CHANS battery rebuild: giving traditional laptops a new life with refactored batteries cells

Back in 2018, when I acquired my legendary ThinkPad X220, I discovered that there is a lady in Ontario, Ms. Chulkova, who does professional battery rebuilding as a side-gig: if you’re in Canada (or the USA, to an extent) you can get your laptop (and other power tools) battery cells replaced by new high-quality cells. This is interesting if you have an electric bike or if, like me, you are a luddite who believes that the last great laptop keyboards were produced in 2011 (before the ThinkPad X230/T430/etc.) and that ultrabooks—with their proprietary slim LiPo batteries and general planned obsolescence—are an ecological disaster.

Continue reading “CHANS battery rebuild: giving traditional laptops a new life with refactored batteries cells”

The Longest Debugging—The journey towards a reliable Linux workstation

I have this curse where I keep finding heisenbugs not only in the software I use, but also in hardware… The difference being that, unlike software misbehavior, hardware issues take me months to figure out. But hey, they say I’m a persistent bastard.

This blog post is mainly a tale about computer hardware, which is a bit unusual on this blog, but it ties back into software and Linux graphics troubleshooting, so there should be something interesting for everybody. Enjoy a solid 20 minutes of reading my walk through the inferno, hopefully you’ll find something insightful or funny in my unreasonably persistent path to hardware salvation. Anyway, the reading time doesn’t seem that long when it took me 10 months to write this post over 37 revisions 😉 Continue reading “The Longest Debugging—The journey towards a reliable Linux workstation”

Renommer un périphérique audio avec PulseAudio

J’ai découvert par hasard qu’on peut faire un clic droit sur le “profil” d’un périphérique dans pavucontrol pour renommer le périphérique. Or, cette fonctionnalité n’est pas disponible par défaut puisqu’il faut un module supplémentaire (sous Fedora, du moins). Pour faire un essai en temps réel:
$ pactl load-module module-device-manager

“34”? Quelle drôle de réponse! Je présume que ça veut dire que ça a fonctionné. Si on réessaie immédiatement la même commande, on obtient:
$ pactl load-module module-device-manager
Échec : Échec lors de l'initialisation du module

Ce qui, si on pense comme un développeur un peu paresseux sur la sémantique, dévoile que le module est bel et bien chargé déjà.
On peut maintenant donner des noms beaucoup plus courts et pratiques à nos périphériques favoris. Par exemple, ma SoundBlaster au nom beaucoup trop complexe:

Ou encore mon bon vieux micro Logitech au nom tout à fait cryptique:

Côté interface, tout ceci est plutôt moche et difficile à découvrir, alors j’ai ouvert un rapport de bug à ce sujet. Ce qui est un peu dommage, c’est qu’on renomme ici les entrées/sorties (sources/sinks) individuellement, au lieu de renommer le périphérique matériel dans sa globalité. Aussi, les versions renommées ne sont pas prises en compte par les paramètres de son de GNOME Control Center.
En tout cas, après ce test concluant, on peut ajouter les lignes suivantes dans le fichier ~/.config/pulse/ pour rendre le changement permanent (je préfère éditer ce fichier dans mon dossier personnel plutôt que le fichier système “/etc/pulse/”, pour que ça persiste après des “clean installs” de distros:
.include /etc/pulse/
load-module module-device-manager

Note: “”, contrairement à daemon.conf (dans lequel j’ai simplement mis “flat-volumes = no” pour revenir au mixage à l’ancienne), n’hérite pas automatiquement des paramètres système, c’est pourquoi j’ai inséré la ligne “.include”.

Pour en finir avec la congestion nasale: comment effectuer le rinçage nasal correctement

Ce billet pourrait aussi s’intituler “Comment en finir avec les allergies et court-circuiter le rhume, la grippe ou la sinusite”, car selon mon expérience la congestion nasale est un symptôme qui empire tout le reste au point de prolonger la maladie. Quand on élimine la congestion nasale et la sinusite, on récupère beaucoup plus rapidement (ou, du moins, beaucoup moins péniblement). Regardez comme cet homme semble heureux:

neti pot sinus rinsing epic dude
Continue reading “Pour en finir avec la congestion nasale: comment effectuer le rinçage nasal correctement”

