El cheapo keyboards and their rubber hills

Regular cheap keyboards (and most expensive ones) use a dome-type switch under every key — it’s a hill-shaped rubber thingie that, when you press, activates a sensor below, and when you release, pushes the key back.

They suck.

Model Ms and their clickety-clickety switches

Almost all Model Ms, in contrast, use mechanical-type switches with springs (yes, there’s a spring under each key). This is the way they work:

1. When you press a key, the spring underneath it opposes an increasing amount of resistance until it sort of “gives way” (the spring actually “breaks” halfway) — and when it does, it stops resisting, clicks loudly and sends the keypress to the computer.
2. When you release the key, the spring recovers its original form, and the key bounces back, with another click.

Many people despise Model Ms precisely for the noise and the springs’ resistance. I’m sure these people aren’t typists or heavy keyboard users — I usually refer to them as ignorants — because those characteristics are exactly why you want a Model M.

And here’s why you want one: faster typing speed, more accuracy, and less frustration.

Model Ms are more accurate

Accuracy determines how fast you type, and how frustrated you get in front of the computer. Make many mistakes a day, you’ll hate your job. Type fast with no errors, you’ll be happy.

When a key on a Model M clicks, you can be 100% sure that the key was registered. If a key didn’t click (because you pressed it too lightly), then there’s a 0% chance the key registers. Dome-type keyboards aren’t that way — there’s no click, so you can’t be sure without watching the screen — and that makes your hands exert extra pressure just to be sure keypresses get through. Mind you, not that much extra pressure — you can’t consciously tell — but the effort builds up and wears your arm muscles.

Give it a shot, if you have a cheap keyboard. Press a key lightly — lightly enough that the key “gives way” but not strongly enough for the computer to detect. You’ll see this is not only possible, but a rather frequent event. That kind of half-ass keypress isn’t possible in a Model M. Either it clicks and it works, or it doesn’t. Ever.

The feedback is more than just click-click — you actually feel it under your finger

Plus, you not only get very clear audible feedback. As stated before, the springs actually give way when the key is activated, so you can tell beyond a shadow of doubt when you activated a key, even using headphones with music at 120dB, because you feel the spring give way, and that is an unequivocal signal that your keypress has gone through. You don’t need to double-check on your screen, ever.

And that feedback is key to computing without frustration

These characteristics of mechanical-type keyboards like the Model M not only (dramatically) reduce typing errors, but they let you determine without even glancing at your screen when you’ve made a common mistake, then quickly backspace over it. This makes a brutal difference if you’re into activities that require uninterrupted concentration or simultaneously reading and typing massive amounts of text. Let me give you just one example why:

Have you accidentally hit two keys simultaneously with one fingertip? I don’t know about you, but it happens to me occasionally — especially when I move my hand back from the mouse.

With dome-type keyboards, you need to check the screen to see if one or two letters were detected (because of the half-ass keypress problem I explained above). Checking the screen before typing again sucks; yes, I know it’s only a couple seconds — however, multiply that by hundreds of mistakes a day, and you get an idea of how frustrating it gets. And every time you do it, you lose your laser-sharp focus on what you were doing.

With a Model M, you can instantly tell and correct that type of mistake — because of the tactile/audible response combo — and then you just hit the Backspace key the appropriate number of times. Two clicks/bumps in your finger, one backspace; Four clicks/bumps, two backspaces. No need to check the screen and interrupt your flow — be it dictation, thoughts or computer programming — just backspace one or two times and continue typing.

And, boy, do they last!

As for their durability… I bought my Model M used from a bank in a used-equipment sale. That keyboard had been used for a decade since 1980 — at a bank terminal on a mainframe-based character terminal system. You know, bank teller monkeys pounding on its keys all day long, nine-to-six, nonstop, for ten years. And it still works like a charm. How’s that for durability?

Test one!

So, if you’d rather spend $80 unstressing on Friday nights and $3 on your keyboard, be my guest. But if you’re into doing your job quickly while avoiding stress in the first place, you should give one Model M a spin for a week (or, if you can’t find any, perhaps Das Keyboard is your next-best bet), then tell me if you went back to your old keyboard. You won’t regret it.

