Meanwhile, a friend of mine just got a response on a PR. Written in nice polite professional prose: “Sure, we can merge that, thanks for this improvement, but would you first please …” and that’s where it spun off into senselessness. It was all politely written, no swearing, but made no sense.

Linus will swear. And read the code. And understand both the code and what you tried to do, and have a decent idea of what’s achievable.

It seems to me that some people really mind the swearing, and do not realise that everyone has faults. Complaining about swearing isn’t a step towards faultless people, it’s effectively pushing us towards other faults, the ones that get no complaints.

Prelude: I worked myself into a state where I needed a two-hour break after about twenty minutes of keyboarding. Two keyboards with different layouts on different desks, changing between them every few minutes, a terrible project, and some ungood stress out of the office. But as luck would have it, that office was in a… well, not in a hospital, but on campus, so got to see a real specialist quickly. A friend in his department just brought me along. He spoke cluefully and I’ve followed his advice in the decades since. In brief: “Pay attention to your body. Learn what hurts and stop doing that. Don’t let anyone at you with a knife.” It’s served me well. I admit to a certain arrogance about it.

A colleague and I bought and tested about 10-15 keyboards, including some very expensive specials, and the ones we ended up using weren’t the most expensive one, which probably would get you into trouble with the bookkeepers if you were to try it. Arrogance helps.

I currently use an 88-key unlabelled WASD with dampening o-rings. Because it keeps me from moving my hands much while I type, and keeps me from looking at the keyboard, and I like the feeling of the keys. You should ignore the first sentence in this paragraph and focus on the second, because the brand name isn’t important, how your body reacts is vastly important. Does it feel okay? Then okay. Not? Then change.

The same applies to your chair and desk, because your body is one. The shoulders are tightly connected to your hands. (The chair comes first, btw. You get a chair for sitting on, then a desk that suits your body on that chair. http://rant.gulbrandsen.priv.no/arnt/ideal-office has more speechifying about chairs and stuff. I speechify too much.)

That’s rich, from a guy who done his best to advance client-server cloud model in his time.

OK, not really happy about the acquisition either, but overall GitHub has been a massive boon to the community in general. It lowered the threshold to collaboration, publishing your projects and facilitated a bunch of dependency fetching ecosystems with much higher availability than was possible before.

I don’t understand the author’s objection to Outreachy. As far as I can tell, they want to fund some interns from marginalized groups so that they can work on open-source. They are not preventing the author from working on open-source. They are not preventing the author from funding interns he approves of from working on open-source. What is the problem?

As I understand things, Jon Skeet is the most successful stack overflow user at doing the things stack overflow users press up arrow for. And yet this blog discusses how the stack overflow company want a library of answers to all programming questions, while he has gatekeeper criteria for “good” programming questions.

It seems there is a disconnect between the goals of the company and the people who press the up arrows.

[related to the subject but tangential to this blog: there is also a huge opportunity indicated by the stack overflow user survey and here, to cater to the rest of the programmer community who are not young white male gatekeepers with a couple of years of JS experience who hate Perl, asm and ObjC]

Yes, the clang team made a real effort to produce helpful messages about a decade ago. I think there’s still scope for improvement, though, I frequently get errors where the line with the actual error is pointed at by the seventh of twelve messages.

the concerns about input on international clients such as those with with diacritic marks are very real. I vividly remember my first time using a *NIX overseas and fighting to get many special characters to work.

of course, one could argue that few users directly input URLs any more, but I would still avoid almost any even remotely special character. over the decades and dozens of languages through which I have survived, I have seen countless URL [en|de]code implementations that choked on even common punctuation. if it isn’t a slash, it isn’t worth it. (only partly being sarcastic)

Let’s see. The reasons have to do with Τεχ, with those character encodings that were used before utf8 took over, because it’s difficult to find on an unfamiliar keyboard, because of something to do with Spanish and Portuguese, and because people misprint it. Right? That’s a fair summary?

I don’t see any point in using tilde. Home directories on multuuser servers have disappeared.

But those are hilariously weak reasons to stop. Τεχ is dead, so are those character encodings, most typing is done on familiar keyboards, if I were to stop doing things because they’re difficult in some other language (even just ones I’ve half-learned for a vacation) I’d have to change both my first and last name, and as for misprints the tilde is a minor detail. Many pages today have more difficult URLs. The URL of the page you’re reading includes the character sequence “dlk8d1” and orgies of “…?…=…&q=…” are normal.

“Too random” means “inappropriately distributed”.

