So we were all talking about Palm Pilots, and someone mentioned that they’re available dirt cheap off eBay.
Sigh. I didn’t need this, but it arrived today anyway.
Frankly, a lot of the IIIxe’s assets hold up well. One hardware button press and the relevant app is open and ready half a second later. My fingers remember how to write Graffifi. No notifications, except calendar event reminders I configured. I’ve used less productive devices all too recently.
I have a 32" monitor and a 27" monitor on my desk. Don’t ask. But although they’re different sizes, they both have the same native 3840x2160 resolutions. The 27" just has smaller pixels.
This is fine 99% of the time. When I’m working away with different windows open on each screen, all is peachy. It’s only slightly annoying that the menu bar on one screen is a tiny bit larger than on the other, but I’m tough. I can live with that.
What I can’t tolerate is that it’s impossible to exactly line up graphics across the 2 monitors. If I flick my mouse cursor from one to the other at the top of the screen – I have the physical screen tops aligned at the exact same height – it flies smoothly over the gap. If I try the experiment near the bottom, the cursor jumps in altitude as it crosses the border because “2000 pixels down from the top” is a farther distance in inches on a 32" monitor than on a 27" screen. And if I dare drag a window from one screen to the other, its position and size change as it crosses over. This will not do.
A friend recently nudged me to look at BetterDisplay Pro, and my life is better for it. It has exactly one feature I care about: the ability to enter a custom screen resolution, which I can then select from the normal Displays system setting. I did these things in order:
Wow, wow, wow. After many long months of mortal anguish, that annoyance is completely gone. My monitors play nicely together as I always wished they would.
Note: Yes, now there’s a slightly different nit, in that it slightly irks me that my resolution has a very strange non-integer scaling factor of 180:103. I don’t care. I can live with it. macOS still sends a 3840x2160 signal to the displays, and the pixels are so tiny that I can’t visibly tell it’s not running at native resolution. Of course, that non-integer scaling might slow the display down very slightly, but this is on an M1 Max system and why pay for the TFLOPS if you’re not going to use them? It’s totally worth the tradeoff.
I think I’m going to upgrade my personal MacBook Air to Sequoia tonight. YOLO!
What sadist invented this captcha?
I’ve been helping on a fun project with some incredibly brilliant friends. I found myself talking about it to a reporter at The Washington Post. The story just came out. My part was crucial, insightful, and far, far down the page:
Once known for distributing hacking tools and shaming software companies into improving their security, a famed group of technology activists is now working to develop a system that will allow the creation of messaging and social networking apps that won’t keep hold of users’ personal data. […] “It’s a new way of combining [technologies] to work together,” said Strauser, who is the lead security architect at a digital health company.
You bet I’m letting this go to my head.
At work: “Kirk, I think you’re wrong.” “Well, one of us was featured in WaPo, so we’ll just admit that I’m the expert here.”
At home: “Honey, can you take the trash out?” “People in The Washington Post can’t be expected to just…” “Take this out, ‘please’.”
But really, Veilid is incredibly neat and I’m awed by the people I’ve been lucky to work with. Check it out after the launch next week at DEF CON 31.
I use the ForkLift 3 file manager on my Mac. Part of my job involves copying-and-pasting the contents of various files into web forms. I made a trivial little shell script so ForkLift can help me:
#!/bin/sh
if [ ${#@} -ne 1 ]; then
echo "Expected exactly 1 filename."
exit -1
fi
pbcopy < $1
Then I created a new “Tool” called “Contents to Clipboard” that calls the script with the name of the selected file.
/Users/me/bin/copy_contents.sh $SOURCE_SELECTION_PATHS
Now I can select a file, select the Commands > Contents to Clipboard menu, and voila! The file’s contents are ready to be pasted into another app.
Good Project Managers are important. As an Application Programmer, they’re your Interface to the rest of the organization. They’re your API. Believe me, you want that interface between you and their managers, unless you like giving status updates and making projection reports to pass around.
In fairness to PMs, we engineers don’t make it easy for them to do their jobs. Does this sound familiar?
PM: How long do you think that’ll take?
Engineer: Oh, I don’t know.
PM: Can you give me an estimate?
Engineer: Uh, a while.
PM: OK. I’ll write “a while” on my report. I’m sure the CEO will love that.
Engineer: Fine. 2 months.
PM: Thanks!
Now, think of all the task management applications you’ve been asked to use. Each claims to be useful for engineers (“put all your requirements and future work in one place!”), but they’re also marketed to PMs (“be able to tell your manager how long the engineers are going to take!”).
Jira is one of these. It’s awful. It’s a code smell, and if you’re interviewing at a company that says “we track everything in Jira!”, ask a lot of follow-up questions to figure out if you genuinely want to work there.
Jira itself is… fine. In isolation, it’s neither wonderful nor terrible. It just is, like a rock, tree, or slime mold. Its main feature — and what makes it potentially evil — is its immense configurability that makes it irresistible to the wrong kinds of Project Managers. A good PM will see it, tweak a couple of knobs, and start putting stories in it for the engineers to hack on. A bad PM will see it, embrace every configuration option available, and saddle Engineering with a 23 step status workflow where each story goes from “Inkling” to “Idea” to “Rough Draft” to “Planning Review Meeting 3” to “Code Complete” to “Ready for Testing” to “Testing” to “Tested” and eventually on to “Done”, “Shipped”, “Finished”, “Announced”, “Demoed”, and then “T-Shirts”. The bad PM will love this because they can give reports like “this sprint is 43.8% done and we’re 97.53% likely to hit our target”, and the engineers will hate this because they’ll spend as much time updating story statuses as they do working. A good PM won’t care about all that and will prefer to use something simpler.