"SmartEco" or "Extreme Eco" projector lamp power saving modes are a trap

Here are some findings I’ve been meaning to post for a while.
A bit over a year ago, I fulfilled a decade-long dream of owning a good projector for movies, instead of some silly monitor with a diagonal measured in “inches”. My lifestyle very rarely allows me to watch movies (or series*), so when I decide to watch something, it needs to have a rating over 90% on RottenTomatoes, it has to be watched with a bunch of friends, and it needs to be a top-notch audio-visual experience. I already had a surround sound system for over a decade, so the projector was the only missing part of the puzzle.
After about six months of research and agonizing over projector choices, I narrowed it down to the infamous BenQ W1070, which uses conventional projection lamps (Aaxa’s LED projectors were not competitively priced at that time, costing more for a lower resolution and less connectivity):

First power-on, with David Revoy‘s beautiful artwork as my wallpaper

In the process of picking up the BenQ W1070, I compared it to the Acer H6510BD and others, and this particular question came up: how realistic are manufacturers’ claims about their dynamic “lamp life saving” features?
The answer is:
bullshit - ten points from gryffindor
For starters, I asked Acer to clarify what their “Extreme Eco” feature really did, and to their credit they answered truthfully (emphasis mine):

Acer ExtremeEco technology reduces the lamp power to 30% enabling up to 70% power savings when there is no input signal, extending the lifespan of the lamp up to 7000 hours and reducing operating costs. 4,000 Hours (Standard), 5,000 Hours (ECO), 7,000 Hours (ExtremeEco)*
*: Lamp Life of ExtremeEco mode is based on an average usage cycle of 45 minutes ECO mode plus 75 minutes ExtremeEco (30% lamp power) mode

Continue reading “"SmartEco" or "Extreme Eco" projector lamp power saving modes are a trap”

Le coût de la sécurité publique

Un constat du jour en m’amusant avec l’outil de visualisation du budget de la ville de Montréal: il semble que le Service de police de la Ville de Montréal coûte plus cher que la somme des salaires de tous les arrondissements! Avec les cotisations de l’employeur par dessus, le SPVM représente ainsi environ 12% du budget total.
dépenses en salaires à la ville de montréal, budget 2016
Quand on rajoute le service d’incendies (qui est également un service centralisé), on constate alors que la « sécurité publique » consomme environ un cinquième du budget total de la Ville de Montréal. Ça peut paraître gros, mais il semble que ce soit dans les normes: dans mon petit patelin précédent (au budget de 30M$ plutôt que 5G$), la Sécurité publique occupait 23% des dépenses.
Quelques souhaits pour le futur:

  • Il est dommage qu’on ne puisse pas creuser à l’intérieur des budgets d’arrondissements avec cet histogramme interactif… il y a un certain flou de ce côté là.
  • L’histogramme modal est un peu problématique lorsqu’on veut conserver une vue d’ensemble, j’aurais bien aimé avoir un système en arborescence dépliante.
  • Un histogramme en rangées plutôt qu’en colonnes pourrait être intéressant, spécialement pour la lisibilité des étiquettes: présentement les étiquettes sont à 45° sur l’axe des X, alors qu’elles pourraient être à 0° sur l’axe des Y. Qui plus est, l’histogramme dans sa forme actuelle consomme relativement beaucoup d’espace lorsqu’il y a peu de colonnes, énormément de « padding » qui pourrait être éliminé dans une orientation « en rangées ».

Continue reading “Le coût de la sécurité publique”

BBC Radio's adaptation of Isaac Asimov's Foundation trilogy

Other than for self-improvement, I’m not a big fan of books (nor podcasts) in general, because of the big time investment required. THIS, however, is such an amazing masterpiece of a radio adaptation that I can heartily recommend it to anyone who has a good grasp of spoken British English (it was produced over fourty years ago by the BBC, after all). After the first episode, I was hooked, and ate through the entire série in a week or two. I found it best listened to while relaxing, with eyes closed to immerse yourself in the intergalactic drama at play.
I’ll smugly say I foresaw a couple of the plot twists (including a big part of the chapters concerning the Mule), but Asimov kept surprising me otherwise.
Besides having very talented voice actors give life to what might otherwise be a bit of a dry story for non-sci-fi connaisseurs, it turns out that the radio adaptation has a special segment about the life of farmers on Rossem. That segment is absolutely hilarious, contrasting heavily with the doomy & gloomy nature of the whole series. It is also fairly philosophical, touching on the question of life fulfilment. The exchange between Pritcher and the Mule, after talking with those farmers, was a great emotional portrayal: you could actually feel perplexity and doubt in Pritcher’s voice, and shock and urgency in the Mule’s.

