
Phasing out Intel Macs?! 🤮
more than half of my customers on MacOs are using Intel Macs… what can I do? never upgrade to keep them happy ?

more than half of my customers on MacOs are using Intel Macs… what can I do? never upgrade to keep them happy ?
Bonjour à tous,
Je pratique la grande voie et, jusqu’ici, j’ai toujours réalisé mes relais sur 2 points avec une longue sangle, en triangulant pour créer un point central.
Récemment, j’ai grimpé avec quelqu’un qui reliait simplement les deux points avec une sangle/cordelette, sans faire de triangulation comme je l’ai appris. J’ai aussi remarqué qu’il installait son relais beaucoup plus rapidement que moi.
Que pensez-vous de cette méthode ? Est-elle sécuritaire ? Quels sont ses avantages et ses inconvénients par rapport à un relais triangulé ?
Je n’avais encore jamais vu cette technique et on m’a toujours appris à trianguler, donc vos avis m’intéressent.
bonjour, je suis sur Marseille , je cherche des soirée danse dans un style pas trop éloigné de celui que je connais, le rock 4 temps.
si quelqu’un a des recommendation je suis preneur
I'm the developer of the app. Feel free to ask any questions (there is an FAQ at the link)
I'm the developer of the app. Please ask any questions. Here is an FAQ: https://fractale.itch.io/zero-loss
jxl_coder 1.0.0 brings a cleaner API for reversible JPEG to JXL compression in Flutter and Dart, with byte-exact JPEG and metadata recovery.
Supports iOS, macOS, and Windows.
From mobile, I can send Codex instructions to multiple connected computers and switch between hosts. On Windows desktop, the app can be a host, but cannot control other machines.
It would be very useful to manage multiple Codex computers directly from the Windows client.
I’ve seen a few comments here saying that using Codex on Windows costs more credits because PowerShell commands are longer.
I’m not totally clear on how credits are calculated. Is it based on input tokens, output tokens, or both?
I understand that PowerShell commands are often more verbose. For example, ls versus Get-ChildItem. But I would assume most of the token usage comes from the code/context sent to the model, plus the model’s response, rather than the command-line syntax itself.
So my intuition is that the difference should be negligible overall.
Am I missing something here, or are people mostly optimizing for something that does not matter much in practice?