free variable

This photo.net query is either pretty hilarious, a sad commentary on the age, or both. (link via markd on twitter) Of course, there are obvious parallels to this question in the audio-nerd world, where novice recordists often ask “which plugin/software will make my recordings sound like artist/producer X” — ignoring performance, technique, instrument, room, microphones, mic placement, and preamplification. For some reason, though, reading this recalled this classic Onion article for me.

Assume for the sake of argument that you regard email both as a necessity and as a regular source of distraction and interruption. You might set your mail client to check infrequently so that you can batch-process messages, but perhaps — due to force of habit or weakness of will — you find yourself clicking “Get Mail” more often than is healthy. How can we solve the distractions of compulsive clicking?

Consider patient-controlled analgesics (PCA), which hospitals use to allow patients to self-administer doses of intravenous painkillers by pressing a button. Of course, it would be dangerous to allow patients to take arbitrary doses of opioids, so the PCA system only dispenses drugs once in any given interval — fifteen minutes, for example. That is, at most one button press every fifteen minutes will actually have an effect — the rest will serve as placebos.

In patient-controlled email, the “Get Mail” or “Refresh” button will check with the server at most once every n minutes, no matter how often it is pressed. To make the user feel better, the standard “checking mail” messages and spinners could pop up and pop away as they typically do. This behavior will dramatically reduce the potential interruption caused by compulsive checking for new messages. (I suspect that some clients and servers actually implement this, although I have no proof.)

The concept is simple: you may believe that you know how often you need to check email, but the client knows better. Perhaps the act of refreshing your inbox is merely punctuation in the middle of some other, more involved tasks — a simple hiccup in your workflow that allows you to go on and do something else unfettered by the thought that there might be some message demanding your attention. With patient-controlled email, you wouldn’t have to fear the arrival of an actual message, causing a comma to turn into a sidebar. It would be simple to support this feature and its configuration as well:

The “check at most” option could default to “as requested” — meaning that the “Get Mail” button might deliver an overdose. In any case, patient-controlled email is compatible with automatic checking and — for some patterns of email use — enhances manual checking.

If you’re using the Safari 3 beta, there’s a cool feature that I don’t believe was available in previous versions: you can examine the properties and styling of DOM elements in a floating inspector panel:

I’m not trashing my copy of CSSEdit, but this is an extremely nice feature!

UPDATE: Here’s some more on this feature.

This consequence of the Engadget Apple rumor fiasco merely indicates to me that too many people are directly managing stock portfolios who have absolutely no business doing so. I mean, what kind of fool would you have to be to engage in a financial transaction based on rumor and conjecture on a website? A random website rumor about Apple is about as reliable as an index card on a urinal reading SONY WILL HAVE TO RECALL ALL CONSUMER PRODUCTS INCLUDING ORIGINAL WALKMAN. I AM TRUSTWORTHY..

I completely agree with the folks at bancomicsans.com, who claim to provide “the source for anti-comic sans propaganda.” Comic Sans is facile, ugly, and immediately casts a pallor over any work typeset with it: “this is the product of an unserious human.” Few fonts make me so angry.

[thanks to H., via M.]

Rolf sends a link to this gizmodo article about some company that calls itself “Amiga” and is introducing new computers next week. Sadly, instead of fondly remembering the machine that made computing fun (and, notably, introduced me to emacs in the mid-1980s), I was instead reminded of when I bought the “Amiga SDK” in 2000, which was a hosted, VM-based development environment that ran on Linux. (It also had some bizarre copy protection scheme that didn’t work on machines with more than one ethernet interface.) I don’t know if this soon-to-be-not-vaporware “Amiga” hardware uses the same SDK or not.

After browsing amiga.com, though, I was reminded of the original bearer of the Amiga name. In the heady days of the mid-to-late 1980s, I used many applications that were way ahead of their time: a real shell on a personal computer, the trackers, Deluxe Music, Deluxe Video, and a wide range of other tools that didn’t have competitors in the rest of the computer world for years.

Fittingly, the current Amiga is also a source for software that has no peers elsewhere, like 2006 Arena Football League Word Search. (I’d be willing to bet that isn’t licensed and would be running the risk of an instant cease-and-desist if it weren’t so far under the radar.)

