Hacker News new | past | comments | ask | show | jobs | submit login
Bootstrap 3 RC2 (getbootstrap.com)
110 points by kefs on Aug 13, 2013 | hide | past | web | favorite | 56 comments

Things are moving quickly in Bootstrap-land. I'm not quite sure I would have called these "Release Candidates" though, given the number of functional changes to the framework that have been happening/are lined up to happen. Still, that's likely due to my interpretation of an RC (feature-freeze).

On a different note: I am a big fan of Bootstrap. Yes, there are tons of complaints from users about sites "looking like Bootstrap", but that's hardly a fault of the framework (at least they look passable). I've always found that some button customisation (easier in 3.0), changing the navbar (color, height, dropdowns) and tweaking the forms is the easiest path to getting a site that doesn't look like vanilla framework. And none of those things require much work or expertise: heck, just a change in height, a different nav-bar font + color go a long way as it is.

Is "looking like Bootstrap" really a problem? Why do webapps/sites really need to look different from one another?

As long as they look good, are easy to use and have recognizable branding, what is the problem? Say HN used bootstrap with a few small tweaks for colour so it was still identifiably HN. Also your webmail client, hotel booking site & social network. What harm? Isn't familiar look and feel comforting to users?

Is there a good reason for the aversion to looking similar or is it just a product of the history of the web, with everyone doing their own thing?

Some sites look nicer than your average bootstrap site, but there are so many with ugly designs that don't scale down to mobile devices well at all. If the worst you ever came across on the web was "yet another bootstrap site" that seems it would be a huge step forward.

Yeah we probably should have called them betas, but I'm fine with using RC—beta sounds way less serious and we want folks to just go to town on this. It's helped a lot.

And glad to hear you're a fan! <3

I agree with the parent. The term "Release Candidate" comes with certain expectations, including but not limited to: feature freeze, API freeze, and few(er) serious bugs. Calling Bootstrap 3 in its current state an RC is misleading to developers and may adversely impact the project's reputation.

With that said, I've been using the 2.x releases on multiple projects over the past year. Thanks for all of your hard work!

The rapid rate of change has definitely made getting Jetstrap to support the latest releases harder, but in general it's been a huge pleasure to use and MUCH easier to extend and customize. Great work! :)

I've been using the WIP for a few months and the changes in RC2 are looking really great. I am curious as to what the ramifications will be on the learning curve of people really making the most of the new grid system and how it can be way more dynamic by calling different columns at different snap points. I didn't really "get it" right away, but after figuring it out I can see how it's a huge step forward.

I've been upgrading one of my apps to RC1 the past 3 days. I think it's great! Lot more easier to get it responsive. I was always having troubles with nested columns on the previous version. Things seem to work a lot more better now! Was about to tackle an issue about getting column structure to remain same on phone devices but looks like RC2 takes care of that with .col-xs !

Taking on the model of moving fast, breaking things no doubt.

I am surprised by some of the inclusions such as overhauled grid system to include four tiers instead of the original three of RC1.

To change something so fundamental does seem more like a beta feature than RC1 and no doubt people who've built 2x to 3x converters will be a little more anxious now.

That said I'm sympathetic to the fact that if you're going to change fundamentals or pivot, now's the time. Looking forward to more releases and feature stability.

My complaint is not the name (though the changelog RC1->RC2 is a bit much). I find it strange to only have the RC downloadable from the main page, the way I see it the currently stable release should be the main download and standard documentation. By doing it this way there is a risk of of users within BS3 and also possibly slower uptake in apprehension of other potentially incombatible changes in upcoming BS3 releases.

Actually they only have the latest git snapshot available from the main page. It's not even RC2.

On your note: The great part about having stylesheets that cascade is exactly that. But luckily for the non-customizers, it seems (at least at first glance) that some of the flatness of BS3 might actually mitigate some of the sameness - perhaps because there are fewer particular styles (e.g. those obviously BS2 button gradients) to distinguish it by?

Yes, there are tons of complaints from users about sites "looking like Bootstrap"

These aren't complaints by users. These are complaints by developers and designers who often feel a little...perhaps...threatened by bootstrap and things like it.

