Відео

Muy buna explicación! Qué pena que la calidad del video sea tan baja y no se aprecie muchas veces el código.

Very inspiring

For reference you can afford have a much smaller picture window and still be in the frame instead of overlapping the text.

Can anyone help me find Jakarta 10 documentation for its API's? In the eclipse project page it redirects to 9.1. Can't find where to learn it.

Awesome Presentation

49:08 > _"when i'm doing upfront design, I still going to use whiteboard & paper"_ nice, thanks

2:33 to 3:38 - C4, and maps wow, nice

bad sound, not clear picture

Awesome presentation!

That's a good one. Thank you!

hubiera sido que lo hubieran hecho en espannol, pq no es por criticar pero el ingles esta bien malo!!!, asi que en espanish puro hubiera sido mejor al 100%, yo hablo ingles y en verdad no entendi casi nada!!!

😀 ｐｒｏｍｏｓｍ

To answer a couple of the audience questions at the end. If multiple mutations are enabled at once this creates what the literature calls a "higher order mutant". These can be interesting, in that it is possible that the mutant will be harder to detect than if just a single mutation had been made. Most likely, the mutant will be easier to detect however (killed by any test that would kill any of the individual mutants). Finding the interesting higher order mutants is an even more expensive process than standard mutation testing, as the number of possibilities explodes. Enabling all the mutation in a class at once has exactly the issue the speaker describes, if the higher order mutant is killed, you have no idea which of the mutations would be detected individually. Trying to reduce computation time by selecting a random sample of mutants as one of the audience suggested is a well established idea. Most early mutation testing systems did this as they were too slow to be workable any other way. This isn't very useful however. If the selection is random (and not reproducable) it is hard for a developer to know if they have fixed identified issues. If the selection is reproducable (eg by supplying a seed), the situation is better, but it still gives much less confidence in the code. All modern system I'm aware either do not support random sampling, or disable it by default. Like pitest, they perform a lot of optimisations which keep analysis times managable, and do not blindly run all the tests against each mutant. As another member of the audience pointed out, this is a development technique. Mutation testing produces information which is useful for developers as they're coding. I wrote up how I recomend using pitest here blog.pitest.org/dont-let-your-code-dry/ A key element for scaling it out to larger codebases and larger teams, is the git integration discussed there.

why do you afraid to mutate? :D

ᎮᏒᎧᎷᎧᏕᎷ 🌷

tremendo video!!

Este video, solamente para oir el vocabulario del programador español, vale 1 millon de views!

Really great presentation! Good job, Burr Sutter! Please do more stuff on vert.x!

???????

👍

Que buen contenido, excelente trabajo !!!

Very informative presentation, thanks for publishing!

Thank you!

Billy Korando presentation links - Slides github.com/wkorando/to-java-17-and-beyond Recommended Books - Effective Java by Joshua Bloch - Release It! By Michael NyGard - Continuous Delivery Jez Humble and David Farley

A must! David está mas joven cada día. El CQRS con DDD rejuvenece.

sound problem :(

Thanks for the presentation. I watched the FizzBuzz TDD example and it seems to be misleading and teaching the incorrect approach. The main issue is the test naming. A test needs to describe a behaviour, serve as a specification, tell a requirement. When a test is called like "for 1 return 1" it doesn't describe the requirement that for numbers not divisible by 3 or 5 we return the same number. Same for other example tests you created. It's also better to start with special cases and name a test like "for a number divisible by 3 return Fizz". So the requirement is clear. Then a second test will be like "for a number divisible by 5 return Buzz". Then for other 2 cases. So in total there should be 4 tests. You can give more input values as you did with parameterized final test but in each test separately. This way you can have a clear list of requirements of what the rules are. No need to remove any of the previous tests. Actually, after all tests were combined into 1 it still was confusing because a description like "return expected number" doesn't tell anything, no explanation of what is expected and a reader has to guess the rules. Another thing is the refactoring should be done after each test and not in the end for everything altogether. You also don't need many hardcoded values in production code. There's a principle saying that tests and implementation diverge, as more specific your tests get - more generic your production code becomes. Having 2 hardcoded examples is enough to generalise code.

How to run kotlin script in Intellij?

Where can we find the BDD project demonstrated. Is there any github repo? Thanks!

very enlightning talk, I just want to do programming

Muchas gracias por el material, muy bueno, tengo una consulta, tengo unos microservicios construidos con Java spring boot y que luego los deployo en WebLogic 12, mi consulta es si trabajo los Microservicios con QUarkus podre deployarlos en Weblogic o es solo para Kubernetes? gracias

Thanks. It was a pleasant feeling to see you guys keeping things interesting for us. Love from Pakistan 🇵🇰

Just in case you are searching for github repo here it is. :) github.com/victorrentea/clean-lambdas-jbcnconf

13:43 I love that. Do you love me? 😍💋 💝💖❤️

Aside from the awesome content, can we please talk about how Victor did an absolut excellent job of actually delivering the information? This is some serious skill in teaching, great job!

0:11autoprofitsystem codz2019.blogspot.com/2019/11/blog-post.html

Thanks Nacho!

Share the code please

Very nycc stuff, following from quite a time jst one doubt in last Template Pattern. What we have multiple methods to fill in Email in future ? Were we passing all those methods reference in parameter ?? (Same doubt with passing Runnable in previous example)

Great talk. Clean architecture is a must for serious apps.

Great talk! I enjoyed the content and QA. Good job!!

Sehr guter Vortrag ... Ich programmiere (denke) ähnlich. Aber es ist recht schwierig sowas im Team abzustimmen ... manch einer programmiert halt eher "locker" und dann hast Du immer das Gefühl die Putze zu sein ;)

Ha first

Any advices to a junior? Who's seniors and colleges are writing legacy code and "clean code" is considered as "the code which presents no bugs at the time it is written". I will appreciate any suggestions.

Amazon really sticking it to Oracle

Would be good to record the speaker's own mike instead of the camera mike. ;-)

Man it sux , when so many vulnerabilities exist, we as developer must fight for allowance to write unit test, i know for the fact that some big companys dont test their code WTF!

One thing about 22:50 that I have problem with, and I see it a lot in my production code, is naming function based on single context. In this case function is named hasDeliveryDateBeforeWarning and date which is passed is named warningDate. But if variable would be named f.e. discountDate, it would look ridiculous if reused anywhere, because that implies that only warning date can be passed to it, just in this context. To make it more reusable and not wonder about what function does on the inside (what is warning date in different context?) just name it hasDeliveryDateBefore(LocalDate targetDate) which would make predicate more reusable in arbitrary contexts. Thats all for clean code nazi, great presentation as always :)

Can we use rest assured api for micro services? kindly advice.

Best!!