I just installed the OS X 10.4.9 update yesterday. Since then, ssh has failed to forward my Kerberos and AFS tickets to the office. Saying this is a big pain is perhaps the understatement of the decade. (It’s thrilling to log in to my office computer and not have access rights to any of my files — it makes me feel like a secret ninja hacker, just like Matthew Broderick in Wargames!) As far as I can tell, this is the default behavior in the version of ssh included with 10.4.9 (bad idea, Apple). Fortunately, this simple solution worked for me:

1. Open Terminal.
2. Using your favorite editor, open the file /etc/ssh_config
3. Uncomment (i.e. remove the “#”) from the following lines:
   • Host *
   • GSSAPIAuthentication …
   • GSSAPIDelegateCredentials …
   • GSSAPIKeyExchange …
4. If a no appears in the … part of any line you uncommented, change it to a yes.
5. Save the file. You’ll need an administrator password.
6. (Hopefully) enjoy functional ticket forwarding again, like before you upgraded.
7. Grimace, since you haven’t tested any of your Audio Units under 10.4.9 yet. Be glad you made a backup.

This seems to make ssh slower, but it also seems to work.

I’ve been following Actiontastic for a while, installing time-limited betas and hoping that the final version would be released soon so that I could buy it and stop worrying about having a decent GTD system on my computer. Well, I can stop waiting, I guess:

Actiontastic will be free and open source. The free (as in “free beer”) part starts tonight. The code (as in “freely available source code”) will follow when the overhead of a new team won’t crush the project under its own weight. Those with experience getting to 1.0 will understand what I mean.

Crosby is also going to release the source code to actionatr, his upcoming productivity tool web service that syncs with Actiontastic. Nice.

Here are some typefaces you will like from the consistently impressive P22 foundry: the Pop Art set, Kells Round, and Franklin’s Caslon. (The Kells link is to veer.com because P22 does not make it easy to link to an individual face in a set.) I particularly like P22′s apparent m.o.: finding good lettering from art and digitizing it.

(I started window-shopping for type after reading this article about Gill Sans. You may find it interesting as well.)

I managed to end the nightmare of crappy rasterized fonts in PDFs after reading this thread. Here’s the short version:

\usepackage[T1]{fontenc}
\usepackage{lmodern}

Apparently, the lmodern package replaces CM-Super, which replaced something else, …, which transitively replaced something that was supposed to solve the crappy raster font problem years ago. I don’t know why I didn’t know about it until now.

Technorati Tags: notes-to-self

If you’re a nerd, either you already read John Gruber’s daringfireball.net or you ought to. Gruber’s writing about popular technology is always intelligent, insightful, and lucid. A regular feature on the site is the “Jackass of the Week,” in which he dissects the ridiculous claims of some foolish technology pundit.

This week, the award went to Gundeep Hora of cooltechzone.com for claiming that Apple is likely to stop making OS X. Gundeep Hora is not as notorious as many of the previous JOTW recipients¹, but I’m glad that Gruber pointed out Hora’s site, as it is pure comedy gold — it’s the webpage equivalent of that wild-eyed guy on the bus who’s always on a tirade about how Antarcticans caused the national debt.

“CoolTechZone” is a collection of ad-impression-friendly short articles consisting almost exclusively of risible, factually-challenged navel gazing.² Consider his recent article “Why Microsoft Should Acquire Linux”. The title alone indicates what sort of argument we should expect from this fellow (although he claims in the article that he merely means that Microsoft should acquire all commercial Linux distributors). Hora’s tense, rhetoric, and ability to discriminate between the hypothetical and actual are deeply confused throughout; furthermore, he presents claims like the following without evidence:

Assuming Microsoft makes such a move, it will obviously concern the Linux community, and rightfully so. What would prevent Microsoft from killing Linux just so Windows could continue to be the dominant OS maker? Nothing, to be honest. I suppose the various Linux distributions that Microsoft may acquire would have to work on that with the software giant.

Well, to be honest, one supposes that they would. But Hora’s closing thought is brilliance in 10-point Arial:

With the amount of resources Microsoft has, and its potential threat to Linux, it only makes sense for the two competitors to merge and keep everyone satisfied.

QED, I suppose.

¹ As of today, Google identifies only 438 pages linking to cooltechzone.com; a cursory scan indicates that the vast majority of these are located on cooltechzone.com.
² Here’s an article template for Hora’s site: I’m making a mildly controversial or foolhardy assertion about something that is unlikely to happen. This might not actually be possible for reason X that I will now ignore. I think that person or persons Y would be happy/unhappy in this event, and person or persons Z would be unhappy/happy. Calm down, person Y/Z! Time to wrap things up. The end.

