Serverless functions aren't ideal for persistent/long running tasks for the reasons you've found. Depending on your expected level of traffic, it might be worth looking into setting up a cheap backend. Even the most discount-pile shared VPS boxes Ive used recently have had pretty acceptable performance for tasks like this where a second or two of extra latency isn't critical.

How does Google cloud platform, and enterprise services fit into this? I'd have assumed they were absolute cash cows in a similar way to how if you look at Amazon's profit breakdown, they're an extremely profitable hosting company that also happens to have a marketplace.

Don't have a link because i'm on mobile, but at least as of a few years ago, google specifically claimed that they didn't use bounce rate as a factor when determining search rankings. It makes sense, imagine how easily you could damage a competitor if all it took was clicking their result and then leaving, over and over again.

If they show ads on their site, putting the recipe at the bottom increases the amount of time viewing them.

I just searched for "dogs" which I would assume has millions of viable hits, and hit next as many times as it would let me.

First time trying I got to page 22 before getting a notice that "many results had been removed as duplicate" and that I could try repeating the search without a duplicate filter.

Trying again without the duplicate filter it maxes out at page 43.

There are 43 pages worth of results on the internet relating to "dogs" apparently.

Slightly different to the comment you're replying to, but I guarantee you've experienced your brain "backfilling" memory before without realizing it.

Have you ever noticed that when first look at an analog clock, the first tick sometimes seems to take noticeably longer than a second? That's because when your eyes move they don't do it smoothly, and you're actually blind for the short period of your eyes moving (called saccade).

So what your brain does to help out, is take the image from when your eyes stop moving, and use it to retroactively fill in the blind spot.

This phenomena is called chronostasis.

I'd strongly recommend looking into webhooks, this is exactly the type of usecase for them. A webhook is essentially a post request sent by a service (in this case stripe) after an event has taken place, so it will make your life easier than trying to implement the same thing yourself.

I'm sure you've probably seen it already but here's a link to the stripe api docs for webhooks, it looks like they do have one that triggers on payment confirmation.

Practicing api could be good via postman?

Postman is definitely a tool you should learn to use well, it will come in handy.

And watching some yt guides?

I don't have any that I can recommend but I'm sure there are hundreds out there.

Any software to actually try it?

There are quite a few free api's available that you can use to learn.

You could start with something simple like fetching basic data from JSONPlaceholder which requires no authentication.

Then you could try something like fetching data from the Twitch API or the Reddit API which will require you to learn how authentication works.

It's less of a hassle than I probably made it sound. Testing it now, it took one or 2 seconds max, and saves the risk of having to install an extension with the permissions necessary to take full page screenshots.

It's mostly shortcuts/muscle memory:

ctrl shift j

ctrl shift p

"screenshot"

down arrow key

enter

You can do it natively in chrome/chromium as well. Open dev tools, press ctrl+shift+p, type "screenshot" and select the "full page screenshot" option.

SQL is good general knowledge regardless of platform because integrating with a database will always be useful. I'm not sure about python being used with either SAP or Oracle, but from my very limited experience with SAP S/4hana cloud, the main languages that were used in the integration suite were javascript and groovy.

Having said that, the best thing to learn is how API's work, regardless of which language you choose to learn. If you know how to work with an API, how to understand responses codes, how to deal with authentication, how to transform data in python, you'll be easily able to apply that knowledge to any other language, java, javascript, c# etc. The most important thing is understanding API fundamentals.

Not sure if this will be helpful, but a lot of IT vendors are always on the lookout for integrations people to implement integrations for enterprise software (SAP, Oracle, Mulesoft) etc. Perhaps that would be something you could look into if you're interested in integrations.

It's a skill like anything else, the more you practise, and the better quality that practise is, the better you will become.

Does the factor of safety not vary by industry and application?

From your description it sounds like you're trying to generate a website from markdown files. If that's the case, the term you'll need to google is "static site generation". Some specific options you might want to look into are nextJS, Eleventy, Jekyll/Hugo, Docusaurus, but there are tons more if none of those fit the bill.

My guess would be that during the count, people keep their hand in a fist. If they haven't decided what they're going to throw when the time comes, the simplest thing is to keep their hand in the same position, throwing a rock.

NextJS supports static site generation. You can drag and drop your folder onto netlify to deploy.

Not OP but understanding the emotional conflict that could arise from their situation really isn't difficult...

You just reminded me of something that they do in navy seal training called "drownproofing" or similar. Candidates have their wrists and ankles bound and have to bop up and down in a 10ft pool, pushing off the bottom like you described.

Have you been going to http://localhost:3000 or just localhost:3000? Some browsers automatically upgrade all connections to https if you don't specify http.

Unless it's been changed recently, that's not possible as far as I'm aware.

Chromium dev team has specifically stated they will not enable such functionality.

See the other answers in the stack overflow linked for work arounds that might possibly be satisfactory.

Alternatively, on chrome/chromium at least, if you open developer tools, press ctrl-shift-p and search for full-page screenshot, it will do the same thing natively.

Any time there is a post about problems with the client, there are always people in the comments eager to say how it would be too expensive or too difficult to fix. I wanted to prove that that's bullshit, and the reason the client doesn't get fixed is that it's annoying, low impact and low priority. So to demonstrate that, I made a tool called LoL Client eXtenDer (LoLXD) that adds various functionality and improvements to the client.

Before anyone gets too excited, LoLXD is intended only as a demo, not a serious application, though I do have the code available if anyone is interested. I can't see any way it would give an unfair advantage to anyone, but even so Riot might ban you for using it, you've been warned.

In no particular order, the improvements I've made are:

1. Language selection/localization selection
2. An Auto-saving rune system
3. Enable replay modes (like in professional broadcasts)
   1. Extended zoom
   2. First person mode
4. Replay loader
5. Improved in-game mute command (/n mute all)
   1. Mutes everything EXCEPT team pings
   2. still able to type in chat
6. New voice comms mode
   1. Toggle mute hotkey
7. Open multiple clients
8. Simulate ping
9. enlarge in-game HUD size
   1. increase Chat/Map size
10. Restart Client UX process

Some other info that might be interesting:

How hard was it to create LoLXD?

Most improvements were fairly straight forward, either only requiring a change to config files/settings or launch options.

The rune autosave system was more complex because I had to rebuild the rune system

How did you add a new command to the game?

The improved in-game mute takes advantage of an existing command that nobody seems to know about or use. If you type /n anythinghere, "anythinghere" will be saved to a file called MyNotes.txt. LoLXD just watches that file to see if "mute all" gets added, and then triggers a macro to send /mute all and /pingmute team.

The reason the macro is a bit slow, is that making it any faster also made it work inconsistently, letters would be skipped etc.

Why use a browser for the GUI instead of making it a standalone app?

1. There used to be a problem where electron based apps (like discord, spotify etc) would freeze when the client was open, presumably because the client is built on CEF
2. Browser is light weight and most people (including myself) would already have chrome/FF open. No need to have a second chrome instance if I were to have made an electron app.
3. This is just a demo to show what's possible, and browser GUI is easy

For anyone curious, the front end is made with sveltekit + tailwind.

You seem familiar...

I also made liveviewsync.

Now, finally, I hope we can stop with the bullshit excuses for the client. Thanks!

twitter link in case anyone wants to be updated on other new projects: @LasTechLabs

They're all pretty easy to use. Cloudflare workers has documentation/examples that should cover your use-case here so that'd be a copy-paste and update with your API URL.

Other than that I'd check their pricing model, depending on how much traffic you're expecting, but again, all are fairly similar (free for the first X amount, then Y for Z amount after).