Request MX-23: Shotwell 0.32.2

[oswald_c]

Could Shotwell be updated to v0.32.2 please?

Changelog - https://gitlab.gnome.org/GNOME/shotwell ... e12207a17b

[FullScale4Me]

In the MX Package Installer, Enabled Repos tab is version: 0.32.1-0.1~mx23+1:

Package: shotwell
Version: 0.32.1-0.1~mx23+1
State: not installed
Priority: optional
Section: gnome
Maintainer: Steven Pusser <stevep@mxlinux.org>
Architecture: amd64
Uncompressed Size: 6,646 k
Depends: dconf-cli, default-dbus-session-bus | dbus-session-bus, librsvg2-common, shotwell-common (= 0.32.1-0.1~mx23+1), libc6 (>= 2.34), libcairo-gobject2 (>= 1.10.0), libcairo2 (>= 1.14.0), libexif12 (>= 0.6.21-1~), libgcc-s1 (>= 3.0), libgcr-base-3-1 (>= 3.8.0), libgcr-ui-3-1 (>= 3.8.0), libgdk-pixbuf-2.0-0 (>= 2.39.2), libgee-0.8-2 (>= 0.10.1), libgexiv2-2 (>= 0.13.0), libglib2.0-0 (>= 2.68.0), libgphoto2-6 (>= 2.5.10), libgphoto2-port12 (>= 2.5.10), libgstreamer1.0-0 (>= 1.0.0), libgtk-3-0 (>= 3.21.5), libgudev-1.0-0 (>= 146), libjson-glib-1.0-0 (>= 1.5.2), libopencv-core406 (>= 4.6.0+dfsg), libopencv-dnn406 (>= 4.6.0+dfsg), libopencv-imgcodecs406 (>= 4.6.0+dfsg), libopencv-imgproc406 (>= 4.6.0+dfsg), libopencv-objdetect406 (>= 4.6.0+dfsg), libpango-1.0-0 (>= 1.18.0), libpangocairo-1.0-0 (>= 1.14.0), libportal-gtk3-1 (>= 0.5), libportal1 (>= 0.5), libraw20 (>= 0.16.0), libsecret-1-0 (>= 0.19.1), libsoup-3.0-0 (>= 3.0.3), libsqlite3-0 (>= 3.5.9), libstdc++6 (>= 11), libunity9 (>= 3.4.6), libwebkit2gtk-4.1-0 (>= 2.25.1), libwebp7 (>= 1.2.4), libwebpdemux2 (>= 1.2.4), libxml2 (>= 2.7.4)
Conflicts: shotwell:i386
Replaces: shotwell-common (< 0.26.2-1), shotwell-common:i386 (< 0.26.2-1)
Description: digital photo organizer
Shotwell is a digital photo organizer designed for the GNOME desktop environment. It allows you to import photos, pictures, images and videos from disk or camera. Shotwell can organize them in collections and in other various ways. The viewer shows them in full-window or fullscreen mode and presents them as galleries or slideshows. The integrated editor can rotate, flip, crop and tag the photos, adjust the colors und remove red eyes. Export is possible to facebook, Flickr or Youtube to share with others. It is able to manage a lot of image formats such as JPEG, PNG, BMP, TIFF and Raw CR2.
Homepage: https://wiki.gnome.org/Apps/Shotwell
Tags: implemented-in::c, interface::graphical, interface::x11, role::program, uitoolkit::gtk, use::organizing, use::viewing, works-with-format::jpg, works-with::image, works-with::image:raster, x11::application

Need to get 35.0 MB of archives. After unpacking 58.6 MB will be used.

[Stevo]

It's not made it into our primary source for backports, upstream Debian, but...well, I'll give it a shot.

[Stevo]

Here's my best shot if you want to try it early! https://drive.google.com/file/d/1dIP64C ... sp=sharing

I enabled face-detection as an experiment, which Debian hasn't done yet, so would appreciate any feedback on that. It bundles some neural network AI stuff when that is enabled.

[oswald_c]

Here's my best shot if you want to try it early! https://drive.google.com/file/d/1dIP64C ... sp=sharing

Thank you very much. I wondered why the zip file was so large, but all became apparent on examining it,

I enabled face-detection as an experiment, which Debian hasn't done yet, so would appreciate any feedback on that. It bundles some neural network AI stuff when that is enabled.

It works... sometimes. On my limited experiments so far, the faces need to be well lit against a darker background to be recognized immediately automatically. Once you manually identify one face on more challenging photos, it gets better at identifying others.

[Stevo]

Here's my best shot if you want to try it early! https://drive.google.com/file/d/1dIP64C ... sp=sharing

Thank you very much. I wondered why the zip file was so large, but all became apparent on examining it,

I enabled face-detection as an experiment, which Debian hasn't done yet, so would appreciate any feedback on that. It bundles some neural network AI stuff when that is enabled.

It works... sometimes. On my limited experiments so far, the faces need to be well lit against a darker background to be recognized immediately automatically. Once you manually identify one face on more challenging photos, it gets better at identifying others.

Thanks! So the neural network is still learning on the fly, apparently.