 return process.stdout.readline()
else:
 os.kill(process.pid, signal.SIGKILL)
 os.waitpid(-1, os.WNOHANG)

This works perfectly well (and mimics the C version pretty closely), but what if the process is already wrapped in a python module and you don't have direct access to the process?

This guy has a good solution:

  def raiseTimeout(a,b):
   raise Exception()
signal.signal(signal.SIGALRM, raiseTimeout) # bind a signal handler
signal.alarm(1)                             # raise the alarm sig in 1 s
some_hang_prone_function()                  # this is the problem function
signal.alarm(0)                             # disable the previously set alarm

My original version used a lambda in place of a proper function, but it turns out that lambda a,b: raise Exception() isn't proper syntax.

Bay Area Kvetch: Pubilc Transit on the Peninsula

2009-02-28T13:59:00.000-08:00

This really does seem to be the only public transit route from my job to SFO:

Apple's USB dongle

2009-02-11T07:39:00.000-08:00

The Ethernet port on my 2006 Macbook died, so I went to Fry's and picked up Apple's USB Ethernet adapter -- you know, the one w/ the imperceptibly fine print on the back explaining "only for use in Macbook Air." That one.

And it worked at first, until I unplugged it for the night and plugged it back in. No amount of rebooting helped, but finally I worked out the procedure:

Under System Preferences -> Network, select the USB Ethernet device and hit the (-) button at the bottom, removing it from the list. This apparently unloads the driver from the kernel.

From the command line, run sudo kextload /System/Library/Extensions/IONetworkingFamily.kext/Contents/PlugIns/AppleUSBEthernet.kext

Back under System Preferences -> Network, press the (+) button and choose "USB Ethernet," then "Apply"

That should do it.

How Not to Create a Dialog Box

2009-01-29T16:05:00.000-08:00

From Aquamacs on Mac OS X:

Or perhaps this was plain OS X Emacs, Aquamacs certainly has the same dialog. "Save this but no more," eh?

My computer is taunting me. Loudly.

2009-01-28T15:17:00.000-08:00

I'm leaving Xoma, and arrived at work yesterday morning to a surprise: my computer is taunting me! And loudly, too.

There followed much perusal of ps lists and greping through the entire filesystem, but I couldn't find the source. Nithin finally suggested I find any file altered in the past few days, which led me to ... Nithin's .bash_history. Which led me to this, from "site-packages/date.py":

#!/usr/bin/env python
import os
import sys
import random as ran

taunts = [
  'Wha' + 't ar' + 'e y' + 'ou do' + 'ing?',
  'go' + 'od ridd' + 'ance',
  'say' + 'anara! suc' + 'ker', 'y' + 'ou a' + 're a dir' + 'ty trai' + 'tor',
  'I tho' + 'ught w' + 'e w' + 'ere frie' + 'nds',
  'a ro' + 'ck wa' + 's mi' + 'spla' + 'ced, ' + 'Kie' + 'ran ha' + 's vani' + 'shed',
  'won' + 'der b' + 'oi h' + 'as go' + 'ne t' + 'o t' + 'he da' + 'rk si' + 'de']

if __name__ == '__main__':
t_index = ran.randint(0, len(taunts)-1)
os.system('os' + 'ascr' + 'ipt -e "set Vol' + 'ume 6"')
os.system(('sa' + 'y "%s"') % taunts[t_index])

sys.exit()

This process was run every 15 min by a launchd plist called "com.apple.spaces.plist," which sounds pretty innocuous. Brilliant!

Xoma's lucky that Nithin Reddy and Jake Rothenbuhler possess more loyalty and general fortitude than I do.

The dread -lcrt0.o error on Mac OS X

2008-12-14T10:46:00.000-08:00

I periodically run into this problem compiling software on the mac:


 ld_classic: can't locate file for: -lcrt0.o
 collect2: ld returned 1 exit status
 make: *** [muscle] Error 1

In this case, I was compiling MUSCLE, a multiple alignment program.

The solution is to track the -static flag down in your Makefile, and remove it. GCC/OSX does NOT like -static.

On the color of bits

2008-12-13T14:13:00.000-08:00

Programmers and mathematicians tend to have a hard time with Intellectual Property law. and Copyright in particular. Fundamentally, when speaking of information, lawyers and math types aren't speaking the same language. This essay manages to capture the difference better than anything I've read: What Colour are your bits?

A Minor Quibble with Javascript, or: Why Automatic Type Coercion is Evil

2008-11-25T14:44:00.000-08:00

Earlier today, I posted a question to mozilla.dev.tech.js-engine, to whit:

isNaN( [1] ) = false. Why?

And got the answer I feared: it's in the spec!

As best I can follow, the spec lays out this sequence for determining whether [1] is, in fact, a number:

isNaN( [1] ) -> ToNumber( [1] ) -> ToPrimitive( [1] ) -> [1].toString() -> "1"

And "1" is coerced into a number.

Automatic coercion, of which this is a rather contorted example, produces counter-intuitive results every time. Language designers should never, ever make it core to their language. That is all.

Berserk, ep 25

2008-10-28T09:34:00.000-07:00

i finished the anime treatment of the Berserk manga last night. overall, it's good: well-drawn and voiced, and full of those morally ambiguities that make for interesting drama.

but the last episode felt more than anything like a bored
father who's gotten sick of telling his kids a bedtime story and
decides to finish it, and finish it fast:

"so then the broken Prince takes was reunited with his loyal soldiers
and fled to a field. and then, uhh, they were all eaten by monsters.
yeah. the end. now GO TO SLEEP!"

the Berserk Manga wikipedia entry suggests a much lengthier, more baroque plot arc centering on an ongoing battle between the survivors of the last anime episode (ie, Guts) and demons, but that doesn't change the fact that supernatural occurrences were pretty subtle until the last couple of episodes, making the series finale pretty jarring.

so i'll echo my brother's advice: if you decide to rent Berserk, you'll sleep better if you stop at episode 19.

