Atemu
24m ago
•
100%
How different is this to gitlab’s open core model?
That's a really good question that I don't immediately have a satisfying answer to.
There are some differences I can point out though:
- Gitlab has demonstrated its commitment to keep the core of their product, though limited in features, free and open source. As of now, BW's clients cannot even be compiled without the proprietary SDK anymore.
- Gitlab was always a permissive license (MIT) and never attempted to subvert its original license terms
- Gitlab-EE's "closed" core is actually quite open (go read the source code) but still squarely in the proprietary camp because it requires you to have a valid subscription to exercise your freedoms.
Is this a permanent change?
It'd be quite trivial for them to do in technical terms: Either license the SDK as GPL or stop using it in the clients.
I don't see a reason for them to roll it back though. This was decided long ago and they explicitly decided to stray away from the status quo and make it closed source.
The only thing I could see making them revert this would be public pressure. If they lose a sufficient amount of subscribers over this, that might make them reconsider. Honestly though by that time, the cat's out of the bag and all the public goodwill and trust is gone.
It's honestly a bafflingly bad decision from even just a business perspective. I predict they'll lose at least 20% but likely 30-50% of their subscribers to this.
Is the involvement of investors the root of this?
I find that likely. If it stinks, it's usually something stinky's fault.
Are we overreacting as it doesn’t meet our strict definition of foss?
They are attempting to subvert one of the FOSS licenses held in the highest regard. You cannot really be much more anti than this.
An "honest" switch to completely proprietary licenses with a public announcement months prior would have been easier to accept.
Atemu
51m ago
•
100%
As with all of their services, the back-end is closed-source.
For the purposes of user freedom, it's not that critical as the back-end merely facilitates the storage and synchronisation of encrypted data. This is different from the bitwarden case where they're now including freedom disrespecting code into the most critical part of their software: the clients which handle the unencrypted data.
Fact of the matter remains however that Proton Pass restricts your freedom by not allowing you to self-host it.
If you are fine with not being able to self-host, I'd say it's a good option though. Doubly so if you are already a customer of their other services.
Proton has demonstrated time and time again to act for the benefit of its users in the past decade and I see no incentive for them to stop doing so. I'd estimate a low risk of enshittification for Proton which is high praise for a company of their size.
Atemu
1h ago
•
100%
Keepass isn't really in the same category of product as Bitwarden. The interesting part of bitwarden is that it's ran as a service.
Atemu
2h ago
•
100%
the fact that the two programs communicate using standard protocols does not mean they are one program for purposes of GPLv3
The fact that they would even think about attempting to subvert the GPL (much less actually pulling through with it) makes me think they have stopped being an open source company a while ago.
Atemu
2d ago
•
66%
It would break a lot, require a new API, and devs reworking a lot of programs.
As I understand it, this would have been a perfectly backwards compatible change. You'd only get the events if you explicitly asked for them.
Atemu
2d ago
•
100%
The Immich app.
Although since it doesn't really function as a full gallery app yet, so I have Fossify gallery installed as a backup to open images in via intent.
I only learned about Aves today and trying it out for the same purpose, I think I like its picture viewer better.
cross-posted from: https://lemmy.ml/post/21519137 > I recently switched from a MBP to a Framework 16 as my primary laptop and one thing I immediately noticed was that I was unable to stop kinetic scrolls in Firefox by laying my fingers onto the touchpad. It'd just slide by unimpeded. You could work around this by counter-scrolling a little rather than holding still which is how I've been coping with it but it's suboptimal to say the least. > (As are many things in the Linux touchpad experience. Linux desktop developers *really* ought to use a macbook for a little to get a sense for how to do this properly.) > > This was caused by Firefox' use of GDK3 to implement its windowing and input needs which does not support hold gestures. > > GDK4 does support them but, as I understand it, a port of Firefox to GDK4 would be a ton of work and there isn't really much desire for it as GDK4 doesn't offer many real advantages over GDK3 as Firefox doesn't use classical GTK widgets or anything and only really uses it for basic input/output primitives. > > A backport to handle hold gestures in GDK3 too was attempted but, in classic GNOME fashion, it was rejected. > > The implementation now somehow gets events from the touchpad directly via wayland somehow from what I could gather but if it works, it works. > > You can try this out in the latest nightly builds.
Atemu
3d ago
•
100%
It was there and "consumed space". It was just a less noticeable icon before (dark grey downwards-facing chevron).
I recently switched from a MBP to a Framework 16 as my primary laptop and one thing I immediately noticed was that I was unable to stop kinetic scrolls in Firefox by laying my fingers onto the touchpad. It'd just slide by unimpeded. You could work around this by counter-scrolling a little rather than holding still which is how I've been coping with it but it's suboptimal to say the least. (As are many things in the Linux touchpad experience. Linux desktop developers *really* ought to use a macbook for a little to get a sense for how to do this properly.) This was caused by Firefox' use of GDK3 to implement its windowing and input needs which does not support hold gestures. GDK4 does support them but, as I understand it, a port of Firefox to GDK4 would be a ton of work and there isn't really much desire for it as GDK4 doesn't offer many real advantages over GDK3 as Firefox doesn't use classical GTK widgets or anything and only really uses it for basic input/output primitives. A backport to handle hold gestures in GDK3 too was attempted but, in classic GNOME fashion, it was rejected. The implementation now somehow gets events from the touchpad directly via wayland somehow from what I could gather but if it works, it works. You can try this out in the latest nightly builds.
