These trolls really exist on Russian social networks, trying to refute any news from the opposition media. And some people (Russian enthusiasts) count and list these bots. But your article does not prove it to me, because some Russians are afraid of vaccinations and there are crazy paranoids in Russian social networks that convince not to get vaccinated, including vaccines that are done by Russian medicine - what do you think this proves? That Russian bots are trying to weaken their own nation? For me, these are not bots, but religious fanatics, religions often reject medicine. You may have noticed that many religions have denied the existence of COVID this year.
It is very sad that the authors of the article either believed in all these accusations against the Russians or are afraid to say that this is nonsense so that they are not accused of supporting the Russians.
I want to note that the Russian government also accuses its political opponents of being foreign agents, using the same rhetoric.
I don't have the general problems with reading, but I do have the problem you mentioned. And I'm also kind of a speed reader, I think the problem is here.
It reminds the index of the first element in an array in different programming languages. Some use one-based indexing, others use zero. Which causes problems (bugs) when porting the program to a language with different indexing.
Modern computers have some sources of entropy that can be used to create a random numbers. For example, the timings of the memory access, race condition of threads. User interaction (mouse movement, delays between keystrokes) can also be used as an input signal to feed a random number generator (which is some hashing algorithm).
My guess is that quantsmooth won't handle squares in the background well, but will be better at the edges (without excessive blurring).
Update: I tested it, it's about 10% quality, jpeg-quantsmooth doesn't handle this quality. At 25% quality - result is fine, starting at 20% and less - not good.
Perhaps the synthesis of all three projects combined into one (taking best of each) could give a great result.
So then...
jpeg2png: overblurs and slow
knusperli: does only deblocking, not removes artifacts
jpegqs: fails to deblock low frequencies (less 25% quality)
Well, I meant better upsampling than bilinear interpolation, which blurs chroma.
"Oh, you can also jointly optimize the YUV-RGB conversion because if the pixel ends up out of range in RGB space it is probably a compression artifact."
Probably not, I suspect that the mozjpeg encoder may use this, so images with black on white text look better (artifact noise go beyond of range).
Actually - JPEG is not lossless even at 100% quality, there is a little noise at +/-1 of pixel value, some rare pixels at 2. Grayscale is not needed, you may skip RGB->YUV conversion and encode in RGB colors, libjpeg can do that.
j19544519 - seems to be the username on this hosting.