Can you elaborate on what you mean by "using this on existing bookmarks"? It might be a misunderstanding. Have you ever used bookmarklets? Bookmarklets are javascript code that you convert into a bookmark. Then when you click on the bookmark, the code will be executed on the page you are currently on.
Ahh, I understand the workflow now. This is great, thanks for sharing. May I ask, with the Ruler example, upon executing the script and seeing the ruler appear on the current site, how could the script then be terminated so that the site returns to operation without the ruler?
There are websites where you can upload an image and have it converted to an .svg. I've used Python (not sure of your technical background) to extract the paths from the .svg and then you can alter/play with them as you need. I cannot think of a one-step solution to your problem. But the path extraction is the way I'd do it given my background and inclinations. I'm not at home but do have a python program that does path extraction from .svg files.
Yes but it wouldn’t look very good - you would see fine zigzags everywhere - they’re never going to blend into a solid black fill, well maybe with a thin brush pen. If you draw directly into a vector software so you have centreline paths you could plot them with a thick marker ?
Given the time it takes Twitter to load (whether that's poorly-optimised JS or deliberate slowing to make their app more appealing) it's usually quicker to just replace the twitter.com in the URL with nitter.net. I have
Yeah, I saw that. The problem is with how he's testing it - he's generating all the fizzbuzz output inside a docker container, then passing that to the host to measure the speed. That alone is very bad for performance, but also logging is not disabled so every single line is getting written to journald as well! I get 1MB/s when I try it.
I'm a bit confused the environment the tests run on is not standardized. Using a different docker image for every entry could completely skew the results. Let alone sometimes using a container and sometimes running the code on the host os. I posted a question about it.
It would make sense to use a consistent approach certainly. Tbh, running all submissions in docker might not be a bad solution as the submitter can configure the image as they want.
I got bored on a Friday afternoon and didn't want to work. I hit as high as 14 MiB/s using Node 18 (including increasing heap memory) on a M1 MBP. Although, there were a lot of times when pv seemed to lose the output (the throughput would drop to 0 but the CPU usage would be really high)
A project like this already exists, check this article on Wired:
Well, I wouldn't upload my Twitter Archive to some website.
is the value based on annual revenue or market cap?
Market cap
Anything by Euell Gibbons.
For "Euell Gibbons" I see results like "Stalking The Wild Asparagus" and similar titles on most shops I tried. Any specific book that is hard to find?
Care Santos’ Crypta (Crypt) is near impossible to find even within Latin America.
Isn't it this one?
So nested function calls in JavaScript …
Even assuming you could avoid the issue with comma-operator confusion with
d(c(b,7),8)
I've had a look, but I don't quite understand the workflow for using this on existing bookmarks, say, within FireFox?
Can you elaborate on what you mean by "using this on existing bookmarks"? It might be a misunderstanding. Have you ever used bookmarklets? Bookmarklets are javascript code that you convert into a bookmark. Then when you click on the bookmark, the code will be executed on the page you are currently on.
Ahh, I understand the workflow now. This is great, thanks for sharing. May I ask, with the Ruler example, upon executing the script and seeing the ruler appear on the current site, how could the script then be terminated so that the site returns to operation without the ruler?
I made it so that a click on the ruler should terminate the script. But the mouse is sometimes not exactly on the ruler. A few clicks should work :)
https://www.reddit.com/r/ProgrammerHumor/comments/9ln0iy/we_dont_do_that_here/
Good point. Took out the two semicolons.
Unfortunately searching for "@
how practical is this though? what's the compatibility rate?
According to CanIuse it is supported by every browser except IE and Opera Mini:
Well done! The chinese (?) letters give it a nice mystic aura.
There are websites where you can upload an image and have it converted to an .svg. I've used Python (not sure of your technical background) to extract the paths from the .svg and then you can alter/play with them as you need. I cannot think of a one-step solution to your problem. But the path extraction is the way I'd do it given my background and inclinations. I'm not at home but do have a python program that does path extraction from .svg files.
So far, the best solution I found is vpype hatched:
Yes but it wouldn’t look very good - you would see fine zigzags everywhere - they’re never going to blend into a solid black fill, well maybe with a thin brush pen. If you draw directly into a vector software so you have centreline paths you could plot them with a thick marker ?
A thick marker would be pretty cool.
Given the time it takes Twitter to load (whether that's poorly-optimised JS or deliberate slowing to make their app more appealing) it's usually quicker to just replace the twitter.com in the URL with nitter.net. I have
This post is actually a follow up to my yesterday's post, where I linked to my bookmarklet which replaces twitter with nitter:
Whats the advantage over just using the Redirect Addon for this purpose which is actually way more versatile?
I prefer bookmarklets for a few reasons:
Don't you have to add a bookmark for every tweet or at least profile page though?
No, this one bookmark works for all Tweets.
Yeah, I saw that. The problem is with how he's testing it - he's generating all the fizzbuzz output inside a docker container, then passing that to the host to measure the speed. That alone is very bad for performance, but also logging is not disabled so every single line is getting written to journald as well! I get 1MB/s when I try it.
I'm a bit confused the environment the tests run on is not standardized. Using a different docker image for every entry could completely skew the results. Let alone sometimes using a container and sometimes running the code on the host os. I posted a question about it.
It would make sense to use a consistent approach certainly. Tbh, running all submissions in docker might not be a bad solution as the submitter can configure the image as they want.
Great. I also replied now. Let's see what comes out of it.
I did the 3 PHP versions and I'm FizzBuzz exhausted now :)
I got bored on a Friday afternoon and didn't want to work. I hit as high as 14 MiB/s using Node 18 (including increasing heap memory) on a M1 MBP. Although, there were a lot of times when pv seemed to lose the output (the throughput would drop to 0 but the CPU usage would be really high)
I think you need to indent the code with 4 spaces to make Reddit display it correctly.
:13383:
https://i.imgur.com/ccntU1w.png
I maintain this chart of small Android phones:
Galaxy A01 Core is missing.
Good point, will add it!