Okay, I guess it could be users if you cater to those groups, but for the overwhelming majority of cases users don't know and don't care. They find a clean, modern-looking site that offers the functionality and information that they need, and that's all they really care about.

I don't think any designer feels "threatened" by Bootstrap. Almost all front-end devs use either Bootstrap or something similar (I'm a fan of Foundation personally).

The complaints probably are from devs and designers, but that doesn't mean users like Bootstrap. Devs and designers complain because we know enough to recognize, and complain about, the source of all the monotonous websites coming out these days.

Bootstrap isn't the source of monotonous websites. It just makes them look better.

Yes it is. It's not the source of BAD websites, nor ugly ones. But as far as just plain monotony, it's Bootstrap that allows bad/ugly/designerless websites to look so similar to each other.

Source of bad design? No. Source of bad taste? No. Source of monotony? Yes, just out of convenience.

For native applications (whether desktop or mobile), building interfaces from a consistent set of components is usually considered a good thing.

Why should every web application look different?

If it was up to me, the browser default styles would be something closer to Bootstrap (and actually implement stuff like the <menu> element), so you can build a decent looking site with touching CSS at all.

> For native applications (whether desktop or mobile), building interfaces from a consistent set of components is usually considered a good thing.

That's a remnant of the past and it's changing quickly. Look at any consumer app developed and released in the past year. Most have custom elements, layouts, etc. The web just did it first and is further along in the "designification."

Native apps all looked the same because they were all designed, built, and used by largely engineers. As computing (both usage and development) becomes less esoteric and more powerful, design will naturally become more expressive and purpose-built. This trend extends back to DOS as well and applies well outside of computing (cars, buildings, clothing, hardware, etc).

Foundation 4 v Bootstrap 3.

Your thoughts?

Foundation 4, definitely. SASS is unreal once you get used to it. I feel like Foundation is a lot more robust while also being a lot less intrusive and zealous about its appearance or configuration.

Bootstrap made a big leap with changing the box-sizing though. It was unusable prior to that.

The similarity of so many websites is exactly what I mean is threatening: It downgrades design, to a degree, in favour of cleanliness and utility. Which, I think, is a good thing. We all go to websites to achieve a goal, not to marvel in their unique take on user interfaces.

To the devs who feel threatened, I am referring to people who currently sit on a large bulk of custom, in-house code that has been often rendered redundant, if not obsolete, by projects like Bootstrap. I would contest that all front-end devs use Bootstrap or something similar, as the overwhelming majority of devs work on projects with very long lifecycles, quite unlike the common HN manifestation of short-cycle apps.

It is a lousy complaint too. With sites like Wrapbootstrap.com, it is trivial to throw down $5 to slightly change the default Bootstrap theme, or even use one of the many free bootstrap templates out there.

Am I the only one that thinks v3 looks terrible?

I like v2 because I could just whip out a little internal site/admin console/etc. and it looked as good or better than most other similar things without changing anything. I have some stuff out there with 5-10 custom CSS rules for the whole project.

Now we've got v3 which has all sorts of wack colors and typography, like someone wanted to go "flat" but gave up halfway through.

> Now we've got v3 which has all sorts of wack colors and typography, like someone wanted to go "flat" but gave up halfway through.

The idea of BS3 wasn't so much "flat" as it was a "remove gradients to allow easier customisation". Previously (in BC2) you had to override a ton of stuff to get "sane" buttons again. BS3 just gives you simple buttons that you can add to immediately. Same goes for the nav-bar. I much prefer it this way.

I can see that, I guess it's just going after a different use case than I was using it for (i.e. the decent designer-less site). It has no real value to me if I have to go through and style everything over again.

Just head on over to Bootswatch and pick out your favorite theme, that's your new bootstrap :p

My favorites so far are http://bootswatch.com/cerulean/

Wow, thanks for reminding me of http://bootswatch.com! I just checked it out again after having not looked for a while and I'm impressed with the v3-targeted swatches.