The limits of Open Source

2008-10-15T10:58:00.000-07:00

One of the arguments made against Open Source in the bad old days, back before IBM started making Real Money on Linux and Microsoft's attitude became schizophrenic, was that Open Source developers don't come up w/ truly new ideas, but simply re-implement old ones.

Hogwash, but I can think of two technologies we should have thought up first:

Dtrace
ZFS

Just saying.

Reading the iPhone

2008-09-24T14:29:00.000-07:00

The iPhone has a wonderfully clear, hi-rez screen, and seems a perfectly usable platform for ebooks -- not as good as the Kindle, but better than the Palm V on which I once read "Don Quixote."

Sadly, Plucker hasn't been ported to the iPhone yet, putting it somewhere behind the 1999 Palm V in functionality.

There are a plethora of eBook readers available for download from the App Store. Most of them cost money; Stanza is one of the exceptions. It's got some decent books available for download, and the library interface is only moderately wonky, but... but. The book reader is TERRIBLE. Rendering takes minutes--not downloading, but rendering--and tilting the screen triggers a complete re-render, meaning that any accidental tilt interrupts your reading for 2-3min. Completely unusable.

The best book reader I've found is the iPhone's built-in PDF viewer, accessed via Air Sharing, an app that was free a week ago but now costs $6.99. Air Sharing allows you to connect to a network share on your iPhone via WebDAV, which initially sounds only mildly interesting, until you realize that files within that share can be opened for viewing on the iPhone.

PDFs rendered this way open fast, rotate even faster, and the viewer state--ie, which page you're on--is saved between Air Share sessions. It would be perfect were it not for the sometimes screen-unfriendly nature of PDFs.

Text and HTML files can also be viewed, and the Text viewer retains state, but render times are prohibitive.

Sync'ing Google Calendar w/ the iPhone

2008-09-17T10:07:00.001-07:00

Do we have a word for the relationship milestone marked by my sync'ing my girlfriend's Google Calendar with my iPhone?

It's pretty easy to do, at least on a Mac.

A loathsome and indefensible lie

2008-09-09T11:49:00.000-07:00

From Livejournal, by way of Pharyngula:

The theory of childhood, also known as child origin, is a damnable, loathsome and indefensible lie. How can any thinking person suppose all humans used to be babies once? ...
The development of children has been well-researched in our six-month study following a sample of one thousand children and adults of various ages. We have conclusively proven that while there are minor changes in features like height and body fat ... incontravertibly still every creature in the study that started out as a child had only slightly more adult features at the end of the observation period than at its beginning.

Two patched protein subtypes and a conserved domain of group I proteins that regulates turnover.

2008-09-08T15:57:00.000-07:00

I'd almost forgotten about this:

Two patched protein subtypes and a conserved domain of group I proteins that regulates turnover.

Kawamura S, Hervold K, Ramirez-Weber FA, Kornberg TB.

Biochemistry & Biophysics, University of California, San Francisco, CA 94158.

Patched (Ptc) is a twelve-cross membrane protein that binds the secreted Hedgehog protein. Its regulation of the Hedgehog signaling pathway is critical to normal development and to a number of human diseases. This report analyzes features of sequence similarity and divergence in the Ptc protein family and identifies two subtypes distinguished by novel conserved domains. We used these results to propose a rational basis for classification. We show that one of the conserved sequence regions in the C-terminal domain of Ptc 1 is responsible, at least in part, for rapid turnover. This sequence is absent from the stable Ptc 2 protein.

Levenshtein distance

2008-09-03T10:13:00.000-07:00

A Levenshtein distance, or "edit distance," is a measure of the similarity of two given strings as embodied in the number of changes--insertions, deletions, or substitutions--required to transform one string into the other.

This is an incredibly useful tool for grouping and ordering strings, and in particular, non-language strings -- we already have a convention for ordering words in English, but if you're staring at 200 protein sequences, alphabetic order doesn't do anything for you.

There's a fast C implementation of the Levenshtein algorithm and though it doesn't seem to have a proper project homepage, it can be found in the Pootle & Translate Toolkit.

I needed to add a wildcard parameter to the Levenshtein algorithm so that the distance between, eg, ABC and AXC is 0, given that 'X' is the specified wildcard character. (Normally, the Levenshtein distance between "ABC" and "AXC" is 1, of course: it takes one substitution to turn one string into the other.)

So I did.

Before I attempted to modify this C version, though, I put together a simple, unoptimized Python version, and the difference in performance was shocking. On my test data set, the C code took 0.3s to run, and the Python version took ... over 3 minutes. It was over 1000x slower.

After adding the couple of if statements and loops necessary for the wildcard code, the C code slowed to 1.2s.

MochiKit, IE6, and DOM manipulation

2008-08-19T15:56:00.000-07:00

I'd recommend MochiKit to anyone who has to write Javascript code -- it lives up to its motto and "makes Javascript suck less." I'd recommend it twice as forcefully to Python coders, as features like the iteration tools are explicitly inspired by Python; and thrice as forcefully to any functional programming -minded Python coders, as it provides handy functions like map and zip, plus short-cuts for partial application and the like.

Bob Ippolito is a smart man.

MochiKit also provides some convenient DOM manipulation tools, such that swapping in a whole table takes only 3 or 4 lines of code:


 swapDOM( $('mytable'),
          TABLE({'id':'mytable'},
                TBODY(null, [TR(null, TD(null, "a cell"))] ) );

(assuming you already had a DOM element w/ id='mytable')

A quick caveat, though: Internet Explorer 6 is picky about tables, and requires that a table created via the DOM have a TBODY or nothing will be displayed.

