• 27 Posts
  • 102 Comments
Joined 1 year ago
cake
Cake day: June 12th, 2023

help-circle
  • KOReader is by far better than the crappy stock firmware from Kobo. While the interface is not the prettiest, it still has a lot of advantages :

    • it adds the ability to browse the filesystem (how do people use an e-reader without folders ?)
    • loading medium to large PDFs takes ages in kobo’s stock UI, while it’s almost instant in koreader
    • there are a bunch of plugins you can add to koreader

    While I really hate Kobo’s stock UI, I still recommend getting one if you like truly owning your hardware. It’s really easy to enable ssh access and then it’s just regular Linux. It’s even possible to run an X server and launch Linux graphical apps on the e-ink display (not quite usable though)



  • The Wikipedia page says the following :

    On January 28, 2015, the ACME protocol was officially submitted to the IETF for standardization.[28] On April 9, 2015, the ISRG and the Linux Foundation declared their collaboration.[9] The root and intermediate certificates were generated in the beginning of June.[29] On June 16, 2015, the final launch schedule for the service was announced, with the first certificate expected to be issued sometime in the week of July 27, 2015, followed by a limited issuance period to test security and scalability. General availability of the service was originally planned to begin sometime in the week of September 14, 2015.[30] On August 7, 2015, the launch schedule was amended to provide more time for ensuring system security and stability, with the first certificate to be issued in the week of September 7, 2015 followed by general availability in the week of November 16, 2015.[31]

    So we’ll have another anniversary to celebrate in nearly a year










  • Hey ! I (superficially) looked up the Belousov-Zhabotinsky reaction and can confirm it is related to the Gray-Scott model.

    The Belousov-Zhabotinsky reaction involves an autocatalytic reaction and chemical species diffusing at different rates, just like in the Gray-Scott model. The main differences are related to constraints of doing actual chemistry instead of simulating it :

    • The speed constant is (roughly) fixed for any given reaction (and temperature). Scientists cannot tune speed constants like I did in the simulation
    • In the simulation, we constantly add some “food” and remove some catalyst. In an actual chemical reactor, there must be an process to achieve this. A real world implementation of the Gray-Scott model would probably use something like a semi-permeable membrane above/below the petri dish. In the Belousov-Zhabotinsky reaction, it is other chemical reactions that ensure “food” gets replenished and the catalyst gets consumed









  • I can recommend some stuff I’ve been using myself :

    • Dolibarr as an ERP + CRM : requires some work to configure initially. As most (if not all) features are disabled by default, it requires enabling them based on what you need. It also has a marketplace with a bunch of modules you can buy
    • Gitea to manage codebases for customer projects. It can also do CI but I’ve not looked into it yet
    • Prometheus and its ecosystem (mostly promtail and grafana) for monitoring and alerting
    • docker mail server : makes it quite easy to self host a full mail server. The guides in their doc made it painless for me to configure dmarc/SPF/other stuff that make e-mail notoriously hard to host
    • Cal.com as a self hostable alternative to calendly
    • Authentik for single sign-on and centralized permission management
    • plausible for lightweight analytics
    • a mix of wireguard, iptables and nginx to basically achieve the same as cloudflare proxying and tunnels

    I design, deploy and maintain such infrastructures for my own customers, so feel free to DM me with more details about your business if you need help with this