nobody?

MMartin 0x522E nobody?

J

James•6/20/23, 1:34 PM

/ cc @celsomartinho

Aadamwathan Hey folks, Adam Wathan from the Tailwind CSS team here! We're hacking on a fun t...

I

itsmatteomanf•6/22/23, 3:44 PM

Hey there! Maybe @celsomartinho or @Vero

can help

Aadamwathan Hey folks, Adam Wathan from the Tailwind CSS team here! We're hacking on a fun t...

V

Vero•6/22/23, 3:47 PM

Hey Adam, welcome to the server! Sure, I've just enabled it for you

J

James•6/22/23, 3:49 PM

That sounds like a super cool use-case. Definitely let us know if you have any questions along the way!

V

Vero•6/23/23, 9:49 AM

^ @celsomartinho

T

Tilon•6/23/23, 4:12 PM

Hi, would the browser-rendering-api basically be able to do similar things like for example browserless.io? Like managed puppeteer instances? We are especially looking for a solution to take screenshots of WebGL based renderings at scale. So we have a website that renders a 3D model and we want to generate preview images (WEBP/JPEG) for those 3D models. Would that work with the browser API?

L

laure•6/26/23, 5:49 PM

would also love to know if there's a timeline for increasing the limits, and where they may end up!

N

n0riskn0r3ward•7/5/23, 9:14 PM

Hmm, just saw the limits after getting 429 errors:

Two new browsers per minute per account.
Two concurrent browsers per account.

Any word on when these limits will be lifted? My whole use case was to run a bunch of browsers simultaneously to build an index for a search engine with rendered html...

U

Unsmart•7/5/23, 9:20 PM

It is in beta currently and free so there needs to be some limits in place

C

Chaika•7/5/23, 9:25 PM

You can use the same Browser to access multiple pages, and you can persist it with Durable Objects, someone in this channel released their code for using a DO and a list of tasks: browser-rendering-api

C

Chaika•7/5/23, 9:27 PM

It's still limited with only two concurrent browsers, but you can queue and reuse the browser

W

white dragon•7/8/23, 6:13 PM

I have been using this API in a worker and it is awesome!

I have a question. Can this be used in pages/functions??

S

skorfmann•7/8/23, 7:15 PM

Is it possible to swap the

await puppeteer.launch(this.env.MYBROWSER);

await puppeteer.launch(this.env.MYBROWSER);

await puppeteer.launch(this.env.MYBROWSER);

await puppeteer.launch(this.env.MYBROWSER); for a "normal" puppeteer for local testing? I'm aware that

wrangler dev --remote

wrangler dev --remote

should be used ,but this causes other problems (e.g. that queues are not working there and durable objects seem to have issues for this env as well)

S

skorfmann•7/8/23, 7:36 PM

ok, so while the normal

puppeteer

puppeteer

puppeteer

puppeteer package doesn't

launch

launch

launch

launch (some node stuff is missing) it's working with

connect

connect

connect

connect to services like browserless. Think that's good enough for me right now. Gonna see if there's some opensource version of a websocket server for puppeteer

S

skorfmann•7/8/23, 7:43 PM

nice, so running this in a docker container https://github.com/browserless/chrome and doing this works great wing

wrangler dev

wrangler dev

this.browser = await puppeteer.connect({ browserWSEndpoint:

this.browser = await puppeteer.connect({ browserWSEndpoint:

this.browser = await puppeteer.connect({ browserWSEndpoint:

this.browser = await puppeteer.connect({ browserWSEndpoint: ws://localhost:3000

});

});

});

});

GitHub

GitHub - browserless/chrome: The browserless Chrome service in Dock...

The browserless Chrome service in Docker. Run on our cloud, or bring your own. - GitHub - browserless/chrome: The browserless Chrome service in Docker. Run on our cloud, or bring your own.

B

BlackForestBoi•7/10/23, 11:18 AM

hey you all, we've been playing with the browser rendering api and it works great, thanks!

A question we have for using it properly, even though in beta:
The 2 browsers per minute spin up is quite limiting. Is there the ability to increase that, and/or when do you roughly expect this enter GA so those limits are not there anymore?

H

Hello, I’m Allie!•7/10/23, 6:06 PM

HHello, I’m Allie!Empty message

B

BlackForestBoi•7/11/23, 12:32 PM

When people at Cloudflare say "soon" what range are we talking about here?

Not trying to get you pinned down just for us to have a rough overview if that is days, weeks or months?

N

NicS•7/13/23, 4:39 PM

https://tenor.com/view/dfk-bloaters-defikingdoms-gif-24254004

Tenor

N

NicS•7/13/23, 4:39 PM

Hey, you got any of them Beta invites?

NNicS Hey, you got any of them Beta invites?

M

meyer•7/13/23, 4:39 PM

N

NicS•7/13/23, 4:40 PM

Thanks

A

Abhishek Mirajkar•7/14/23, 4:07 PM

Hey all,

I'm trying to implement ServerSideRendering using Puppetter in workers,
Does anyone know if this would work?

also can anyone enable beta access for my account id please ? cce05b73fe5b9a4bd24dcc4e56b28cfb
It would be nice to have your inputs

Thanks

Original message was deleted

V

Vero•7/18/23, 4:08 PM

Heyo, can you please try again?

P

Pascale•7/19/23, 6:57 PM

https://github.com/cloudflare/puppeteer/pull/15 (PDF support) broke our wrangler deployments, but it outlines: "Adds back support to page.pdf() using Cloudflare Workers Streams and Buffers support." to use node:stream and node:buffer.

Any way to make it work with wrangler deployments?

