Not an Asus fan

Dear Asus,

Please put less shitty chipset fans on your rather expensive motherboards. It’d be nice if these things lasted more than about a year.

/signed/ a customer who was replacing a shrieking fan at 5AM this morning, bodging in a 40mm fan in place of your 38mm fan made of unobtanium, using hot glue, tin snips, something to melt plastic, and a liberal supply of invective

Posted in Uncategorized | 2 Comments

It’s called fdisk because…

In days of yore —

“When was Yore, Daddy?”

“It was the age before we had terabyte hard drives. Really, it was before we had any hard drives at all.”

“Is that old, Daddy?”

“You bet your sweet bippy, it was.”

— when rocks were young and you could count the number of megabytes on your computer on one of your hands — honestly, you didn’t even need all of a hand, or even a whole finger — I was chatting with some friends at school about how long a disk copy took.

“Just a few seconds, right?” I was a little starry-eyed. Okay, utterly naïve.

“Oh no. I’ve heard it takes a couple of minutes.”

That was unbelievable to me. Disks (eight inch floppy disks in those days, if you must have the truth) were supposed to be fast. Definitely faster than the audio cassette tape I was planning to store my programs on, when I got my own computer, which would be soon.

“I don’t believe you,” I said.

And we tromped off to the computer store, where they had a microcomputer named SWTPC, and we implored the owner of the computer store (Poor Richard’s Calculator Shoppe, in a town in northern Colorado and yes we had paved streets) to copy a disk. And since it was a slow day and nobody [in their right mind] was buying SWTPCs, Richard took pity upon our curiosity and loaded up two disks and did a copy, and lo, we saw it copy a disk in four minutes, accompanied by a crapload of seeking and other grinding noises (because this was a SWTPC, and they had issues).

I was pretty disappointed. Not so disappointed that I didn’t want a pair of floppy disks for my own computer, which I would have soon, somehow, but that interminable four minutes definitely knocked floppies off of the performance pedestal I’d put them on.

“What’s a pedestal, Daddy?”

“A pre-prepared disappointment.”

“What?”

“It’s like you expected Band class to be tons of fun, with everyone jamming away on their instruments and the Band teacher grooving on the podium with his baton, but what was it really like?”

“Awful blatting noises.”

“Yeah. None of you could play a note in tune, much less play anything together, and you’ve gotta practice scales and simple tunes that are really boring, and I happen to know that the band teacher is borderline suicidal, not from any specific knowledge but because they all are, teaching fifth graders to squawk and go hornck! in close synchrony will drive anyone to the brink. See? Onward.”

Let’s go forward a few decades and visit the scene of the actual rant now.

Problem: Wife has a laptop with an SSD that is nearly full.

No problem! I have a spare SSD that is larger. We’ll just clone that disk. Shouldn’t take more than a couple of hours. It’ll be glorious.

Three weekends later I have admitted defeat. The dragons have done me in. I have written file systems and that shit is hard, but copying the sectors containing a file system? School kids do that. But what I have encountered is serial madness. This is one of those problems that is not supposed to be hard, but somehow it is. When we were not being vigilant, while we were playing in the sunny afternoon chasing butterflies and filling-up terabytes without a worry, complexity was doing push-ups.

Tool A claims, no matter how I instruct it, that the file system on the source drive wasn’t shut down properly. There’s an option to have it not do this check, but it’s a damned lie and the tool keeps checking and failing anyway. It has another mode, a “raw” mode, that appears to work, but after several hours of blinking activity lights the clone does not boot.

Tool B — which I paid dear money for, and which has an excellent reputation in the disk copying industry — makes a clone without any fuss. In fact, the tool sports a fancy user interface and makes important-looking animations and pronouncements during the copy. It does everything but strut. I am impressed, this is quality stuff and I’m happy to have spent the money on this quality tool. Naturally, the copy does not boot.

Tool C, provided by the drive manufacturer, makes some sincere promises, but the FAQ has some scary looking workarounds for things that shouldn’t need workarounds. The copy it makes doesn’t get high enough to crash.

Tool D says that the copy will finish in 24 days. Also, its UI makes my eyes bleed.

My malware scanner vaporizes tool E milliseconds after it is downloaded. Memento mori, I guess.

Well, there’s always booting into Linux and doing the old DD if=something of=something bs=somethingbig and then maybe hex-edit the partition tables after it runs off the end of the source drive. This is my wife’s computer and she’s worth it. For kicks, I look at what partition tables have mutated into. They were a miserable hack back in the early 80s, and time has not improved them. I wince and close that web page.

