I liked Pivotal Tracker. I liked the tracker Abhijit and I used for aox. That's all. Of all the task trackers I've used over the years, those are the only two I liked.
So of course I had to write my own eventually.
Mine is a single file, easily edited, in almost this format, except that mine has special support for links.
I need to keep an eye on bugs in remote trackers, issues, pull requests, etc. and look at them every few weeks, so I wrote code that acts like a compiler. When I run that in the editor as though it were a compiler, the task tracker inspects the links, and if one needs attention, two things happen:
- It opens the relevant link in a browser.
- It shows an
error message
and the editor moves the cursor to the note containing that link.
That's it. That's really the only feature. I like it. git handles everything else. If I want to know when I added something, git blame tells me. If something is done, I delete that note and git commit.
I needed to reset the password of my account with a largish company recently, 14k employees, so I sent the following request: I need to delete my account. Apparently it's the only way to reset the password and use my new phone. Sounds rather dysfunctional, doesn't it?
Their reply conformed that they were pleased to confirm we have now actioned your request.
Dear support team: I don't want to have my prejudice boosted. Please don't use action
as a verb. Thank you.
I didn't expect that this would be common enough to see it myself:
What it means is that IPv4 has become (or is becoming?) less reliable than IPv6 as a transport. The web service (a nice professionally run place) detects that my
NAT gateway has tried to do something nasty, and blocks some functions from its IPv4 address. Thousands of people use that NAT gateway, because my ISP has more customers than IPv4 addresses and puts all of us behind one IPv4 NAT gateway.
IPv6 doesn't have this problem — each customer gets a unique IPv6 prefix so only the malevolent customer is blocked, the block won't affect thousands of bystanders.
China has a million unicode domains (慕田峪长城.网址 is the tourist site for the Great Wall), Japan has barely more than zero. That difference puzzled me until I saw the Japanese magazines below (posted by Derek Guy). […More…]
I replaced my old Mikrotik hardware recently. The oldest AP was almost fifteen years old, and Mikrotik still delivered OS upgrades for it: Fantastic. I'm a fan.
But I replaced it. My new setup involves three APs (my home has a very difficult layout) and four SSIDs. Setting it up was a little too tricky, RouterOS 7.14 is substantially different from both of the older approaches (capsman
and interface wifiwave2
). Mikrotik's documentation mostly explains it, this posting explains it differently. […More…]
I backed the Jelly Max on Kickstarter as soon as it was announced. It's larger than the old Jelly phones, which I liked but ultimately stopped using.
I wrote about the Jelly 2 that it fits in every pocket I have, even in my tightest jeans, it runs the apps I need, and if it makes me spend less time on Twitter, that's fine.
Fine, all true, all correct. Why did I eventually switch back to my old Xperia XZ1 Compact?
Part of the story is that the nice guys who build LineageOS for the xz1c kept providing upgrades (here's Android 14), another part is that the bezels on the Jelly 2 ought to be narrower, a third part is that I couldn't find a good phone holder for my bike. I had to switch from the Jelly 2 to the xz1c before a weeklong bicycle trip and… never switched back. […More…]
Following an unfortunate sequence of events, my mail server is currently reachable only via IPv6. It took me a few days to even notice the lack of IPv4.
IPv4 connectivity should be back in a day or two. We fixed the root problem last night.
What I learned from this involuntary experiment: Lack of IPv4 access for a mail server is real problem in 2024, it's something that has to be fixed, but it's a surprisingly small problem.