Copying files and using the computer should not be painful rocket science

You see, modern operating systems don’t just let programs blindly use the disk without arbitrating who gets to go first. Linux is no exception — the disk I/O scheduler attempts to balance even request distribution between process against maximizing throughput (the actual amount of data delivered).

We in the desktop camp couldn’t care less about throughput, of course. OK, we do care a bit about it, but if we’re copying a 40 GB file from disk A to disk B, we sure as hell will want to use our computer productively while the copy is going on, don’t we?

Linux CPU scheduler: actually kinda nice

While the Linux CPU scheduler does a somewhat non-pathological (you could even say a good) job, the disk scheduler sucks SMELLY, MUSKY NUTS.

I/O schedulers… I demand a refund!

I learned to tolerate the slow application startup times in my last computer (a not-too-slow 768 MB + 320 GB disks Athlon XP machine). I guess “userspace is pathological” like Ingo (or Con) said. But I could never understand the two big psych breakdowns my machine experienced frequently:

Big file copies

Copying a big 2 GB file between two disks is enough to throw the desktop into an unresponsiveness fit that lasts for a good few seconds after the file has been finished copying, makes mouse pointers jumpy, and task switching impossible (you can switch among tasks, however you won’t see anything but a grey rectangle while the system is busy attempting to swap a few pages in from disk into the application’s physical memory space, lagging behind hundreds of requests in the disk request queue).

Backups!

rsync processes: don’t get me started please. Every time I run a backup, it trashes the disk caches and pushes applications to swap. I don’t know exactly how, but after running an rsync of my whole /home partition, the system stays in a state that could be accurately described as “between stupid and lazy”. Actually attempting to start an application while the rsync process is reading from disk… why bother — Akregator takes over a minute just to show up!

Why I classify them as pathological cases

Both cases almost unversally exhibit the same symptom, which I’m going to term “the grey rectangle of death” — applications don’t redraw their screens until each and every essential code/data page has been loaded from swap; and that, my friends, can take almost 30 seconds for some applications.

These two pathological cases unfortunately find themselves replicated in almost every modern Linux desktop. As far as I recall, the disk IO scheduler works in fixed priority rings — applications that hog the disk don’t get demoted in priority, and in fact I think they get an interactivity bonus for constantly blocking on I/O. (At least) a desktop machine should never do this — it should make requests for interactive applications (defined as “applications that haven’t caused a page fault or disk access in a long time”) literally jump the queue in front of disk hogs. Yes, that would kill server throughput — but we’re talking about a desktop machine, aren’t we?

ionice as the solution?

Get real, dude. ionice is feasible for an automated backup procedure, but I don’t want to type a UNIX command every time I drag and drop an episode of House M.D. to file in the right folder. That’s a job for the computer. And, ideally, a job for the disk scheduler, not an external program that reacts slowly using out-of-date information.

Give me responsiveness, goddamnit. I don’t care if the file copy or backup drops to 20 KB/s, so long as Konqueror starts in under two seconds. Both the file copy and the backup will finish on their own time.

My quick fix

Here is it. Should make your machine more responsive. Getting a new machine may also help.

I still maintain I shouldn’t have had to get a new computer in order to solve mitigate disk starvation issues. And yet, I did:

Good news for me — I got myself a PowerEdge 1U server with 2 400 GB SATA disks, RAID1ed, as a desktop machine. Dual Core x64 and all the bomb. It was kind of hard to set up using a KUbuntu Live CD (no LVM2 or mdadm tools included, and please note that all my filesystems are in LVM volumes now).

But it screams. It’s brutally fast. Minus, of course, audio, decent video, DVD, and TV capture card, all handily virtualized through my home network for the time being. Thanks, PulseAudio!

Finally, OpenOffice warm-starts in under three seconds. Before that, I dreaded receiving a PowerPoint presentation — and now I find them tolerable! And multi-gig file copies/rsyncs have become a bit faster.

