cleaning up the code and adding minor improvements to my waterBoard project, which as a javascriptless flatfile anonmouse imageboard, it has lots of fancy css3 stuff like a full screen gallery, thumbnail expander, dynamic forms, hide images, inline replies... i just figured it out and gave up
i've decided to give the regex parser another try, i think i have in my head a certain design concept that just wasn't gonna work and i should just settle for slightly less elegant solution before i figure other way out that wrap everything in <p> and then runs the regex for each paragraph individually, which is relatively resource intensive, but properly parsing user generated posts for rundown / bbcode is primary feature of the board so who cares.
i'm currently using md5 for media hashing, 32 char filenames are long enough lol but there are security concerns md5, at first i figured it doesn't matter because the point of the media hashing to to prevent uploads of pre-existing media, so if someone tried to upload a false file it wouldn't matter, but assume you are trying to upload a pdf of a leaked document, well a hacker could make a dummy file with the same md5 and prevent you from uploading it... as obscure as something like that is, it's not acceptable for an anarchist space
so perhaps better way is use the sha3 hash as the database indexer ( that jives also with tag based or) and just save the og filename+dt-uploaded which could be manually renamed to ensure descriptive filenames, would also like auto-save alt text into exif data it's currently saved to .txt files under uid of file
Embed Notice
HTML Code
Corresponding Notice
- Embed this notice
🇵🇸 single use plastique 🏴☠️ (mook@possum.city)'s status on Thursday, 28-Nov-2024 07:19:03 JST🇵🇸 single use plastique 🏴☠️