As noted in the OP, “use by” and “best before” (etc.) dates are vastly different from each other. “Best before” is very often ad hoc like “a year from the date of manufacture”, but if you’re past a “use by” date on your perishable food, you should approach it skeptically.

Climate Town has a good video on it and why this system exists as it does.

Those “last two data points” are two months. For OS X, there’s some upward movement the last two months (about 3%, which isn’t a lot given it spiked late last year from 8.3% to 14.07% to 8.2% in a span of three points), while Linux is practically flatlined with a very light upward incline just the last month.

To the extent StatCounter is useful to begin with, the Linux point is functionally meaningless while the OS X increase of a few percentage points may or may not be significant; trying to divine a trend from twitches in those two points is reading noise. Look at the “unknown” line, meanwhile, and see how it begins mirroring Windows’ decline around mid–late 2025. It rose from 11.3% in September 2025 to 21.5% in June 2026. That represents actual, meaningful, lasting change.

Okay, but StatCounter’s graph (edit: here) clearly shows that “unknown” is what’s cannibalizing most of that – and this is specifically measuring web traffic.

“Lemmy celebrates the rise of bots polluting web traffic stats because they didn’t bother to read the article beyond a misleading headline that reaffirms their biases.”

Unrelated to the question, but I think the other commentor got it right, so I’ll say this instead. If you’re doing GasBuddy, it would be amazing if you also added data to OpenStreetMap where applicable. We don’t keep track of the current price, because it’s supposed to be semi-static infrastructure, but we do keep track of payment methods accepted and types of fuel offered.

We keep track of other amenities too like compressed air and vac stations, but just adding the payment, fuel, etc. data to the existing station object would already be above and beyond and so appreciated.

That said, if not, it’s still cool knowing that you’re also out there helping to crowdsource knowledge.

How is it burying the lede when anyone with basic reading comprehension knows that’s what the headline’s obviously suggesting? Like sorry your reading comprehension is apparently trash, but that doesn’t mean GamesRadar did something wrong.

«Works for me» is seldom a good advice.

That wasn’t the advice; the advice was “try offline tiles and see if that helps.” In their case, it didn’t, but it’s just covering arguably the most likely potential cause. The fact it wasn’t their problem doesn’t mean it wasn’t worth asking.

“Takes seconds” seems like a strange experience. I remember OsmAnd years ago performed like that for me – clearly, painfully loading in the individual tiles. But nowdays, it smoothly transitions between LoD and has no problem smoothly scrubbing over e.g. a major city.

I’m using offline vector maps, for context.

Edit: Trying online tiles again, I’m assuming your problem can be resolved by switching to offline. You can have the offline data update automatically such that you don’t have to worry about it. CoMaps and Organic only use offline maps, which is why they’re similarly snappy to offline OsmAnd.

This is under Configure Map > Map Source, and you only need to download for the region (usually e.g. a province) that you’re using.

Something I’d suggest if OsmAnd feels too cluttered for you is to change the settings; OsmAnd lets you change a lot, but one of the ones I do is to change “Map Style” to Osm-Carto.

Carto, for context, is the vector map that you’d find by going to the OSM website. Much cleaner color scheme, imo.

Yeah, OsmAnd is really good; it’s what I use as my daily driver. CoMaps/Organic to me feel too limited, but some people may like that.

(I use Vespucci for editing on Android.)

Would switch Organic Maps to its fork CoMaps. (See: this open letter)

And I would never recommend Brave as the first choice; it’s run by a shady corporation and reinforces Chromium’s hegemony.

I think it’d also be reasonable to add ProtonMail to email and Mullvad to VPN since you can have multiple.

Formally on reddit?

Yes. I even write semi-formally on Mastodon for PCSX2 (or wrote; on hiatus) within its constraints; a PR team attempting to apologize for something will normally resort to formality regardless of the forum.

You’re giving off major “A 10-page essay before AI-assisted writing? As if!” vibes. I’m sorry basic PR etiquette is inconceivable to you personally. The Wendy’s Twitter account is that way if you want to soothe your preconceptions about PR on social media.

I write formally all the time; smells like a PR team writing formally.

Ah, yes, reaffirming that someone raising an issue is correct to have done so is the telltale sign of an LLM. Couldn’t just be basic professional writing etiquette that LLMs were trained forwards and backwards on; it has to have been written by an LLM.

You don’t actually write formally very much, do you?

Sky Map appears to be vibe-coded to hell and back.

Yes. An LLM is just a model trained on a large corpus of text, which DeepL absolutely is. It uses a transformer architecture – which literally could not work without it.

Nah, I’m with them. The average Lemmy user is stupid as fuck and willing to believe anything that even vaguely conforms to their biases, and I 100% believe the majority of upvotes are from people who thought this screenshot is real.

Several weeks ago, there was this post in a news community to a BBC News article. It was a real article; no tricks, correct headline. But the link was a 404. When I found it, it was upvoted about 20–0. I downvoted it as obviously nonfunctional but also commented remarking that it’s a 404 and giving the OP the correct link. When I came back to that post a couple hours later hoping to upvote a fixed link, the link was unchanged, it was upvoted 50–1 (my downvote), and it had one comment (mine) upvoted 1–0. (Edit: I checked, to preemptively clarify, that this wasn’t a “me” problem.)

The lack of scrutiny Lemmy applies as a collective is fucking appalling, and the level of introspection they have about it is somehow even worse.

“There’s more than a 50% chance that Mary-Kate and Ashley Olsen were sexually molested by multiple (you decided to pluralize it; not me) Epstein affiliates in Hollywood” is a massive stretch that you present zero evidence for and confidently assert. It’s becoming increasingly clear with your responses that you genuinely have a problem.

Did I make you uncomfortable with consuming the lives of the Olsen twins like a cheap and disposable drug?

Dude holy shit, seek help. I have never once in my life given an iota of a shit about the Olsen twins or anything they’ve been involved with. I might’ve seen one episode of Full House by coincidence, and I played an nth-hand copy of Mary-Kate and Ashley: Sweet 16 – Licensed to Drive for the PS2 two decades after its release (when the twins were nearly 40 and the publisher and developer had been defunct for years) for about 30 minutes as a gag. This is the first time I can remember seeing a picture of the Olsen twins in years, and I only responded to your comment because it’s so insane (I genuinely knew so little that I assumed I’d completely missed out on a scandal and tried in good faith to find out what happened).

Your willingness to assume anything without evidence as long as it feels correct to you is staggering.

Removed by mod

my SPECULATIVE STATEMENT.

Pfffffffffffffft. What a chode. “I hedged with ‘most likely’ bro so you can’t poke holes in the ridiculous bullshit I’m asserting with zero evidence.” You’d be really good at JAQing off. “Were the Olsen twins sold into sex slavery on Epstein’s Island? We can’t say for sure, but it’s impossible to rule out.”