OK, I’ll stop with the cognitive dissonance and just say: I guess it sucks to be not me

This is a screenshot of me watching TV. You’ll notice that a faint overlay shows up above the TV. This overlay is the software installer displaying installation progress. Thanks to this, I don’t have to interrupt neither work not pleasure to monitor long-running tasks:

How did I do this? I just dragged the installer window above the TV window, then I depressed the Alt key on my keyboard and pulled the mouse wheel down a few notches. Instant see-through window!

Even on an old PC (graphics card: generic GeForce 2 MX), Beryl can hold its own. This video was shot at 640×480 because the CPU could not capture video fast enough at higher resolutions, but the same experiment runs fluidly at 1024×768 when not capturing video.

Check the window miniatures. Perfect for monitoring a program (say, the Firefox download manager) on a different side of the desktop cube!

The most amazing thing you can do with Beryl is open a TV window (yes, it gets along fairly well with TVTime), maximize it, put your Web browser on top of it, and make the browser semitransparent. You can watch TV and browse the Web simultaneously, no fuss, no muss!

Oh, almost forgot: You can pimp up your Beryl at beryl-look.org! Tell your friends!

Believe it or not, Beryl works just fine in low-end computers. Though this video was captured at 640×480, Beryl does an average of 30 frames per second at 1024×768 on a computer with a 5-year old GeForce 2 MX, and is impeccably fluid at 800×600.

The window wobble plugin is disabled for a perceptible performance boost. The computer’s CPU is an AMD Athlon XP 2400+.

Яudd-O: dude relationships are easy
Яudd-O: it’s the fighting that’s hard
Яudd-O: I know that sounded sophistically easy
Яudd-O: but it’s the truth
Яudd-O: starting a relationship takes effort
Яudd-O: from then on, it all works itself out on its own
Яudd-O: why? cuz solving a problem is EASY if the two involved want to solve it
Яudd-O: it’s conversely HARD to solve if only one of them (or none) wants to solve it
Яudd-O: and if that’s the case, then why go on? it would be an exercise in futility
Яudd-O: the precondition to make a relationship survive is, thus, the WANT of two people to solve an impasse
Яudd-O: (tangentially to this conversation, there are certain things, in my opinion, that simply DON’T warrant even trying to solve, such as cheating for example)
Яudd-O: once there’s a mutual WILL, everything else falls into place
Яudd-O: so you see, the important part of a relationship is maintaining the WILL
Яudd-O: and how do you maintain the WILL? simply through CHALLENGE
nono, you misunderstood the concept of challenge
Яudd-O: challenge is all about making the other person want you, value you and miss you
Яudd-O: sure there’s nothing wrong with sharing, but in order for her to SHARE with you, she needs to WANT to SHARE with you first

So, there you have it. Feel free to believe this.

• I haven’t had a virus, worm or hacker in my computer.
• My computer hasn’t had a single unexplained crash or blue screen.
• I’ve never installed anti-virus software.
• The same Linux installation I’ve had I’ve maintained. All I ever do is upgrade periodically as new OS releases go gold, but I’ve never done a fresh re-install “because the computer just got slow”.
• I’ve never been asked by the computer to “register” with a company or “acquire” a “license” to operate it.

Of course, “boring computing” leaves me with plenty of time to:

• Develop software and earn money.
• Play.
• Listen to music all night long, thanks to the best ausic player ever invented: Amarok.
• Go out and visit friends and girls.

¿Y para qué necesito dos monitores? Pues para esta maravilla que están viendo. Ya la tengo instalada en mi computador y funciona tal y como se ve. Pero con dos monitores es mucho más vacán!

It’s joy.

¿Qué toca?

Pues hacer un megavolumen a partir de una partición de 20 GB en un disco viejo, y una partición de 120 GB en otro disco. ¡MD (Multiple Devices) de Linux al rescate! Uno los dos discos con mdadm, y formo un solo gran volumen: /dev/md0.

Hasta ahí, todo bien. ¿Pero qué pasaría si se me chorean la computadora? ¡Pues que tendrían mis respaldos!

