ux

Thank you for pointing this out. I added a note about this.

It's funny because I'm actually actively using the atime: when I need to know what's the last video file I watched in a series, I use `ls -ult [--time-style=full-iso]` or `stat -c '%x %n' *.mkv | sort -r` to show files by order of access time: last accessed (played) will appear first.

> A unique feature of JPEG XL is that it is possible to recompress existing JPEG images (of which there are a lot out there!) to a JPEG XL file that is on average about 20% smaller, without introducing any loss. In fact, the bit-exact same JPEG file can be reconstructed from the JPEG XL file.

I wasn't aware of this and this is actually crazy, and freaking brilliant from a migration point of view. I wonder how that was achieved. I didn't care so much about JPEG XL, now thanks to Google derping I want it everywhere.

Alternatively, you can use http://bugmenot.com/ (doesn't work that often I must admit)

Numerical stability is a nightmare for the winding number algorithm. I've spent about 2 years on it already, but I believe I've managed to make it work. I used the paper from Eric Lengyel which worked it out with quadratic and extended it to cubics. I kept the topology classification strategy, but had to work out a different solution for the analytical resolution stitching. I'm planning to write about it; this serie of math articles are actually building up the foundations I need to explain it.

I'd be happy to discuss it though.

Yes it's a great video, my contribution is modest in comparison. Watching it again it appears some information unsurprisingly overlap (typically the lerp/mix thing!), but I think I’m still covering different aspects of it.

Yay, that means I'm a genius!

Joke aside, yeah various people pointed that out to me so I quickly added an edit at the end of the post. Thanks :)

I'm not sure why I missed the combinations(..., 0), I'm pretty certain to have tested in a previous iteration (maybe because I wasn't using math.prod() initially); code and text adjusted.

For the map yeah I thought it wasn't recommended by Python style but I personally prefer it so I'm adopting that change as well.

For the sign change, your initial comment (before your edits) was proposing the (1, -1)[i & 1] trick, which I actually like better than negating all the roots. At least for clarity in understanding the blog post (which mentions the sign juggling), I think it's better to keep it that way. I added a link to your comment for anyone looking for further improvement.

Thanks for the all the suggestions :)

I wrote something similar a while back: http://blog.pkh.me/p/15-the-music-classifying-nightmare.html

> Unfortunately for me, I find external sources (blog posts, news articles, etc) often the most relevant and insightful; and Twitter deliberately filters them out when discovering new people.

Right, so what the author is looking for is... RSS feeds? Twitter is simply a poor-man marketed RSS feeds aggregation with all kind of flavored noises.

To me it feels like the author is not interested in what the people have to say, but more like a network of shared resources on specific topics. Well then, maybe reddit is a generic answer, and HN or lobste.rs for specialized topics.

Great post, learned a bunch of things. Only one blog post, and no RSS (yet?) unfortunately.

Thanks

> Does restic provide commands to automatically resolve these errors?

Yeah, a few. Here is a dedicated thread on the topic: https://github.com/restic/restic/issues/828#issuecomment-706...

You forgot to add a vignette filter :)

Also, the vignette preset for the curves filter is a bit over-the-top, you may want to get a smoother preset (the vignette filter can import Photoshop preset files, so you can look if you can find a free preset somewhere). Alternatively you can find out strategic keypoints within GIMP and translate them to FFmpeg parameter (a bit more hassle I admit), but note that the interpolation between points might not be exactly the same.

The key word here is "good health". If you start shifting from "living as long as possible" to "living with the most dignity up until the end", you end up questioning euthanasia, but also how people live post work-life: typically, how people are going to live decently through their last stage post 60s, can they enter a stage where they are relieved from the stress, pain, etc. You can't just say "well let them die earlier" if you don't provide some form of hope of relief for their last years.

Today I'd argue that the growing lifespan is used as an argument to make people work longer, which means in worse physical conditions. The "Work until your death" motto (the curse of people living in poverty) is not great for a healthy and attractive society. Questioning it ends up in a heated political/societal/philosophical discussion pretty quickly. And we're definitely not heading in the healthy direction: not because we don't want to shift the medical institution, but because we need a more profound society change. From a purely economical PoV, it's great that people die before around the beginning retirement, from a societal PoV it's a disaster.

Yes I'm aware. But we can agree it would be welcome in the standard, isn't it?

Is there any plan to deal with the locale fiasco at some point?

Some hints on what I'm referring to can be found here: https://github.com/mpv-player/mpv/commit/1e70e82baa9193f6f02...

Unrelated, but I also miss a binary constant notation (such as 0b10101)

This reminds me of https://www.ghacks.net/2012/09/12/blizzard-watermarking-wow-...

