Blackcap Blog

Published at 17:35 on 23 July 2017

Some years ago, I was renting a room in a house. In the utility room was a box of a product called “Dirtex,” which could be used for, amongst other things, a floor cleaner. Because I needed to clean a floor that day, and it was handy, I tried it. It worked wonderfully.

A glimpse at the ingredients showed that it was mostly “sodium sesquicarbonate,” a compound new to me at the time. It’s basically a double salt of sodium carbonate (a.k.a. washing soda) and sodium bicarbonate (a.k.a. baking soda).

I couldn’t find that product after I moved to Bainbridge Island, but the grocery store here sells both washing soda and baking soda, and I have a gram scale. So it was a simple matter to weigh out 286 grams of washing soda (being a decahydrate, it has a high molecular weight) and 84 grams of baking soda, and mix the two.

No, that’s not making true sodium sesquicarbonate unless I dissolve and recrystallize the result, but given that I’m just going to be dissolving it in a bucket of mop water, it makes no difference to the resulting solution. And yes, I’m sure there’s a little bit of variation as the powders separate and settle, but mopping the floor isn’t a precision science. It works well enough.

Which, to the best I can recollect, is about as well as the commercial product worked, which in turn is quite well indeed.

Published at 11:47 on 28 June 2017

A few months ago I had the opportunity to use an iPhone. Unbelievably, the thing took eight keystrokes to simply hang up an in-progress call. Eight! I am not making this up:

1. After 30 seconds or so of idleness, the phone locks itself due to security measures. (The phone for some reason considers itself to be “idle” even though it is actively in use for a call at the time.)
2. Given virtually all calls last longer than 30 seconds, that means you must first get the attention of the now-locked iPhone. Press the one and only actual mechanical button offering tactile feedback the device has (1 keystroke, 1 in total).
3. It is now time to enter the unlock code for the phone (4 additional keystrokes, 5 in total, and counting).
4. Despite the device being a phone, and a phone call being actively in place, for some reason you are now in the phone’s default mode, which has nothing to do with making or managing telephone calls. Tap the icon that puts the phone in phone mode (1 additional keystroke, 6 in total, and counting).
5. Despite there being a phone call actively in place, when you enter phone mode you are placed in the mode where you can make an additional call, not for managing the existing in-progress call. You must manually select the current call (1 additional keystroke, 7 in total, and counting).
6. You are finally now presented with the desired icon to click on that will end the call. Click on it (1 final keystroke, grand total of 8).
7. Congratulations! You have at long last managed to hang up.

By the time that’s all done, odds are at least 50-50 the other party has long since hung up already and the call has timed out before you could hang it up.

Why would I want to have a device that packs so many non-phone duties into itself, and implements its total set of duties so poorly, that using it for its primary intended purpose is then severely compromised? The nearly 40-year-old 2500DM set on my desk never has firmware to update, will never radically and unexpectedly change its user interface, and has a set of hook switch buttons that are always there waiting for me to use them on a moment’s notice whenever I want to hang up on a call. Even the cheapest flip phone has an END button that’s always there waiting for me to use it. Neither phone decides in the midst of an in-progress call of all things that it’s “idle” and now needs a password to be unlocked.

The killer came when I realized that this is an iPhone, and Apple has a well-deserved reputation for the best-designed system software. That is how the best smartphone on the market implements its user interface. The other smartphones are almost certainly worse.

Published at 14:36 on 9 March 2017

Introduction

I’m working on my on-again, off-again digital clock project, and one thing I learned before I broke off before is that the number and density of connections made soldering very impractical. So wire-wrapping it was. Some basic points:

1. Wire-wrapping is spendy. Just getting a batch of sockets, a spool of wire, and an entry-level tool cost me around $100.
2. Because of that, I can’t consider it worthwhile for less-ambitious (read: less complex) projects.
3. You’re on your own, basically. There’s no instructions with the tool I ordered. I presume that’s typical for most tools.
4. There’s an incredible variety of tools and wires out there.

So, with all that, here’s what I’ve learned so far.

Get a Less-Expensive Tool

Unless you’re really going to get into wire-wrapping, don’t get one of those gun tools. The base price is spendy enough, and they’re even more spendy than that! The gun tools (both electric and hand-powered) require both a bit and a collar, and both of those items are costly as well. Expect to spend $200 to $300 or more just to get a tool capable of making connections. Ouch!

Instead, opt for a “screwdriver” type tool. I’d recommend getting one made by a well-known name brand, because you are going to be making hundreds of connections on the typical project, given the size and complexity needed to make wire-wrapping make sense.

Use 30 Gauge Wire

Specifically, 30-gauge, Kynar-insulated, silver-plated wire. It’s basically the standard. (Just to make it interesting, they sometimes call Kynar by other names: PVDF, polyvinylidene fluoride, or polyvinylidene difluoride.)

Get the Right Tool

The tool you need is governed by both the wire gauge and the style of wrapping you’ll do. You already know the gauge; the wrapping style I’d recommend is the “modified” style that involves wrapping about a turn of insulated wire around the terminal posts before the bare wire starts. This provides a degree of strain relief, which minimizes the chance of things breaking while you wire your project.

