• 0 Posts
  • 18 Comments
Joined 6 years ago
cake
Cake day: August 21st, 2019

help-circle
  • I’ve been trying helix and the built in LSP integration is excellent. Out of the box it’s capable of a lot when paired up with a good language server. The equivalent LSP integration on neovim takes some fat plugins.

    I will say helix seems immature in a few areas. I noticed it’s missing tab layouts present in vim. Also might it not be impossible to resize splits? There’s also no linewise selection mode.

    Helix out of the boz is close to my 15 years of accumulated vim config, but it seems like it’s not quite there yet.






  • Roast brussel sprouts:

    • quarter brussel sprouts until you have enough to cover the bottom of a cast iron or carbon steel oven safe pan without crowding; a glass casserole dish will also work
    • toss in cooking oil of choice, salt, pepper, herbs of choice, e.g. herbs de provence
    • 420F for 30 minutes, stirring at 10 and 20 minutes, optionally sprinkling with shredded parmesan at 20 minutes

    End result should be crisped on edges. Probably the side dish I’ve made more than than any other.




  • Corporate social media is a vastly more effective, bidirectional successor to legacy corporate media and thus is used to the same ends, i.e. manufacturing consent for broad goals of the bourgeois state. But it can also shape behavior. Facebook published a study years back that they know which candidate individuals are likely to vote for, and they demonstrated that delivery of apolitical voting reminders at election time has a statistically significant outcome on whether the test group ends up voting, and so it follows Facebook has an ability to influence election outcomes.

    Beyond narrative management and behavioral control, corporate social media also has well documented ties to the u.s. intelligence apparatus.

    Shoshana Zuboff makes an interesting case in her book that surveillance capitalism is a new business model in which the behavioral data of populations is harvested and sold by social media companies as a new raw material and ultimately used to predict and shape consumer behavior. There’s not a fundamental difference between use of such information for marketing or for political manipulation.

    Capitalist social media should be destroyed and I believe most people are fools for using it. I do believe that a form of social media has value and would/should exist in a just society that is accountable to regular working people.



  • You want images not on local app server disk; do you want to host the files yourself (I.e. on VM disk)? Or are you open to a managed service like s3? What disk consumption are we at? Number of files? What kind of request volume and data throughout are we talking? How fast are those things growing?

    Depending on your hosting, S3 may well be less expensive, and it will def be simpler.

    If you’re forced to DIY you could look at seaweedfs, which is an HA horizontally scalable blob storage cluster which exposes an S3 compatible interface.

    Either way, someing like Pictrs might be nice over some kind of S3.


  • My novice take: the waveforms on the right are sufficient to recreate analog stereo.

    Digital audio signals are fundamentally represented by a series of numbers with a frequency corresponding to a fixed sample rate. For example: 16-bit unsigned integers at 44.1kHz. Each number is a sample and represents only a magnitude/intensity in time, forming a time domain waveform. Tones and pitches, and everything else one can hear are represented by these numbers in varying arrangements. Check out the waveform for a sine or triangle tone: the waveforms literally depict those shapes.

    The optical analog waveform is encoding the same information as the sampled digital one described above. I’m not sure of the mechanism by which an optical system translates the analog waveform into sound, however.