The most important function to write performant next.js apps 

The maintainer of bearded theme is cookn up my own webdevcody theme with my brand colors 🤣 I’m going to suss it out to see if people hate it or not

@@WebDevCody that's great lol

From nextjs 15 it wont be the default behaviour.

Would that be a security issue?

Often you can’t use fetch to get data from a database. You usually use a database driver.

How people missunderstand such things. React has nothing to do with fetch. Its behaviour that nextjs introduced and made it default which is sooo wrong. Fortunately they remove it as a default and instead it will be opt in in nextjs 15

@@Vantivify You're the one unfortunately misunderstanding. React ALSO extends the fetch API. That is why in the video he is importing from React and not Nextjs. This video is about request memoization which is a React caching layer. You are talking about the nextjs data cache layer, that is what is being turned off in nextjs 15. Look up request memoization if you don't believe me.

Yeah I only run getCurrentUser inside server components and often I’ll show different client components if the user is authenticated or not. Sometimes I’ll pass in the user info into the client components, but I often find it better to just try to conditionally show things in the RSC

@@WebDevCody If you're doing that why is the request running 6-7 times? Isn't it supposed to run on the initial request of the server component only?

​@@nasko235679 I believe it's running 6-7 times because we're looking at the outputted console.log statement from a function, whereas by default Server Components will usually look to memoize duplicate fetch requests instead (i.e. if the same fetch request is called multiple times from different components). So therefore, if your function had to do a lot of heavy calculations immediately after receiving that fetch data, then the cache idea would be a perfect opportunity to use it, to help ensure those calculations are only run once.

Cache is specifically for your RSC, context is for your client components. It’s different imo

gengar!!

Klack.app

Cody did a video on the app ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-AJDSZcv27JI.html

Curious as well

See this what I do then recall through middleware for certain pages I’m not sure if it’s right or wrong but I do it 😅

If you’re on a page that uses a cookie the page will always be dynamic

Klack app

Error lens

Maybe a custom with linear switches. Maybe tactiles.

he should be using the macbook keyboard with a software that adds the sound of key press on the recording. he has a video showing it

Bearded theme feat web dev Cody. It’s not out yet

​@@WebDevCodyohh, nice one.

RevalidatePath or revalidateTag but those are unrelated to what I just discussed

All my buttons and forms are always client components. I’m not sure I understand what you’re asking

@@WebDevCody if it's client side just make a context and all your client component check from the context

Idk how to answer this honestly. The cache is per request. Sometimes you need to verify values on each request on your backend. Sometimes different RSC in your tree might need the same data during your SSR, so you’d cache it server side

This is a different level of cache, this cache (request memoization) get evicted after the request ends. NextJS has 4 levels of caching, I know, extra confusing. You are talking about the Data cache.

​@@hello19286I understand but I'm asking about this specific use case since we know probably the user data won't change unless maybe u change your profile or do certain actions then you refresh the user data by fetching the user again. Isn't this a common practice among next.js devs to have a front end state store for such use cases ? Fetch once, share anywhere and refresh on need.

I mean, you should really be checking and validating the users, tokens or sessions every single request. For example, if someone invokes your server action, you need to make sure that the Paul request has a valid non-expired session so you can’t really trust that any other way, other than the look it up every time.

No clue, you’d need to check their docs or look to see if they call cache somewhere

Yes, it’s hard to answer this question. You cant connect to a database from your middleware because it requires serverless compatible libraries. Many database drivers are not serverless friendly, such as a postgres driver. This means that you either need to expose an api endpoint on your app and do a fetch request from your middleware to your api (which is slow and defeats the purpose of middleware running on edge), or you need to find a driver that will work (or find a way to expose your database over https publically which I’d say is a security concern). My personal opinion is to not use middleware because of added complexity and to just verify sessions on your RSC directly.

Not sure what you mean

Context is often still needed for various things. I guess it depends on what you need it for. Often you’ll fetch the initial value in your RSC and populate the context value using it so the client component can use it when it hydrates

@@WebDevCody makes sense, i see RSC has replaced a lot of the old fetch paradigms

You can’t access context values in RSC, but yes you could pass it down as props if you want

Web dev Cody theme

​@@WebDevCody Can i wish theme link 😂

i am also using convex for backend

Take that up with the react team 😉

Yeah

Sorry, many people like it

You’ll get your reward later tonight

@@WebDevCody woah

i was not expecting this when trying to learn react

@@WebDevCodyAwww! Cuddles and a cup of cocoa later tonight.

@@Innesb 🤣

It will implement RSC very soon so it doesn't change anything to be honest

Verifying the jwt in the middleware is also fine. You shouldn’t need to call cache in the middleware. If the token is invalid just redirect the user to another page. I think you’d still need to call a method to get the user session inside your RSC.