I’m currently listening to Express from the album “The Roots Of Dubstep” by Bbbb

Sorry if this is obvious: I had wanted to try this for a while, but wasn’t sure it would work. It does, so I’m writing about it.

I was running out of disk space on my notebook, with often-hilarious consequences. (It should not be surprising how poorly most modern operating systems deal with having absolutely no free space and needing to page.) The most obvious advice OmniDiskSweeper gave me was hard to take: well over a third of my disk was devoted to music and photos. I chose to ignore this advice and the law of diminishing returns for as long as possible, deleting other random files, archiving infrequently-used documents, and so on, but I grew tired of having to decide what to trash before running some experiments.

I rejected as “too nerdy,” “too inconvenient,” or “too expensive” the options of storing my music on a networked file system (WebDAV, AFS, or Samba were all contenders); setting up a DAAP server and “sharing” from home (probably over an ssh tunnel); using Amazon’s S3 service as paid-for network storage; or buying a new internal notebook drive. Instead, I took the plunge and moved my iTunes Library to an external drive. I don’t haul the drive around, I just keep a selection of music on my iPod. It works pretty well and was reasonably easy; I am documenting the steps I took because there are a few things that may save you some time if you want to do the same.

1. First, I plugged in my iPod and set it to “manually manage songs and playlists.” This is important, since I want to be able to plug my iPod in to charge it even when the external drive is not available, and I don’t want some overzealous syncing to erase my music. I ejected the iPod.
2. Then, I needed to copy the music to the external drive. This was easy: I just changed the library location in the iTunes preferences to point to a folder on the external drive and selected “consolidate library.” (“Consolidating” your library moves all music files to the library folder specified in your preferences.) Now I had two copies of my iTunes library: one on my notebook drive, and one on the external drive.
3. Because I have some paid-for downloads (and a bunch of ripped personal CDs that would be a massive pain to re-rip), I made another backup to a different disk before I wiped out the copy on my notebook.

It works pretty well, actually. If you haven’t tried this yourself, you may be pleased to learn that it’s possible to use the iPod as a music source from within iTunes — so if you’re at a desk most of the day, you can leave the iPod docked and just listen to its music on iTunes. There are only a couple of drawbacks:

• Having to manually manage playlists is sort of a pain. (related: Having to remember to eject the iPod is sort of a pain.)
• My iPod is not nearly large enough to hold all of my music. (It doesn’t do photos, so I don’t have to worry about those.) This is not a huge problem, but it does mean that I’ll need to devise some strategy so that I don’t listen to the same 13gb of music exclusively for the rest of my life.
• As far as I can tell, it’s not possible to burn CDs directly from the iPod in iTunes. This means that I will be less likely to burn throwaway CDs to play in the car. This is probably not a big deal.
• To actually sync my iPod, I need to daisy-chain it off of the external drive.

All told, I think these are reasonable tradeoffs.

I’m currently listening to Backward from the album “Memories Of The Future” by Kode 9 and The Spaceape

Site notes

Some change at my web host has caused all of my redirection links (i.e. those http://willbenton.com/go/to ones) to cease working. I’m going to try and fix it soon. You may have noticed that I changed the theme a bit — now, you can see recent comments on the right. So if you ever experience a shortage of tirades from baton-twirlers with poor reading comprehension or bartering offers from Japanese-whisky enthusiasts with poorer reading comprehension, you’re in luck. (Note that I am referring to proponents of Japanese whisky, not to Japanese proponents of whisky; also note that, as of today, this site is the #3 Google hit for “Nikka whisky.”)

The “latest photos” are temporarily gone while I rework my photolog to use Gallery. The old photolog will remain accessible at photo-classic.willbenton.com. Of course, since the majority of my recent exposures are strictly of WT, you can see the latest kid pictures on my flickr photostream.

Analogy neutrality

The EFF’s Mike Godwin has a nice analogy for net neutrality: taxi service in NYC. Of course, everyone knows that phone companies are worse than the Nazis.

1999 vs. 2006

It’s probably time to update the infamous waterskiing photo:

Well, at least the incompetent fellow has his youth (and most of his hair).

Here’s a nice wiki about Ableton Live.

Cool Mac OS X discovery of the day: the pbcopy and pbpaste shell commands, which provide shell-script access to the system-wide pasteboard (aka “clipboard”).

Tom at Music Thing links to this absurdly awesome Kraftwerk video from 1975. (1975!) I like Kraftwerk a great deal; however, I am reminded that electronic music is not only an expensive hobby but an extremely dorky one. (Wait for the “lapel” discussion from the narrator.)