#### Better image search with AI-generated images detection Last week we got quite clear and [direct feedback](https://kagifeedback.org/d/4972-ability-to-filter-out-obvious-ai-image-websites-from-image-search) that we need to do something about AI-generated images that started creeping up in the image search. You asked for it and we delivered! We are enhancing our image search with a new ranking algorithm and an AI image filter. These updates prioritise authentic, human-created images and downrank AI-generated content. Users can also opt to exclude AI-generated images entirely. Learn more about Kagi's AI image filter here: [https://help.kagi.com/kagi/features/exclude-ai-images.html](https://help.kagi.com/kagi/features/exclude-ai-images.html) [](https://kagifeedback.org/assets/files/2024-10-17/1729194595-631881-image-search-ai-labelf8dce12e.png) We will continue actively improving this feature based on your feedback. ### Improvements and bug fixes - Ability to filter out obvious ai image websites from Image search [](https://kagifeedback.org/d/4972)[\#4972](https://kagifeedback.org/d/4972) [@Thibaultmol](https://kagifeedback.org/u/Thibaultmol) - Snaps not working with automatic quick answer [](https://kagifeedback.org/d/5022)[\#5022](https://kagifeedback.org/d/5022) [@Jesalx](https://kagifeedback.org/u/Jesalx) - No margin between clear and question mark [](https://kagifeedback.org/d/5031)[\#5031](https://kagifeedback.org/d/5031) [@azdanov](https://kagifeedback.org/u/azdanov) - Link for most blocked domains missing in new kagi website [](https://kagifeedback.org/d/4746)[\#4746](https://kagifeedback.org/d/4746) [@sw](https://kagifeedback.org/u/sw) - Quick answer [](https://kagifeedback.org/d/5061)[\#5061](https://kagifeedback.org/d/5061) [@TheBOT](https://kagifeedback.org/u/TheBOT) - Dark mode doesn't adapt [](https://kagifeedback.org/d/5013)[\#5013](https://kagifeedback.org/d/5013) [@greyfivenine8244](https://kagifeedback.org/u/greyfivenine8244) - Support time shorthands on timer widget [](https://kagifeedback.org/d/3444)[\#3444](https://kagifeedback.org/d/3444) [@sefidel](https://kagifeedback.org/u/sefidel) #### The Assistant - We've enhanced the Assistant mobile experience with several updates, including improvements to the thread view and header navigation - Kagi Assistant v2 mobile threads menu [](https://kagifeedback.org/d/4725)[\#4725](https://kagifeedback.org/d/4725) [@frin](https://kagifeedback.org/u/frin) - Kagi assistant gpt LLM not working [](https://kagifeedback.org/d/5055)[\#5055](https://kagifeedback.org/d/5055) [@ash](https://kagifeedback.org/u/ash) - Assistant Tooltips are Hard to Click [](https://kagifeedback.org/d/5037)[\#5037](https://kagifeedback.org/d/5037) [@silvenga](https://kagifeedback.org/u/silvenga) #### Celebrating Our Community At Kagi, we’re all about community, and nothing has made us prouder than seeing the incredible photos our members are sharing online of their Kagi swag! Whether you’ve posted a picture of your free t-shirt or a sticker proudly displayed on your laptop, we want to take a moment to say thank you. Your support means the world to us. **The story:** We were thrilled to celebrate our first big milestone: [reaching 20,000 loyal members](https://blog.kagi.com/celebrating-20k) who’ve stuck with us through this adventure. To show our appreciation, we've sent out free t-shirts and sticker packs to those first 20,000 supporters. Earlier this month, these members started receiving their packages and excitedly shared photos across various socials. Here’s a snapshot of just a few: [](https://kagifeedback.org/assets/files/2024-10-17/1729193189-984221-kagi-collage.png) For anyone who missed out, don’t worry! You can head over to the [Kagi Store](https://store.kagi.com/) and snag your first-edition Doggo t-shirt. We’ve got four fantastic designs, each one representing our determination to build a better, more ethical and more human web. #### Help us reach our next milestone: 50,000 members As we gear up to reach 50,000 [members](https://kagi.com/stats?stat=members), we’d love your help in spreading the word! We don't rely on ads to drive our growth; instead, we depend entirely on word-of-mouth marketing and community recommendations. This grassroots approach means that every conversation you have about Kagi, every post you share showcasing your swag, and every mention you make to friends and family has a real impact. Your enthusiasm helps us reach new members who share our vision for a better web. Let’s make our next milestone one to remember. #### Where to find and tag us on the social web: - [KagiHQ on X](https://twitter.com/KagiHQ) - [KagiHQ on Mastodon](https://mastodon.social/@KagiHQ) - [Kagi.com on Bluesky](https://bsky.app/profile/kagi.com) - [Kagi.com on Threads](https://threads.net/@kagihq) - [KagiHQ on Instagram](https://instagram.com/kagihq) We also have a [Kagi Community Discord server](https://kagi.com/discord) where you can join our community for real-time discussion about the product and chat with other Kagi users. As we move forward, we want to reiterate our commitment to continuously improving Kagi for all our members. Your [feedback](https://kagifeedback.org/) and engagement are invaluable as we strive to create an even better experience. Thank you for being a vital part of our journey - your support inspires us every day, and we can’t wait to see where this adventure takes us next!
