Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • 10:16
    AregevDev commented #974
  • Oct 15 21:30
    jansol commented #1060
  • Oct 15 21:30
  • Oct 15 21:29
    jansol synchronize #1060
  • Oct 15 21:28
    jansol synchronize #1060
  • Oct 15 21:24
    fintelia commented #1060
  • Oct 15 20:56
    jansol synchronize #1060
  • Oct 15 20:53
    jansol opened #1060
  • Oct 15 11:58
    jansol commented #1039
  • Oct 15 02:34

    HeroicKatora on master

    Remove redundant code Merge pull request #1059 from J… (compare)

  • Oct 15 02:34
    HeroicKatora closed #1059
  • Oct 15 01:34
    Travis JohnTitor/image@f26d338 (remove-redundant-code) passed (32)
  • Oct 15 00:51
    Travis JohnTitor/image (master) fixed (31)
  • Oct 15 00:36
    JohnTitor edited #1059
  • Oct 15 00:00
    JohnTitor opened #1059
  • Oct 14 15:13
    HeroicKatora commented #1057
  • Oct 14 10:19
    birktj commented #952
  • Oct 14 09:56
    est31 commented #952
  • Oct 13 14:49
    fintelia commented #1058
  • Oct 13 14:45
    fintelia labeled #1058
HeroicKatora
@HeroicKatora
@cooperspencer Opening the webp and then saving the resulting DynamicImage as png should work
Andreas Wachter
@cooperspencer
@HeroicKatora thanks. May I ask how exactly? I opened it with:
let img = image::open(&blub);
what do I have to do after that?
sorry for the stupid question, but I am still getting used to Rust...
@HeroicKatora ok found it. I forgot the .unwrap().
The only think I see is that it is in grayscale, but I read that the library doesn't fully support webp.
Is it ok for me to ask if the library will fully support webp in the future?
HeroicKatora
@HeroicKatora
That's already in the works, you can track progress here image-rs/image#952
Some bug leads to wrong colors, that needs to be resolved but it can fully handle the image structure itself.
HeroicKatora
@HeroicKatora
It should also be noted that there is a minor difference between 'full webp' and 'full vp8'. webp is the container format and vp8 is the actual image data encoding (more information on Google's own page. While the PR mentioned works on decoding colors from vp8, the riffcontainer permits other data streams as well (EXIF metadata, animations, ..) and I don't think there is a plan to support these extensions.
That's probably more information than you bargained for ^^
Andreas Wachter
@cooperspencer
no, that information works well for me. Thanks. As a matter of fact, my pictures are vp8. So I will just wait for the new version of image and then I will change my program accordingly. Again thanks for the info and keep up the good work
HeroicKatora
@HeroicKatora
The exact next version (0.22) might not yet include that change but should have the required interface so that it can be added in a 0.22.1, without another SemVer breaking change. Different parts progressing at different speeds :smile:
SnoozeTime
@SnoozeTime
Hi there,
Is there any ongoing effort for this image-rs/image#940 ? If not, I might try to tackle it if you accept PR/provide code review :)
HeroicKatora
@HeroicKatora
A PR would be appreciated, a few requested features simply stall due to limited available time. When ready, request review on the PR and someone should come along. If not, ping me I guess.
And by a few features I mean a large portion of them ^^
SnoozeTime
@SnoozeTime
Great thanks!
SnoozeTime
@SnoozeTime
alright :o , modifying the dynamic image to include 16bits images requires quite a lot of work
HeroicKatora
@HeroicKatora
Yes, unfortunately :|
The best way to start would probably be to explicitely not involve DynamicImage but only expose this directly through the reader.
DynamicImage is a bit of a mess that needs cleanup of its own
SnoozeTime
@SnoozeTime
I see, so I can allow loading the ImageBuffer, Luma<u16> directly via whatever codec loader. Then, when the DynamicImage is cleaned, we can allow loading 16bit grayscale with the nice api
I've successfully loaded and saved PNG grayscale 16-bits, but I need to clean up the mess a bit :D
HeroicKatora
@HeroicKatora
:+1:
Yuki Okushi
@JohnTitor
Hi, are there any plans for the next release of imageproc? imageproc::drawing::draw_text(_mut) doesn't seem to work well with the dependencies image v0.22.1 and imageproc v0.18.0 (image v0.21 or master branch of imageproc works fine). This issue has been fixed in master branch of imageproc, I think.
HeroicKatora
@HeroicKatora
Not yet. imageproc performance is substantially lower if we avoid all the unsafe functions (the GenericImage::unsafe_get_pixel etc) and so we're discussing undoing their deprecation. Kind of unfortunate, it seems we overestimated the replacements and llvm.
:/
Yuki Okushi
@JohnTitor
Oh, okay, thanks!
HeroicKatora
@HeroicKatora
Ugh, Travis build broke on a timeout downloading crates. At least this shows that the notifications work!
Yuki Okushi
@JohnTitor
Hmm, webhooks also picks up my fork repository?
HeroicKatora
@HeroicKatora
It picks up anything that runs on Travis, connected to the repo. So if you have an active PR it will run Travis and should appear here
Not sure why canceled appears after the merge, not a big deal though
Yuki Okushi
@JohnTitor
I canceled the image repo's jobs to go through other repo's jobs
Uh, I think it's better that the fork repositories are not shown.
HeroicKatora
@HeroicKatora
I'm looking into whether this is possible.
Yuki Okushi
@JohnTitor
We're using CircleCI on image-rs/image-gif? It seems failed several times due to no config.
HeroicKatora
@HeroicKatora
Should be built on Travis as well. https://travis-ci.org/image-rs/image-gif
Ah, it still refers to PistonDevelopers in the docs.
It should be noted that there is no core maintainer for image-gif, currently
Yuki Okushi
@JohnTitor
Can you turn off the CircleCI jobs?
HeroicKatora
@HeroicKatora
Where is it building, anyways?
Yuki Okushi
@JohnTitor
It seems CircleCI is still alive. https://circleci.com/gh/image-rs/image-gif/tree/master
HeroicKatora
@HeroicKatora
I find no button to do this.. Confusing
HeroicKatora
@HeroicKatora
Yes it does, those are indeed webhooks (per repo...).
Anyways, I'll fix the documentation references etc.
Yuki Okushi
@JohnTitor
Okay, thanks!
Yuki Okushi
@JohnTitor
Sorry to bother again but CircleCI is also enabled on inflate https://circleci.com/gh/image-rs/inflate/tree/master
HeroicKatora
@HeroicKatora
Okay, fixed
HeroicKatora
@HeroicKatora
I've just been contacted by fuzzit.dev about setting up automatic fuzzing. I'll relay more information as it becomes available to me.
Yuki Okushi
@JohnTitor
Sounds good
Kevin R. S.
@BourgondAries
Hello! Submitted a PR for a recent regression since v0.22. image-rs/image#1031
HeroicKatora
@HeroicKatora
Thank you for supplied fix. Did that ever work correctly?
Kevin R. S.
@BourgondAries
It worked correctly in 0.21.0
May I suggest we run cargo fmt -- --check in travis? There's some trailing whitespace in the code
HeroicKatora
@HeroicKatora
There's #1022 and #1024 already, though it doesn't seem to show any trailing whitespace in the PR diff.