Wow. That’s really all I can say.

Related: buy “Possessed,” in which the Balanescu Quartet performs string-quartet arrangements of Kraftwerk songs. It’s not quite the Kronos Quartet doing “Purple Haze” — it’s better.

I’m currently listening to Trans Europe Express from the album “Minimum-Maximum” by Kraftwerk

This post from The Daily WTF definitely reminds me of some abject coding horror I witnessed when doing some maintenance programming on a consulting gig once. The “C/C++ tips” alone is worth the price of admission, but the fact that the programmer used strcpy instead of strncpy is icing on a terrible, terrible cake.

It looks like Apple has fixed the “stuttering powerbook” bug. I posted a tirade about this earlier, and I’m glad to see it’s taken care of. (It is still pretty embarrassing that they apparently let a race condition slip in to a kernel driver, though.)

I’m currently listening to Unspoken from the album “Rounds” by Four Tet, and it isn’t skipping.

schedule.willbenton.com is where I’ve published my iCal for the last couple of years. It’s worked out pretty well: I update something in iCal, it gets webdavved on over to the server, and the PHP iCalendar script does an admirable job of making it look ok. PHP iCalendar will happily generate dynamic web pages for every day, week, month and year from the UNIX epoch until 2038. As one might imagine, Google goes after this like Cookie Monster would a box of Oreos:

The above image is taken from the web statistics for schedule.willbenton.com for this month so far (i.e. 12 full days). Google sucked down 9.15gb of bandwidth, presumably trawling for activities I’ve engaged in in the previous 36 or upcoming 32 years. By comparison, in that period of time, four actual distinct humans checked my schedule (presumably these are my colleagues, trying to set up lunch), consuming a total of 1.87mb.

The real kicker is that I had a robots.txt file; unfortunately, it excluded web spiders from “willbenton.com/schedule” but not from its alias “schedule.willbenton.com.” Since the Google spider seems to have gotten much more aggressive lately, you should make sure you have a robots.txt file installed if you use PHP iCalendar. (There’s also a fairly recent version of PHP iCalendar available that fixes some security bug.)

One wonders how many of Google’s octuple-kajillion indexed web pages are someone’s electronic schedule from 1973.

Technorati Tags: google, ical

I’m currently listening to I’m currently listening to I’m currently listening to I’m currently listening to I’m currently listening to I’m currently listening to

The first time I encountered the infamous, bloody-ear-inducing PowerBook G4 “looping audio” problem, I thought it was part of the “glitch” music I was listening to at the time. (This genre is best described by Andrea, who calls it “music that hurts me,” “music that sounds like yelling,” and “did you make this?”) When it recurred during some Mendelssohn, with substantially greater ill aesthetic effect, I was more annoyed.

This is a ridiculous bug for Apple to have allowed into a shipping product, especially one that is otherwise as functional (and as likely to be used for audio-intensive tasks!) as the Powerbook. Today, John Wiseman posted about this issue and linked to powerbookdefect.info, where Apple users affected by this bug can sign what is certain to be an ultimately-ineffective petition about the matter.

Hey, I signed it anyway. Until then,

I’m currently listening to O Superman (For Massenet) from the album “Talk Normal: The Laurie Anderson Anthology (Remastered)” by Laurie Anderson

Technorati Tags: mac, osx, powerbook

UPDATE: this appears to be a “feature” of Blogger and not of Safari. The original post is preserved starting with the next paragraph.

The most recent version of Apple’s Safari browser has a potentially frustration-saving feature: if you enter some data on a form and attempt to leave the page (whether by closing the browser window/tab or by navigating to another page), it will give you a chance to avoid losing whatever you’ve typed in to the form:

This is a great idea; while the contents of substantial web form submissions (e.g. weblog or forum comments) are rarely important per se, it is still infuriating to lose something you’ve typed. Unfortunately, this feature interacts poorly with another Apple frustration-avoider: AutoFill. AutoFill isn’t perfect, but it has cut down on the number of times I’ve had to enter various details, like my home address and identifying information on web comment forms.

The problem is that Safari will alert you whenever you try and leave a page with some information on a form, independently of whether or not that information had been “AutoFilled” by Safari! So if I visit someone’s Blogger comments page, Safari will AutoFill my name and URL into the comment submission form. If I was merely reading comments (as is the case the vast preponderance of the time), then I will close the tab or go to a different page — but not before dismissing a browser warning. This mistake shouldn’t have made it into a shipping product, but it should be easy to fix.