A useful task manager will be somewhat opinionated. It will almost, but not quite, do what everyone wants, and will annoy everyone equally with the few things they think it does wrong. A tar pit of a task manager will claim to be everything to everyone after customization, meaning that a few people will think it’s heavenly and everyone else will despise it with the heat of a thousand suns.
Jira is one of the bad ones. If you run across someone who adores it, tiptoe away quietly and quickly.
By today, July 29, 2021, Intel has shipped the last of its Itanium processors, the last holdout of a rough decade of their history. You’d be forgiven for not having heard of this unusual CPU as they carved a niche of a few supercomputers in the early 2000s and some legacy mainframe holdouts.
In 1994, Intel and HP looked around and saw a wide variety of successful server CPU architectures like Alpha, MIPS, SPARC, and POWER. This annoyed them and they decided to make a new CPU that no one would want to use. To these ends they invented an instruction set architecture that was impossible to program efficiently, planning that future compilers would be clever enough to make software run acceptably well. (This never happened because it turned out that anyone smart enough to write these compilers would rather be doing almost anything else.)
In 2000, Intel launched the NetBurst Pentium 4 CPU. It had serious design compromises that would hypothetically allow CPUs to run at upwards of 10GHz. Since these beasts could fry an egg at 3GHz, it was good that they never came anywhere near 10GHz as the heat would likely be sufficient to induce nearby hydrogen atoms to fuse.
Customers begged Intel to release a 64-bit Pentium-compatible CPU. They refused because they knew this would canibalize Itanium. Why write software for a weird and uncommon architecture if you could use something like the terrible x86 instruction set you already knew, but better?
In 2003, AMD launched their 64-bit, but Pentium-compatible, Opteron CPU. Everyone stopped buying Intel CPUs for a while. Within a few years Intel made their own 64-bit, but AMD-compatible, CPUs to avoid entirely losing the desktop and small server market. They were right earlier: almost everyone immediately embraced AMD’s instruction set and no one but HP wanted anything to do with Itanium.
And then, for a long time, nothing much happened. That’s happy news when you’re talking about earthquakes or tornados, but not so hot when you’re talking about sales of processors you spent a few billion dollars developing.
In 2015, HP admitted defeat and launched a line of mainframes using AMD’s 64-bit instruction set so developers could write and test software on systems that cost both over and under a million dollars.
Intel was contractually obligated to keep Itanium limping along but it was apparent their heart wasn’t in it. In 2019 they accepted the inevitable and announced that Itanium would be officially dead as of today. The final batch of CPUs was built on a 32nm process when everyone else was on to 10nm, 7nm, and 5nm designs.
Goodbye, Itanic. You were a strange, unloved little detour, better known for the good designs you killed than for any successes of your own. Few will miss you.
Ironically, in 2020 Apple launched their own desktop-class CPU that wasn’t compatible with more common Intel or AMD designs. The difference was that Apple’s M1 was actually nice and fast, both for developers and end users.
Someone invented a new standardized way to opt out of telemetry for command line applications. This is a horrid idea.
The existence of the setting establishes “tracking is OK!” as the default, and makes opting out the responsibility of the end user. With this in place, if a company collects the names of all the files in my home directory, it’s my fault for not tweaking some random setting correctly. (For technical types: don’t forget to set the “don’t track me!” variable in your crontabs, or else they’ll run with tracking enabled! Be sure to add it to your sudoers file, or now root commands spy on you!)
If this should exist at all, it should be in the form of a “go ahead and spy on me!” whitelist, with all telemetry and other spyware disabled unless explicitly enabled. Then it becomes the responsibility of each application’s author to encourage their users to enable it. Or better, get over the bizarre and radical notion of enabling spyware in command line utilities.
Progress bars suck at predicting how long things will take. I’ll tell you what I want (what I really really want): a system-wide resource that receives a description of what the progress bar will be measuring and uses it to make an informed estimate the entire process’s duration. For example, suppose that an application installer will do several things in series, one after another. Perhaps an explanation of that process could be written in a machine-readable format like this:
vendor: Foo Corp
name: My Cool App installer
stages:
- Downloading files:
- resource: internet
size: 1000 # Number of MB to download
- Extracting files:
- resource: disk_read
size: 1000 # Size of the downloaded archive file, in MB
- resource: disk_write
size: 2000 # Size of the extracted archive file, in MB
- Copying files into place:
- resource: disk_read
size: 2000 # Now we read the extracted files...
- resource: disk_write
size: 2000 # and copy them elsewhere.
- Configuring:
- resource: cpu
size: 100 # Expected CPU time in some standard-ish unit
Because I’ve used the progress bar resource before, it knows about how long each of those things might take:
Ta-da! The whole installation should run about 44 seconds. When the installer runs, instead of updating the progress bar manually like
update_progress_bar(percent=23)
it would tell the resource how far it had gotten in its work with a series of updates like
update_progress_bar('Downloading files', internet=283)
...
update_progress_bar('Copying files into place', disk_read=500)
update_progress_bar('Copying files into place', disk_write=500)
...
update_progress_bar('Configuring', cpu=30)
The app itself would not be responsible for knowing how what percent along it is. How could it? It knows nothing about my system! Furthermore, statistical modeling could lead to more accurate predictions with observations like “Foo Corp always underestimates how many CPU units something will take compared to every other vendor so add 42% to their CPU numbers” or “Bar, Inc.’s website downloads are always slow, so cap the Internet speed at 7MB/s for them.” Hardware vendors could ship preconfigured numbers for new systems based on their disk and CPU speeds where the system can make decent estimates right out of the box. But once a new system is deployed, it gathers observations about its real performance to make better predictions that evolve as it’s used.
We should be able to do a much better job at better job of guessing how long it’s going to take to install an app. This solution needs to exist.