Copilot is one of two LLMs I’ve briefly tried. It was noticeably better than Gemini was at the time, but still seemed entirely pointless. Nothing it (or Gemini) offered to do were things I wanted help with. I enjoy research and writing, so why would I outsource those things and burn down an acre of rainforest in the process?
It’s an older expression I’ve never really understood or liked, but here you go (help from gemini) “” The expression “It’s a gas” likely originated from Irish slang, where “gas” meant “a joke” or “something frivolous”. This usage gained traction in the 1920s and 1930s within the African American community during the Jazz Age, and James Baldwin used it in print in 1957. It was further popularized by the hippie movement in the 1960s. Here’s a more detailed breakdown: Irish Slang: The earliest documented use of “gas” to mean “joke” or “frivolity” is in Irish slang, appearing in print by James Joyce in 1914. Jazz Age: This usage of “gas” spread within the African American community during the Jazz Age of the 1920s and 1930s. James Baldwin: The phrase “It’s a gas” appeared in print for the first time in 1957 in a short story by James Baldwin. Hippie Movement: The expression was further popularized by the hippie movement of the 1960s.
Thats not the only argument against inline images: 4.4.3 Why doesn’t text/gemini support inline images? This is a deliberate decision made in direct service of the guiding principle of user autonomy (see answer 4.2.2), specifically the idea that text/gemini documents should have no way to trigger additional network requests. Images are one particular case where this principle also overlaps substantially with our guiding principle of user privacy (see answer 4.2.1). So-called “tracking-pixels” have been a standard tool of the internet’s surveillance marketing complex for many years. These tiny, invisible images abuse the web’s behaviour of automatically downloading image files from arbitrary third-party servers to effectively trick your computer into “pinging” surveillance servers, reporting your movements as you explore the web. gemini://geminiprotocol.net/docs/faq-section-4.gmi
I love my iPhone, I’ve been using them for like 12 years (I loved Android but switched when physical keyboards stopped exiting cuz I liked iOS’s keyboard a lot, then fell in love with most of the rest of the OS and how long I could keep my phone for before replacing.) The one thing I absolute hate is the App Store. It’s insane. It’s so counter to the rest of Apple’s “simple stuff that works” … “ethos”. I use VERY few apps, I only have a couple installed on my phone. Say I want one, though. I open the App Store right now. Main page it dumps me on: 75% of the screen is a Tinder ad (I’ve never used a dating app and I’ve been with my partner for 15 years.) The rest of the screen space below that is a Google Gemini ad (I don’t use any Google products or services anymore, wtf Apple?) I click on the “Games” tab below and it’s an EXTREMELY busy, loud, colorful assault on my senses. Absolutely not! The “Apps” tab isn’t any better, huge ads take up 50% of the screen and below that, GET TIKTOK AND TINDER. Uhhhh no, never. Okay, let’s go to search. Click, and it gives me a search box at the bottom with some suggested apps and games I might like. I can appreciate that Apple’s data collection is allegedly a lot lighter than other companies, because none of the suggested shit is anything I would ever touch in my life. Let’s download one of the very few games I have on my phone: Ascension, a deck building game. I type “ascension game” and the first result is an AD FOR A WEIRD MEDICAL APP THAT DOESN’T SAY ASCENSION ANYWHERE AND HAS 70 REVIEWS WHYYYYY?! An absolute mess compared to the mostly great layouts of their other programs on the iPhone. I do not love the latest iPhoto UI change but I can deal with that. The App Store is fucked.
Gemini puts content first! It’s more hackable than Markdown—the parser for my gmi-to-html JS implementation gmi-web uses a one-line regex. This makes it super easy to publish content to either gemini:// or https://. The protocol is only interesting to me in-so-much as it’s a network based on text/gemini and so aggregates lots of good content to read.
That’s good advice, but I beg to differ about the perspective on browsers. There are a very few browsers that only render content. Most do much more: tabs, bookmark management, cookie management, password management, plugins/extensions/add-ons, history management, JavaScript, downloads management… they’re full-on mini desktops, and they do much more than just render content. And all of this - while useful and desirable to many people - costs, in compute and especially in memory. Unless you’re running an Electron app, odds are that your the browser is the single largest consumer of memory on your computer at the monogamy moment. If I run Firefox, it even tops Factorio with multi-planet factories. Unfortunately, Acid2 compliance is very complex, and the content of many websites is inaccessible without JavaScript, so the idea of just something like Evince for HTML isn’t pragmatic. However, having an engine that only renders CSS and XHTML could still be useful. Many sites are either JS-free, or the JavaScript only adds functionality that might be irrelevant to the content: commenting and feedback support, for example. Gemini has failed, but a really pared down browser can still be valuable, and a fair portion of the web is still browsable without JavaScript. I think OP’s question is entirely rational, and practical. An example that illustrates my point is epub, which is just xhtml and assets in a zip: images, yes. JavaScript, no. I can imagine a wrapper that does the networking to fetch assets, bundles the allowable ones into epub, and then runs an epub renderer. It would be an order of magnitude smaller, and cleaner, than even one of the minimalist WebKit browsers like luakit, surf, or vimb. I run the smallest browser I can, and only open Firefox when I hit a site I both a) want to see, and b) requires JavaScript to be at all functional. Most online shopping stores fall into this category, but banking’s another. I don’t begrudge the more demanding requirements of those sites, but I don’t want the needless resource consumption of Firefox when the sites don’t require it.