Atemu
4d ago
•
100%
In what regard?
Atemu
4d ago
•
100%
Statistically, yes.
::: spoiler spoiler (This is a Joke.)
In simple terms, Large Language Models predict the continuation of a given text word-by-word. The algorithms it uses to do so use a quite gigantic corpus of statistical data and a few other minor factors to predict these words.
The statistical data is quite sophisticated but, in the end, it is merely statistical; a prediction for what is the most likely word given a set of words based on previous data. There is nothing intelligent in "AI" chat bots and the like.
If you ask an LLM chatbot a question, what actually happens is that the LLM predicts the most likely continuation of the question text. In almost all of its training data, what comes after a question will be a sentence that answers the preceding question and there are some other tricks to make it exceedingly likely for an answer to follow a question in chatbot-type LLMs.
However, if its data predicts that the most likely words that come after "What should I put on my Pizza" are "Glue can greatly enhance the taste of Pizza." then that's what it'll output. It doesn't reason about anything or has any sort of storage of facts that it systematically combines to give you a sensible answer, it merely predicts what a sensible answer could be based on what was probable according to the statistical data; it imitates.
If you have some text and want a probable continuation that often occured in texts similar to it, LLMs can be great for that. Though note that if it doesn't have any probable continuation, it will often fall back to an improbable one that is less improbable than all the others. :::
Josef Bacik who was working on this said: > I fixed up all your review comments, but yes we don't care about this internally > anymore so it's been de-prioritized. I have to rebase onto the new stuff, > re-run tests, fix any bugs that may have creeped in, but the current code > addressed all of your comments. Once I get time to get back to this you'll have > a new version in your inbox, but that may be some time. (Note: This was back in April.)
Atemu
1w ago
•
100%
Measure resource usage during play. What is the bottleneck?
Atemu
1w ago
•
100%
This has not been the case for at least a year or so thanks to graphics pipeline libraries.
Shader comp also only really manifests in frametime spikes, not generally high average frame times.
Atemu
1w ago
•
100%
Snaps are 👌
Such an obvious feature in retrospect. I added !ddgr to DDG back when I used it because reddit-exclusive results were so commonly useful but adding a bang for each and every website twice does not scale. (I don't need to do that very often these days due to Kagi's personalised ranking but it still happens.)
I wish snaps would stay in the @ syntax though as that'd make them much easier to edit. Perhaps that could be added as an option as it does hurt discoverability a little. @kagihq@mastodon.social
Hello, Kagi Fans! We know it's been a few weeks since our last update, and there's a good reason for that. The entire company took a break from the usual routine to fly to Barcelona, Spain, for our annual retreat - a chance to reenergize and refocus for the exciting journey ahead. [](https://kagifeedback.org/assets/files/2024-10-05/1728090499-177024-kagi-team.jpg) But don’t worry, it wasn’t all fun and no work. In fact, we’ve got a pretty exciting update for you today. The retreat gave us fresh inspiration, and we’re ready to bring some of those new ideas to life starting right now. ## Launching the official Kagi Android app We're delighted to announce the launch of the [official Kagi Android app!](https://play.google.com/store/apps/details?id=com.kagi.search&hl=en_US) This marks an important milestone in our mission to make Kagi Search more accessible and encourage more people to try it out. [https://play.google.com/store/apps/details?id=com.kagi.search](https://play.google.com/store/apps/details?id=com.kagi.search) The app allows users to begin using Kagi Search immediately right from the first launch (**no account needed**!). For those with existing subscriptions, signing in is seamless, enabling access to all Kagi features. Additionally, the app includes **native homescreen widgets**, providing one-click access to search directly from your homescreen. [](https://kagifeedback.org/assets/files/2024-10-10/1728592468-556003-image.png) If you use an Android device, go check it out! We've simplified the process of sharing Kagi with your network. Additionally, a recent [EU ruling](https://www.android.com/choicescreen/dma/) presents a significant opportunity for Kagi. Google is now required to include any search engine that meets [specific criteria](https://www.android.com/choicescreen/dma/searchengine/), such