Video

601 Harris Views

Petra Junctions

As a web designer, I felt like this for a long time, but then I realized nothing I create will be rotting in a landfill for a thousand years either.

I feel this. Probably gonna quit for this reason.

Love this one. You guys have some of the best education videos AND music!

Awesome idea. I mean the shorts in general.

Abshire Orchard

I'm pretty sure I read somewhere that you can technically turn off the garbage collector and manage your own memory. I haven't messed around with that at all though

Turcotte Greens

002 Laurie Mountain

Skiles Ferry

Rita's content is amazing! Great interview to let us get to know the genius behind the videos! 💯

did anyone figure out how to iterate over aws regions with a for-each in terraform?

This is real world stuff, folks. This is how Big Finance uses IT beyond the work-a-day CRUD routine.

The level of detail shared in this ~4 minute video is great and as always Bill's style of teaching really makes it enjoyable to watch.

Such a great explanation

Zig to rule them all

I have concern about this approach: why middleware of the Web request processing shoud deal or have a knowledge about "transaction" for the database? When we use the approach that Bill suggested we are coupling Web request processing and Database "internal" details? Is this a right approach? Can be Database or Repository layer be responsible for all transaction management?

I was hoping to hear exploration and other uses of blockchain other than financial industry. Great interview btw

Incredible!

One of the best in class tutorials about go lang, i ever watched. This is recommended for computer science enthusiasts.

Awesome video! Saved me hours. Thanks :)

Excelente video Bill! Gracias!!

Could you please explain the benefits of sharing the transaction tx inside the context? Except for making the transaction tx available to any part of the program that has access to context.

If it uses Makefile, why not just use instead? It's like you''re making a knife using a knife...

New lengthout hair and beard, same bouncy-feet Bill K energy.

This is a Wonderfull content. Thanks Kelsy

Nice explanation! What about debugging async code? I tried it, but it never actually traces through the code. (Debugger just jumps to the end of async functions)

this seems lik eit starts in the middle of something but ihni where it starts..

Are they all there in the playlist? It doesn’t seem like it’s all the same series

I took this class at Ardan Labs. It was a really good class and definitely worth the money.

I am with Bill on this one. I am just a bit more extreme LOL.

sound is very low compared to others video in youtube.

cdebug exec 344dedd442e4 ls ln: /proc/1/root/.cdebug-5d382faa: Permission denied opensearch container

I like the speaker‘s presentation “Konrad Reichiiiiii” 0:27

Hey, I tried running the following snippet of code: ``` MaxInt64 = 9223372036854775807 RelativelyLargeConst = 9223372036854775808 ``` and encountered the error: `cannot use RelativelyLargeConst (untyped int constant 9223372036854775808) as int value in argument to fmt.Println (overflows)` Can you please explain the reason?

thank you, very useful

what about if we have many roles/words in the jwt payload ? every http call will decrease its performance...

In my opinion the need to evaluate code locally depends on several factors: - amount of changes; - how well you know the codebase; - the level of use of helping tools like SonarQube. I usually jump to local evaluation as needed. And my best code reviews always start from getting acquainted with functional requirements ;)

I wish there was a 'super like' option for this.

In truth, I don't understand how developers may forget to use sync primitives in a conc app and why that is a selling point for mature language like Rust. It sounds like bullshit; even if you forget, there are tests and code reviews after all.

Nice. It would help if you made the code window a little larger and/or used a slightly larger code font.

what was that?

2:06 vs 5:12

I very often find in someone else's code when an error is returned without its description. As a result, I spend a lot of time searching for where it happened and its causes.

I was about to write a different comment halfway the talk until Trusted Errors came up. I like to think of them as Expected Errors. Server is down, No rows in the database, I like my error values to have 2 error values. A front facing error value and a developer value, that is always nil in expected Errors. Non nil developer errors should not only be logged but raise some notification callback. Have a separate front facing error, makes it easier to translate those, if your front face is in multiple languages. Now my problem with golang Errors, and only Errors is that they should support exhaustive handling. What do I mean by that. When checking an error, I should be able to do (or actually forced to do) something like switch t:= Errors.Type(err); t { multiple AND exhaustive } If there was a mechanism to force deal with all expected errors returned by a function, we could easily split errors in Expected and Not expected, making error handling more predictable and robust. Non expected errors, will raise a dev error AND a 500 server generic error in case of a web api. That way also we would have full error propagation and not cover errors "as they happen" This is my only "problem" with errors in golang. I love that errors are values, but they should be enumerable and force exhaustive handling, which would be of course bypassed by handling all errors as Trusted. For example, we will do the exhaustive handling in our code Boundaries, were our code interacts with external apis. Once it gets in, values returned by OUR functions should be considered trusted and handled.

This topic is a bit complex, I understand, but there is also a lack of objectivity in this presentation.

Thank you for great video.

Having a separate context and returning an error from your handler is a mistake. Best stick to the std lib handler signature so you can use standard middleware. The separate context also opens you up to split context issues where developers don’t know which to use. We went down this road early in our go migration and refactoring back to the std signature was the best thing we did since.

ok, then don't use the code generation... You are 100% being unproductive and it is 100% time wasted... You can still think about the next section of your code while not having to waste time on generic strings you are writing over and over...