Python class inconsistencies

2008-07-17T13:01:00.000-07:00

Update:

Masklinn's correction is obviously correct -- this isn't primarily a class v. instance variable problem, it's the difference between the operators I'm using on the instance variables.

So, an assignment (k.t = ...) points the attribute at a wholly new object, while accessing one of the attributes' methods actually alters the attributes' state.

(Which is where the confusion over mutability comes into play, but it was still confusion on my part.)

My original posting:

This makes a modicum of sense if you have a basic grasp of the distinction between mutables and immutables in Python, but it still seems like a mess.


>>> class K(object):
>>>    t = (1,2,3)
>>>    l = [1,2,3]
>>>
>>> k = K()
>>> j = K()
>>>
>>> k.t = ('a','b','c')
>>> j.l.append( 10 )
>>>
>>> j.t
(1, 2, 3)
>>> k.l
[1, 2, 3, 10]

Do you see it?

Class K defines two attributes, a tuple t and a list l.

If you instantiate K twice (k and j), then change the t attribute of one and the l attribute of the other, the change to l will be shared between instances, while the change to t will only effect the instance.

Whether an attribute belongs to the class or the instance depends on whether it's type is mutable or not!

There's a fix in one case: you can use __init__ to make mutable data instance-specific:


>>> class K(object):
>>>    t = (1,2,3)
>>>    l = [1,2,3]
>>>    def __init__(self):
>>>       self.l = [1,2,3]

Now changes to k.l won't effect j.l

However, I can't find a way to accomplish the opposite--create a class attribute for an immutable data type--w/out resorting to __get_attribute__ magic.

Pyrex for performance and obfuscation

2008-06-14T12:19:00.001-07:00

I've recently been asked to obfuscate a bunch of Python code. Encryption is one possibility, but the user needs the key along with the encrypted code in order to run the code, so this is really just a round-about form of obfuscation. And if multi-billion dollar (and rather unsavory) industries can't get this right, I'd rather not even try.

One novel form of obfuscation is compilation to C-code, a task made relatively simple by Pyrex and, more recently, Cython. Both projects are mainly intended to ease the integration of C libraries with Python; both accomplish this by compiling native Python code into a .so shared object. This .so file should, in turn, be slightly harder to decypher than Python bytecode.

Pyrex isn't as actively maintained as Cython, but it is available via Macports, so I'm using Pyrex for now.

Pyrex appears to work by first translating your Python code into C, then compiling this C against the Python libraries. Unannotated Python objects remain PyObject * pointers -- it's quite possible that the Python interpreter, or VM, or whatever lies underneath, is still doing most of the heavy lifting with Pyrex-translated code; I can't make any sense of it.

But Pyrex also allows you to write Python-like code that gets translated to native C, with all the implied performance gains. As a simple example, I've done a naive implementation of the Fibonacci sequence in plain Python and in Pyrex' C/Python intermediary. Here's the file, called "pyrex_fib.pyx":


cdef _cfib( int i ):
 if i < 3:
  return 1
 else:
  return _cfib(i-1) + _cfib(i-2)

def cfib( i ):
 return _cfib( i )

def pyfib( i ):
 if i < 3:
  return 1
 else:
  return pyfib(i-1) + pyfib(i-2)

_cfib and pyfib are the same function, w/ _cfib implemented in Pyrex C notation; cfib is a wrapper around _cfib. (Native C functions can't be called directly from Python and must be wrapped.)

"pyrex_fib.pyx" is compiled to a Python-friendly .so file via distutils; here's the contents of "setup.py" -- lifted from Michael's Guide to Pyrex:


from distutils.core import setup
from distutils.extension import Extension
from Pyrex.Distutils import build_ext
setup(
name = "PyrexGuide",
ext_modules=[
 Extension("pyrex_fib", ["pyrex_fib.pyx"])
 ],
cmdclass = {'build_ext': build_ext}
)

The compilation is accomplished via python setup.py build_ext --inplace, but note that bugs can result in the rather cryptic error message error: Pyrex does not appear to be installed on platform 'posix'

I also wrote a plain Python version, "py_fib.py":


def pyfib( i ):
 if i < 3:
  return 1
 else:
  return pyfib(i-1) + pyfib(i-2)

This enables me to compare Pyrex-translated Python code stored in a .so to the same code stored in a regular Python module, and compare them both to the "native" version.

I do this comparison via a simple module that imports both forms and runs them, timing each invocation. Here's the output:


kieran@host:~/tmp/pyrex$ ./time_fib.py
Sat Jun 14 12:15:07 2008
pyrex_fib.cfib(40) = 102334155 in 9.6s
pyrex_fib.pyfib(40) = 102334155 in 121.5s
py_fib.pyfib(40) = 102334155 in 98.0s

Interestingly, the Pyrex-translated Python code is about 20% slower than the regular Python; presumably, it's not benefiting from various interpreter optimizations. The "C" implementation blows them both out of the water.

Pyrex looks great for wrapping C libraries for Python, and might serve for code obfuscation, but the major limitation is the difficulty of moving non-scalar data types between C and Python: it wouldn't have been easy to return a Python dictionary from my cfib routine, for example.

Ouroboros

2008-05-30T10:34:00.000-07:00

Circular dependencies will break your Python code!

Given some module A that depends upon module B (ie, import B), and given a module B which depends upon module A, you'll get this rather cryptic error message:

ImportError: cannot import name A

This works at any remove, of course -- the circle could stretch through 1,000 modules, but once you forge that loop, you're toast.

The solution is to refactor your code to put routine from A needed by B into a separate module C that depends upon neither.