as having an app with **over 5,000 installs**, in the default list for Android and Chrome — [install the app](https://play.google.com/store/apps/details?id=com.kagi.search) today and help us meet the criteria! Stay tuned for more exciting updates coming soon! ## Introducing Snaps [Snaps](https://help.kagi.com/kagi/features/snaps.html) are an innovative search operator that allows for site-specific searches directly from the search bar. By simply typing `@` followed by the designated [site bang](https://help.kagi.com/kagi/features/bangs.html), you can limit your search results to a specific website. For example, to search on Reddit or HackerNews, you can enter: - `@r winter boots` which translates to `winter boots site:www.reddit.com` - and `@hn vim emacs` to `vim emacs site:news.ycombinator.com` This feature was made possible thanks to the contribution of our user [@tuesday](https://kagifeedback.org/u/tuesday) ( [](https://kagifeedback.org/d/1231)[\#1231](https://kagifeedback.org/d/1231)) on [Kagifeedback](https://kagifeedback.org)! This will work with any bang that Kagi support, including custom bangs you define. [Kagi Bangs](https://github.com/kagisearch/bangs) are open source, we welcome your contributions! ## Universal Summarizer for Chrome [Universal Summarizer](https://kagi.com/summarizer) is one of the most-loved Kagi treats. The extension is now available for Chrome - **[Download it here!](https://chromewebstore.google.com/detail/kagi-summarizer/dpaefegpjhgeplnkomgbcmmlffkijbgp)** We had to split it out of Kagi Search extension for Chrome due to rules of the Chrome store. Kagi Universal Summarizer is available to Firefox users via [Kagi Search for Firefox](https://addons.mozilla.org/en-US/firefox/addon/kagi-search-for-firefox/) and is natively built into the [Orion browser](https://kagi.com/orion). ## The Assistant - 30 day update It's been 30 days since we launched the Assistant, and we're excited to share a range of new updates with you. **Upload functionality**: We're excited to announce that you can now **upload a variety of local files directly to the Assistant**! This new feature allows you to **add multiple files in a single prompt**. Here's a list of supported file types: - Text files: .txt, .md - Rich text documents: .pdf, .docx, .pptx - Tabular data files: .json, .csv, .xlsx, ... - Images: .jpeg, .png, .webp, ... - Audio files: .mp3, .wav, .m4a, ... [](https://kagifeedback.org/assets/files/2024-10-11/1728670037-180584-image.png) Additionally, you can **add a URL** to have its contents summarised or to pose questions about it. [](https://kagifeedback.org/assets/files/2024-10-11/1728660555-886127-image.png) **Stop button** We've introduced a Stop button that allows users to immediately halt the Assistant's streaming output. [](https://kagifeedback.org/assets/files/2024-10-08/1728393487-484837-image.png) **Mobile improvements** And also made several enhancements to our mobile experience, and there's even more on the way! [](https://kagifeedback.org/assets/files/2024-10-11/1728668293-62610-image.png) ### Plus many other Assistant improvements - Give assistant a PWA manifest [](https://kagifeedback.org/d/4734)[\#4734](https://kagifeedback.org/d/4734) [@Thibaultmol](https://kagifeedback.org/u/Thibaultmol) - Add Stop button for AI assistants [](https://kagifeedback.org/d/4034)[\#4034](https://kagifeedback.org/d/4034) [@andrew-rivera](https://kagifeedback.org/u/andrew-rivera) - Multiline assistant queries on mobile [](https://kagifeedback.org/d/4693)[\#4693](https://kagifeedback.org/d/4693) [@fxgn](https://kagifeedback.org/u/fxgn) - Added `Llama405b` model to the list of available models in the Assistant - Assistant v2 Text Input Form increases with use [](https://kagifeedback.org/d/4799)[\#4799](https://kagifeedback.org/d/4799) [@DaJonas](https://kagifeedback.org/u/DaJonas) - Session Link for Kagi Assistant [](https://kagifeedback.org/d/4772)[\#4772](https://kagifeedback.org/d/4772) [@melomac](https://kagifeedback.org/u/melomac) - Assistant V2 initial prompt text field doesn't scroll [](https://kagifeedback.org/d/4739)[\#4739](https://kagifeedback.org/d/4739) [@Thibaultmol](https://kagifeedback.org/u/Thibaultmol) - Kagi Assistant UI Issues [](https://kagifeedback.org/d/4729)[\#4729](https://kagifeedback.org/d/4729) [@emptyjar](https://kagifeedback.org/u/emptyjar) - Browsing to [https://kagi.com/assistant](https://kagi.com/assistant) when not logged in results in 404 [](https://kagifeedback.org/d/4765)[\#4765](https://kagifeedback.org/d/4765) [@cmart](https://kagifeedback.org/u/cmart) - \[firefox\] text box in assistant text box can display funny at wrap point [](https://kagifeedback.org/d/4002)[\#4002](https://kagifeedback.org/d/4002) [@sbeckeriv](https://kagifeedback.org/u/sbeckeriv) - Universal