I’m currently listening to Cantata No. 104, “Du Hirte Israel, höre”: V. Aria (Bass) from the album “Bach: Cantatas 21, 34, 46, 56 & 104” by Alois Pernerstofer, Hugues Cuenod, Jonathan Sternberg, Lorna Sydney, Vienna Chamber Choir, Vienna Symphony Orchestra & Wilhelm Stracker

The absurdity of “list price” or “manufacturer’s suggested retail price” is well-documented in many retail arenas, but rarely is it as bad as in the music-production gadget market. The “MSRP” is simply a number that the manufacturer has pulled from thin air in order to let their retailers present the illusion that they are offering the product in question at an excellent price. This is especially on items like electronic music gadgets or guitar equipment — I recall being a young teenager and marveling at the selection of guitars available at a local store for “50% off.” When items are routinely priced at 40-60% of “list,” the list price is not a particularly meaningful metric.

The situation has improved somewhat since I first became interested in musical gadgets. Many manufacturers, instead of providing a “list” price, now provide an estimated street price or their (enforced) “minimum advertised price.” The ubiquity of the internet has made retailers more honest (especially when a prospective customer can visit prepal.com or eBay from their cell phone in a showroom). However, today I turned up an especially egregious example of this phenomenon:

What a deal! (Or should I say “What a cheap?”) This guitar lists for $3000, but I could buy it today for only $1149 (er, plus $20 for shipping and at least $44 for a case, which Carvin will make you buy if you want the guitar shipped). There’s only one problem: Carvin only sells direct. You can’t buy this guitar from a retailer; rather, Carvin makes the guitar, sets the price, and you buy it from them. There is no “list price” because there is no “list.” No prices are “suggested;” rather, prices are set by the manufacturer, who is also the retailer.

Few regular readers will find this amusing, but this article about Behringer “tube” preamplifiers was pretty hilarious, as was Chris Randall’s scoop of an upcoming Behringer product announcement.

In the last four months, my computer — a formerly-reliable powerbook — has been a bigger crash factory than the Zildjian plant. Amazingly, it seems to crash most frequently when I am doing actual work; notably, when using the following applications:

• Prolog and iTunes (at the same time)
• OmniOutliner, OmniGraffle, or Keynote
• Eclipse

This is a royal pain:

I’m currently listening to nothing so my computer doesn’t crash while I am running these queries.

Sadly, my computer does not crash when I am using any of the following applications:

• Safari
• Ableton Live
• Apple Logic
• ecto

As one might imagine, this provides a serious disincentive to avoid procrastination.

I’m currently listening to So Gehst Du Nun, Mein Jesu, Hin from the album “A Book Of Chorale-Settings For Johann Sebastian” by Johann Sebastian Bach

Roland’s TB-303 Bass Line instrument, released in 1982, did a pretty terrible job of emulating a bass guitar, but managed to provide one of the most recognizable sounds in electronic music when used in a manner inconsistent with its labeling. If you already knew what a “303” is, then you already know this sound; if not, you probably knew the sound anyway — just not by name. The uninitiated can hear a sample (466 kb MP3, 27 seconds) complete with characteristic resonant lowpass filter sweep.

If you’re interested in an explanation of the 303 phenomenon that might feel at home in an academic discipline with one or both of “critical” and “studies” in its title, download Nate Harrison’s fairly po-mo film Bassline Baseline [link via MT]. (It’s not clear how ironic, meta-ironic, or “post-ironic” this film is supposed to be, but it is both somewhat entertaining and somewhat informative — whether or not you’re familiar with electronic music clichés.)

One major downside to the TB-303 and its myriad software emulators is the fairly dismal step-based programming method. (At least this author, a putative “classically trained musician,” considers it to be dismal.) With step-based programming, you choose the “step” (generally corresponding to a 16th-note) and set the pitch, the octave, and whether this note is a rest or not and whether it is accented, tied to the next, etc. It may be an OK interface on a hardware device, but it is inexcusable in software. (Given my preference for the convenience and reliability of software and given the absurd expense of vintage hardware, I don’t own a real TB-303; hence my complaint.)