Leopard emacs is broken...

2008-05-28T10:02:00.000-07:00

Which means that GNUplot is broken under Leopard, and with it, scipy.

Fortunately, there's a solution:

sudo mv /usr/bin/emacs-i386 /usr/bin/emacs-i386.backup sudo /usr/libexec/dumpemacs -d emacs --version emacs

It's quite beyond me why the emacs shipping w/ Leopard is broken out of the box, but can be repaired via the dumpemacs command, but there it is.

Redirecting from mod_python

2008-02-14T10:40:00.001-08:00

The sort of thing I'm inclined to forget:


def handler(req):
    req.headers_out['location'] = 'http://www.modpython.org/'
    req.status = apache.HTTP_MOVED_TEMPORARILY
    req.send_http_header()
    return apache.OK

Care of the mod_python FAQ.

mod_python, Leopard, and the trouble w/ 64 bits

2008-01-21T09:01:00.000-08:00

mod_python is broken under Leopard, at least on 64-bit capable systems.

Actually, this is only half true. Leopard ships w/ Apache 2, but more importantly, it ships w/ a 64-bit capable Apache 2. Because it's the 64-bit version that runs on capable systems, any modules linked by Apache need to be 64-bit capable.

Apache is 64-bit capable:


kieran@bali:~$ file /usr/sbin/httpd
/usr/sbin/httpd: Mach-O universal binary with 4 architectures
/usr/sbin/httpd (for architecture ppc7400):     Mach-O executable ppc
/usr/sbin/httpd (for architecture ppc64):       Mach-O 64-bit executable ppc64
/usr/sbin/httpd (for architecture i386):        Mach-O executable i386
/usr/sbin/httpd (for architecture x86_64):      Mach-O 64-bit executable x86_64

And mod_python, built by simply running ./configure and make, is 64-bit:


kieran@bali:~$ file  /usr/libexec/apache2/mod_python.so
/usr/libexec/apache2/mod_python.so: Mach-O 64-bit bundle x86_64

So that much works. And basic mod_python functionality is intact. The problem comes when you try to compile python libraries that depend upon native C/C++ code. Neither command line options nor the CFLAGS environmental variable can force python distutils to build 64-bit friendly libraries.

I haven't figured out how to force distutils to cooperate yet, but in the meantime, rebuilding the entire software stack in fink creates a usable environment.

The first step is to simply ask for mod_python:

stop apache: sudo service org.apache.httpd stop

install the necessary packages: fink install libapache2-mod-python-py24

accept the lengthy list of dependencies

Installing the PHP module wasn't so easy; I wound up building it by hand. First, make sure the Postgres include files are installed, assuming you need Postgres:


fink install postgresql82-dev

./configure --with-apxs2=/sw/bin/apxs2 --without-iconv --with-mysql=/sw/ --with-pgsql=/sw/

Magsafe Connector Disassembly

2008-01-14T10:55:00.000-08:00

The chord nearest the Magsafe plug on a modern Apple laptop is under a good deal of mechanical stress, and mine finally gave out; the outer layer of wire looked like loose steel wool. So I decided to take the plug apart:

Notice that black bit on the PCB in the left-hand image? That's a chip, identified by the following text:

A3
2100
613A1

It most likely controls the switch between the orange "charging" LED indicator and the green "powered" LED. However, I wonder if this chip doesn't hold the answer to another mystery: why won't the Macbook battery charge off the Magsafe airline adapter?

For further information on the subject, I'd suggest Stuart Schmitt's guide to hacking the MagSafe cable for use w/ a standard DC transformer.

mod_python under Leopard

2008-01-08T15:22:00.000-08:00

OS upgrades break my workaround for compiling mod_python under OSX.

This is a good guide for fixing mod_python under Leopard, but with a caveat: if you've upgraded your system from Tiger, the "missing symbol" problem might persist. Run otool -L /usr/libexec/apache2/mod_python.so -- if you see a reference to /Library/Frameworks/Python.framework/Versions/2.4/, that's your problem. I simply renamed /Library/Frameworks/Python.framework to /Library/Frameworks/_Python.framework and rebuilt.

Update: I take it all back; mod_python essentially doesn't work under Leopard, at least on a 64-bit capable system. See my Jan 21 post for more information.

Merry Christmas, San Francisco Style

2007-12-17T15:07:00.000-08:00

Macy's on Union Square, Saturday, 15 December 2007. But a small portion of the March of Santas:

(About the photo: my apologies, but the camera on the Sidekick 3 is almost worthless. 1.3 megapixel, sure, but you've gotta drop the rez to VGA to smooth out the noise ... what a mess.)

Re: the EMBOSS bioinformatics toolset, and the IEP utility in particular

2007-11-27T13:32:00.000-08:00

The iep.c codebase is supposed to predict Isoelectric Point for an amino acid sequence. And it does this, but it also manages to serve as an object lesson in bad C code:

AjPFile, AjPStr, and ajint types: your very own custom integer type?? Really?
Put everything in global variables! Thus, you wind up w/ void functions that take 0 parameters and do something incredibly important; the challenge is to track down what that important thing might be! (Every programmer should be required to write code in a purely functional language for at least 3 months. By law.) Take void embIepPkRead() , for example.
Functions that simply wrap other functions. Code as treasure hunt!

Biologists shouldn't code.

If anybody is interested in the basic algorithm used by EMBOSS IEP to calculate Pi, it basically calculates the charge of the protein iteratively across potential pH's until it hits 0 charge, or the upper and lower bounds converge. The primary function can be found in embIepPhConverge. ajint *c holds a histogram of amino acid counts, double *K is an array of dissociation constants for each amino acid, and double *pro is the moneymaker -- it gets updated w/ a proton count for each amino acid as different pH's are tried. The loop starts w/ max 14.0 and min 1.0, sets pro for the midpoint pH, then sets either the upper or lower bound to the midpoint, depending upon whether the calculated charge is positive or negative.

