floo

got that fixed, now. thanks again for pointing it out!

thanks for pointing that out. should have checked availability of view animation timelines.

will make sure it falls back to fully visible text.

https://developer.mozilla.org/en-US/docs/Web/CSS/Reference/P...

thanks! none of the audio is AI generated, it's all human narration, mostly sourced from librivox. (credits show when clicking the info icon)

the idea of learning languages from AI didn't quite sit right with me. but that might be something to circle back to.

integrating learner dictionaries does sound like a fantastic idea. will definitely explore that!

oh that looks amazing. can't believe I didn't find these during research. thanks for pointing them out!

cool idea. haven't really explored dialects yet. gonna see if I can find any latin american recordings. thanks for the suggestion!

thanks, that's a really good point. having some beginner friendly books for each language is definitely a goal.

are you looking for stories in a specific language?

glad to hear it!

went through quite a few iterations of aligning text to speech. found that ai transcription was really good most of the time but would hallucinate quite a bit towards the start and end of books. which I think might be related to those models being partially trained on audiobooks, and only having the book text itself, without any of the intro or credits.

in the end I landed on extracting text from ebooks, using rule based and language specific segmentation, and espeak based alignment. pretty basic, but it worked wonders in terms of reliability and accuracy.

if you are looking to generate audio from ebooks this is probably not too helpful. it is something I tried to avoid. something about learning a languages from generated audio didn't sit right with me haha.

hey HN! this is my attempt at language learning with audiobooks. it synchronizes text to speech, and shows translations. the audiobooks themselves are all public domain.

got all of the audio alignment, translation, and asset generation working on my gaming computer. pretty happy with the pipeline, except for the sometimes subpar translations.

if anyone is interested in the details I am happy to write them up!

if you are into language learning, I would love to hear if this could be useful to you!

hey HN! been working on this project on the side for a couple years now.

as a kid i got an audiobook along with a text copy to learn english. i found that to be a very nice combination.

i have been trying to learn more languages, but found that purely reading tends to worsen my pronunciation, but my listening comprehension is not good enough for audiobooks.

so i took some public domain audiobooks, synchronized them with text, paired it up with translations in some of the most spoken languages, and put it all on a website.

as this is for learning, i am avoiding ai generated audio entirely.

looking for feedback and happy to discuss details!

Because kg is the fundamental unit of mass and kW is typically used for electric motors.

Same reason you wouldn't use m²/s³ even though that's also technically correct.

Looks complete to me.

For context: I own licenses for both Affinity Designer, and the full Affinity 2 suite.

Just tried the new affinity application for a couple hours and it's pretty great. Personas are now studios and as far as I can tell features from all apps are now integrated into one.

Giving this away for free is insane value and I am very glad to have this as a photoshop alternative.

Yes.

Possibly M1 hardware optimisation for x86 emulation?

Appreciate seeing some actual benchmarks. But it is a bit silly to round the results to seconds when they are in the single digit range.

Bitcoin has all of the problems listed in the article and is, arguably, not yet providing the social value that blockchain could.

Not sure how a high market cap is a counter argument to that.

If I understand this correctly this was mostly about Google getting an unfair advantage over the ISPs.

Which wouldn't be the case if everyone loses access to the IP + DNS request info.

According to apple [1] ISRG Root X1 is available all the way down to iOS 10. So that would be every iPhone since the 5.

As far as marketshare goes iOS 13 and 12 make up 94% of devices [2]. So I am guessing its an insignificant amount of actual users.

[1] https://support.apple.com/en-us/HT204132

[2] https://developer.apple.com/support/app-store/

The title seems to imply that this affects all WhatsApp users, when it really only affects those putting a link with their phone number online.

Feels like clickbait.

https://faq.whatsapp.com/general/chats/how-to-use-click-to-c...

Exactly: https://brand.systemd.io/

Baffling how he didn't manage to find that on google.

From the linked article:

This interface will not grant you access to the users filesystem. Instead you will have a "virtual drive" within the browser sandbox. If you want to gain access to the users filesystem you need to invoke the user by eg. installing a Chrome extension. The relevant Chrome API can be found here.

Security reasons apparently. This is news to me too.

https://www.diagrams.net/blog/move-diagrams-net

For me gmail loads and becomes interactive in under 2 seconds on an empty cache. I did not expect it to be that slow for other people.

That said I do practice inbox zero, so it might just be spending time on loading a lot of mail?

You can watch 4K Netflix in Edge, which I would call browser. But the point still stands, only works with supported hardware and proprietary software.

https://help.netflix.com/en/node/13444

More like 3% when you combine linux and mac. That can make it hard to justify the cost.

Still bs for an existing game with platform support from unreal though.

https://store.steampowered.com/hwsurvey/Steam-Hardware-Softw...

Well, I don't know the details. But AFAIK canvas drawing is GPU accelerated.

So I would guess you are effectively fingerprinting the combination of browser and GPU. And that does not sound like its easy to fix on the browser side.

I would really appreciate full disclosure upfront. Hiding advertisement for neptune-ml behind a misleading title seems rather dishonest.

Other than that this hits some valid points that also apply to other parts of computer science. Just versioning code has often not been enough in my experience and proper experiment management sounds like a great solution!

I do like using JWT. But its point is to offer flexibility. If you fix the entire header i.e. use a single signature method, you might just as well concat that signature directly.

In other words if you stop utilizing JWT, you won’t have JWT specific problems.

Well yeah. But you, presumably, only have to do that once. After that you get 5ms per account, which seems to be the important takeaway.

Do you have any sources for the coal burning rate?