I consider writing something. How hard can it —

I do a clean install of the OS on the new drive, hand the laptop back to her and say, “Sorry, nothing worked, you’ll have to reinstall all that stuff. Here’s the old drive, you can copy your old files off of it.”

It’s really for the best.

“But Daddy, why didn’t you just edit the GPT entries to–“

“Have you practiced your instrument yet today?

Posted in Uncategorized | 19 Comments

Oh, yes they can

Oracle, oh Oracle
Oh have you seen Oracle
Oracle’s EULA lately?

Don’t reverse that buggy code
Lest you make poor Maddi ‘splode…

(… you can learn a lot, from Oracle!)

— after Groucho Marx

link to discussion on HN

 

Posted in Uncategorized | Leave a comment

Fresh Oz

New Ozric Tentacles! (yay)

 

Posted in Uncategorized | 3 Comments

Newton Power

I stopped at a surplus computer hardware store today and found a bin with a couple hundred of these:

newtonBatteryHolder

So, if your Newton MessagePad 100 needs a battery holder, I know where you can go (RE-PC in Tukwila, WA). I have no idea how these things wound up there.

Gaston Bastiens was the General Manager of Newton when I was there. He was not well liked, and he kept making crazy decisions. For instance, he decided it would be good idea to charge Newton developers a percentage of their profits (unheard of at the time, and lunacy for a struggling platform). And: When it was clear that the Newton wasn’t selling all that well, and unwanted Newts were stacking up in warehouses, Gaston had Apple buy parts to make another 80,000 units. I won’t even get started about the sea of T-shirts we were swimming in (there’s a big difference between “Buy $5000 more promotional shirts” and “Buy 5,000 more shirts”).

There were a lot of Newtons sitting in warehouses.

Also, many Newton Fax modems. Many, many of them. God only knows how many he ordered. There could well have been more Newton Fax modems than there were Newtons to plug them into.

One of our developer support folks (Bob E) found out that it was possible to order a palette of surplus Apple products delivered to your office. Hey, the stuff had no hope of being sold, so if someone had a use for it . . . Bob ordered a big palette of Fax modems, and when he shipped stuff off to developers (hardware, manuals, etc.) he would pour in some Fax modems as packing material.

—-

We had a dev who liked to do anagrams. He maintained (somewhat tongue in cheek) that anagrams uncovered the true nature of things.

For myself, for instance: Dandy Loner

For Gaston: Neat Bong Assist (which is one of the reasons that a Newton easter egg is to write “Neat Bong”, select that text and hit the Assist button).

For one of our whip-cracking dev managers, a two parter: “Donate us a gun” (because) “Us at a dungeon“. The year we spent shipping Newton was a hard, hard year . . .

 

Posted in Uncategorized | 3 Comments

On Reviews

My first job performance reviews, first as a dishwasher and borderline incompetent kitchen robot for a minor restaurant chain, and then writing video games for Atari, were short and sweet. Every year or so I would sit down in my manager’s office and he’d say, “We think you’re doing a great job. Here’s a pay raise.” That was it. It only took a couple of minutes. We didn’t have a long talk about objectives, or chat about what had gone right or wrong, or make a strategy for the next year.  There were no head games or bullshit politics. Just: Everything’s fine, no problems. Here, have some money. I still consider this to be a state of innocence, and possibly an ideal. It was stress-free and I had no worries. I had a similar experience working as an intern for the federal government: Every once in a while they’d increment my GS rating and I’d get a little more money (not much, it was the government).

At Apple I was introduced to the torture of writing a self review.  This was an genius application of laziness on the part of management: Each employee had to write down what they had done, be honest about their accomplishments and failures, and hope for the best. Apple actually seemed to care what you wrote, but in the end it was all classic backward-looking stuff. “You did good, no problems. Here, have some money.”  Like washing dishes or writing video games or being a very junior government drone, there was nothing about objectives or future plans. (This was in the 80s and 90s, and I don’t know what reviews at Apple were like under return-of-Steve or if they changed in the post-Steve era. It’s possible they use chemical interrogation now. I could totally see that).

At Microsoft I sweated through years of similar self-reviews before learning the truth: By the time you were done with all the self-purging and 3AM panic attacks and had finally excreted the perfect gem of a critical but not /too/ critical self review, well, by that time all the numbers had been decided a month earlier in secret meetings and whatever was coming your way, candy or stick, was already a done deal. Short of your running nekkid through the lobby or installing Linux on your workstation, your review wasn’t going to change. You could write pretty much anything (“All work and no play make this developer a dull boy. All work and…”) and the only time it mattered was when you went looking for a new position inside the company and your new management got curious about the kind of drivel you’d written about yourself. The self review itself was “post-hoc, ergo propter hoc” bullshit, a substrate for your manager to justify what the secret meetings had decided. And at its core, the process was about humiliation and sowing self-doubt.