Java from the command line and GUI dependence

2007-11-16T16:40:00.000-08:00

Calling a .jar file from the command-line in headless mode--say, from a web-server--seems to pose a problem under OS X Tiger, in that java requires communication w/ the GUI system. (This leads me to believe that the problem applies under Linux as well.)

The solution is the "-Djava.awt.headless=true" parameter; for example, here's how I call the Batik SVG rasterizer:
/usr/bin/java -Djava.awt.headless=true -jar /Library/WebServer/Documents/py/archive/batik-1.7/batik-rasterizer.jar -w 1000 -h 1000 -bg 255.255.255.255 xxx.svg

I knew I didn't trust that newfangled Java.

Mac OS X Mirror: The RAID for disks that never fail

2007-09-24T15:32:00.000-07:00

I built a mirrored array (RAID-1) on a Mac OS X machine, and a disk failed. Which is why you mirror disks in the first place.

The instructions for re-creating a mirror basically tell you to drag a replacement disk into the array -- presto chango, you're done! Except, I can't. Because when I originally created the array, I didn't dig through the options and enable "Auto-rebuilding".

Or, as Tom said, "Did you go into 'advanced' and deselect `corrupt randomly`?"

Macbook evisceration

2007-09-21T22:51:00.000-07:00

Dumping a glass of water on your Macbook keyboard is double-plus ungood, but I got a lucky break: when I flipped it over to dump the water, I flipped it towards the DVD drive. The machine was dead, to be sure, but the water went to the optical drive instead of the motherboard.

If this happens to you, I'd recommend this guide to Macbook disassembly, and access to a desiccator.

Archive

2007-09-15T18:52:00.000-07:00

The post preceding this one is archival: the past few years of technical notes, minus the past couple months' worth. Those months were lost when my previous host went away.

2007-09-15T18:50:00.000-07:00

Thur, 31 May 2007

SMB under OSX

Wonderfully annoying: when you mount a Samba share under OSX, only root and the user who mounts the share can see the files, irrespective of file permissions and ownership. So, if you want a web server to see a SMB mount, here's the magic: sudo -u www mount_smbfs //username:password@server.domain.com/share /share

Note that /share must be owned by www, in this scenario

Wed, 16 May 2007

Image Magick hints

Convert x.gif to grayscale x.png: convert x.gif -type grayscale x.png

Isolate the red channel from RGB image x.jpg: convert x.jpg -channel R -separate x.png

Tues, 8 May 2007

Python whinging

A couple of quick notes, for my own reference. 1. 16-bit grayscale images can be created from a binary string slike so: Image.fromstring("I", (250,250), s, "raw", "I;16B") 2. the python zipfile modules is SLOW.

Tues, 8 May 2007

The vagaries of coincidence

The .cxi files generated by Applied Biosystems 8200 FMAT imaging software are NOT Clemex format images, though they share a file extension and are both used in biology. No. Instead, .cxi files are RAW image data.

Sat, 7 April 2007

Ugly shell tricks

One particular shell incantation I've managed to remember for some years now: 2>&1

This redirects both STDERR and STDOUT to one channel, and is handy for piping things sensibly into less. But care of O'Reilly's, "Unix Power Tools," here's a discussion of why dumping the resulting stream into a file isn't intuitively done:


One of the common questions about the Bourne and Korn shells is why only
the second command will redirect both stdout and stderr (13.1) to a file:

   $ cat food 2>&1 > file
   cat: can't open food
   $ cat food > file 2>&1
   $

Although lots of sh manual pages don't mention this, the shell reads
arguments from left to right.

  1.  On the first command line, the shell sees 2>&1 first. That
means "make the standard error (file descriptor 2) go to the same place
as the standard output (fd1) is going." There's no effect because both
fd2 and fd1 are already going to the terminal. Then > file redirects
fd1 (stdout) to file. But fd2 (stderr) is still going to the terminal.

  2.  On the second command line, the shell sees > file first and
redirects stdout to file. Next 2>&1 sends fd2 (stderr) to the
same place fd1 is going - that's to the file. And that's what you want.

Mon, 2 April 2007

Recommended to me by Tom: The Software Project Survival Guide

Fri, 30 Mar 2007

On installing linux on your cheapo Linksys router

The Linsys WRT54G router can easily be had for $50 and converted to a half-decent Linux router. Most discussions of the Linsys WRT54G advise against buying the newer versions (v5 or v6), since they've cut the available flash RAM to 2M. However, it CAN be made to work: first, only use v23_SP2, not the newer SP3. The instructions are pretty good, but omit one important detail: when time comes to upload the firmware, you're computer's ethernet card has to be set to 10Mb/s / half-duplex. Auto-detect won't work, at least on my Intel Macbook.

Tues, 14 Mar 2007

On JSON, Python, Javascript, and types

A Python server and Javascript client interact pretty well, BUT: associative arrays in Javascript are a fragile hack. To create an associative array (or "dictionary," or "hash table") in ECMAscript/Javascript, you treat a generic object, which necessarily has keys and values, as that's what an Object is, as an associative array. But the keys are STRINGS; so a numeric associative array is now keyed off of STRINGS, not INTs. Therefore, if your JS client receives a dictionary keyed off of ints, those ints will automatically be converted to ASCII representation. If this dictionary is then sent back to the server, Python will maintain the string key representations, and everything goes to hell.

