search

Devographics / surveys

YAML config files for the Devographics surveys
44 stars 8 forks source link

Editor and tool incompatibilities or issues #18

Open romainmenke opened 2 years ago

romainmenke commented 2 years ago

In my opinion CSS is moving very fast lately (nice 🎉 ) and because of efforts like interop 2021 these new features ship in all browsers at roughly the same time.

I would like to avoid that tools and code editors start to lag behind on critical features. This can give the false impression that a feature isn't ready or safe to use.

A good example is @layer which is unusable in VSCode without a bunch of warnings.

Can we ask users which features are causing the most issues in popular editors and tools?

SachaG commented 2 years ago

That's interesting but it may be a bit too niche to ask about?

romainmenke commented 2 years ago

In which way is this niche?

Just a repeat of the compat issues question but focussed on developer tools would already be very valuable.

SebastianZ commented 2 years ago

I assume @SachaG means "niche" in a way that the answers would be very specific and only target specific tools. So the outcome of that question might not overlap much.

People should rather create bug reports for their tools instead.

The collected data would only be valuable if the editors of those tools actually see the answers and take action on them. For IDE developers this probably isn't the case most of the time.

Though I still see some value in that for tools strictly tied to CSS like pre- and post-processors, CSS frameworks and CSS-in-JS libraries.

Sebastian

romainmenke commented 2 years ago

means "niche" in a way that the answers would be very specific and only target specific tools. So the outcome of that question might not overlap much.

This I understand. The question and options will need to be carefully picked for this to be valuable in any way. Any maybe we do not find the right question or options, that would be fine.

People should rather create bug reports for their tools instead.

Most people don't do this. As a maintainer of a couple of dev tools I am always surprised at how many people tweet about a thing being broken vs. actually open an issue.

From my perspective it seems that there is a barrier for most developers between having a problem with a tool and reporting this problem.

The collected data would only be valuable if the editors of those tools actually see the answers and take action on them. For IDE developers this probably isn't the case most of the time.

I am here, I would see the results and I will do my absolute best to make sure answers are brought to the attention of maintainers in an actionable way.

I co-maintain the polyfill-library and postcss-preset-env and contribute to countless other projects. Because I use VSCode, this includes the bits that relate to CSS. There are others like me.

A lot of the questions related to tooling read like a popularity contest. With few exceptions like : https://2021.stateofcss.com/en-US/technologies/pre-post-processors/#pre_post_processors_happiness

This (in my opinion) is not very interesting as it isn't actionable.

Maybe my original proposal isn't ideal and better, more specific questions can be found that focus on tools and IDE's?

SebastianZ commented 2 years ago

People should rather create bug reports for their tools instead.

Most people don't do this. As a maintainer of a couple of dev tools I am always surprised at how many people tweet about a thing being broken vs. actually open an issue.

Same experience here.

A lot of the questions related to tooling read like a popularity contest. With few exceptions like : https://2021.stateofcss.com/en-US/technologies/pre-post-processors/#pre_post_processors_happiness

This (in my opinion) is not very interesting as it isn't actionable.

Maybe my original proposal isn't ideal and better, more specific questions can be found that focus on tools and IDE's?

Maybe there should be a question about which tools people use to author their CSS, in the first place. And a second question could ask how happy they are in general with their tools. Then there could still be a way to ask for specific pain points or incompatibilities to new CSS features.

Sebastian

romainmenke commented 2 years ago

Maybe there should be a question about which tools people use to author their CSS, in the first place. And a second question could ask how happy they are in general with their tools.

I am unsure how this can be asked without just measuring the popularity of tools and IDE's. I want to avoid pitting tools against each other in a question.

In the previous survey there was only one question (that I could find) that listed different browsers and this was without making this competitive : https://2021.stateofcss.com/en-US/environments/#browsers

romainmenke commented 2 years ago

which features are causing the most issues in popular editors and tools?

"popular editors and tools" This I meant as one group without specifying it as individual items.

Not a matrix of features and tools, but just a list of features and "does it work in your stack". Whatever "stack" might be.