Sep 18, 2020 8 tweets 2 min read Read on X
Does the value 252 seconds (or 1/2400 week) ring a bell for anyone in the context of GPS? I'm trying to reverse-engineer data written by a GPS device, I was able to make sense of most of the fields, but there's something which seems to be a counter wrapping every 252s. 🤨 •1/8
Context on what I'm trying to do is here: — but not really important here. What's relevant is that I have data from a GPS device and I'm trying to understand its format. I was able to decode most of it (date, time, lat, lon, speed, bearing…), … •2/8
… but a few fields still escape me. Irritatingly, there doesn't seem to be a sub-second timestamp in my data. On the other hand, there's this weird set of values, time counters of sorts, which I'll call: a (4 bytes), b (1 byte) and c (2 bytes) — in this order: … •3/8
… the value b is incremented every second, mod 252, i.e., it wraps around every 252s. (At least over short time periods: over longer ranges its period doesn't seem to be exactly 252s.) WHY 252s? I have no idea. This is 1/2400 week, but this doesn't really help. •4/8
(I mean, I know weeks are used by GPS, but I don't know about 1/2400 week, and I wasn't able to find anything which goes up to 2400.) As for the value c, it is incremented every time b wraps from 251 to 0, so it keeps count of 252s increments. •5/8
But c seems to wrap to 0 every day or so (not every week which would sort of have made sense), at approximately but not exactly 20:00 UTC, and not even exactly the same time each day. This makes absolutely no sense to me. •6/8
As for a, it seems to count days, except not really. I would like to think that a is incremented when c wraps back to 0, but I don't have proof. Over short periods, a seems to increment every day, but over more than a year, it increased by more than it should have. •7/8
The whole thing is extremely weird, and I have no idea why anyone would want to include a 252s counter in GPS data but not include a subsecond timestamp. (OTOH, I also wonder why anyone would want to XOR the entire record with a seemingly random byte. Makes no sense!) •8/8

• • •

Missing some Tweet in this thread? You can try to force a refresh

This Thread may be Removed Anytime!

Twitter may remove this content at anytime! Save it as PDF for later use!

# More from @gro_tsen

Jul 20
OK, this 🔽 blew up (and I'll need to mute the thread). I can't respond to every reply individually, but there are a few common ones that deserve specific attention. So here are a few followup points. 🧵⤵️ •1/15
First, there are number of things that I forgot to mention. Like how many homeless people were forcibly removed from Paris ahead of the games, or students being told to vacate their residence because they would be used as lodging for the games. •2/15
Or how the price of metro tickets has been suddenly doubled(!) for the games. Admittedly this is more of a tourist tax, because locals typically have monthly cards or could buy their tickets well in advance, but this is still crazy from a supposed public service. •3/15
Jul 19
I think non-French and non-French-speaking readers should be told about how utterly insane things are now in and around Paris because of the Olympics are. The opening ceremony is one week away and they've already barred any access to the Seine! 🧵⤵️ •1/19
The different levels of security perimeters are incomprehensible, but basically all areas within ~100m of the Seine are now completely off-limits. You can only go there, even by foot, if you live there or work there, and you need to get a special permit in advance. •2/19
Getting the permit is no sure thing: apparently some people have had their application denied (no explanations given), so they're not allowed to go where they work, or even where they live. Nobody knows what's supposed to happen to them. •3/19
Jun 24
En tant que Français et Canadien, je pense qu'il est nécessaire que je dise un mot au sujet du fait que les binationaux sont soudainement considérés comme suspects en France, presque des ennemis de l'intérieur. 🧵⤵️ •1/22
Je suis né en France d'une mère française, et j'ai vécu quasi toute ma vie en France, je parle français depuis que je sais parler, j'ai fait mes études en France, mes amis vivent en France, je m'intéresse à l'histoire de France, au droit français, à la langue française… •2/22
Bref, je suis non seulement légalement français, mais, que ça me plaise ou non, je suis aussi culturellement français, et, je n'ai aucune honte à le dire, j'aime profondément ce pays, au moins autant que Monsieur Bardella. Même si en ce moment j'ai du mal à en être fier. •3/22
Jun 13
Une chose fascinante avec le regard que la droite et le centre portent sur la gauche française, c'est que la gauche du passé leur semble toujours plus acceptable que celle du présent: Gambetta, Briand, Blum, Mendès-France, on conteste très rarement leur héritage, … •1/11
… souvent même la droite reconnaît explicitement que c'étaient de grands hommes et affirme que la gauche les aurait ensuite trahis (Sarkozy en 2007 par exemple: «la gauche d'aujourd'hui [n'a] plus grand-chose à voir avec la gauche de Jaurès»). •2/11vie-publique.fr/discours/16640…
La gauche du présent, en revanche, ou du passé proche, paraît toujours détestable: la victoire du Front populaire de 1936, et l'alliance des socialistes et radicaux avec les communistes, n'a pas été accueillie avec enthousiasme par le centre-droit d'alors, … •3/11
Jun 11
J'ai calculé les coefficients de corrélation des résultats entre toutes les listes des dernières européennes, et je trouve ça très intéressant pour se faire une idée de la proximité idéologique entre elles (notamment, de qui se rapprochent les «petites» listes). 🧵⤵️ •1/14

(Désolé pour ce dump sous forme d'images: si vous voulez le résultat sous format texte, il est là: — la même page comporte le script Perl abominable que j'ai utilisé pour le calculer.) •2/14gist.github.com/Gro-Tsen/e54a8…
La source des données est là: (il faut fournir ce fichier csv en entrée à mon script Perl pour produire le résultat). •3/14 data.gouv.fr/fr/datasets/re…
Nov 26, 2023
Let me say a few things about “random” versus “generic” sequences (binary sequences, say). A random sequence of 0's and 1's everyone has some idea of what that looks like: take a fair coin and flip it repeatedly. But what in the world is a “generic” sequence? 🤔🧵⤵️ •1/21
There are many flavors of randomness and correspondingly many flavors of genericity, but the overall theme seems to be that a random object is one that doesn't belong to any “easily explicitly described” set of measure zero, … •2/21
… whereas a generic object is one that doesn't belong to any “easily explicitly described” meager set. What does all of this mean? Well, “easily explicitly described” depends on the flavor we're talking about: but clearly we need some kind of restriction, … •3/21