I have been trying multiple at this point, though I am not sure I found one I am happy with. I tend to use SearXNG, but I tend to make too many searches too quickly and it causes various search engines to block me for a bit. I thought I should share another search engine if it helps: LibreX / LibreY Here’s a few more search engines: Marginalia Search - A search engine that favors text-heavy sites and punishes modern web design - I just tried this one, and it’s very handy. Definitely recommend, though the results may not be relevant depending on your search, but sometimes you may find another similar site you would want to bookmark for later. wiby - A search engine of curated websites, useful for finding small blogs Mojeek - Proprietary but uses its own web crawler MWMBL - A free, open-source, and non-profit search engine At this point, I have been unsatisfied with search engines because the internet it just filled with SEO optimized crap. I made a habit of bookmarking every interesting or useful thing I find. I rely on sites like Lemmy/Lemmygrad, Reddit, Hacker News, etc. to find interesting links. I also look for awesome list git repositories (such as this, this, this, and this) (and this and this to search for more) and Reddit/Lemmy wikis to find useful resources. If I find (semi-)trustworthy news sites like Grayzone, AntiWar, etc., I will keep a tab open to use that as my news feed instead of Reddit. If I want to look up a news story or opinion on a search engine, I tend to add the word “Marxist” or “socialist” and it sometimes gives me results I want. If I want to find help online, I either find the documentation web page or I will just add the word “reddit” since I still rely on reddit question posts to find answers. Bringing Back The Minimal Web - A blog post about an idea I agree with and that I want to see manifested One thing I would wish if we reach a socialist society is for everyone to be provided the option to make their own website (multiple options available for various levels of tech savviness, from self-hosting on one’s own dedicated server to “simplified web designing tools with presets” for the least tech savvy) subsidized by the government. Then a search engine can be built to prioritize those sites, especially those that people used to make important informational resources. I also like browsing Gemini websites, which requires a browser add-on or an entirely different browser. You may learn about it here: https://geminiprotocol.net/ I wish there were more Marxist Gemini capsules, though I hope to make Gemini equivalent mirrors of important Marxist websites someday. I also wish there to be an org-mode protocol to make websites as org documents for emacs. There’s this person’s blog post that mentioned a similar idea, which I found from this discussion on Reddit. I find that visiting niche communities on niche platforms helps with finding quality content and resources, thus I tend to enjoy looking through old/new protocols and small websites/platforms before they are flooded with commercialized crap and are abused by users that don’t respect the sites’ rules/etiquette/culture. Sidenote: If you like to sail the seven seas, you can self-host Jackett to use as a proxy server to query torrents.
Gemini is basically an alternative web, so it has its own protocol, own server and browser software, and an own content language (heavily inspired by Markdown). Everything is built to be much more lightweight, so certainly in the spirit of solarpunk. But big downside is that “the thing you want” is honestly probably not on Gemini at all, because it does not share content with the HTTP web. That does mean that all the ads and bullshit are gone, too (and cannot be implemented beyond static images, sponsored articles and I guess, ASCII art). But yeah, at this point, it’s most useful for reading (and writing) blog posts, and having fun with the technology itself. It does have a lovely community for that.
I really just want web browsers to die, and be replaced by one of the slimmed down options like gemini, gopher, or some markdown viewer. The web just keeps getting increasingly bloated and ad-ridden, and filled with popups. Web browsers are as complex as entire operating systems now, so only 2 orgs (google and mozilla) have the resources and expertise to build a browser, and mozilla might throw in the towel eventually, leaving the internet as one big google ad. IE move viewing of mostly static content into these simple variants like gemini, and move dynamic things to local apps with API access.
My guess is that the most expensive single component would be the lidar. Prices on lidars can be well over $100k. When I worked with lidar about 5 years ago, IIRC a Velodyne 128 was $160k. These robots would probably be using a 32 though, which is probably going to be less than 1/4 of a 128. Also, Velodyne and Ouster merged since I last used lidar. Ouster does in-device sensor fusion, which likely takes a significant load off the CPU and potentially GPU, meaning these robots may be able to get away with lower spec CPU and GPU. It appears that Ouster now does object detection, which is another reason these could get away with lower spec GPUs (assuming they’re using Ouster) Obviously there’s a lot of speculation in my response, but since there’s no teardown of the robot, and without spec sheets or a BOM, all we can do is speculate.