summarizer can't find transcript: yet finds transcript for slightly different URL edition [](https://kagifeedback.org/d/4451)[\#4451](https://kagifeedback.org/d/4451) [@leftium](https://kagifeedback.org/u/leftium) - Assistant history vanishes temporarily after branching [](https://kagifeedback.org/d/4686)[\#4686](https://kagifeedback.org/d/4686) [@Value7609](https://kagifeedback.org/u/Value7609) - Assistant v2. Custom LLM settings not persistent. [](https://kagifeedback.org/d/4672)[\#4672](https://kagifeedback.org/d/4672) [@flokerama](https://kagifeedback.org/u/flokerama) - New Assistant Beta looks bad in Safari [](https://kagifeedback.org/d/4675)[\#4675](https://kagifeedback.org/d/4675) [@Rudotel](https://kagifeedback.org/u/Rudotel) - Assistant does not respect lens when searching [](https://kagifeedback.org/d/4756)[\#4756](https://kagifeedback.org/d/4756) [@double](https://kagifeedback.org/u/double) - Assistant - Broked Fonts [](https://kagifeedback.org/d/4837)[\#4837](https://kagifeedback.org/d/4837) [@silvenga](https://kagifeedback.org/u/silvenga) - Assistant not respecting multiple lines. [](https://kagifeedback.org/d/4764)[\#4764](https://kagifeedback.org/d/4764) [@tombert](https://kagifeedback.org/u/tombert) - Opus 3 interacting poorly with internet access [](https://kagifeedback.org/d/4804)[\#4804](https://kagifeedback.org/d/4804) [@artemvmin](https://kagifeedback.org/u/artemvmin) - Assistant (V2): make the 'send' button in the chat gray as long as the llm is still 'typing' [](https://kagifeedback.org/d/4738)[\#4738](https://kagifeedback.org/d/4738) [@Thibaultmol](https://kagifeedback.org/u/Thibaultmol) - Claude models are confused about internet access in Assistant with Web Access [](https://kagifeedback.org/d/4843)[\#4843](https://kagifeedback.org/d/4843) [@Nashwan](https://kagifeedback.org/u/Nashwan) - Model name is intersecting the assistant answer on mobile [](https://kagifeedback.org/d/4694)[\#4694](https://kagifeedback.org/d/4694) [@fxgn](https://kagifeedback.org/u/fxgn) - Completely innacurate assistant context [](https://kagifeedback.org/d/4667)[\#4667](https://kagifeedback.org/d/4667) [@tinkling6961](https://kagifeedback.org/u/tinkling6961) - Assistant - Inconsistant "New Tab" Behavior [](https://kagifeedback.org/d/4909)[\#4909](https://kagifeedback.org/d/4909) [@silvenga](https://kagifeedback.org/u/silvenga) - New Assistant - Asking to Search [](https://kagifeedback.org/d/4794)[\#4794](https://kagifeedback.org/d/4794) [@silvenga](https://kagifeedback.org/u/silvenga) - New Kagi Assistant has lost the ability to parse line breaks [](https://kagifeedback.org/d/4778)[\#4778](https://kagifeedback.org/d/4778) [@orb](https://kagifeedback.org/u/orb) - Impossible to copy the code while the response is still being displayede [](https://kagifeedback.org/d/4848)[\#4848](https://kagifeedback.org/d/4848) [@kfbme](https://kagifeedback.org/u/kfbme) - Assistant.v2 firefox issues with scrolling to bottom [](https://kagifeedback.org/d/4662)[\#4662](https://kagifeedback.org/d/4662) [@zut](https://kagifeedback.org/u/zut) - Assisant 'internet=off' is ignored [](https://kagifeedback.org/d/4965)[\#4965](https://kagifeedback.org/d/4965) [@Thibaultmol](https://kagifeedback.org/u/Thibaultmol) - The buttons on the bottom of a thread (copy, regenerate, info) shouldn't be inactive depending on how far you scrolled [](https://kagifeedback.org/d/4831)[\#4831](https://kagifeedback.org/d/4831) [@chris_20017](https://kagifeedback.org/u/chris_20017) - Assistant v2 Text Input Form increases with use [](https://kagifeedback.org/d/4799)[\#4799](https://kagifeedback.org/d/4799) [@DaJonas](https://kagifeedback.org/u/DaJonas) - An unknown error has occured in assistant [](https://kagifeedback.org/d/4988)[\#4988](https://kagifeedback.org/d/4988) [@oNco](https://kagifeedback.org/u/oNco) - Assistant mode adding redundant newlines on paste [](https://kagifeedback.org/d/4845)[\#4845](https://kagifeedback.org/d/4845) [@pravinxor](https://kagifeedback.org/u/pravinxor) - Code font is giant on phone for assistant [](https://kagifeedback.org/d/4797)[\#4797](https://kagifeedback.org/d/4797) [@tombert](https://kagifeedback.org/u/tombert) - Universal Summarizer dropdown values unreadable in Dark mode [](https://kagifeedback.org/d/4749)[\#4749](https://kagifeedback.org/d/4749) [@Christoph](https://kagifeedback.org/u/Christoph) - Assistant V2beta character display error [](https://kagifeedback.org/d/4673)[\#4673](https://kagifeedback.org/d/4673) [@OxygenX](https://kagifeedback.org/u/OxygenX) - Assistant thread UI buttons on Firefox mobile don't work [](https://kagifeedback.org/d/4805)[\#4805](https://kagifeedback.org/d/4805) [@tinkling6961](https://kagifeedback.org/u/tinkling6961) - Paste an image in research assistant to chat about