allows alot of new things. dithering, contrast, saturation, even imbuing the image with more of a certain color. This takes either the result of the image completion from the text-to-image plugin or a canvas.
example page: https://perchance.org/alloimgexample
various things to improve such as allowing multiple tasks flawlessly, and figuring out why the canvas being in the iframe prevents it from changing. future updates will come :)
example is of plugin here: https://perchance.org/alloimg
my learnings from this for pluginmakers: i decided to make my own stuff that needs to run as coherent constant code in to a subobject of window. I will update my other plugins that add something to window to check if they have started yet to a safe place within my base object. Basically none of my own plugins will sprawl globally and will all be accessible under window.allo. Also I figured out to load an external script thru the return of the html text at the bottom of the output, which may be handy.
You must log in or register to comment.
btw, you can ‘prettify’ the javascript code that you have written with this website: https://prettier.io/playground/
What you have is a little bit unreadable since there isn’t any indentation even if there are ‘if/elses’.
makes it ironically barely readable at all now to me since im used to typing with way less indents, lol. but i prettified it so you can read it
i never ever split something().thing().thing() in to 3 separate lines like in the prettified version lol
I never either lmaooo, there is an option there to increase the width (
--print-width) so it wouldn’t be added into a new line.
@Alllo This is a neat plugin! Question, is this the ordinary CSS’s filter function but it’s “post-processing” so it could be used for the text-to-image outputs and the canvas, not just for HTML images?
Also, you should’ve written “alloimg” instead of “allo.img” in the list of filters in the example page 😄
I think they want the namespace ‘allo’ as a prefix for their plugins so that’s why it is laid out like that. I think it is using the Jimp Javascript library looking through the code.
it isn’t ordinary css, it is Jimp, which is an imported javascript library that does a bunch of the common processing options of higher end image processing except in pure javascript. Back when i was learning code in droidscript, jimp was a droidscript plugin and i found it cool how it was entirely in JS yet did these things. So this is a port in of what they were porting in. Jimp takes image data, in this case from a blob created from the canvas element returned when a texttoimage is created, and outputs it to a file or canvas, in this case back to the same canvas. Jimp has a few more resizing and rotating and masking and concating images options not seen here.
and it is allo.img because i have actually created a global subobject on window, so window.allo.img to hold all these img manip functions. If I were to make a bunch to manipulate sounds, or pixi.js sprites, or three.js models, i would organize them like window.allo.sprites or window.allo.sound. then anything imported with one of my plugins that has to do with window will be in the neat window.allo box and not touching nor interfering with anything else. Because window doesn’t need to be typed, the code comes out just allo.img.saturate. like vioneT said it’s a namespace for what i make
![[plugin] alloimg ~to postprocess generated images](https://lemmy.world/pictrs/image/8f017998-1067-45e4-b95b-37a9f72674ef.jpeg){kind=link}