memo 📎

#artspam #yaoi source

@xinit @xChaos @Colman @Gargron

@organicmaps rebranding is seperate from attribution. attribution is important, but attribution is not the issue here. It's fair and reasonable to require attribution, but rebranding is an expected result of forks and downstream projects free licenses permit the licenser to restrict the use of their name and logo as long as it doesn't practically limit the four freedoms, however the inverse, requiring derivitives to include your branding is a violation of the freedom to modify or redistribute modified copies. a core tenant of the freedom to redistribute, with or without modification, is the right to do so without permission. requiring the user to seek your conditional permission to perform a modification is a violation of freedom 3, making it not free @opensourceopenmind @manankanchu

@organicmaps Any white-labeling or rebranding use requires explicit written permission from the Organic Maps team this is not a free license. you may feel it's a reasonable and justified clause, and i wont argue against that, however it is factually incorrect to call it free in the sense of user freedom @opensourceopenmind @manankanchu

@mkb@mastodon.social @rysiek@mstdn.social i like the way the https://selfh.st/ newsletter does it, they has a paid RSS feed without ads, and a free RSS feed that only shows the article title and a link to the website

@lina @organicmaps @manankanchu @opensourceopenmind how does it infringe on the rights you’ve listed? https://wetdry.world/@memoria/116325754814885427 tl;dr yes rebranding is forbidden under the normal license yes which is okay per your words no they may use whatever license they wish, that's their prerogative. it's still not a free license

@organicmaps @opensourceopenmind @manankanchu nothing you said addressed the problem i explained, which is your incorrect claim that the data license is a free license you conflate attribution and branding while sidestepping the free license problem. the data license may be a valid license, however that does not mean it's a free license. requirements for attribution are allowed in free licenses. your provision for persisting branding with conditional permission to modify makes it not a free license if you won't address the free license issue and continue to change the topic to different matters, then there's no reason to continue this conversation

@nixCraft Note: setting a birth date that makes the caller appear older than 150 years is rejected with EINVAL, as the kernel does not support vampires or other immortal entities at this time.

@lina @organicmaps @opensourceopenmind @manankanchu hi, i edited paragraph 3 to make it less ambiguous and more direct. " im probably misunderstanding but are you saying that "attribution always required = free" and "attribution requirement can be removed by request = nonfree"?" No. required attribution is good and is ubiquitous in free licenses attribution is seperate from branding the issue is the conditional permission to modify branding. the FSF recognizes licenses that require branding to be changed to avoid confusion between forks and derivative projects, but the FSF does not accept branding being used as a tool to effectively hinder or limit the user's freedom to modify and redistribute without permission.

@santi sure, i use inference in a few ways: karakeep - tagging bookmarks, semantic search immich - face detection, semantic searchpaperless-gpt - document titles, tags, and OCRlibretranslate - language translationsSpeakr - voice to text transcription, tagging, summaries, semantic searchaudiomuse - sonic analysis on my music collection to generate sonically similar playlists and track queues as for LLM models: i really like the IBM granite4 models, specifically the 7B hybrid model (granite4:7v-a1b-h). It's hands down the best text-only model for it's CPU and memory (4.2GiB) requirements. Gemma3:4b is an all around good model for it's size, and can output text from text and image inputs. it's a pure transformer model so it's heavier to run than hybrid models, and 4B models do tend to go off the rails faster and more frequently. qwen2.5vl:3b is the best image to text model i can run on my system. qwen3vl:4b is significantly better, but i can't reasonably run it with an M1 ultra you could probably run the largest of these models and have it complete inference instantly :neocat_laugh:

@simonzerafa @pluralistic @tante Pop a power meter on that LLM adorned PC and let us all know what the power usage looks like with and without your chosen LLM running on a typical task challenge accepted! :D my laptop uses about 6w per hour when idling, and 25w when playing games or running inference I'd attribute the difference as about 19w per hour of inference my 900W microwave uses 15w per minute so microwaving a frozen burrito for two and a half minutes is equivalent to two hours of inference (or games) on my laptop also, that burrito was frozen. refrigerator wattage varies widely, but an average hourly running wattage of 150w is nominal at 150w the freezer takes almost 8x watts per hour more than the laptop inference, and the freezer runs 24/7/365! most of my inference tasks complete in about 30 seconds at about 0.16 watt per inference job. thats almost 940 inference jobs (assuming 30s average) equivalent to 1 hour of refrigerator running wattage

@duc @dev @kolya Pixelfed has had the ability to turn off replies for a long time and they seem to be doing okay

@mochabeau ... ... ... sincerly, :3

@espidev@fosstodon.org awesome! so happy this project has been resurrected

@freeplay sorry the millions in nonprofit donations need to pay the executives, there's not enough to make the app good

@peribotsarah@babka.social @thomholwerda@exquisite.social exactly, fedora isn't a single distribution, it's a dozen distributions sharing infrastructure. there are also derivatives preloaded with software for games, astronomy, graphic design, audio work, security, etc dismissing a whole ecosystem because someone(s) unilaterally created their own derivative preloaded with ai software is asinine