One of the main issue is no error checking on this kind of "allocation" (it's the same as alloca()). This issue also somehow exists with a fixed size buffer, but at least you get a much more predictable outcome (you have full control on the stack usage, unless you have a user controlled recursion).

I always found disturbing the way a mixing "art" with this kind of subject. It's giving a much more positive feedback, as in "it looks cool" or the feeling of emotions you get in a movie (because it doesn't look real).

Let's not forget war is freaking horrible and there is nothing worth idealizing one way or another about it. That's my personal belief, but I don't think art has its place here. Any form of heroic spectacle are truly misleading about what the people and soldiers truly endure.

coursera has some; I typically have locally the images-2012 one, but you also have things like dsp-001 which is a bit more advanced. Generally speaking, coursera has some good material in many related domains.

This "a" is the main reason I use "Clean" here (distributed in clR6x12-ISO8859-1.pcf.gz by xorg)

Unfortunately, that's a bitmap font, so not always appropriate depending on your monitor.

This kind of map actually needs some minimal classification. There are actually flavors between free and open wifi.

For instance, how would you categorize a wifi in an airport which needs you to fill a random form/survey (not paying anything) to access the Internet? Same question if it is then limited to something like 10 minutes (and you then need to pay, or just trick on your macaddr/dhcpname/...)? In these case, you can actually access the Internet for free, but they are extremely annoying.

Note: these kind of wireless network are actually very common in some places

You can backport it easily, and it will be present in 3.0.1. Ask on IRC or the dev ML.

Yeah, sorry about that. It was indeed done in a hurry.

I think the main highlights are:

- The API/ABI break (implied by the major bump)

- The many improvements in the native AAC encoder making it the recommended one (libaacplus and libvo-aacenc are removed)

- A ton of filters were added

- Many ASM optimizations that weren't mentioned in the Changelog (it will take a while to make highlights on those, I don't remember them)

Hopefully a proper news will be posted soon. Sorry again.

I'd be surprised if you had actually read the blog post I wrote (pkh.me) :)

> most of the libavcodec and libavformat work has been done in libav

Well, I believe this is a bit more complex. While it might be true for API changes and cleanups, FFmpeg has dozens of additionnal codecs and formats, as well as way more assembly/optimization code.

> What you should know is that the current codebase of FFmpeg is a huge mess compared to libav

Honestly, this is pure FUD and it's very surprising to read that from you. You seem to have no idea about the cleanups of FFmpeg because we do not advertise them (they are not relevant for the users). What is so much a huge mess? This is very insulting for the work done by FFmpeg developers these last years. Hey, even libmpcodecs (the MPlayer filter wrapper) is no more. A lot of work was also done on the documentation. Seriously, please stop doing FUD like this.

> Choosing one or the other gives you different set of features

Not exactly true.

OK so basically we have 26 sounds (http://typedrummer.com/audio/[a-z].mp3), an accel marker '(' (only 1 scale?), and a decelerator ')' (only 1 scale?). Anything else?

And since everyone feels like sharing his own crap, I'm doing it as well http://typedrummer.com/j5c05o

Edit: it's funny how Firefox (37) gets completely insane when it hasn't the focus on the tab anymore

Can you show an example of an animation with more than 256 colors?

As you said, the hack to get more than 256 colors is to redefine a sub rectangle with a new palette. But that's also the mechanism used to make animations. Are you able to get fluid playback with that?

Note: I'm not talking about new frame being just a new sub rectangle, more about N sub rectangles for the same frame, then a delay event, then M sub rectangles for the new frame, etc

Yeah, but the article is already quite long so I omit the "details" on that...

Basically, it's just to specify a complex filtergraph: in the case of -vf you're just passing one video stream which gets filtered by successive filters. In the case of -lavfi (or -filter_complex if you prefer) you can feed multiple inputs (and get multiple outputs). In this case, the second input is the palette image.

You can find more information on http://ffmpeg.org/ffmpeg-filters.html#Filtering-Introduction

ImageMagick is very popular, but apparently the algorithms in pngquant gives excellent results. Don't forget to include it if you do comparisons with the different solutions.

Also keep in mind that the compression efficiency and quality depends a lot on the content and how you play with the encoding options (how surprising...). I tried to cover most of the cases in my post, but make sure you are "fluent" with the other tools as well to make an objective comparison.

Overall, I think FFmpeg trump-card is that it includes the whole stack; timing, decoding, filtering, encoding, everything is in.

> (convert -delay 5 -loop 0 $tmp_dir/ffout*.png $2) >& /dev/null

So using ImageMagick for this... How is this memory wise BTW? (all frames are in memory at once or not?)