LibreWolf vs Brave - Between a rock and a hard surface
So I'm trying #LibreWolf but it seems like it consumes more memory than #Brave. On the other hand, the former feels faster than the latter. For comparison, I opened both with nothing but "data:text/plain," page (effectively empty page, emptier than about:blank). Librewolf uses 410MB, Brave uses 190MB.
Despite that, I'll try to stick to LibreWolf. But I'm trying to also stick to things such as #Geminispace and #Gopher, bc the web is getting increasingly heavier. Sure it's fun to #develop using #HTML5 and #JS APIs and I'll continue to develop my projects, but using it daily is a whole other story..
For fun: a very instructionless instruction...
A linguistic salad from an untamed unconsciousness
From my stream-of-(un)consciousness:
Je nyet rén salam, mwt maha fogo, because tutti ist dagen Los Dämonen.
Je нет 人 سَلَام, 𓏏𓅑𓎟/𓅓𓏏𓅱 महत् fogo, because tutti ist dagen Los Dämonen.
(I'm not person of peace, [but from] Greatness Fire Mut/primal waters/death, because all are dawning of the demons)
Seems like Claude is either trained to detect its potential hallucinations, or there's another moderation layer which adds a mechanism to detect potential hallucinations.
Indeed, it hallucinated the names. Would be good for dadaist poetry or inspiration for stream-of-(un)consciousness poetry, but it's really bad for my current logical endeavor.
There is an ActivityPub extension called ForgeFed that's implemented by Vervis and Forgejo (the gitea fork that Codeberg is based on), but it is still far from feature complete.
Forgejo – Beyond coding. We forge.
Forgejo is a self-hosted lightweight software forge. Easy to install and low maintenance, it just does the job.forgejo.org
GitHub - gobengo/distbin: distributed social pastebin using Activitypub
distributed social pastebin using Activitypub. Contribute to gobengo/distbin development by creating an account on GitHub.GitHub
Aqui está uma descrição de texto alternativo para a imagem fornecida:
A imagem mostra uma conversa em um aplicativo de mensagens, onde o usuário pergunta se o DeepSeek consome menos água que o ChatGPT. A resposta afirma que sim, pois o DeepSeek foi projetado com foco em eficiência energética e sustentabilidade, utilizando técnicas de otimização para reduzir o consumo de recursos, incluindo água. A água é usada principalmente para resfriar os data centers que hospedam os modelos. A resposta também menciona que o consumo exato de água pode variar dependendo da infraestrutura e das práticas de resfriamento. A tela mostra opções para iniciar um novo chat, enviar uma mensagem e pesquisar.
Fornecido por @altbot, gerado usando Gemini
Marcelo likes this.
like this
Testing post via Raccoon.
Formatting test: bold italic underline. (Funny.. underline got html tags) ~~not~~
this is code ? "Huh... Interesting, it's not there when I access through web version" : (this is quote ? "Nah, it's just the formatting for quoteing" : "It's just a pair of grave accent surrounding the text and making me think that it'll format as code"
does it allow sections?
maybe.
- it may allow
- it may not allow...
- ...lists.
- Neither does ordered lists
- Just like this one.
What about multiline code?
process.stdout.write("Test...")
process.exit(0)
Testing only.
Stroudy
in reply to Daemon Silverstein • •A Motorhead fan group?
New To Friendica reshared this.
skribe
in reply to Stroudy • •Lemmy is the Fediverse's version of Reddit.
join-lemmy.org/
like this
Daemon Silverstein, Stroudy and Mark like this.
reshared this
New To Friendica and Mark reshared this.