It would have been nice to see errors working, from what I see you can’t throw or return an error in these stream promises.
@EliSpizzichino23 сағат бұрын
How to upload the new doc for Svelte 5?
@StanislavKhromov22 сағат бұрын
Hi, the docs became too large to use the content.json method, I have tried slimming them down, you can try the content_optimized.json file from here, it fits in Projects context (barely): github.com/khromov/svelte-ai-context/tree/master/scripts
@SRG-Learn-Code2 күн бұрын
Great aproach, and since sveltekit allow you to csr, ssr and render statically, I guess a good approach would be to prerender the "website" part and stream the "app" part, if that's your deal.
@SRG-Learn-Code2 күн бұрын
great topic and great teaching style. Also I love that you record in 4k and you don't increase the font size to ridiculous amounts. You earned my like, subscription and bell.
@wentallout3 күн бұрын
ah yes the npm migration script type beat
@IvanKleshnin3 күн бұрын
You're completely right that there's more than a single type of web apps and we'll eventually have different frameworks to cover different needs of these approaches. The web is too complex and diverse to have one "perfect" swiss-knife tool. I would also note that API + classic SPA client approach is not going anywhere, especially with advent of languages like Mojo that are incomparably more performant than any VM can be. If your project involves ML, the easiest and the most maintenable approach might be to have your BE in pure Python/Mojo and your FE in TS . With some SPA like Svelte plus a bit of SSR if necessary. Mixing backend languages is a recipe for troubles.
@tommyeliassen43205 күн бұрын
Thanks for making this video, that was really enlightening. My bundle sizes will get smaller and fewer from now on :)
@devgauravjatt5 күн бұрын
Svelte js ❤🎉
@stepan_rybalka5 күн бұрын
Thanks for the advice for refactoring the bundle size
@mikejohneviota92935 күн бұрын
🎉🎉🎉
@Stefan-vz7op5 күн бұрын
Cool video. Thanks for sharing it
@StanislavKhromov5 күн бұрын
Thanks for watching! ✨
@jeffreyturkelson70516 күн бұрын
How often do update the code base file in the project itself / when is it necessary?
@StanislavKhromov5 күн бұрын
I usually update the code before starting a new chat. Changes you perform inside a chat are remembered.
@jakubthesmiley8 күн бұрын
whoa!
@mikeeomega8 күн бұрын
Hahahahaha vibing with this.
@BhideSvelte8 күн бұрын
Amazing 🎉🎉🎉
@budivoogt4919 күн бұрын
Useful video Stanislav! Helpful to understand there are bundling benefits without reworking to Svelte 5 standards.
@ЦветомирБановски13 күн бұрын
Hello Stanislav, wonderful video to begin with. I have a question: would it be possible AI-Digest to be tested with Llama3.1?
@StanislavKhromov5 күн бұрын
Thanks for your comment! I'll try to make one!
@runs961318 күн бұрын
great contribution!! And I feel no one else can answer this question better than you. Its has been two months, what are the alternatives you see now in comparison to proposed package? is ai-digest the only package that does this job better?
@StanislavKhromov16 күн бұрын
Thanks for the feedback! Aside from the Claude Projects workflow from this video (which I still use every day and have great success with), two new tools that have been released since this video are Aider and Cline (aka Claude Dev). There are upsides with these tool (more automated) but also downsides (harder and more expensive to add docs and the whole context). I am working on a video comparing the pros and cons of these tools, if that sounds interesting, feel free to subscribe! ☀
@deucebigs986019 күн бұрын
Thanks! Interesting strategies to get it to count that I'll keep in mind for the future.
@mikejohneviota929321 күн бұрын
more capacitor svelte 5 content plss 😢
@mikejohneviota929321 күн бұрын
hello sir🎉 new follower
@99dynasty29 күн бұрын
The type of language you give a smaller model for chain of thought would have to be researched and appropriately worded. It’s highly likely you could get a very similar increase in performance if you understood the exact limitations of that model
@fguerraz29 күн бұрын
I mean no offence but your prompt was not formulated in correct English for llama, so maybe that’s why it didn’t do great. “Count the number of the letter r” is not a great formulation
@StanislavKhromov29 күн бұрын
Hi, thanks for watching and for the comment. Even if we try to write in more correct english (eg: "How many times does the letter 'r' appear in the word 'strawberry'?") neither Llama 3.1 nor Sonnet can answer it correctly. For example, Llama 3.1 8b told me: "Let's count! The word "strawberry" contains two occurrences of the letter 'r'."