it [](https://kagifeedback.org/d/4028)[\#4028](https://kagifeedback.org/d/4028) [@hadario](https://kagifeedback.org/u/hadario) - The Assistant input box isn't cleared when submitting a prompt using the microphone if the device keyboard has been used [](https://kagifeedback.org/d/4923)[\#4923](https://kagifeedback.org/d/4923) [@laiz](https://kagifeedback.org/u/laiz) - Copy the search prompt to the assistant prompt when "The Assistant" button is clicked on the search page [](https://kagifeedback.org/d/5005)[\#5005](https://kagifeedback.org/d/5005) [@webwalrus](https://kagifeedback.org/u/webwalrus) ## Search improvements and bug fixes - New users can now register for Kagi using their GitHub credentials - Sort news articles by time [](https://kagifeedback.org/d/142)[\#142](https://kagifeedback.org/d/142) [@Kai](https://kagifeedback.org/u/Kai) - "News" doesn't use the browser preferred language [](https://kagifeedback.org/d/589)[\#589](https://kagifeedback.org/d/589) [@cbaconnier](https://kagifeedback.org/u/cbaconnier) - Suddenly images are always using SafeSearch, even when disabled. [](https://kagifeedback.org/d/4560)[\#4560](https://kagifeedback.org/d/4560) [@ShadowKGI](https://kagifeedback.org/u/ShadowKGI) - Exclude site from search works very inconsistently [](https://kagifeedback.org/d/4627)[\#4627](https://kagifeedback.org/d/4627) [@dantheman](https://kagifeedback.org/u/dantheman) - Inform user on misspelled words [](https://kagifeedback.org/d/3029)[\#3029](https://kagifeedback.org/d/3029) [@null](https://kagifeedback.org/u/null) - No code snippets in Europe [](https://kagifeedback.org/d/3559)[\#3559](https://kagifeedback.org/d/3559) [@NevevrAlak](https://kagifeedback.org/u/NevevrAlak) - Duplicate search results on case-insensitive sites (e.g. GitHub) [](https://kagifeedback.org/d/2071)[\#2071](https://kagifeedback.org/d/2071) [@pawo2500](https://kagifeedback.org/u/pawo2500) - Prefixing @ (the at symbol) to a query doesnt give a results page (redirects to home) [](https://kagifeedback.org/d/4951)[\#4951](https://kagifeedback.org/d/4951) [@ChristenGottschlich](https://kagifeedback.org/u/ChristenGottschlich) - Summarizer "ask questions about page" chat cannot determine authors of papers from PubMed [](https://kagifeedback.org/d/4594)[\#4594](https://kagifeedback.org/d/4594) [@Nashwan](https://kagifeedback.org/u/Nashwan) - Phone number recognized as a fedex tracking number [](https://kagifeedback.org/d/4541)[\#4541](https://kagifeedback.org/d/4541) [@Ronzino](https://kagifeedback.org/u/Ronzino) - Searching "Hunter S. Thompson" gives at-a-glance results for Oscar Zeta Acosta [](https://kagifeedback.org/d/4714)[\#4714](https://kagifeedback.org/d/4714) [@UndarkAido](https://kagifeedback.org/u/UndarkAido) - Visit website button is not clearly visible in light mode [](https://kagifeedback.org/d/4550)[\#4550](https://kagifeedback.org/d/4550) [@astronaut](https://kagifeedback.org/u/astronaut) - ‘Showing results for’ has weird padding [](https://kagifeedback.org/d/4644)[\#4644](https://kagifeedback.org/d/4644) [@sefidel](https://kagifeedback.org/u/sefidel) - Language Error in Popup [](https://kagifeedback.org/d/4899)[\#4899](https://kagifeedback.org/d/4899) [@greyfivenine8244](https://kagifeedback.org/u/greyfivenine8244) - Safe search and safe image search are not about privacy [](https://kagifeedback.org/d/4892)[\#4892](https://kagifeedback.org/d/4892) [@slamor](https://kagifeedback.org/u/slamor) - Add British English (en_GB) locale [](https://kagifeedback.org/d/4615)[\#4615](https://kagifeedback.org/d/4615) [@sefidel](https://kagifeedback.org/u/sefidel) - IP Address Widget Fails to Resolve IPv4 Address [](https://kagifeedback.org/d/4239)[\#4239](https://kagifeedback.org/d/4239) [@silvenga](https://kagifeedback.org/u/silvenga) - Wolfram info button goes on top of search suggestions [](https://kagifeedback.org/d/4678)[\#4678](https://kagifeedback.org/d/4678) [@sefidel](https://kagifeedback.org/u/sefidel) - US news result mentioning Poland caused response to be in Polish [](https://kagifeedback.org/d/4589)[\#4589](https://kagifeedback.org/d/4589) [@yokoffing](https://kagifeedback.org/u/yokoffing) - Custom bang names with colons fail to render past the colon on the edit page [](https://kagifeedback.org/d/4777)[\#4777](https://kagifeedback.org/d/4777) [@kenk](https://kagifeedback.org/u/kenk) - Inform user on misspelled words [](https://kagifeedback.org/d/3029)[\#3029](https://kagifeedback.org/d/3029) [@null](https://kagifeedback.org/u/null) - FastGPT No Longer Cites Sources [](https://kagifeedback.org/d/4840)[\#4840](https://kagifeedback.org/d/4840) [@fish](https://kagifeedback.org/u/fish) - Universal Summarize does not summarize Yahoo Finance article [](https://kagifeedback.org/d/4821)[\#4821](https://kagifeedback.org/d/4821) [@yokoffing](https://kagifeedback.org/u/yokoffing) - Summary bang for specific language [](https://kagifeedback.org/d/4954)[\#4954](https://kagifeedback.org/d/4954) [@Albi](https://kagifeedback.org/u/Albi) - Redirects parses URLs incorrectly when leaving out protocol [](https://kagifeedback.org/d/4619)[\#4619](https://kagifeedback.org/d/4619) [@utopia](https://kagifeedback.org/u/utopia) - Display title of summarized website [](https://kagifeedback.org/d/4834)[\#4834](https://kagifeedback.org/d/4834) [@Wojtek](https://kagifeedback.org/u/Wojtek) - Repeated podcast results [](https://kagifeedback.org/d/4683)[\#4683](https://kagifeedback.org/d/4683) [@menturi](https://kagifeedback.org/u/menturi) - Scroll bar stutters in Kagi Summarzier [](https://kagifeedback.org/d/4628)[\#4628](https://kagifeedback.org/d/4628) [@Fracture](https://kagifeedback.org/u/Fracture) - Prefixing @ (the at symbol) to a query doesnt give a results page (redirects to home) [](https://kagifeedback.org/d/4951)[\#4951](https://kagifeedback.org/d/4951) [@ChristenGottschlich](https://kagifeedback.org/u/ChristenGottschlich) - Broken currency formatting Thai Baht (฿) [](https://kagifeedback.org/d/4970)[\#4970](https://kagifeedback.org/d/4970) [@MX5](https://kagifeedback.org/u/MX5) - Advanced Search info truncated on mobile [](https://kagifeedback.org/d/4779)[\#4779](https://kagifeedback.org/d/4779) [@jessemerriman](https://kagifeedback.org/u/jessemerriman) - Select elements don't work in Kagi settings (Safari 15.6.1) [](https://kagifeedback.org/d/4565)[\#4565](https://kagifeedback.org/d/4565) [@carl](https://kagifeedback.org/u/carl) - Clickable images on Changelog [](https://kagifeedback.org/d/4691)[\#4691](https://kagifeedback.org/d/4691) [@dix](https://kagifeedback.org/u/dix) - Quick Answer does not always link sources [](https://kagifeedback.org/d/4883)[\#4883](https://kagifeedback.org/d/4883) [@menturi](https://kagifeedback.org/u/menturi) - AI refers to system prompt [](https://kagifeedback.org/d/4942)[\#4942](https://kagifeedback.org/d/4942) [@maskmonarch](https://kagifeedback.org/u/maskmonarch) - Ability to filter out obvious ai image websites from Image search [](https://kagifeedback.org/d/4972)[\#4972](https://kagifeedback.org/d/4972) [@Thibaultmol](https://kagifeedback.org/u/Thibaultmol) - Editing Custom Bangs cuts off value of Search Engine name field after first word [](https://kagifeedback.org/d/4886)[\#4886](https://kagifeedback.org/d/4886) [@Christoph](https://kagifeedback.org/u/Christoph) - Line height not correct [](https://kagifeedback.org/d/4898)[\#4898](https://kagifeedback.org/d/4898) [@greyfivenine8244](https://kagifeedback.org/u/greyfivenine8244) - Leaving the image gallery/view by pressing the X or esc requires several clicks/taps [](https://kagifeedback.org/d/4839)[\#4839](https://kagifeedback.org/d/4839) [@laiz](https://kagifeedback.org/u/laiz) - Searching with site:.TLD yields results from other TLDs [](https://kagifeedback.org/d/4880)[\#4880](https://kagifeedback.org/d/4880) [@campfire](https://kagifeedback.org/u/campfire) - The !scoop Bang has weird behaviour when left empty [](https://kagifeedback.org/d/4963)[\#4963](https://kagifeedback.org/d/4963) [@LucasOe](https://kagifeedback.org/u/LucasOe) - Repeated podcast results [](https://kagifeedback.org/d/4683)[\#4683](https://kagifeedback.org/d/4683) [@menturi](https://kagifeedback.org/u/menturi) - !cocoapods search is broken [](https://kagifeedback.org/d/5011)[\#5011](https://kagifeedback.org/d/5011) [@sidwolf6583](https://kagifeedback.org/u/sidwolf6583) - 502 Server Error whenever I try to modify a bang [](https://kagifeedback.org/d/4960)[\#4960](https://kagifeedback.org/d/4960) [@umar](https://kagifeedback.org/u/umar) - Clear button doesn't work on android app [](https://kagifeedback.org/d/5002)[\#5002](https://kagifeedback.org/d/5002) [@agustinfitipaldi](https://kagifeedback.org/u/agustinfitipaldi) - Viewing individual search usage in a duo/family plan [](https://kagifeedback.org/d/4597)[\#4597](https://kagifeedback.org/d/4597) [@devilkin](https://kagifeedback.org/u/devilkin) - Dates are often wrong in search results & filters [](https://kagifeedback.org/d/2603)[\#2603](https://kagifeedback.org/d/2603) [@Vapid](https://kagifeedback.org/u/Vapid) - Improvements to Kagi changelog [](https://kagifeedback.org/d/4913)[\#4913](https://kagifeedback.org/d/4913) [@laiz](https://kagifeedback.org/u/laiz) - Lenses description tweaks [](https://kagifeedback.org/d/4606)[\#4606](https://kagifeedback.org/d/4606) [@laiz](https://kagifeedback.org/u/laiz)
Atemu
1w ago
•
100%
I'm a bit apprehensive towards LLM-generated things in general but I also use this from time to time when I just want a simple fact or get a feeling of whether the page is at all worth visiting.