Escape from Stockholm — getting to Arlanda with night bus 592 or 593

Rule number one: trust Google Maps. Rule number two: do not believe the locals’ knowledge of their own transportation system.
Stockholm is this big city that prides itself on having a great public transportation system. It sounds great. Until you find out at 1AM that everything shuts down at night. So here you are, in central station, with SL employees telling you “It’s closed, sorry. Get to the airport tomorrow”. That obviously won’t work with a flight that leaves at 6.
Let me restate my disbelief here: a city with four friggin’ airports has no 24/7 public transportation to get to its biggest airport. Except it does, but nobody knows about it, or they don’t believe it works. Bus 592 and 593 take you from some parts of downtown to Stockholm-Arlanda airport.
The reason I’m bothering to write a blog post about this is that I want to put this information out there; it was impossible for me (with a fair amount of googling skills) to be absolutely certain of the fare/pricing/payment system for bus 593 and bus 592. The SL’s website doesn’t tell you, Google doesn’t tell you, wikipedia and wikitravel don’t tell you, and all the locals I asked were convinced that it is impossible to travel to Arlanda with normal tickets. They will tell you to try to take the train instead.
I am living proof that it is possible, and I’m putting this blog post out there to let the rest of the world know when they search the Internet for two very simple questions:

  • “Can you go to Arlanda with one ticket on a blue SL card?”: Yes. The night buses are really local buses, and since you’re boarding them in the (central) zone A, a normal ticket is enough. I had a blue RFID transportation card with only two tickets left; I used one for the subway, realized the transportation offering was rubbish, ran in ten directions asking various locals along the way, ran to the 593 bus stop I initially planned to go to, and confirmed there that it works fine with one normal ticket. So this turns out to be the cheapest way (also arguably the simplest) to get to the arlanda airport, although it takes a bit longer due to the many bus stops.
  • “Can you pay with a credit card on the bus?”: No, I don’t think so. I did not see any cash or credit card machines on the bus, only SL “access” card machines.

Continue reading “Escape from Stockholm — getting to Arlanda with night bus 592 or 593”

Privacy does not exist — it never did

Warning: this is certainly a depressing post.
The NSA/PRISM/big brother scandal of late didn’t surprise nor shock me. The discovery that “the NSA probably silently circumvents/broke all our crypto and hid backdoors everywhere” is not really a discovery to geeks who have given the whole system a bit of thought — spies are spying on us, and they’re not telling us that our GPG keys or disk encryption are not sufficient to guard against them? No sh**, Sherlock!
What kept me relatively optimistic about the whole thing is that running a fully open-source software stack on your computer gives you long-term protection, in theory. Sure, it probably has backdoors planted by the NSA right now (you may now begin the witch hunt), but the increased awareness and outrage from recent events means (hopefully) that we’ll weed out those vulnerabilities with more audits, ruthless efficiency and an almost fanatical devotion to security.
But then, on the privacy side of things in general, some much more worrying thoughts have started resurfacing in my mind.
signs - tinfoil hats
For starters, let’s just take the case of WiFi networks. Until today, it somehow did not truly occur to me that no matter how security-conscious I am, no matter how “safe” my encryption is considered to be or how clever my wifi passphrase is, it doesn’t matter at all: as soon as a friend or family member logged into my wireless network once, it’s game over: Google (or Apple, or whoever) knows your passphrase, which allows circumventing the entire system.
I say that I’m surprised this realization had not occurred to my mind until now, because this is just the logical extension of what I’ve been thinking for years about everything else, about GMail, social networks and addressbooks. Network effects means that you do not have any “opt out”, only the illusion of it. Even if you boycott them and avoid giving out any information, your friends and family (you know, normal people) will happily enter this information into the system. Didn’t fill in your birthdate, phone number or door passcode number in your profile? It doesn’t matter: somebody else added that info right next to your name and email address in their own addressbook which, unless they are as paranoid as you, is stored (or synchronized) online.

Cat going nuts
Pictured: me going “AAAaaaaargh”

Even if we go full-tinfoil and use only privacy-intensive, security-audited, local-storage-only open-source software, we are still compromised… because humans are socially interconnected and technology has become integral to the world as we know it. I don’t know where to go from there, so maybe there’s a huge flaw in my analysis that I’d love to hear about.