The nationwide fuckup in the US is zoning rules that block commercial venues from residential regions, which means people cannot step outside their front door and get groceries in a 1 block walk. People are forced to travel unwalkable distances to reach anything, like food and employment. Which puts everyone in a car. Which means huge amounts of space is needed for wide roads and extensive car parking, generally big asphalt lots, which exacerbates the problem because even more space is wasted which requires everything to be spread out even more, putting resources out of the reach of cyclists. Making the city mostly concrete and asphalt also means water draining problems where less of it makes it into the soil and groundwater, and it means the city temp is higher because of less evaporative cooling from the land mass (Arizona in particular).

This foolishness is all done for pleasant window views, so everyone can have a view of neighbors gardens instead of a shop front.

Europe demonstrates smarter zoning, where you often have a shop on the ground level and housing above it. You don’t need a car because everything is in walking or cycling distance. But you more likely have an unpleasant view.

By size, you are referring more specifically to area. Area while neglecting population is inversely proportional to population density¹. But even apart from that – how does that support the claim that it’s sensible to disregard cities and just look per capita nationwide? NYC should be compared as a single whole city against other cities of comparable population density. Area does not matter as an independent variable on its own. What would the point be to blur NYC into a nationwide track per capita?

BTW, NYC has a subway system. I’ve used it a few times and it was not even close to being overcrowded but maybe I had lucky timing. Are you saying more track is needed there?

¹ population density: heads per m²

Subways are pretty much exclusively built in the cities

Not just any city. Dense cities. Cities that are so densely populated that it would be /impossible/ for every person to move around in a car. Countless US cities are not even close to crossing that threshold. It just makes no sense to look at nationwide per capita on this. Only a city by city comparison of like with like population density is sensible.

(edit)
There is a baby elephant in the room that needs mention: US cities are designed with shitty zoning plans. They are designed so that each person on avg needs to travel more distance per commute to accomplish the same tasks (work and groceries). This heightens the congestion per capita. So ideally we would calculate daily net commute distance needed per capita plotted against subway track per capita for cities of comparable people per m². Which would embarrass US city mayors even more.

From the article:

See the U.S. flatlining in transit miles per capita

A devil’s advocate would rightfully argue that that’s expected given the much lower average population density of the US – the same factor that made it a struggle to get broadband Internet to everyone in the US. Bizarre to use a nationwide per capita as a basis for mass transit comparisons. It should be a city-by-city comparison that groups cities by comparable population density. US cities would likely still come out behind and embarrassed, but more accurately so.

Consider the marketing angle – instead of saying “the US is losing” (which diffuses responsibility and makes plenty of room for finger-pointing), instead say “@conditional_soup@lemm.ee’s city lost its ass in the bi-annual city infra competency competition”. Then that mayor has some direct embarrassment to pressure action.

We did some behind the scenes changes with the firewall setup which will make it easier to identify and block scraping attempts and other such abuses of our technical setup.

Why is scraping being treated as abusive? I have actually been on the look out for a tool the scrape Lemmy servers so I can grab a copy of all my own messages, replies thereto, and parent messages of my comments. slrpnk can go down at any time. It would also be useful to be able to grep my own content to recall where I discussed something and to know where to expand on past discussions.

fwiw, here is an emacs version:

https://codeberg.org/martianh/lem.el#headline-11

I think what would be most useful would be a usenet→lemmy gateway, so that rich catalog of usenet clients can be leveraged on Lemmy.

One other Lemmy instance is on 0.19.4, and that also has the same problem:

Ungoogled Chromium: the create post form erases the fields after tabbing out of them. Tor Browser (firefox based): no issues with the create post form.

Now slrpnk is on 0.19.5, along with two other Lemmy nodes I tested this on. The create post form is working in this version with Ungoogled Chromium.

timeline selector broken

There is another problem that was introduced with version 0.19.4 and still persists in 0.19.5: There are four possible timeline views: subscribed, local, all, moderator view. That selector is broken in Ungoogled Chromium 112 but not in Firefox-based browsers. In UC, I click “moderator view” and the button highlights as I click it, the page refreshes, but the selector does not stick. It is trapped in the “local” view and only shows the local timeline.

This problem is replicated in other 0.19.4 and 0.19.5 instances.

update

Actually there is still a problem with 0.19.4 w.r.t. post creation. And this affects Firefox too: if I use the cross-post feature to copy the post elsewhere, the form is populated just fine but then I have to search for the target community at the bottom of the form. As soon as I select the target community, the whole rest of the form clears. I have not yet tested this specific scenario in 0.19.5.

I started with the “Browser Default” theme.

Which browser did you try to replicate it in?

i just switched to hanubeki cold. Same problem.