@bambanxАй бұрын
do you have discord ?
@StanislavKhromovАй бұрын
You can ask questions in the official Svelte Discord, i hang out there.
@bambanxАй бұрын
also suscribed and liked
@bambanxАй бұрын
my friend thanks for the video, is a bit advanced for me, no sure why you do some things like change adapter node ... or Which are the benefits of google domains vs cloduflare? Why on the docker file you use node 19.7 spline and in local you use nvm 18? Why in the docker file is exposed in port 3000 ? And not port of the production as I know is 4173 ? thanks you so much
@StanislavKhromovАй бұрын
Thanks for your comment! I'll try to address your points in order: 1. Google Domains does not exist any more, but Google Domains and Cloudflare are different, Cloudflare is a way to both manage domains and deploy your app, while Google Domains and other domain sellers just manage your domains. For an alternative to Google Domains, check out Porkbun. 2. If there is a discrepancy in Node.js version, I think it's a mistake, it should be the same in nvm and in the Dockerfile 3. Changing to adapter-node is required to deploy via Docker, you cannot use another adapter. 4. The adapter-node default port is 3000. Port 4173 is for Vite preview. Hope that helps!
@bambanxАй бұрын
@@StanislavKhromov thanks you so much for your answer, i was trying for a while without success and i foudn this on github and worked for me the dockerfile is: FROM node:18.7.0 WORKDIR /app COPY package.json package-lock.json ./ RUN npm ci COPY . . RUN npm run build ENV PORT 80 EXPOSE 80 CMD ["node", "build"]
@bambanxАй бұрын
what do you think about that setup ? thanks you
@StanislavKhromovАй бұрын
Sure, that looks fine. Since I made this video I also wrote a dedicated Docker tutorial for Kit that explains some of the details in the Dockerfile I used: khromov.se/dockerizing-your-sveltekit-applications-a-practical-guide/
@bambanxАй бұрын
@@StanislavKhromov just curious why it works on port 3000 like you did and in 80 as the code i showed you ?
@mykelfashonhous1037Ай бұрын
Hello Dear I hope this message finds you well. I'm Emon, a professional digital marketer with a focus on KZbin video SEO and Google Ads. Are you struggling with subscriber and video views for monetization? I wanted to reach out and share how my expertise can help enhance your KZbin channel's visibility and drive more engagement. I was analysing your KZbin channel.
@StanislavKhromovАй бұрын
Addendum: If you want a workaround to make streaming promises work for users without JavaScript, try the approach in this blog post. The downside to this approach is that the initial SSR page load will not have streaming, but every navigation after that will have streaming, as it will be a client side navigation using the SvelteKit router: geoffrich.net/posts/conditionally-stream-data/
@Yuliya.VerbenkoАй бұрын
The name of the browser suggests thoughts about Ukrainian roots, because in Ukrainian "STRIMKIY" = "FAST" - means something that is characterized by great speed, dynamism (about movement, any action, any process) ;) :)
@priyanjitdeyАй бұрын
Quality content
@liam_kАй бұрын
Currently making an offline first app, could you recommend a way to have a client side database (maybe indexed db) that syncs when internet is available?
@StanislavKhromovАй бұрын
Hi! There are a few project like that out there. I like Convex for this. Many people use Firebase but personally I don't like it. For self hosting, check out RxDB. Also check out "Storing data" on localfirstweb.dev/ they have tons of other projects you can use.
@asweare349128 күн бұрын
PGLite is a good Local first DB but you'd have to take care of syncing to the server. ElectricSQL is developing this and offers a full syncing service but I haven't used it.
@deucebigs9860Ай бұрын
Love this tutorial! I really appreciate your teaching style.
@prashlovessamosaАй бұрын
Great I am learning Sveltekit Thanks for sharing this
@KevinMacKenzie61Ай бұрын
Maybe I missed it, but when/why would someone use it?
@StanislavKhromovАй бұрын
It speeds up the initial page response so users see your site faster and then you can use streaming for secondary or slow data!
@KevinMacKenzie61Ай бұрын
@@StanislavKhromov Thank you!
