@nickchomey That's interesting. I have an old moto X I could test that I didn't dig up (I thought about it, but I figured the newer Tecno Spark 8C would be in much wider use than a Moto X today).
I feel like I also owned some kind of Moto G phone at some point, although I think I gave it away and don't have it anymore, but I can see about that as well.
@andrewrk On the Tecno Spark 8C, this uses a fair amount of CPU up front (4.7s), but interaction is (relative to other sites using the same device) decently quick when you tap any target or scroll one of the pages if you scroll a page that has enough content to scroll (I didn't test the Itel P32 because the battery is basically dead and it needs to charge for a long time to be usable and I don't generally keep it charged)
I think that's really interesting. I'll add a bit about this in the post.
@julian@nodebb It's much faster than Discourse, but also slower than all of the php forums tested. Just trying it on the M1 and the Tecno Spark 8C, I get 0.3s/0.4s on the M1 and 3.4s/7.2s on the TS8C.
It's quite usable after the initial load on the TS8C. Clicking on links and loading individual threads is slower than on the php forums, but scrolling, clicking, etc., all work as expected, which can't be said for Discourse or many other "modern" sites.
It's the only site that has an LCP* that's much faster than the LCP (3.4s vs. 7.3s for the TS8C) so there might be a small optimization that could be done that would hugely improve LCP? The content mostly shows up, but then stuff ends up moving around a bit or something?
I didn't try to debug this, but it might be worth looking into seeing if there's an easy way to hugely cut LCP on slow devices since I'm told that this is a ranking signal for Google search.
@slightlyoff That would be great! You definitely know way more about this area than I do and I'd love to work with you on something. I'm not sure what, but we can probably figure something out :-)
@mpt Thanks! "constraint" comes from a quote, so I don't think I should change that, but I fixed the typo and added an explanation for holdback.
I'm not sure about making a post on line length. I might do it at some point, but (some) people get so mad when I bring it up. For some reason, people respond with the intensity that's normally reserved for a culture war issue.
@danluu somewhat confusing graphical issue on https://danluu.com/butler-lampson-1999/ where Software engineering/RPC/... show up in "Maybe" when they should be in "No". This is because each <tr> has two <td>s when it should have an empty one in the middle.
@cis_female Thanks! This seems like yet another Hugo markdown parser bug (I looked at my markdown for this and it indicates there should be an empty column).
I guess I can fix this by taking the HTML and copy+pasting it into my post and then fixing the bug in the table.
chat, what's your recommended way of learning the theory behind PID controllers and autotuning? i need one that doesn't involve a 150 page treatise on z-transforms and poles in complex plane. i just want to write a script that makes heating in this house tolerable
There've been a number of cases where a company was caught violating the GPL. AFAIK, there were no meaningful consequences in any of the cases I'm aware of.
Now that this has been tested and companies know that they can get away with open source license violation, why should we expect that companies take these licenses seriously? Maybe Amazon and Google will, but say we're talking about some random $5B company
I don't have a lot of confidence that they'll try to comply.
Just for example, Wix blatantly violated the GPL, taking and using quite a bit of WordPress code and even initially calling themselves WixPress (https://ma.tt/2016/10/wix-and-the-gpl/). From a cost/benefit standpoint, AFAIK the cost of this was 0 and the benefit was a billion dollar a year company that's worth many billions.
At most, the cost was that a few nerds might not use Wix out of principle, but those nerds aren't the target of Wix anyway.
I'm doing mobile performance benchmarking on various slow phones (ones sold in "developing" countries), on a low-latency high-speed internet connection (a computer on WiFi in the same location gets ~1Gbps with a few ms latency).
Using default themes, which blogging platform do you predict has the BEST performance?
I'm doing mobile performance benchmarking on various slow phones (ones sold in "developing" countries), on a low-latency high-speed internet connection (a computer on WiFi in the same location gets ~1Gbps with a few ms latency).
Using default themes, which blogging platform do you predict has the WORST performance?
I'm doing mobile performance benchmarking on various slow phones (ones sold in "developing" countries), on a low-latency high-speed internet connection (a computer on WiFi in the same location gets ~1Gbps with a few ms latency).
Using default themes, which forum platform do you predict has the BEST performance?
I'm doing mobile performance benchmarking on various slow phones (ones sold in "developing" countries), on a low-latency high-speed internet connection (a computer on WiFi in the same location gets ~1Gbps with a few ms latency).
Using default themes, which forum platform do you predict has the WORST performance?
I guess I can read the source code for wpt, which must have a way to automate this, but is there, documentation, some API name I should look up, or a simple example there?
I wonder if we'll ever go back to the level of verification effort Intel used to put in. I suspect not in my lifetime. Historically, AMD was much worse than Intel w.r.t. serious stability bugs and, rather than AMD catching Intel, Intel seems to have caught AMD.
Intel has not caught up to Nvidia in bugginess, but maybe that's something to look forward to in the next decade.
@rygorous It's hard to tell about prevalence from the outside, but I feel like getting a part that hadn't undergone sufficient manufacturing testing used to be extremely rare.
The cases I used to hear about (25 years ago) were things like, you overclock and overvolt your CPU for a year and then it starts being unstable, but those CPUs were never designed to be used that way, whereas now it's expected that many users will get "gaming" (or w/e) boards that do this and the parts won't work.
@rygorous There's a huge amount of pressure to not test enough since tester time is money and pushing parts into higher yield bins is also money.
I guess a related thing is that it used to be the case that almost all parts sold had a huge amount of margin and could be pushed significantly harder without many issues, but trying to squeeze the last bit of nominal performance out of chips has led to more chips being sold at or beyond their actual capability?
But, maybe 90% of the time, when I look for something from the 90s, there doesn't seem to be an easy way to get a copy of it.
I've heard that Google has copies of basically everything, but they don't share them (just the opposite, they removed the ability to query their cache), so most of this stuff is effectively lost.
On the abbr text, it looks like Chrome has a 1024 character limit. I'm not sure that I'll fix this one, but I'll try to keep this in mind for future posts since most people use Chrome.
I wonder what the long-term rate of "working-age adults who report 'serious' difficulty remembering, concentrating, or making decisions" is going to be.
This rate had been declining basically forever for the obvious reasons, but this has turned around since 2020, also for the obvious reasons.
@hyperpape I don't know why but, on my polls, Mastodon users tend to be pessimistic.
I think a not totally off the wall mechanism would be some kind of compounding issue or the belief that there's a threshold affect and a bunch of people are right on the edge of the threshold, but I think the general pattern is that Mastodon voters are super pessimistic compared to Twitter or Threads voters, so I agree that's probably the main factor as opposed to some belief about a particular mechanism.
Nice post by Gradon Hoare (original creator of Rust) about his preferences in language design vs. the Rust team and the resultant differences that fall out of the differences in preferences: https://graydon2.dreamwidth.org/307291.html.
It's rare to see someone who has a deep understanding of the trade-offs involved in major technical decisions write up their thoughts, so I thought this was a really great read even though I'm not really a "language person".
