vrfy, for those few who don't remember nineties email debugging tools, reports on an email address. I run it occasionally and made a git repo ten years ago, github:arnt/vrfy.
Today I converted the code from K&R C and updated the code with support for the most important new RFCs (5321, 6531, 7505, 8200).
Most apps tell the user to check this and that when there's a network problem, any network problem. The web browsers I've seen offer the same evasive blah-blah for any problem except a failing TLS certificate.
I have an app that's mine alone, though. It shares my values and its behaviour is guided by my pet peeves, including being annoyed by messages that just say can't reach server
. So for it I wrote a class to analyse a network problem and offer a less generic message.
My analysis is quick and will offer a diagnosis after at most 2.5 seconds. My reasoning here is that when something takes a little longer than usually, there may or may not be a problem, and it's reasonable to start analysis while continuing to try. After a couple of seconds more, there very likely is a real problem, and the app should offer the user a decent explanation. That gives a couple of seconds from when a problem appears likely until the explanation must be ready.
It's cheap, in particular it doesn't cause a lot of network traffic. It sends and receives packets, but not so many as to slow down existing up/downloads. This is because I want to start the analysis before I'm certain that there is a problem. […More…]
I was searching for something else when a really old mail message caught my eye.
It was a message from dtrump@tpalace.com
, with whom I was not previously familiar. That dtrump
address sent me two messages in July and August 1997, both about an income opportunity
with which I could get rich
.
I'm not sure the messages are genuine. They're unsigned, of course, and the chain of Received fields is difficult to analyse reliably at this point. I also don't know to whom tpalace.com belonged at the time (the oldest information I could find is from 2008).
If there existed a host named tpa-us14d4.tpalace.com with IP address 111.243.5.2 in 1997, I would guess that the spam really came from, uh, tpalace, whatever tpalace was.
I buy less and less hardware as time passes. But I couldn't resist the See.sense Icon3 bike rear light. It's really, really bright and has fancy features: It should be able to brighten when a car approaches from behind, answering the car's headlights as it were, without wearing out its battery.
The light charges via USB C, badly. None of my cables can charge it from a USB C port, it only charges from USB A ports using A-C cables. This may be due to faulty use of the CC1 and CC2 lines in the Icon3. My Anker 737 says it charges at 1.1W from USB A.
It shone very nicely when I used it on a transalpine bike trip this summer, and wasn't at all bothered by rain. The car drivers behaved as if they could see me well.
Update: The charging problem is clearly due to a hardware error in in the Icon3. I bought a USB-Chyna adapter and when I use that, the Icon3 charges correctly from all of the USB C chargers I've tested.
I implemented the IMAP UIDONLY extension in Archiveopteryx a while ago. A couple of notes:
It was really simple; two lines of code changed to six lines starting at line 1154, plus a few lines to announce the extension etc.
It didn't bring any any performance improvement, because generating the EXISTS response is mandatory. With UIDONLY Archiveopteryx counts messages in order to report how many there are in an EXISTS response. Without it, Archiveopteryx makes a list of messages. The loop is the same.
It ought to improve performance for the case where a connection opens a large mailbox and another connection already has opened that. This seems rare with Archiveopteryx. It's more important than it seems though: Opening really large mailboxes can take many seconds. I decided to leave the code in.
There's so much new text on the web that the valuable old writing is difficult to find. I know that Ali Davis wrote an excellent description of various kinds of answers, with a very lucid example of how some answers open the space of possible conversations and others restrict it. I'm 99% sure she posted it to improvisation.ws, and while archive.org has others of her postings, that one is gone and I can't find it.
Sigh.
So I want to quote this little gem before it becomes impossible to find, from a blog mostly about economics:
The one revelation for me was Austria, especially rural Austria. At first I wondered if what I was seeing was an unrepresentative sample—tourist areas that looked richer than usual. But no, the whole of rural Austria looks extremely impressive—like a Swiss travel poster. It’s makes America’s built environment look shoddy by comparison.
Everything looked high quality and attractive. The trains, trams and buses all looked brand new. The roads had no potholes and were well designed. The drivers were all skilled and the traffic flowed smoothly. Doors and windows seemed as solid as a bank vault, not the flimsy Home Depot crap you get in America.
Yes, I know that America is richer than Austria. I know that there is more to GDP than the quality of the built environment. Services matter (don’t ask about their food). Size of houses matters. But I’d also argue that quality counts for something, and probably gets overlooked in GDP comparisons. If an Austrian (or German) told me that America seemed poorer than their home country, I would not argue with them. It’s all subjective.
The residential buildings in rural Austria were quite attractive, quite unlike the ugly new houses being built in America. I can’t blame America for lacking the beautiful baroque buildings of Vienna and Salzburg; we are too new to have any of those. But why is even the modern architecture in Austria so much better.
At some point I'm going to take my office chair to Salzburg to get about half of it replaced. An long-day bicycle trip each way, with the chair on a trailer. No doubt I will stay at my favourite hotel there, one that's a perfect example of the attitude he describes in that blog posting: Money was spent on something good and while the goodness somehow counts for something, counting it is difficult and not done.
That hotel is just pleasant. There's a bourgeois solidity that isn't visible in photos (on booking sites or here) and isn't visible in economic statistics either.
Meanwhile, here's a photo from a chain hotel in a rich American city. The door looked like plastic to my eyes, the frame too, and then there's the door handle.
debconf: cannot find a question for
means that the package database and debconf's database have lost sync. Packages are supposed to add and remove questions when they're installed/uninstalled, and mistakes can happen.
The solution is simple: Run /usr/share/debconf/fix_db.pl as root. fix_db.pl deletes any questions that related to deleted packages.