Entonces toca cifrar los respaldos. Linux al rescate de nuevo, con lo último en criptografía transparente: LUKS. Creo un mapeo (mapping) LUKS cifrado con Twofish (uno de los algoritmos simétricos más seguros para cifrado) encima de /dev/md0, llamado backup. Ahora tengo un volumen cifrado /dev/mapper/backup.

(Ignoremos por un momento el hecho de que perder la computadora equivale a perder no sólo los respaldos, sino también los volúmenes en vivo que uso. También ignoremos el hecho de que la llave de seguridad para LUKS está almacenada en uno de los volúmenes en vivo.)

Finalmente, utilizo el poco conocido dirvish para hacer los respaldos. Dirvish es fantástico (pero medio dificilón de configurar): el primer respaldo es un respaldo completo, mientras que cualquier respaldo subsiguiente es un respaldo estilo incremental sobre el primero.

Automatizándolo con Vixie cron, todos los días a las 3 de la mañana, tengo respaldos diarios de los últimos 7 días. Además, cron me envía gentilmente el informe del respaldo por correo electrónico, todos los días.

Resultado: respaldos seguros, hechos a dos particiones diferentes en dos discos diferentes. La única cagada monumental es que si me chorean el computador, también se llevan la llave privada para el volumen de respaldo hehehe. Y los respaldos.

Para contrarrestar esta lamentable posibilidad, lo siguiente en mi shopping list es un computador viejo con unos 300GB de capacidad total en disco, conectado via WiFi a mi computador principal, pero escondido en el clóset de mi cuarto o en el techo. Así, si la policía viene a mi casa a llevarse mis preciados MP3, no los perderé.

Este es el informe diario de cron:

Starting backup disk
Assembling linear array
mdadm: /dev/md0 has been started with 2 drives.
Opening LUKS-encrypted volume
Command successful.
key slot 0 unlocked.
Mounting backup
Expiring old backups
Backing up
04:59:03 dirvish --vault root
05:16:50 dirvish --vault home
06:45:55 done
Stopping backup disk
Unmounting media
Stopping LUKS-encrypted volume
Disassembling linear array
Syncing disks

Yo, que tenía 64 Kb/s de ancho de banda, hoy tengo 200 Kb/s, ¡por el mismo precio! He recibido reportes de amigos con 128, que ahora tienen 400.

Es de notar que, en mi caso particular, he decidido desperdiciar 10 Kb/s y mantener el tráfico firmemente debajo de 190 Kb/s mediante Wonder Shaper, para mantener la velocidad de respuesta interactiva simultáneamente con grandes descargas de archivos.

¡Felicitaciones a SatNet! Así, mantendrán la fidelidad de los clientes al máximo.

Ah, cierto: si las descargas de BitTorrent van lento para Uds., puede ser que SatNet esté haciendo traffic shaping de BitTorrent. Consíganse un cliente BitTorrent que acepte conexiones cifradas, y su problema estará resuelto.

Sigh. Smart told me it has to download over 900MB of RPM packages to update my system. I’m on day 2, and it still hasn’t finished.

But there is a bright side to this story:

1. Smart is smart (pun intended). So far, smart has been the only automatic package manager (among those I tested in the last 2 years) that didn’t choke on my installed package base. In comparison, at least yum choked.
2. Smart is fast. It’s nearly as fast as apt. Forgetting for a moment that it balloons to more than 100MB of memory while resolving dependencies, it’s way faster than yum, which insists on connecting to the Internet and downloading a myriad small files.

Ya está!

Well, I just found out that MythTV is even more useful - beyond my wildest dreams. How? Simple, I always record Dawson’s creek reruns and watch them in the afternoon, when I have some leisure time. But sometimes that leisure time won’t be long enough to watch an entire episode…

…and that’s where MythTV’s playback speed function comes to the rescue. I’ve been watching the last week’s Dawson creek reruns at 1.3X the original speed. No chipmunk voices, no jerkiness in the video, just plain Dawson creek, but way more action-packed! After 1 minute of watching TV like that, the pace of the show feels completely normal again. Plus, no commercial breaks, so I get to watch every episode in about 25-30 minutes, and follow the story without interruptions.

The only thing I regret is, at that speed, the ballads from Dawson’s creek sound like country music and acoustic punk.