Robert Korulczyk

Results 275 comments of Robert Korulczyk

It works fine for Symfony: ![](https://poser.pugx.org/symfony/symfony/v/unstable), while it is pretty similar situation. So it looks like it does not entirely depends on `version_normalized`. Could this different behavior between my package...

`dev-master` has now `"default-branch":true` in package JSON - I guess it should be used to internally normalize this branch to `9999999-dev` (this is how Packagist handles it - default branch...

Yeah, it refers to changes described in the same[ blog post](https://blog.packagist.com/composer-and-default-git-branches/). Previously `dev-master` was normalized to `9999999-dev` (which make it always the latest unstable branch), now it is not normalized...

Packagist fixed versions ordering on their website, but API response has not been changed: https://github.com/composer/packagist/commit/d313cb83984a0cc2123d91c51d159fd626d51311. I guess that Badge-Poser should use similar solution to handle edge cases like https://packagist.org/packages/friendsofphp/php-cs-fixer correctly.

There is something wrong with this cache. For https://packagist.org/packages/rob006/flarum-lang-polish I have `v0.3.14` on https://poser.pugx.org/rob006/flarum-lang-polish/v/stable, while `v0.3.15` was released 3 months ago. So if it is cache, it is outdated by...

BTW: Examples above for version and downloads count works fine already, but monthly downloads are still outdated: 24 - https://poser.pugx.org/rob006/flarum-lang-polish/d/monthly 59 - https://poser.pugx.org/rob006/flarum-lang-polish/d/monthly?1

This package is defined in this way in Bower registry: https://bower.io/search/?q=rangy If this is incorrect, you should report it there. Asset packagist only maps bower registry to composer packages.

It is here: https://asset-packagist.org/package/detail?fullname=bower-asset/rangy-1.3 , so you should be able to use it.

From styling perspective it is better to add this class on the most outermost wrapper. If we want to add this on `ContainedAlert`, then IMO we should get rid of...

I'm not sure about this. After https://github.com/yiisoft/yii2/pull/20077 `Url::isRelative()` becomes significantly slower (BTW: I think it would better to ensure that we search for `://` only before `?` instead of relying...