Taken together, the right tool is the Jonard WSU-30M. It’s not much larger than a small screwdriver and costs over $30. Cheaper than the guns by far, but still, ouch! It’s a name-brand tool, and it’s well-known enough that there’s (incomplete, but still better than no) instructions for using it on the web.

An added plus is that the WSU-30M can remove connections as well as make them, so you don’t have to order a separate removing tool. The latter is a must, as Murphy’s Law says you will sooner or later make a wrong connection.

How to Use the Tool

There’s some instructions here. Alas, they’re missing one of the most important things: how much wire to use to account for wrapping and slack between two terminals. After some frustration, I arrived at the distance plus 7 cm (sorry, I’m not going to convert that to inches; I prefer working in metric because the math is easier).

Published at 22:00 on 29 December 2016

This. (It’s not the best-designed web site. Click on the three dot-dash symbols at upper left if it seems to end mid-work without showing the whole thing.)

Yes, Derrick Jensen is something of an asshole (check out some of his rants about anarchists and transgendered people if you don’t believe me). No, I don’t buy his claim that the only viable alternative to civilization is stone-age tribalism.

But, that said, the guy (and his co-authors) does have some valid promises about this civilization being so destructive that it must be ended, the sooner the better, in part because the official mechanisms of power are pretty much useless for the purpose of arresting and reversing the destruction.

Update: Note that this endorsement of the book with the title Deep Green Resistance is not an endorsement of the organization by the same name. The latter appears to be dominated by a cult of personality around Derrick Jensen as much as the Revolutionary Communist Party is by a cult of personality around Bob Avakian.

Published at 10:26 on 19 October 2016

Taleo is a software-as-a-service (SAAS) package that some business’s personal departments use. As the title of this post implies, it sucks. I’m hardly alone in having this opinion, either. Just type “Taleo sucks” into your search engine and see.

I used to put up with its suckiness (clunky menus, duplicated data, bad browser compatibility, excessive use of crap Javascript, inability to view job description and application form at the same time, etc.) in the name of doing a more diligent job search. No more; if a link (or redirection) to Taleo happens during an application process, it’s game over.

I’m hardly alone in having this policy, either; if you read some of the hits you got in your search engine exercise above, you’ll find that others act as I do.

What tipped the balance for me was the realization, at the start of this current job search, that I have never received as much as a preliminary phone screen from any of the dozens of firms I put up with Taleo to apply for jobs at. That’s right, never. Not once. It’s as if my data vanish into a black hole.

My theory is that Taleo sucks not only for the applicant, but also for the person on the other end. Why shouldn’t it? Bad design is generally not confined to just one or two places in a software system; if it exists, it tends to be pervasive. As such, personnel departments also generally avoid using it. Thus, the way to land a job at a Taleo-using company is via some other channel.

But why should I? By the virtue of choosing Taleo, they’ve demonstrated their organizational incompetence by choosing to waste money on a demonstrably bad product. And I have no interest in working for incompetent organizations.

Published at 10:54 on 23 July 2016

I’ve become involved in a discussion about self-driving cars on an Internet forum, and the expected techno-utopian who thinks they will be a sea change which obsoletes mass transit has popped up. To summarize a past post, no, they won’t:

• They will increase the capacity of existing freeways, but that’s nothing that building or expanding freeways doesn’t currently do, and such increased capacity falls victim to the induced demand problem. There is absolutely no reason to expect that the extra capacity furnished by self-driving cars will magically function differently than that furnished by more traditional means.
• It’s actually worse than the above. By automating driving, self-driving cars will make driving easier. When you make something easier, the natural expectation is that people will do more of it. In other words, self-driving cars will themselves be a significant new inducer of demand for road space.
• Some of the worst congestion is found on surface streets in commercial areas, where there isn’t much (or any) room for more cars on the existing streets. Self-driving cars will do nothing to alter this fact; automation does not repeal the basic laws of geometry. Quite the contrary; by increasing the capacity of freeways to deliver vehicles to such areas, they will exacerbate this congestion.

In fact, there’s a good chance that self-driving technology will make mass transit more relevant:

• As demonstrated above, it will make congestion tend to get worse, thus increasing the need for alternatives to personal vehicles.
• One of the largest costs for transit agencies (in fact, it’s typically the largest cost) is the labor cost for operating the transit vehicles. Self-driving buses will therefore enable significant savings for such agencies.
• Labor costs are a big part of the reason why only big buses and not minibuses are used (the latter save very little money for agencies, because labor costs and not equipment or fuel costs dominate). By enabling minibuses to economically replace larger ones, self-driving buses will enable better, more frequent service to low-density areas that are difficult to serve well with traditional mass transit.
• Self-driving cars will further help the ability of transit to serve low-density suburbia by making it easy for riders to get to train stations without requiring huge, expensive park-and-ride garages (which tend to fill up early). Riders can send the car back home after it drops them off; this won’t exacerbate congestion because travel will happen in the opposite direction, and residential streets (unlike freeways and commercial streets) typically don’t suffer congestion issues anyhow.

On top of all that, there’s energy efficiency and the need to reduce fossil fuel use. Self-driving cars still require several tons of steel to often carry just a single person around. The general wastefulness of the personal automobile remains unchanged.