Live data from Hacker News

Why bother with a CSS build process?

lucidchart.com

1–10 of 39 posts

Re: Why bother with a CSS build process?

#2
1. You are exposing your source code to anyone who might be interested.

I think that exposing your CSS source code with comments etc is more beneficial than detrimental. It helps us have a more open web and fosters an environment for new developers to learn from what's already great.

However, the overhead of always sending down the full CSS probably isn't worth it since so few people are actually interested. http://daneden.me/max-css/ suggests including a .max version of your css alongside the .min for this very purpose, which I think is a great idea.

Re: Why bother with a CSS build process?

#4
Is it just me, or is this image solution overkill? I'm curious, how is introducing this complexity to the build process better than just using a sprite? SASS + Compass automatic sprite generation kind of solves the problem for you (not a SASS vs LESS flame, just an example).

I took a peak at your website's source, and it's strange that you micro-optimize the image while at the same time you have 6 separate uncompressed js files and you didn't bother to serve your assets from a separate cookie-less domain. What's the reasoning here?

btw since when is exposing CSS a problem? I don't get what OP means in that point.

Re: Why bother with a CSS build process?

#5

People who are actually concerned about exposing their amazing Javascript and CSS to the world probably don't have very amazing code to begin with.

People who aren't concerned about exposing their amazing Javascript and CSS to the world probably don't have very amazing jobs to begin with.

Re: Why bother with a CSS build process?

#6
post #4

Is it just me, or is this image solution overkill? I'm curious, how is introducing this complexity to the build process better than just using a sprite? SASS + Compass automatic sprite generation kind of solves the problem for you (not a SASS vs LESS flame, just an example). I took a peak at your website's source, and it's strange that you micro-optimize the image while at the same time you have 6 separate uncompress…

We do use sprite sheets for almost all of the icon-sized images (there are hundreds) in the application. And those sprite sheets are big enough to not get inlined in the CSS. The images that do get inlined are generally odd-sized--though I suppose that automatically generated sprite sheets could manage that for us as well.

Exposing CSS is much less of an issue than exposing Javascript; that list was speaking about JS and CSS generally ("why a build process"). The post goes on to describe why CSS specifically is worth paying attention to.

Re: Why bother with a CSS build process?

#7

1. You are exposing your source code to anyone who might be interested. I think that exposing your CSS source code with comments etc is more beneficial than detrimental. It helps us have a more open web and fosters an environment for new developers to learn from what's already great. However, the overhead of always sending down the full CSS probably isn't worth it since so few people are actually interested. http://d…

CSS source maps would also be a good solution.

Re: Why bother with a CSS build process?

#8
post #4

Is it just me, or is this image solution overkill? I'm curious, how is introducing this complexity to the build process better than just using a sprite? SASS + Compass automatic sprite generation kind of solves the problem for you (not a SASS vs LESS flame, just an example). I took a peak at your website's source, and it's strange that you micro-optimize the image while at the same time you have 6 separate uncompress…

Way to edit your post when I was midway through responding :-)

We're relatively unconcerned about our website's (home page) source. Our actual application can be seen at https://www.lucidchart.com/demo. There you'll see a single large Javascript file served from Cloudfront, plus a few third-party-hosted files that we couldn't reasonably CDN (like Google Analytics and some other third-party APIs). There are also three CSS files. One is the bulk of the data (194KB over the wire) which is sent over the CDN. Then there is a separate file for jQuery UI (which will soon be removed) and one that is shared CSS with other parts of our app.

Re: Why bother with a CSS build process?

#9
post #8
post #4

Is it just me, or is this image solution overkill? I'm curious, how is introducing this complexity to the build process better than just using a sprite? SASS + Compass automatic sprite generation kind of solves the problem for you (not a SASS vs LESS flame, just an example). I took a peak at your website's source, and it's strange that you micro-optimize the image while at the same time you have 6 separate uncompress…

Way to edit your post when I was midway through responding :-) We're relatively unconcerned about our website's (home page) source. Our actual application can be seen at https://www.lucidchart.com/demo . There you'll see a single large Javascript file served from Cloudfront, plus a few third-party-hosted files that we couldn't reasonably CDN (like Google Analytics and some other third-party APIs). There are also thre…

> We're relatively unconcerned about our website's (home page) source. Our actual application can be seen at https://www.lucidchart.com/demo. There you'll see a single large Javascript file served from Cloudfront

oh, ok. now it's not so weird :)

Re: Why bother with a CSS build process?

#10

1. You are exposing your source code to anyone who might be interested. I think that exposing your CSS source code with comments etc is more beneficial than detrimental. It helps us have a more open web and fosters an environment for new developers to learn from what's already great. However, the overhead of always sending down the full CSS probably isn't worth it since so few people are actually interested. http://d…

This. If it weren't for sites exposing their sources I wouldn't have been able to teach myself what I know/tinker around when I was just a kid with a dream. This just seems excessive, and one of the employees's responses below about people who show their code not having good jobs seems exactly like the response I'd expect from someone with this mindset. Maybe I'm not understanding this concept, but I'm with Dan on this one; we should be using our code to teach others.
Post reply on HN