• 0 Posts
  • 39 Comments
Joined 1 year ago
cake
Cake day: June 1st, 2023

help-circle

  • I’m far from an expert, but I don’t know of rclone doing versioning, or a continuous sync like syncthing. Also haven’t used proton, so take my thoughts with a grain of salt.

    Stage 1 Run rclone config to set up the proton remote. rclone config should take you through a wizard and will eventually ask you to authenticate somehow with the remote. Once that is done and saved, you’ll exit the rclone config wizard and be back at the command line.

    Then you would run a test command like: rclone ls :

    If it worked, you should see a list of files/folders on Proton. If not, you’ll have to go back to rclone config and edit the remote to fix whatever went wrong.

    Stage 2

    Test out copying the folders with a command something like: rclone copy localfile/folder remotename:remotepath

    Do some testing to get the hang of the command, but it is pretty straightforward.

    Stage 3

    I don’t know how many files or how big the files are, but I assume not too many and not too big. I also don’t know which version of Linux you have, but I assume you have access to systemd, cron, or both.

    You’ll make a basic shell script that runs the command you practiced in stage 2. Easy peasy, put it in a text file with a shebang at the beginning, make it executable, and give it a go. It should run exactly how it did when you typed the command out manually.

    Finally, you will write a systemd timer or a cron/crontab entry to execute that script at some frequency.

    So just to summarize:

    1. Setup the proton remote in rclone using rclone config
    2. Test out copying files to proton through rclone
    3. Write a basic shell script that runs the command to copy files from the desired local folders to the desired proton folders.
    4. Use one of the tools on Linux that lets you schedule the execution of scripts to automate running your copy to proton script as frequently as makes sense to you.


  • This is not about quality and costs

    It is about quality and cost for the majority of purchasers that worry about meeting a budget. Virtually anybody making purchase decisions on some sort of surveillance system will grapple with that issue. My point is that we all tend to want the best performance for the least cost, and breaking that habit for the less tangible purposes of domestic security or human rights somewhere else is why we will continue to see these articles about Hikvision/Dahua cameras getting deployed at times and in places they probably shouldn’t.















  • There isn’t off the shelf software to run things like Reddit, and the work to make that happen is pretty staggering. That isn’t to say there isn’t frivolous spending there - I have no idea.

    Lemmy has been developed since 2019 and the software crumbled when network-wide users spiked into the ~75,000-ish monthly range when some vocal Reddit users sought greener pastures over the app/api issue last year. A lot of talented new developers contributed scalability fixes that were obvious to them (but not obvious to the main devs), and we now have the largest Lemmy server handling ~10,000 monthly users without crashing. The work that has gone into making Lemmy, an open-source Reddit alternative written in Rust (vroom vroom) handle the waning spike of Reddit users fleeing, was substantial. Look through the lemmy github issues discussions page and merged closed contributions/discussions for that journey. Those people were largely contributing time and expertise for nothing in return. Imagine paying a market rate to all of the people who contributed substantial time into the betterment of Lemmy. By the way, Reddit was open source: https://github.com/reddit-archive/reddit

    Takeaways so far: this is a hard problem, even today with faster software and hardware - and Lemmy needed a diverse set of contributors to get its largest server stable at 10k monthly and ~50k across the network.

    Reddit had 46,000,000 monthly active users in 2012, ~7 years after launch. Reddit has 330,000,000 monthly active users today. My guess is that Reddit employs a lot of smart software engineers that are needed to contribute solutions that allow the site to serve an ever-growing user count without major outages with new features rolled in. Meanwhile, the vast majority of Reddit users will never pay a thing to Reddit and it isn’t a good platform to deliver advertising through.

    My point: It is easy to gloss over the staggering amount of work, talent, and skill that goes into supporting a site that operates at this scale. Reddit is around the 10th largest site in the US (8th if you exclude search engines) and 12th globally excluding search engines.