bywo / react-flexbox-layout Goto Github PK

React 15.3.0 introduced new warnings https://facebook.github.io/react/warnings/unknown-prop.html
Opening the ticket to discuss and solve them.

I'm using react-flexbox-layout on a project on which I'm not using lodash.
NPM is giving me the usual warning about peer dependencies.

I wonder if lodash shouldn't just be a project dependency only and get removed of peer dependencies on package.json

We could avoid explicitly listening to window's resize event and just automatically start listening/stop listening depending on the amount of registered components with the engine.

The idea would be that when there is at least 1 component, we listen to the resize event, and when we have no components (because they deregistered), we stop listening.

We can even check if we're in a DOM to support server side rendering, and all that can be handled transparently inside the engine module, without explicitly calling it, or having a side effect when requiring it.

What do you think? Any drawbacks you see?

Hi!

In our project we've been using this library and have the requirement for server side rendering. We've been thinking about approaches to not depend on document but at the same time make React not bail because of different markup being rendered in the server and in the client.

So with @rafaelchiti we've thought about the strategy to move the vendor differences that may cause different markup to static css classes that handle the vendor differences in the css stylesheet.

What do you think about the approach? Is there any concern you see in advance? Would you be willing to accept a PR with an attempt on that?

With latest react 15.1.0 available, had to force react on my project to stick to 15.0.1 because of the fixed dependency to 15.0.x on react-flexbox layout.

Wondering if there was any reason for react-flexbox-layout not to support the latest.

Forked the repo and gave it a quick try moving versions to ^15.0.1: master...holden-caulfield:master

Tests seems to run well but not sure if that is enough. Happy to create a PR for that change, but thought about asking first if there was a bigger reason not to do this.

Thanks!

Right now is forcing me to use 0.13

In the CSS the flex definitions contain no units when the basis is "0". This breaks in IE 10 & 11. The recommendation is to use % to force a unit that is unlikely to be stripped away by CSS minifiers.

https://github.com/philipwalton/flexbugs#4-flex-shorthand-declarations-with-unitless-flex-basis-values-are-ignored

Lodash 4.x support IE9+ as well.

This was suposed to be taken care of in #21, but webpack-dashboard management of dependencies, which I believe comes from webpack ouput itself states otherwise:

Is there any chance anyone else could confirm this? @byronmwong @gianu @rafaelchiti

Would be nice to have the same warnings we have when setting a style with inline styles, for those things that can not be overwritten.
We need this for class names since we are using css modules.
This way we can send a className to HLayout or any of the other components for changing the things that are safe to change but warn the user for those things that are not safe.
We could use computedLayout in dev for achieving this.

I've found this issue a few times already where having a flex item that has inside a long word without spaces will cause the text to overflow, explanation:
https://groups.google.com/forum/#!topic/mozilla.dev.platform/0ZmJwayNsCY

Issue: https://jsfiddle.net/cLxdjuLn/

Wondering if you @byronmwong know something about this? Im not sure I fully understand this but ended up putting already 3 min-widths in my code for fixing, wondering if that makes sense or if it should be part of the library? maybe you are more familiar already with this issue.

Thanks!

Now: * { box-sizing: box-sizing }
Should be: * { box-sizing: border-box }

Hey!, I saw there is a closed ticket where we were doing flex: 1 1 0 and that was breaking the layout completely in IE11 since the flex basis part was unintless and caused a but in IE10 + 11.
I got a question, based on this ticket https://github.com/philipwalton/flexbugs#4-flex-shorthand-declarations-with-unitless-flex-basis-values-are-ignored the suggestion is to use 0%, wondering why we ended up using 'auto' to be honest auto makes sense but I bumped into an error in my layout in IE11 that gets fixed by doing flex: 1 1 0%, I need to do more research but wanted to get the conversation going to understand the rational when closing the original bug.

Thanks!
Rafa.