for all #blind users who use @cubic's veescribe, it is currently down. he pushed an update earlier this afternoon, and now it is acting all haywire. note: it is not down in its status, it is down as in it cannot connect to the Gemini API which is either to do with the key itself, or the fact he tried modifying things. for the blindsoft discord server, there is a backup called Ali images debug, normally utilized for debugging purposes, which is now handling as a backup for the bot. I am deploying this bot to the g4p VPS now. note: eleven soft Ali images is still working fine. if anyone receives any problematic errors, please contact me directly. #discord #bot #python #blind #blindsoft #seediffusion @cubic
for people wondering, there is a reason why #blindsoft image describer (both python and js) barely use embeds for text. the reason is that if you have a tts bot that you need to read text especially from bots, it will not be able to collect embed messages because TTS bots do not read json formatted embed text. at least, most tts bots, on voice channels. also, when i'm reading an image description, I do not want to have to go and constantly scroll through embeds to find what im' looking for. I want to be able to just have it ready. plus, I already get complaints that the image descriptions look like long paged essays, I wouldn't want users to complain more. note: I will be working on a future implementation of @alexchapman's galacticord to see if I can get it to read json formatted embed text, though I am currently not sure how possible or impossible it will be
as a result of veescribe not being able to act as a user application because of no slash commands, for you #blind fokes in servers that have it, veescribe is now able to do simage (scanimage). what it does? if a person turns it on (must have manage server permission) it will automatically scan any image in a #discord #server . for anyone looking to see the commit, see this note: simage will only be able to scan images in which the bot can view channels.note, you can invite veescribe here: https://discord.com/oauth2/authorize?client_id=1460990962892406906&permissions=0&integration_type=0&scope=bot+applications.commands .
#blindsoft #blind #seediffusion #discord #images
Add scanimage (simage) command and automatic image description logic · averlice/blindsoft-image-describer@d5b083e

discord bot to describe images using the gemini API - Add scanimage (simage) command and automatic image description logic · averlice/blindsoft-image-describer@d5b083e

GitHub

hello. we are aware that elevensoft Ali images is currently unable to process .gif images at this time. i'm sure even you #blind fokes know what a gif is, so I won't use my time explaining it, but we are aware of it.
at the moment, there is nothing we can do about it on the Gemini end because Gemini does not support .gif files.

solution?

the best solution nright now is to convert the image from a gif to jpeg or PNG file, which are ali's main supported filetypes.
note: veescribe has the same problem, which, again even with python, neither me or @cubic can do anything about. sometimes later, I will see if I can come up with an alternate fix, but it is on the back burner at the moment. #blind #bsid-js #js #blindsoft