@jayrope zu Vimeo: das ist ein professioneller Account der Universität, und das Streaming richten die auch ein für uns. Vimeo ist sicher weit weniger problematisch als YT. Streaming über PeerTube ist noch in der alpha/beta Phase, das ist gerade erst in der letzten Version hinzugekommen, und da habe ich Null Erfahrung, und es käme auch auf den Server an. Die Aussage zu diode.zone ist jedenfalls, dass ohne ein CDN sicher nicht viele Leute gleichzeitig den Stream in hoher Qualität abfragen können.
Invitation to the online Signale Lecture on "Algorithms that Matter" #almat - our concluding project presentation, featuring a preview of the #AlgorithmicSegments #SoundArt parcours video documentation, and the new Meta Exposition interface.
Wed 21 April 2021, 18:00 CEST (UTC+2)
It's also a much more involved endeavour than forking GNU Social and creating Mastodon. git is a much more complex existing infrastructure, there are other concerns such as security that apply less in the case of a micro blogging service.
I have just discovered the project, so can't say for sure, but it appears to me that it seriously lacks human resources invested and engaged in the project. The project is scattered across twelve-and-a-half different websites and forums. And there are too many stakeholders each with their own goals and governances. They would need to find both an institution like Framasoft, a public funding, and at least one hosting platform that would collaborate on creating an MVP. It's not like you have one developer and they can grasp and extend the codebase of a project like Gitea.
@jayrope GitHub is not a problem here. There are enough serious competitors to have interest in developing a federation for git-based software hosting. We also didn't need consent or collaboration from Twitter to create Mastodon.
@jayrope Yes, but it needs an effort to interconnect GitLab, Pagure, Launchpad, Gogs/Gitea, SourceHut. Otherwise not much is gained from switching platforms.
@jayrope it's not so much about the sizes of instances. GitHub works really well as a social platform, the main problem here is its almost-monopoly and corporate business model. What I wanted to say is, this kind of libertarian underwriting - there are no entities, just peer atoms - appears to me quitly strongly ingrained in these types of projects ; almost all the distributed, decentralised source hosting initiatives somehow come out of a stallman'ish reading. I would start from the opposite site. It's great if your protocol has a solution to broadcasting merge requests across nodes, but the whole point of connecting to each other is to build communities, to discover things, etc. GitHub is way more than a platform for coders, it allows many people to collaborate that would not describe themselves as coders; you can work on texts and collections there, you can come as a software user that doesn't write code, etc. If you don't start your design with this, if you cannot even create a code of conduct for your project and think of the question of moderation, it will never take off.
“Hypothetically, a federated GitLab could fill all the requirements, however, federated services cannot be offline-first and don’t offer sovereignty over user's identity. Users are tied to specific instances and thus subject to some of the same drawbacks as centralized services.” ( https://docs.radicle.xyz/docs/understanding-radicle/why-radicle )
I disagree. Having instances with a particular group of peers, codes and culture is the strong point of federation. The #Radicle design shows the flaws of purist free software and purist p2p ideology with an impoverished idea of social interactions. Operating Tor or Bittorrent is the most asocial activity I can imagine. I truly hope #ForgeFed ( https://talk.feneas.org/c/forgefed/10 ) will succeed at some point.
I also don't see why federation means it cannot be offline. There is no reason a federated git client couldn't cache the status and activities. There is a bit too mucher prepper aesthetics here.
Max Mathew's MUSIC V (Csound predecessor from the late 1960s) source code: https://github.com/vlazzarini/MUSICV
Ok - seems not the original code, but adapted to newer Fortran. But that means it actually runs as far as I understand.
Here is a new text-sound-image piece I made with Nayarí: Divagation 2
@exquisitecorp thanks for both links, definitely helpful. I also saw that the API is actually no so complicated, once you understand REST and have an OAuth token. So I think I will be able to go on from here.
I guess I should put circle markers on the glass slide directly, otherwise if the slides move, the separate circles don't help much
blast from the past - I worked with Hough transform in 2016, although the line based one. There is a variant, circle hough transform. Found a ImageJ plugin that I hope I can just extract the algorithm and run it headlessly. Seems to work despite the perspective distortion of the circles.
@jayrope I want to be able to undo slight camera movements over time (weeks, months), so I'm taking the four cross-hairs as fixed, and to be used for post processing, in order to be able to crop the same portion of the photo now in four months.
CSI = crime scene investigators
some CSI vibes there. Anyone know a Java or Scala image processing library to quickly do the perspective alignment / equalisation of the markers?
Sound Artist, Composer, Researcher
The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!