Tues, 6 Feb 2007

De gustibus non disputandem est

Fri, 12 Jan 2007

Cell phones & Disappointment

Ok, Apple's iPhone has been unveiled, to much hoopla. But wait: it doesn't support 3G! That's livable, because it's not a cell phone, it's a COMPUTER, right? It runs OS X! But it's NOT: it can't run the software I'd like it to. I guess I'll have to buy a relatively ugly linux-based phone.

Fri, 1 Dec 2006

the cut command

somehow, i've never seen the cut command before. very useful: it returns the n'th field from a line, where the -d option can set the delimeter (eg, the default of tab, or "|", or whatever you want). handy!

Tues, 21 Nov 2006

A thought on the Macbook keyboard, with specific regard to Ubuntu Linux, the Parallels Virtualized Machine, and the <F1> key

By default, the <F1> key isn't a function key; it adjusts screen brightness. Nor is the <F3> key a function key: it mutes the speakers. And so on. In order to access the F functionality of those keys, hold down Fn. That is all.

Fri, 10 Nov 20006

Multi-value inserts in Postgres

the basic syntax is as follows--and it's ugly:

copy xha_param(param_name,param_datatype) from STDIN delimiter '|';
   Target identifier|var
   Campaign name|var
   Injection date|var
   Injection type|var
   \.

Thurs, 2 Nov 2006

Griping about Python

There really aren't many: Python feels more like a "real" language than a sloppy scripting language, and, frankly, I like it. I'll restrain from doing my "it lacks functional language feature X" routine, as that's not very interesting, and to Python's credit, it has more functional features than you'd expect--including currying as of 2.5!

But: I've got a couple of small complaints.

Mutable types, eg, lists, should return the object in question when a class method is run. For example, list.sort() doesn't return the sorted list, but ... nothing at all. so you can't do something like ",".join(list.sort())

The pass statement. This is a by-product of Python's whitespace-as-syntax design, but it doesn't feel NECESSARY.

The lack of a tertiary operator. This has been discussed at length elsewhere, so I won't belabour the point. [UPDATE: one has been added: A if CONDITION else B -- 10 Nov 2006]

More to come.

Mon, 2 Oct 2006

mod_python on Mac OSX

First off, compilation of mod_python for the Mac OSX defaults--apache 1.3 and python 2.3--is described at http://www.modpython.org/pipermail/mod_python/2004-August/015994.html.

Next, apache config is discussed at http://www.dscpl.com.au/wiki/ModPython/Articles/GettingModPythonWorking. the important bit is the AddHandler python-program .py line: all the other guides only mention the newer syntax, which won't work w/ apache 1.3 (or, more accurately, the older version of mod_python, required for apache 1.3)

Fri, 18 Aug 2006

Hammerhead, web server load tester

i used to use Hammerhead for stress testing web servers, and it was great. but now i've switched from Debian Linux to Mac OSX, and fink doesn't include a hammerhead port ... much fun ensues, getting it to compile. suck, sucky autoconf files. first, add the line target_os=freebsd before the case "$target_os" in block in configure. next, run ./configure --host=x-y-freebsd. then, edit config.h and replace #define HAVE_SSL 1 with #undef HAVE_SSL. edit hammer.mk and remove -pthreads. last, set LIBS=-lresolv in hammer.mk. jesus.

Update: ultimately, i gave up on hammerhead and used time wget.

Wed, 9 Aug 2006

More on Postgres on Mac OSX (fink)

to stop postgres, issue this command: /sw/bin/pg_ctl stop -D /sw/var/postgresql-8.0/data/ ... though i imagine your milage may vary depending upon your exactly version

Tues, 8 Aug 2006

PostgreSQL on Mac OSX Tiger

I used fink to install postgres. The only major problem here is the postgres user, which is added by fink but unusable, as its login shell and home directory are set to /dev/null. Worse, OSX is non-traditional in its account setup, so you can't just edit /etc/passwd, or /sw/etc/passwd-fink. Instead, use NetInfo Manager, which can be found in /Applications/Utilities.

Mon, 17 July 2006

A New Way to Waste Time

Actually, it's an old way, made new again: DOSBox for Mac OSX. And Syndicate, one of my favorite games from the early 1990s.

Thurs, 29 June 2006

The find Command

I finally bothered to read the find man page. Here's how you recursively grep for string: find ~/ -type f -exec grep -l string '{}' \;

Wed, 21 June 2006

Fun with mysqld

FARW left for Seattle today, to the best of my knowledge.

Anyhow, two thoughts about mysqld: in my.cnf, setting bind-address=0.0.0.0 allows connections from all hosts. Also, to enable a user to connect from anywhere, when she is created, give her the name 'somename'@'%' -- that percent sign is the key, here.

Tues, 20 June 2006

Yet another reason to loath Windows

Shiho's Windows XP machine once again Autoupdate'd (didn't i DISABLE that?!?), which renders the machine inoperative--it just reboot automatically ad infinidum, and gives the error message
STOP: c000021a {Fatal System Error} Windows Logon Process system process terminated unexpectedly with a status of 0xc0000018
The fix can be found here.