C++’ new library has distributions, and it’s great. If you want randomness that looks like how people’s height varies randomly, with that hump around the average, C++ gives you that, it’s called “normal” (and there are variations of that, yes). If you want randonness that looks like the number of customers entering a shop per minute (or I suppose requests to an API per second, absent crontabs), C++ gives you that, it’s called “poisson” (usually). No need to reinvent 19th century statistics just because you model something that isn’t uniformly distributed.

Consumers buy devices based on features and buzzwords and little icons on the packaging. A wifi radio costs pennies and scarcely more than 1mm³. Says the salesman: “Well, this one supports Wifi and DLNA so you can use it as a controller for lots of smart TVs, and this other one does not”.

I run several sites on wholly unremarkable hardware. As in, rails on the cheapest I could possibly find. As long as you have a cache in front (apache mod_cache, nginx or varnish) and your backend produces a Cache-Control header, your software will handle ~all load peaks. High traffic level involves repeated requests and the cache takes care of precisely those.

So my answer is: Just add a cache, make sure your pages are cache-friendly, then stop worrying about this. Worry about the content instead.

There’s a story about this, from the lavalamp shop, where there’s a configuration file for services that says “this needs x gigabytes of ram and benefits from being on the same switch as shard y of service z”. The needs can be quite complicated, so the configuration language eventually grew turing complete.

From that point you could do anything in the cloud in order to determine where in the cloud to do anything. Very powerful, but when you think about it, do you prefer something optimal or something suboptimal and understandable?

The same issue comes up in many other guises. Do you prefer a programming language that lets you do anything, or one where the compiler can reason about the code? Pointer aliases in C is a classic case, but configuration files too are useful objects of reasoning. I once wrote a great feature that reasoned about the configuration file before restarting a service and would say “the service will go down and not restart, because during startup, suchandsuch”. It was very useful, and it was doable precisely because the configuration file was not turing complete.

So I’ve been encouraging people to switch to thinkpads.. but since they’ve gone off trying to replicate the Macbook, what hardware do people recommend?

The T460 I think is probably going to be my last thinkpad if they stay going this “no FRU” path.

I have an old router and a Soekris 4501.

Moving tens of megabits per second was common twenty years ago and still is today; I retired the router a week or two ago after (much?) more than a decade of service. A decade of not having to reconfigure… I love that.

The 4501 (partly by chance, partly by design) is a marvellous time server. Accurate and reliable time keeping requires the NTP server to have a good model of the hardware it’s running on, and the Soekris boxes have very simple, easily comprehensible hardware, and a correspondingly simple bios. And then some luck with the choice of… I think it was the main crystal.

There are quite a few widely used projects on Github, Rails comes to mind. I can believe that few people who write ruby have participated in rails development, or another major gem, but not seen a successful project?

Linus is saying: Make the happy path the only path whenever you can, that way you have only one path to think about and to test. Each if() increases the number of paths through the function.

He’s also saying a little more, but that’s the core, and it’s not subjective: Count the number of paths through a function, low is good. Now, some 4s may be better than other 4s and then you have to discuss subjectively, but that issue happens seldom. Usually you’re looking at 2 vs 12 or something else crude.

Those who like this kind of thing might want to reread Christopher Seiwald’s https://www.perforce.com/resources/seven-pillars-pretty-code — the “fight changes of indentation” bit is excellent advice and generally leads to few paths through the function, and the “exceptions to the top” shortens many of the paths that cannot be eliminated.

I read books by turning off the things that can interrupt me. At home I put the children to bed and do not own a TV, in the office I have a reading chair positioned such that I cannot see the screens. I choose what to concentrate on, see?

The benefits of e-books are such that… in effect you haven’t read a book in years?

I invoice for things I wouldn’t have bought on my own accord, or didn’t want to buy. Do not invoice for my office (which is in what used to be the neighbour flat in my case).

In short: If I want to buy it and keep it if I quit, I pay. If I don’t but someone at work says to buy, I invoice.

You don’t need a meeting room with AV, but you do need something better than laptop microphones. A minimac or laptop with a Jabra microphone is good. You may find it helpful to put a minimac near the coffee machine or in the spot where you’d typically run your standups (at standup height, perhaps). If the remotes can call in and the box autoanswers, that may be good. If you can see the team, you can also see who’s not there, and that’s a great deal better than bothering people in IRC with questions like “can you look and see if Kim is in the office?”

Make sure to start on time. Same time each day, not a minute before or after. Starting ten minutes late is OK when you can see the team anyway, but the remotes will sit on their thumbs waiting and you as team lead won’t notice that you’re irritating part of the team. Having the standup box autoanswer when a remote calls in helps with that.

Watch out for people who mumble too far from the microphone. It’s not necessary to hold the microphone, but mumbling five meters away won’t do.

Be prepared to tweak.

Jabber chat rooms save messages in the manner you want, but client support is scarce. YMMV.