Also another problem I forgot to mention: the main timeline is trapped on the local view. When I click on all, subscribed, or moderator view, there is no effect. The button highlights in parallel to the local button, but does not stick. It just refreshes the local timeline.

not sure what you mean by “doesn’t prevent using the function entirely.” Do you mean the function is available for those who install another browser? My current theory is that the create post function is 100% broken for Ungoogled Chromium users, but probably operational for people running 3rd party clients or Firefox. Considering Ungoogled Chromium is based on the single most popular browser (Chromium), vanilla Chromium users are probably also impacted, which seems serious.

The form should probably display a loud warning about this. Someone might opt to write the body of their post as a very first step, before the title. They could put a substantial effort into writing a long text and then see it all vanish the instant they click the title field. Nothing is worse than seeing one’s work thrown away like that.

I would roll back to the previous version if there is no fix for this.

(edit) Are there any stats kept on user agent strings? It might give a clue on the number of users affected.

For the past few days, the “create post” form is usually broken. After entering a title I hit tab to the next field and the title is instantly erased upon shifting the focus to another field. The form is unfillable.

I was able to post once after this behavior started for no apparent reason. No idea how to reproduce that.

It’s worth noting that submitting comments in existing posts is not a problem.

Browser: Ungoogled Chromium 112

(edit) after more thought, I think that one time I was able to post was probably done using Tor Browser, not UC.

What’s interesting about this is #LemmyWorld uses Cloudflare, and CF was involved in a CP scandal. You might be tempted to report the CP to Cloudflare, but it’s important to be aware of how CF handles that. CF protected a website that distributed child pornography. When a whistle blower reported the illegal content to CF, CF actually doxxed the people who reported it. Cloudflare revealed the whistle blowers’ identities directly to the dubious website owner, who then published their names and email addresses to provoke retaliatory attacks on the whistle blowers! Instead of apologizing, the CEO (Matthew Prince) said the whistle blowers should have used fake names.

deleted by creator

dedicating an entire community to the stance seems a bit much

That’s crazy talk. There are so many dynamics to non-voting that I’m not even convinced that 1 community is enough. This one community has to accommodate:

• advocacy of not voting (Anarchy promotion)
• problems voting & countless forms of voter suppression (my situation)
• voter turnout discussions and consequences of not voting in nations where voting is obligatory (Belgium and Australia)
• hopelessness of voting in regions where voting merely legitimizes dictatorships (China, Russia)
• alternatives to voting (e.g. consumer actions)
• Non-voting as a means to compel positive change (e.g. 97,000 Michigan dems and pro-Palestinians threatening to not vote unless Biden changes his policy in Israel)

I’m not an anarchist but found it interesting to read their rationale for opposing voting and elections.

I’m personally very pro-voting

Then you should be very interested in the 2nd bullet. There are LOT of non-voters in the US and it leads to scumbags taking power (e.g. 2016 POTUS). You should want to see people discussing the low voter turnout problem that helps the republicans.

You can see below in this thread herein that someone got >10 downvotes for a perfectly reasonable and civil post by people who merely disagreed with their comment. Automatic censorship would be overly interventionalist. I would not want to participate in a forum that auto-censored like that. Downvotes did their job – pushed the msg low on the page for reduced visibility.

Thanks for the explanation.

Sounds like an anti-feature because the general assumption would be that the mod of that channel intervened in that specific comment. Ideally it would say something like “removed in bulk cleanup of malicious user’s content” and link to whatever modlog might capture it.

I’m a bit confused because what you originally described sounded like a way to disable autoplay. But your followup sounds more like blocking links to hosts.

If I were to run my own instance, blocking known autoplay hosts would work to the extent they are known, but then running my own instance would require having an unlimited connection. The autoplay problem is in part due to having a limited connection.

Appears so.

$ curl -i https://lemmy.ml/pictrs/image/8122d979-9b5a-4c2b-bced-a685a1e05030.jpeg
HTTP/2 200 
server: nginx
date: Mon, 22 Jan 2024 11:55:17 GMT
content-type: image/jpeg
vary: Origin, Access-Control-Request-Method, Access-Control-Request-Headers
last-modified: Tue, 24 Oct 2023 08:09:58 GMT
cache-control: public, max-age=604800, immutable
access-control-expose-headers: accept-ranges, date, last-modified, content-type, cache-control, transfer-encoding
accept-ranges: bytes
referrer-policy: same-origin
x-content-type-options: nosniff
x-frame-options: DENY
x-xss-protection: 1; mode=block

I don’t understand how that’s possible. If someone posts a link to news.sky.com and a viewer follows the link, they are taken to an external site outside the control of the kbin instance. I suspect that setting you describe could only affect what happens if a video is attached or embedded into a post.

Someone posted this link, which seems to imply 5 blocks audio and video both: https://support.mozilla.org/en-US/questions/1321021