I went through three or four revisions of the performance system at Microsoft, and it became clear that the changes didn’t really matter much. The only invariant that counted was that unless you had a good manager, a real fighter ready to lay down honor and do personal combat for his or her direct reports, you got screwed. The tweaks to the review system just altered the details of how you got screwed, or how you came under special screwtiny [sic], and did nothing to fix the underlying screwedness of the review system’s philosophical base.  The road to success at Microsoft was to put yourself under a champion.

It gets more fucked up than that, but you’ve already heard most of it from various articles on Slate, et al, and the threads on Mini Microsoft. There are truths buried in the Mini Microsoft comments, though they are submerged by rants, name-calling and taunting.

I quit Microsoft over two years ago, and it took a whole year to get some perspective (I wrote a lot of this soon after quitting, and I’m quite happy I never published it; many of the paragraphs simply did a crescendo into incoherent ASCII screams of frustration and anger). I think that many of Microsoft’s technical failures in the last decade can be root caused in a review system that rewarded bad behavior, put the wrong people in positions of power, mis-identified the people that Microsoft should have kicked out, and caused the wrong people to get sick of things and leave. Maybe the new review system does the job; I keep hearing good things.

Posted in Uncategorized | 7 Comments

Zap

Holy crap.

Hola-crap.

I think . . . they rather missed the point.

I’m surprised that only 14 percent of Zappos employees quit. Just wow. Rotating Christ on ball-bearing roller skates wow. This constitution thing is kind of awesome, in its own sickening and twisted way. I’ll bet it’s about as much fun to work in this structure as it is to get a root canal in North Korea.

Posted in Uncategorized | 4 Comments

The New Corporate Campus Curse

Atari. 3COM. Apple (1992). Borland. Silicon Graphics. Sun (twice). Others that I have forgotten.

Facebook?

We’ll see.

Personally, I’d rather work in a barn.

Posted in Uncategorized | 6 Comments

Dear Oracle

Dear Oracle: If you ever install a directory whose name contains an ‘&’ in my system’s PATH environment variable again, I will travel directly to the Bay Area and it won’t be for fun. What were you thinking, that people wouldn’t notice? That it looked pretty? That writing out A-N-D was too hard? Are your directory names chosen by Marketing? (“Can you make that directory a little more puce? No, way too puce now. More, you know, edgy and confident puce, not blatant developer puce.”)

Of course, it would be nice if the Windows command interpreter wasn’t such an awful embarrassment and could do things like string quote escaping without losing its mind. I heard some of the political backstory to why cmd has been left to languish for the past two decades and was not surprised (summary: old and crusty NT dev, with matching territorialism and ego). Oh, and the answer to “cmd is crappy” is not PowerShell, thanks. Who said that? We can go outside and settle it, if you want.

Not really sure what the answer is. For instance, every time I write a bit of bash I want to throw up a little. The only shell programming language I didn’t actively hate was the MPW Shell, and that is decades dead now; MPW sure had some flaws, but parts of it were really nice.

But those ‘&’s don’t work anywhere, not matter how fucking puce they are.

Posted in Uncategorized | 13 Comments

m-x start-them-early

Since it was a national holiday, I took The Gibber into work for the day. I sat him down at a desk near me with a laptop and got him on the guest network. He played some MineCraft, and after a while I suggested he work on his book report a little.

“Here, use this,” I said, launching a program. “First, you can just type stuff and it appears. Then, control F goes Forward, control B goes Back, and control N and control P go the next and the previous lines. You can use the arrow keys, if you want.”

He typed some stuff and tried it out.

“Okay, now Control U and a number does the next thing you type that many times. Type Control U and thirty and the letter A.”

AAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

“Okay, Control U by itself just does four. Another control U multiplies that by four. And another.”

BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB

“Works for the commands that move around, too. Try it out.”

BBBBBBBBBBBBBBBB|BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB

“And the Meta key (and now you know what that ESC key is for) makes things go _bigger_, so you can jump around by words and paragraphs…”

—-

I show him C-X C-S and C-X C-C, and a little more, and five minutes later he’s happily typing away.

A cow-orker comes by and asks what I’m doing.

“Teaching him Emacs.”

“What?”

“No son of mine will use vi.”

“Oh.”

Posted in Uncategorized | 5 Comments