dumblob

Results 1305 comments of dumblob

Any news on this? I have the same issue as @fcladera described in https://github.com/equalsraf/neovim-qt/issues/95#issuecomment-248464656 (the `set title` trick after `nvim-qt filename` also works for me). I'm using the latest development...

@zvirja I didn't try it on the command line, but in two steps: 1. `nvim-qt c:\log.txt` 2. (in the freshly opened nvim-qt window) `:set title` So, the issue is not...

I've just reported a "subissue" of this one on https://github.com/neovim/neovim/issues/8038#issuecomment-367348125 . But it seems it's a more general issue and fits exactly here. Any news on this?

> Hah, @Verdagon who wrote those is a friend, and I am familiar with his work... That's awesome! The world is so small... >but Lobster already optimizes reference counting by...

I definitely raise my hand for making encoders support "smart quantization with dithering and high-fidelity transparency" (and possibly other techniques) as **pngquant** does. It allows for user-settable level of quality...

@kodonnell thanks! I'm really interested in lossy compression, so your findings ignited my further interest in the downscaling idea. I'm though a non-believer in SSIM and thus would want to...

@kodonnell sorry for the late reply. I would imagine something like [ISSIM-S](https://www.spiedigitallibrary.org/journals/journal-of-electronic-imaging/volume-25/issue-06/063015/Improved-structural-similarity-metric-for-the-visible-quality-measurement-of/10.1117/1.JEI.25.6.063015.full?SSO=1 ) (further details about SSIM huge limitations: https://www.researchgate.net/publication/283153178_Limitations_of_the_SSIM_quality_metric_in_the_context_of_diagnostic_imaging and https://ieeexplore.ieee.org/document/9337182 ). As for pictures anything from your smartphone...

There was also this tiling proposal - maybe just read the whole image in terms of a sequence of tiles instead of sequence of pixels would do the trick?

> that would work, but it is strictly harder for streaming encoders. Is it? I mean, the simpliest solution (to ignore tile boundaries for optimization/compression and optimize/compress only each tile...

Some more good images to test against: https://github.com/jhcloos/test_images