zopfli

More info: https://solar.lowtechmagazine.com/about/the-solar-website/#h... https://github.com/lowtechmag/solar/wiki/Solar-Web-Design https://homebrewserver.club/low-tech-website-howto.html IMO the only optimization they 'missed' is to pre-compress the static files and images on disk, since it's cheaper to decompress on-the-fly if the client doesn't support gzip. The could use zopfli to "extra compress" the most frequently accessed files, and use this as a solar dump load to usefully burn extra power when the batteries are full (perhaps compressing with gzip -1 when Pelican re-builds the static site, or even avoid pre-compression altogether during the build process. https://blog.llandsmeer.com/tech/2019/08/29/precompression.h... https://github.com/google/zopfli https://github.com/kurtmckee/pelican-precompress

One thing that's excellent about zopfli (apart from being gzip compatible) is how easy it is to bootstrap: git clone https://github.com/google/zopfli.git cc -O2 zopfli/src/zopfli/*.c -lm It just requires a C compiler and linker.

Also https://github.com/google/zopfli and https://github.com/google/brotli

Depends on the format I'm targeting. PNG: zopflipng https://github.com/google/zopfli/blob/master/README.zopflipn... AVIF: libavif https://github.com/AOMediaCodec/libavif JPG (for backwards compatibility when there's no AVIF): mozjpeg https://github.com/mozilla/mozjpeg

zopflipng typically beats pngcrus and optipng (on Linux at least) but by default it drops auxillary PNG chunks [0] which can result in browsers (and other applications) using a different color space, causing the resulting images to look more washed out than the original. To prevent this you need to explicitly pass --keepchunks=cHRM,gAMA,pHYs,iCCP,sRGB,oFFs,sTER to zopflipng. cwebp seems to have a simmilar issue when starting with png files. Sucks that color space support is still so inconsistent. Unfortunately zopflipng (and most other tools) don't have APNG support, keeping only the first frame :| [0] https://github.com/google/zopfli/issues/113

Going away from Zlib proper but staying with its format, there’s libdeflate[1], an implementation that’s substantially faster on modern machines but can’t stream (although I see no reason why it couldn’t, probably just for simplicity due to its origins as a baseline fast compressor[2]), and of course Zopfli[3], a glacially slow compressor that’s somewhat (and occasionally a lot) better. [1] https://github.com/ebiggers/libdeflate [2] https://encode.su/threads/2482-libdeflate-a-new-optimized-DE... [3] https://github.com/google/zopfli

You can also use Zopfli to get a bit higher compression at the expense of longer compression times: https://github.com/google/zopfli

I tried zopflipng [1] with the very expensive option suggested from the usage and got 2,493 bytes (original 2,740 bytes, nowhere practical as it took 3 minutes) so you don't have to sacrifice the palette, though I'm not sure if 16 or even 4 color palette will significantly alter the visual. [1] https://github.com/google/zopfli

https://github.com/google/zopfli for the latter

> PNG is basically zip with some extra tricks. Sure, if with "some extra tricks" you mean "four relatively simple scanline filters[0] that nevertheless Paretro Principle[1] the heck out of more complex forms of lossless compression". Especially Paeth is amazing in how it makes gradients so much cheaper than naive approaches with such a simple formula. Similarly relying on DEFLATE is smart too, as that means any improvement to that format can be safely applied to PNG as well, like with zopflipng[1]. [0] https://www.w3.org/TR/PNG-Filters.html [1] https://github.com/google/zopfli/blob/master/README.zopflipn...

Since most of Brotli's improvements over its ancestor LZ77 are due to its large, hardcoded, text-corpus dictionary [1], most of the algorithm's strengths would be wasted on binary data like images. Zopfli, from the same people, is a DEFLATE encoder, so it can be used in PNG [2] and this has already been added to some optimizers, e.g. AdvanceCOMP [3] [1] https://gist.github.com/klauspost/2900d5ba6f9b65d69c8e [2] https://github.com/google/zopfli/commit/337d27f25ef15a6cf34f... [3] http://www.advancemame.it/doc-advpng.html

It does decompress the source PNG and build a new file from the pixels. You can see the source here. https://github.com/google/zopfli/blob/master/src/zopflipng/z...

I expected more attention to detail for a project this big. They are using the new logos in a few places now. All suffer from various problems. The one on the Google Official Blog (http://googleblog.blogspot.co.uk/2015/09/google-update.html) has hinting problems most noticeably on "l" symbol (zoomed version – http://i.imgur.com/1kRZB5o.png). The same with the logo on the Google Translate (https://translate.google.com/) and with the presented Google Logotype on the Google Design blog (https://design.google.com/articles/evolving-the-google-ident..., https://g-design.storage.googleapis.com/production/v5/assets...). The black SVG logo at the bottom of the last page has some unkempt figures and a leftover line. On a positive note, the logo image (http://googleblog.blogspot.ru/2015/09/google-update.html) is well optimized with Zopfli (https://g