Category Archives: News

Article note: Good. They affirmed the only sane regime, as long as the "No, you can't point big dynamic LED signs at passing drivers" rules are applied in a content-neutral way, you can forbid those obnoxious blinkenlight signs.

Article note: Attempt to conceal the scale of the user revolt? Too many users running scraper and/or deletion scripts? Attack from overzealous angry users? Wider hosting issues? Algorithmic ranking freaking out because so much is private? Rank incompetence? We may never know.

Article note: Ol' Ted is an interesting figure, especially because of how much prominent thoughtful technologists tend to have such a "Wellll... he's not wrong" attitude about him.

Article note: The _case_ for decentralization is easy, and a huge portion of cultural and technical internet people agree. The _implementation_ is incredibly hard on both technical and social grounds. Old-school single-topic forums with a benefactor (which HN sort of is) are subject to the distortions and whims of the owner. Commercial ones are subject to distortion by the forces trying to turn it into a profit vehicle (and typically their advertisers). Any system that accepts money from users has to deal with payment processors and the moralizing BS from the political wings they impose (and uh... this is the thing that cryptocurrency was _supposed_ to handle but as soon as it became primarily a vehicle for speculation and scams, it stopped being suitable for that purpose. Shut up cryptobros, it didn't work, and it's largely your own fault). But federated platforms still have a (likely smaller) funding problem, have a way harder moderation problem ("People on that other instance have different culture and ideas than on ours! Unacceptable!" is already causing minor chaos with Mastodon), and will inherently end up exposing some of the structure to users which ...people have/had a hard enough time with things like G+'s circles or posting into appropriate subreddits, knowing about sites/servers is a shockingly big ask (even though everyone is acclimated to email and websites). As tech folk we all naturally think the ideal is that every person or small group runs their own instance that interacts with others via a protocol, but even given that someone builds a system cheap and easy to host, that's _way_ beyond the vast _vast_ preponderance of likely users. RSS is one-way and way easier to do, and there are lots of nice canned hosted options... and it's still a weak ecosystem that only a small fraction of us cling to post-google-reader.

Article note: Wow, I expected a trainwreck, but this is even worse than I anticipated. 14 total posts from spez in the whole thread over the course of several hours. None of them answer any of the major questions beyond clearly stating that they're going to turn user activity into money or die trying. Some of them contain slander that were replied to (again) with receipts. And all the answers are rightly so downvoted it's easier to find them via https://old.reddit.com/user/spez than in the thread.

Article note: Kinda hoping I find a more productive use of the time I spend idle-scrolling on my phone after this (and the upcoming blackout of most of the site) forces me to break the routine. The first party reddit Android app is a collection of dark patterns that I refuse to engage with, and their mobile browser experience is nasty to an degree that seems intentional. I'm sure some new vice will fill in, but the alteration will be interesting.

Article note: This general trend of tech companies trying to profiteer off of being the middle-man for distributing user-created content, while simultaneously preventing the creators from monetizing (or users to access) content outside channels they can skim is not going to make for a good time. It's clearly driven by the drying-up of VC money and increasingly obvious over-extension of the advertising industry, but it's also obviously non-viable. Maybe if we're lucky it'll drive some decentralization and break up the chokepoints (eg. Youtube, Reddit, Twitch), but the economics and user experience to do that well are hard to get right.

Enlarge / Visual examples of the kinds of "burned in" sponsorships that Twitch's new guidelines won't allow. (credit: Twitch)

A recent update to Twitch's guidelines restricts many kinds of overlaid ads that streamers themselves can sell and insert on top of their own content. And while Twitch now says it "missed the mark with... policy language" that was "overly broad," many major streamers are still up in arms over a policy they say could severely affect their ability to make money from content on the popular Amazon-owned streaming platform.

In a June 6 update to its Branded Content Policy guidelines (that Twitch says it will start enforcing on July 1), the platform said that so-called "burned in" videos, display banners, or audio ads are not permitted on the platform. These ad formats are popular with many major streamers, who sell a portion of their screen real estate as part of overall branded stream sponsorship deals.

While basic sponsor logos will still be allowed to be shown on these streams, those sponsorships "are limited to 3% of screen size" under the new guidelines. That could be a problem for some disabled streamers because, as blind streamer Steve Saylor pointed out, that size "isn't accessible for low vision users as it is too small for overlays and even smaller on mobile."

Article note: Well that's fucking cool. I've had a chance to page through a couple old copies of Computer Shopper, and the gigantic shitty newsprint and laissez faire typesetting and just deep deep "of it's time"-ness is hard to match. Online won't quite capture the full experience, but I'm very glad to hear folks are making a real effort at preservation before they all finish degrading away.

Article note: Really interesting piece. I played with "early" rust a little, and it was a very obviously philosophically different design than what it has progressed into. This is the how and why.

Article note: Grammatically pleasing AI spew is exploiting the same mental shortcut as the well established "Anything looks credible if it's typeset in LaTeX" or "Printing resumes on nice paper for a subconscious boost" effects. We use secondary/contextual indicators, which are historically proxies for "How much time, expense, and/or expertise was involved in preparing this document" to judge documents. It's not even that different than printing eliminating the quality of the scribe's work as a tell. Which is interesting.