I limit myself to facts that are trivial to verify or should be very hard for an LLM to get wrong.
If I need to look up some function for example, I'd very quickly find out if it's just hallucination. I also expect them to return the right thing if I expect the result to statistically be the most common thing.
Though if I'm being honest, I'd probably prefer a function to just get an expanded preview with the entire preview text offered by the website for this use-case (or an equivalent plain text extract from the website).
Atemu
2w ago
•
100%
Back when I tried it it was a lot worse for my purposes.
I'd recommend you try both though.
Atemu
2w ago
•
100%
The most important features when handwriting IMHO are selection tools and then being able to manipulate the selected strokes.
Write implements a multitude of selection tools such as lasso which most tools have but much more useful to me were ruled selection which selects based on lines on a ruled paper and path selection which selects every stroke you touch with your selection stroke.
You can then move the selected strokes in a ruled manner, so for example I'd select a whole line of strokes and move them down a few lines. This is incredibly useful and brings many of the freedoms we enjoy in editing text on a computer to handwriting.
Re-flowing using stroke divisors is an amazing feature in theory but I've never been able to make it work reliably enough for my purposes, so I personally disabled that particular feature.
The undo/redo dial is also pretty neat.
Once you actually try to take real notes or solve some mathematical problems, you'll really come to appreciate such features and will dread using any note taking application supposedly made for handwritten notes that does not implement such features.
Atemu
2w ago
•
100%
While that's certainly true, using NixOS usually does not involve many advanced concepts or requires you to understand them.
You can set foo = bar in a .conf file without knowing what a variable is either.
Atemu
2w ago
•
100%
I don't know about rnote but Xournalpp was very underwhelming when it came to actual handwriting features back when I tried it.
Atemu
2w ago
•
100%
It was an old Fujitsu Q755, not something I'd recommend you buy.
Had a wacom tablet built into the touch screen though which is the only thing I'd watch out for.
Atemu
2w ago
•
100%
Proprietary and closed source.
I always wondered why as they never sold it or had any kind of business model around it.
Atemu
2w ago
•
100%
(nixos more or less requires you understand programming syntax for writing your system config)
It's technically not a real programming language but an expression language. The difference is that the former is a series of commands to execute in the specified order to produce arbitrary effects while the latter is a declaration of a set of data. You can think of it like writing a config file i.e. in JSON format.
The syntax isn't really the hard part here. You can learn the basics that comprise 99% of Nix code in a few minutes.
The actually hard part is first figuring out what you even want to do and then second how the NixOS-specific interface for that thing is intended to be used. The former requires general Linux experience and the latter research and problem solving skills.
cross-posted from: https://lemmy.ml/post/21154325 > Write is a handwriting app that works on a lot of platforms including Linux which cannot be said about most handwritten note-taking applications. > > More information and demo: https://github.com/styluslabs/Write/ > > I've used it for uni on a Linux tablet/convertible and it worked really quite well and has some nice convenient features for note-taking. > > The UI looks like it's from android 4.something though ^^' > > What I really appreciate about it is that its storage format are plain SVG(Z) which are extremely compatible. All you need to view your scribbles is an SVG viewer (i.e. a web browser) which basically every computer with a GUI has. Their website is in fact mostly just the output of their own app.
Write is a handwriting app that works on a lot of platforms including Linux which cannot be said about most handwritten note-taking applications. More information and demo: https://github.com/styluslabs/Write/ I've used it for uni on a Linux tablet/convertible and it worked really quite well and has some nice convenient features for note-taking. The UI looks like it's from android 4.something though ^^' What I really appreciate about it is that its storage format are plain SVG(Z) which are extremely compatible. All you need to view your scribbles is an SVG viewer (i.e. a web browser) which basically every computer with a GUI has. Their website is in fact mostly just the output of their own app.
www.planetside2.com