I'll copy it, for posterity:

The newest Critical Patches can be the reason for the error, please uninstall KB893066, KB890923, KB890859 and KB893086 ---------------------------------------- Please follow the procedure suggested by microsoft to solve BSOD problem from the latest windows update. 1. Insert the Windows XP startup disk in your floppy disk drive or insert the Windows XP CD in the CD drive or in the DVD drive, and then restart your computer. Note When you receive the following message, press a key to start your computer from the Windows XP CD: Press any key to boot from CD Note Your computer must be configured to start from the CD drive or the DVD drive. For more information about how to configure your computer to start from the CD drive or the DVD drive, see the documentation that came with your computer or contact the computer manufacturer. 2. When you receive the Welcome to Setup message, press R to start the Recovery Console. Note Multiple options will appear on the screen. 3. Select the Windows XP installation in question. Note You must select a number before you press ENTER, or the computer will restart. Typically, only the 1: C:\Windows selection is available. 4. If you are prompted to type an administrator password, do so. If you do not know the administrator password, press ENTER. (Typically, the password is blank.) Note You will not be able to continue if you do not have the administrator password. 5. At the command prompt, type cd $ntuninstallKB.........HERE THE NUMBER OF THE PATCH.....$\spuninst, and then press ENTER. Note After you complete this step, you cannot stop the removal process. 6. At the command prompt, type batch spuninst.txt, and then press ENTER. 7. After the Ptach is removed, type exit, and then press ENTER oo Uninstall the Next please uninstall KB893066 KB890923 KB890859 KB893086

Tues, 9 May 2006

Whoa.

William Gibson's books & short stories, though mostly in cyrillic. that can't be legal.

Tues, 9 May 2006

SF, post global-warming

care of boingboing, it's SF after a 7m rise in sea levels! or, it should be, once their server recovers. things look suprisingly dry; we'll loose SOMA, but the outer sunset, where i live, is suprisingly unaffected.

quick note on using google maps to find GPS coordinates: when you first text search for a spot on google maps, your "page link" URL will just have your search string. to fix this, zoom in, then look at the "page link" URL--the coordinates are right there.

Thurs, 9 Feb 2006

Accessing all 4G of memory on an AMD64 system

Setting up linux on an AMD64 system. The BIOS has some kind of option which can reserve the upper 1G of memory for PCI DMA iff you have 4G loaded, leaving you w/ 3G accessable--but, oddly, it only demands this 1G tithe if you have all 4G installed. Turning this feature off in the BIOS generates an unusable system spewing IOMMU errors, UNLESS the pci=nommconf option is passed to the kernel. Ugh.

Tues, 10 Jan 2006

On Subversion

Switching to Subversion (svn) in lieu of CVS. A couple of
thoughts:

On debian, CVS comes w/ administrative niceties, eg, it creates a /var/cvs, owned by group src ... svn just comes w/ binaries -- best to create an svn user & group, add a /var/svn, etc

Use fsfs. When the repository is created, the syntax becomes svnadmin create --fs-type=fsfs /var/svn/REP_NAME

If you want to commit, checkout syntax looks like this: SVN_SSH="ssh -l kieran" svn co svn+ssh://your.fileserver/var/svn/REP_NAME

Tues, 5 April 2005

vsftpd & the linux kernel "security options":

So, there i was, working on a new system, and having NO luck getting vsftpd, the "very secure" ftp daemon, working. lots of these sorts of error messages from the ncftp client:

OOPS: cap_set_proc OOPS: vsf_sysutil_recv_peek

It turns out that i'd compiled the kernel w/ "different security models" enabled, and needed to load the "capability" module. weird.

Mon, 14 Mar 2005

Almost a year since my last post. In deference to the technical-fix purpose of this page, I'll describe my recent struggles w/ a failed terabyte RAID:

I have 5x Maxtor 200g drives crammed into one machine, most of them sharing channels on a Promise IDE card. The whole thing adds up to 0.75TB after parity, ext3 journal, etc. One drive failed catastrophically. This caused the other drive on that channel to irredeemably corrupt a number of sectors, which meant two drives were, in the eyes of the kernel, bad.

Bad bad bad. This means the RAID is _dead_. The fix went like this:

purchase a new drive of exactly the same model, and copy the drive w/ bad blocks onto it via dd if=/dev/bad_drive of=/dev/new_drive conv=noerror,sync . The conv portion tells dd to continue when it hits a bad block, and pad any missing data w/ the appropriate number of zeros.

reinitialize the RAID w/ the copy in place of the bad drive. in my case, this went mdadm -A /dev/md0 -f /dev/hda1 /dev/hde1 /dev/hdf1 /dev/hdg1, where -f tells mdadm to continue w/ only 4 drives. now you should have a functional device.

fsck the RAID

replace the missing drive--in my case, the drive w/ the bad blocks was fine once reformatted; only attempts to read the wonky blocks caused trouble

at this point, everything seemed to be ok, but after a bit, the kernel started remounting the drive ro after finding problems w/ the journal. so, i recreated it: tune2fs -O '^has_journal' /dev/md0 removed the journal, turning the partition into an ext2 partition, and tune2fs -j /dev/md0 recreated the journal.

ugh.

Tues, May 25, 2004

I'm extraordinarily happy w/ my newest laptop, a hand-me-down IBM Thinkpad 600e donated by Inf. It's amazing how useful a PII 400mhz w/ 192meg, a 1024x768 TFT, and 6gig drive still is. The only complication was getting sound working under linux: though lspci reports a "Cirrus Logic CS 4610/11," using ALSA drivers, `modprobe snd-cs4231 port=0x530 irq=5 dma1=1 dma2=0` did the trick.

Sun, Feb 29, 2004

Ahhh, now I can experience one of life's most satisfying treats from the comfort of my home--or, at least, my therapist's office. It's the virtual crack house, under construction as part of a theraputic technique that uses simulated experiences to help people come to terms w/ the real ones. In this case, the virtual crack house is supposed to help addicts learn to supress their cravings. Good fucking luck, there.

Sat, Feb 14, 2004

I've spent various portions of the past month trying to get mod_caml, the ML analog to apache's essential mod_perl, working under debian stable. And I might finally have solved it.

The big hurdle is OCaml's instability, as a language: the chances of a package written w/ version 3.07 of the compiler working on a system running 3.06 are pretty low. And version 3.06 was the backport, ferchristsake. So I've created my own OCaml backports.