@king-zu3ihАй бұрын
this is one round trip if you make a second request to load data. you need to repeat a step to verify auth
@BhideSvelteАй бұрын
amazing
@adriablancafortАй бұрын
Looking forward for the svelte 5 release.
@zsmainАй бұрын
That's informative, thank you. What are you using for the back-end?
@StanislavKhromovАй бұрын
Thanks! The app I showed briefly at the beginning is local only, it's using Capacitor Preferences, there is no backend. My previous app (Appreciation Jar) used Postgres and Pushpin for realtime updates. You can find some more info on my blog about the setup: khromov.se/building-a-privacy-friendly-self-hosted-application-architecture-with-sveltekit/
@zsmainАй бұрын
@@StanislavKhromov What backend do you recommend for someone who needs it?
@StanislavKhromovАй бұрын
Check out the blog post, there is a summary of the backend tech I like!
@zsmainАй бұрын
@@StanislavKhromov thank you, I like you blog, most probably I find there what I'm looking for.
@deucebigs9860Ай бұрын
More svelte 5 yes!
@StanislavKhromovАй бұрын
Svelte 5 is the future!
@dbug64Ай бұрын
Nice! I like the sound of your keyboard, what do you use?
@StanislavKhromovАй бұрын
Thanks! It's a Logitech MX Mini Mechanical with brown switches!
@NabekenProG87Ай бұрын
View #8 lets goooo! Keep it up ❤
@deucebigs9860Ай бұрын
Another great video! Thanks algorithm and Stanislav!
@deucebigs9860Ай бұрын
Thanks for making a svelte specific video. Subscribed and looking forward to more. Gonna explore your other videos now!
@blarviniusАй бұрын
Thanks! I will definitely give ai-digest a try tomorrow! I was doing all of that manually - copying codebase into project knowledge text files. It was exhausting work. 🤖🦕👽
@default_youtube_profileАй бұрын
I managed to get this to work, however the commit hash is not shown on the caprover but if we deploy usince caprover cli the commit has shows on the caprover app deployment page.
@StanislavKhromovАй бұрын
You are correct, to get the deploy to show the hash in the CapRover list of deploys you'll need to modify the GitHub workflow, save the commit hash as an output, then in the last step deploy the output hash instead of :main like the workflow does now.
@_butterfly1432 ай бұрын
Want video about today's technology trends and update ❤...pls 🙏
@FM-tq2gs2 ай бұрын
Great video! As I am still a beginner in webdev, I think that I will stick to "normal" SvelteKit with SSR and CSR, and learn local-first technology later if I need it. By the way, would you consider optimistic updates to be a kind of (very) light local-first method?
@StanislavKhromov2 ай бұрын
Thanks for the comment! Optimistic updates are a great way to make your app feel better to use, and it definitely fits with the "no spinners" criteria of local first!
@kirandev12 ай бұрын
How much messages will you have with almost 100% project knowledge filled? Based on the video it is about 10 messages. Am I right?
@StanislavKhromov2 ай бұрын
Right now it's around ~15 message every few hours.
@RenanFrancaАй бұрын
@StanislavKhromov how do you handle that limitation in your workflow? BTW, amazing video keep it up!
@StanislavKhromovАй бұрын
@@RenanFranca I spend some time writing the prompt as detailed as I can, this avoids misunderstandings and usually you can get a small feature done in 2-3 prompts. So you can do quite a lot if you try to be economical. If you need to ask questions that don't require lots of context you can use apps like Jan or Msty and connect to the Anthropic API which does not have as heavy limits.
@RenanFrancaАй бұрын
@@StanislavKhromov Thank you very much for the detailed answer! Great idea! Have you tried using Claude 3 Opus in your workflow? Is it worth it? Don't you use ChatGPT?
@StanislavKhromovАй бұрын
As of today, Opus scores worse on most coding benchmarks. I don't use ChatGPT because it scores worse than Sonnet currently and doesn't have anything equivalent to Projects. Hopefully they will improve!
@science_mbg2 ай бұрын
thanks a lot , very useful. I also added --ignore function in my REPO because I didn't know how to submit in yours. But superuseful, Thanks a lot, now I also got claude pro to try these extra codebase option.
@StanislavKhromov2 ай бұрын
Hi, you can use the .aidigestignore file to ignore parts of your project. You are welcome to create an issue on GitHub if you still have issues: github.com/khromov/ai-digest