S

Shahid khan•7/19/23, 6:59 PM

Can I create genesis block in cloudflare and mine it

P

Pascale•7/19/23, 7:00 PM

right.. yeah..

Thanks a lot.

S

Shahid khan•7/19/23, 7:00 PM

Can I create genesis block in cloudflare and mine it

Original message was deleted

A

Ashley•7/20/23, 10:53 AM

I just came across this one, all the guides / blogs are I think wrong, as they all use node_compat = true in them

And I had to switch to use the compatibility_flags

A

Ashley•7/20/23, 10:54 AM

I'm trying out the browser, using a simple bit of code:

const browser = await puppeteer.launch(env.BROWSER);
const page = await browser.newPage();
await page.goto(`https://www.some-page.com`);
const html = await page.content()

const browser = await puppeteer.launch(env.BROWSER);
const page = await browser.newPage();
await page.goto(`https://www.some-page.com`);
const html = await page.content()

const browser = await puppeteer.launch(env.BROWSER);
const page = await browser.newPage();
await page.goto(`https://www.some-page.com`);
const html = await page.content()

const browser = await puppeteer.launch(env.BROWSER);
const page = await browser.newPage();
await page.goto(`https://www.some-page.com`);
const html = await page.content()

And I'm getting

Trace resource limit exceeded; subsequent logs not recorded.

Trace resource limit exceeded; subsequent logs not recorded.

Trace resource limit exceeded; subsequent logs not recorded.

Trace resource limit exceeded; subsequent logs not recorded.. I'm guessing if you're doing browser stuff, you probably need to switch your worker types to unbounded?

A

Ashley•7/20/23, 10:59 AM

Although realistically I can't see why, as it's just getting a page and returning the content, which might take a little bit to get the page but it's not CPU time, and I'd be amazed if getting the page content was taking up a lot of CPU

A

Ashley•7/20/23, 11:05 AM

Interestingly, if I load a simpler web page, it completes

A

Ashley•7/20/23, 11:16 AM

Yeah, just seen this:

Median CPU Time 51.7 ms

Median CPU Time 51.7 ms

Median CPU Time 51.7 ms

Median CPU Time 51.7 ms so will switch this one to unbound, ty!

A

Ashley•7/20/23, 11:26 AM

I'm definitely being thick, but when I go to change the worker to unbound, and agree, it says I don't have an active subscription... but I can't for the life of me work out where to enable that subscription? As under billing -> subscriptions there's no option to add any

And I can't for the life of me find it anywhere in the dashboard!

A

Ashley•7/20/23, 11:26 AM

I'm being thick, there's literally a plans item on the left menu under workers

A

Ashley•7/20/23, 10:32 PM

Not sure if a bug or I'm doing something wrong, but the following code (amended the URLs) results in

Buffer is not defined

Buffer is not defined

Buffer is not defined

Buffer is not defined, which is a bit odd (same for .text() and .json()):

const [res] = await Promise.all([
            page.waitForResponse(res => res.url().includes('something') && res.request().method().toUpperCase() != "OPTIONS"),
            page.goto('https://example.com', {waitUntil: "domcontentloaded"}),
        ]);

console.log(await res?.buffer())

const [res] = await Promise.all([
            page.waitForResponse(res => res.url().includes('something') && res.request().method().toUpperCase() != "OPTIONS"),
            page.goto('https://example.com', {waitUntil: "domcontentloaded"}),
        ]);

console.log(await res?.buffer())

const [res] = await Promise.all([
            page.waitForResponse(res => res.url().includes('something') && res.request().method().toUpperCase() != "OPTIONS"),
            page.goto('https://example.com', {waitUntil: "domcontentloaded"}),
        ]);

console.log(await res?.buffer())

const [res] = await Promise.all([
            page.waitForResponse(res => res.url().includes('something') && res.request().method().toUpperCase() != "OPTIONS"),
            page.goto('https://example.com', {waitUntil: "domcontentloaded"}),
        ]);

console.log(await res?.buffer())

Effectively I'm trying to extract data from a network request that's made during the second page load, and using

pages.on

pages.on

pages.on

pages.on was behaving very strangely so gave this a whirl instead

A

Ashley•7/20/23, 10:37 PM

(I have

compatibility_flags = [ "nodejs_compat" ]

compatibility_flags = [ "nodejs_compat" ]

compatibility_flags = [ "nodejs_compat" ]

compatibility_flags = [ "nodejs_compat" ] in my wrangler.toml)

A

Ashley•7/20/23, 10:48 PM

I feel like this is possibly some Pupetter quirk I'm not aware of, but:

page.on('response', async(response) => {
            const request = response.request();

            if (request.url().includes('something')) {
                console.log(123);
                const text = await response.text();
                console.log(text);
                console.log(456);
            }
        })

page.on('response', async(response) => {
            const request = response.request();

            if (request.url().includes('something')) {
                console.log(123);
                const text = await response.text();
                console.log(text);
                console.log(456);
            }
        })

page.on('response', async(response) => {
            const request = response.request();

            if (request.url().includes('something')) {
                console.log(123);
                const text = await response.text();
                console.log(text);
                console.log(456);
            }
        })

page.on('response', async(response) => {
            const request = response.request();

            if (request.url().includes('something')) {
                console.log(123);
                const text = await response.text();
                console.log(text);
                console.log(456);
            }
        })

It will log 123 many times, but nothing after - never the text, nor the 456. Quite confused by this one

R

robert•7/21/23, 2:07 AM

Hey folks, excited to try out this new API, any chance I could get access to the beta?

Similar Threads

Similar Threads

Similar Threads