I’ve decided I’m going to try and incorporate more 303 sounds into some new compositions, so I designed a nice little grid to make programming patterns slightly less onerous:

This setup lets me notate pitches in a familiar “piano roll” view, with a line extending to the next cell for portamento or ties. The “up” and “down” boxes provide a means to indicate octave transposition, and the boxes along the upper left of the pattern provide space to record pattern-specific metadata. A pattern in this form is simple to visualize, play on a keyboard, or sight-sing, but is also straightforward to translate to the step-programming interface. Finally, six of these little grids, sized to allow reasonable markup, fit comfortably on a sheet of US Letter paper, along with room for some additional metadata (song title, etc). If such a sheet sounds useful to you, feel free to download a CC-licensed PDF (15kb), suitable for printing (or framing).

Technorati Tags: 303

I’m a little late to the train on this one, but Google Talk (Google’s “voice chat” service) supports iChat natively. Nice. I guess if you aren’t going to release a Mac client for your internet service, the next best option is to make your service work with an application that every Mac user already has.

NB: I’m happy to offer gmail invitations — which are required for Google Talk — to the intersection of my readership and the set of all people worldwide who does not yet have a gmail account. (Since the cardinality of the latter set is certainly quite small and probably even smaller than that of the former, I am not liable to run out of invites.)

Actually, the “make your service work with what users already have approach” is almost certainly preferable to the “give them a new application to make the service work approach”. Think about how much nonsense we deal with because some application (or technique, or consumer electronics device) only does 70% of what we want and full capability is only available via a hacked-together patchwork of tools?

I say “hacked-together” because our digital tools (unlike their physical equivalents) are rarely designed to cooperate; each was designed by someone who believes that it is providing a 100% solution. (Some tools, notably software tools in the Unix world, have been designed to do one thing well and play nicely with other tools, but this principle is less frequently observed in the desktop-application and consumer-electronics markets.) As a result, a “solution” to the entire set of needs, which has been cobbled together not from capable primitives but from one or more insufficient “solutions,” is liable to be inconvenient, incapable, or both.

I’ve noticed the hacking-together-poorly problem to be particularly pronounced in the home-theatre arena. Think simply of input switching for a system; imagine you have a satellite or cable set-top box, a DVD player, a VCR, and a couple of video game consoles. (This isn’t a particularly esoteric setup, I believe.) In all probability, you’ll have several different types of video outputs (component, s-video, composite, and perhaps even RF) and several types of audio outputs (analog RCA, digital coax or optical). You will want a simple means to timeshare your television and speakers between these devices, and you will want the process of switching between them to be as painless as possible.

Unfortunately, “as painless as possible” is a foolish dream. The receiver will offer some switching mechanism and the television monitor will as well — but it is almost certain that neither will be adequate to solve the whole problem. The television will likely not handle the digital audio formats; the receiver will either not have a sufficient number of inputs or it will not have a sufficient number of the right kinds of inputs. What to do? At this point, you have to add a third-party switch. So now, you have to keep track of three states when you want to switch: the input selectors on the television, the receiver, and the switch. The switch also probably won’t handle your digital signals (unless you have a really expensive switch!), so you will either have to use multiple digital inputs on your receiver (flipping separate buttons to change audio and video), or resign yourself to being out of luck and reverting to 2-channel analog.

re: “really expensive”: I’d be happy to be wrong about this, but it appears switches that handle digital audio signals are expensive enough that one should probably consider purchasing a new receiver with more inputs before purchasing such a switch

The question is, why is this the case? Of course, there are two possible answers: money and money. But how expensive can it be to add another set of S-Video, optical, or coax inputs to the back of a receiver? Depending on how well the complexity of the switching circuit scales with additional inputs, I can’t imagine that the cost would be that much greater than the cost of extra jacks. (Even in “finished product” configurations, those are cheap.)

Sure, the commodity consumer-electronics market is cutthroat and cost-driven, but is there really no demand for a receiver (or television) with a sane number of inputs? Where is the outrage? Most importantly, why do people put up with manufacturers providing 70% of the functionality that reasonable consumers might need without providing any means either to expand the capabilities of a device or to enable a device to cooperate with others?

What one really needs in this situation is a modular stereo. You connect the inputs, switches, outputs, surround decoder, and power amp, adding and removing components and I/O as you need them. There’s a solution for home-theatre — I’ll leave tackling the problems of software as an exercise to the reader.

Dabblers and Blowhards is a great dissection of the sort of narcissistic twaddle that Eric Raymond brought to netwide prominence. (Dabblers and Blowhards focuses on Paul Graham’s auto-idolatrous shenanigans, but mentions Raymond’s oeuvre as well.)

Oleg Paraschenko hacked up find(1) to support XPath queries. And he wrote some of it in Scheme!