How? I first added the testing source repository to /etc/apt/sources.list: deb-src http://ftp.us.debian.org/debian/ testing main non-free contrib. I then used the "apt-get build-dep packagename" command to get install those dependancies which could be found, and "apt-get source --compile packagename" to build those which needed backporting.

Mon, Jan 5, 2004

My species astounds (care of Bruce Sterling's State of the World Interview).

Sun, Jan 4, 2004

Wonder what your Latitude and Longitude are, but don't wanna buy a GPS?

Thur, Jan 1, 2004

The NYT lists "The Futile Pursuit of Happiness" as their most emailed Magazine article of 2003, and i can see why: it's probably provoked more conversation than any other article i've read lately. (here's a cached copy, in case it falls back into pay-to-read land)

Tues, Dec 30, 2003

Now that I've got a DVD+/-RW burner, it'd be nice to understand the difference between the two standards. DVD+RW seems particularly suited to general computing, as it supports true random write access, but that doesn't tell us what the underlying difference is. Apparently, "[the] +RW format uses phase-change media with a high-frequency wobbled groove that allows it to eliminate linking sectors," while "-RW uses groove recording with address info on land areas for synchronization at write time." Oh, well, that clears everything up.

Unrelated: 802.11b/g cards w/ a 1 MILE range???

Sat, Dec 6, 2003

Alright, a bit of politics: is it possible that the Neocons who pushed us into war buy into the crackpot idea that Saddam was not only a genocidal tyrant, but singly responsible for every terrorist attack on the US in the past decade, including, get this, the Oklahoma City bombing? Oh my...

Sat, Nov 29, 2003

I've been certified an Anonymous Coward!!!

Sun, Sep 21, 2003

Just set up MySQL, requisite for b2evolution, a PHP-based blog engine my gf chose, apparently w/ one criterion in mind: it needs to support multiple, discrete blogs. I'm not yet certain how well it does this, as it's pretty out-of-the-box. The MySQL Manual was was pretty handy.

unrelated: I really need to document how I moved a Windows 2000 Pro installation between drives w/ minimal pain. The key here is Sysprep.

Sun, Sep 14, 2003

Dashboard: the open source world's answer to MIT's Haystack project. Both purport to automagically connect associated ideas from within one's individual data stream, thereby reminding one of that URL for the Government Information Awareness site while perusing, say, a discussion of the corrupting effects of campaign donations. (Ok, that's a pretty odd example)

So, how to get it working on a Debian unstable system? I'm not sure yet, but reasonable directions are available at this wiki. That'll refer you to an apt source, which is really handy. Except that the gtk-sharp provided isn't new enough, so you wind up having to compile from CVS. Which sounds fine, BUT: I couldn't get the CVS version to provide gnome-sharp.dll, so i first had to install the .deb version, then compile & install. Not quite sure why that worked. THEN, you need the at-spi package, though noone mentions this. And you need libgtkhtml3.0-2. Ok, but after all this work ... well, no, it still doesn't do anything. Goddamnit.

Sun, Aug 31, 2003

ok, a list of links re: achieving some sort of 'net telephony between either my gf's win2k or linux box. first off, the problem w/ using speakfreely, wonderful little unix/windows utility which worked wonders back in '99 but has apparently been killed by NATs: MacOS X doesn't have a /dev/audio.

next idea, care of inf: GnomeMeeting, which is apparently compatable w/ M$' NetMeeting and runs under OS X. Or so they say.

this entry is much closer to my original idea for this blog... in that i'm really writing this for google's sake.

Fri, Aug 8, 2003

GAR. Much wrestling w/ [Debian's version of] vsftp, the "very secure" FTP daemon. No idea if it lives up to it's name, but the version packaged w/ Debian Stable, 1.0.0, has at least one signifigant shortcoming: whatever the man page says, you must call the anonymous FTP user, which gets an entry in /etc/passwd, ftp. It's hard-coded in tunables.c, so, you could alternately hack @ that and recompile.

Wed, July 30th, 2003

Wrestled, successfully, w/ two different wonky technologies over the past few days: 1. port forwarding w/ Linux 2.4 iptables routing, and, 2. the 2.6.0-test1 kernel.

The voodoo necessary for Port Forwarding looks like this: iptables -t nat -A PREROUTING -p udp -d $PUBLIC_IP --dport $DESIRED_PORT -j DNAT --to-destination $INTERNAL_IP

Oh, and there's a nice flush script, too.

Now, onto 2.6: everything SEEMED to be working--X came up, mozilla loaded w/out a hitch--but there was one glaring omission: xterms. No login prompt to speak of. No virtual console, no ssh login, no xterm. An entry in /etc/fstab instructing init to mount a new fake fs, type devpts, under /dev/pts brought my xterms back, while virtual consoles required that I enable "VGA Text Console" under "Graphics Support -> Console display." That's it. What do I think of 2.6 thus far? Well, by and large, I haven't noticed a difference, except in one very tangible category: responsiveness. Linux on the desktop (read: X) just got a lot friskier.

One other rather big linux-related problem, though I'll address this some other time, is my recent discovery of the fallability of linux' software RAID implimentation. It's Achilles heel is simple: bad blocks. Now, everyone's experienced a bad block or two, and, so long as their numbers stay in the 1-2 range, they don't pose any serious problem. It's only when they begin to multiply that you've got yourself a paperweight for a disk. However, linux' RAID implimentation is a perfectionist: it can't tolerate a single bad block. The second one is detected, the whole disk is booted from the array. This is a Bad Idea (TM).

Machine readable format.

2007-09-10T17:22:00.001-07:00

Much of my time is spent cajoling clunky technologies into submission, and I occasionally document the process for myself and any desperate search-engine users who stumble across this page.

The rest of you can move along, there's nothing to see here.