I see this reason a lot. Thing is, Bootstrap was never designed for your use case. It was always a framework to get started, meant for developers to build their designs upon.

Think of Bootstrap as a model car kit. Bootstrap 2 was a Porsche that had just enough paint that people ran with it in their showcase, but that was never Bootstrap's intention. You were meant to complete the kit before showing off. Version 3 is a Tesla, but this time no paint.

> for developers to build their designs

Don't you see a problem right there? In my opinion, it should at least come bundled with an optional style (the old Bootstrap2 style would be fine), so that we could at least use it for fast prototyping. In its current state, it's kind of useless since I'm gonna need a designer from day 1 anyway, because it looks terrible by default.

Here's the corresponding issue on Github: https://github.com/twbs/bootstrap/issues/8706

> it should at least come bundled with an optional style (the old Bootstrap2 style would be fine)

It will. :)

v3 will ship with a v2-ish theme that you can optionally enable.

Is there a way to communicate that to users better? I don't think a lot of devs understand that it's easier to build up on a flat button as opposed to trying to overwrite stylesheets on a gradient button. Maybe something in the features for the final release that helps developers understand that it's not flat for flat design, it's flat because it's easier to build on top of.

True, we could be louder about that. I'll get a tweet out about it shortly. To be honest I totally forgot about even mentioning it anywhere but in a few @ replies and issues on GitHub.

And once we add the theme, it'll become clear :).

Thanks for the feedback! <3

Bootstrap 3 is only in RC? From the way they redirect all inbound links to the BS3 front page rather than the BS2 targets people (like me) thought they were clicking to see, you'd think it went gold.

What's worse, it's not really RC. More like Beta. They keep adding/removing features, and even changing fundamentals like the grid system. I'm glad I didn't start using RC1, I imagined something like this could happen. Now I'm in doubt of using Bootstrap at all.

This is driving me crazy. Every single link/search result on the internet redirects you. It's not even a 301/302. It's a slow meta-refresh. Just long enough to tease you.

can someone explain me why bootstrap continues to be NOT compatible with it previous versions? whats the reasoning behind changing -large into -lg other than cosmetic?

also, I am not a fan of !important. To many times it feels "screw it we dont know how properly cascade styles so just overwrite everything with !important". Its especially problematic when you want to overwrite !important with !important :|

Naming changes were for consistency across the framework.

We only use `!important` on utilities where specificity would make them impractical. Quick floats, toggling by viewport size, etc make it necessary unfortunately. When it comes to every other component though, we don't use it at all.

Major version release enable us to break backward compatibility. Other motivations aside, I see no reason to not do it in attempts to do something better.

> Naming changes were for consistency across the framework.

Can you elaborate? Bootstrap framework? so somewhere else you has mood swing and started calling 'xs' and now you want to update the rest of your framework? OR you talking about something else?

> Other motivations aside, I see no reason to not do it in attempts to do something better.

I have been in IT only last 28 years, but until now its been a normal thing that the new version is compatible with old one. Check newest MS Word you can still open Word 97 documents.

What you doing here is you are discouraging those who trusted your framework from spending hours and hours rewriting their codebase to match your newest release. Most will not achieve that, so internet will be infested with outdated bootstrap websites. One could say "nothing wrong with that, noone force you to trust bootstrap", and they will be right.

On that note, anyone can suggest alternative Framework with more mature approach to new release compatibility?

>Check newest MS Word you can still open Word 97 documents.

It's not really a fair comparison at all. One is a stylesheet, the other is a desktop office application.

>but until now its been a normal thing that the new version is compatible with old one

Not true at all, look at Python 2->3, PHP 4->5, jQuery 1.x->2.x. Major releases often break backwards compatibility.

>from spending hours and hours rewriting their codebase to match your newest release

Design is not something you have to rewrite to 'match the newest release', there are no security vulnerabilities in a style sheet. It is fine to stay in Bootstrap 2.

They are easy to refactor (search and replace .btn-mini to .btn-xs in .css and .html files) and if you're depending on betas to remain consistent you're not being realistic. I've done quite a bit in Bootstrap and I appreciate the small names, doubly so when they are consistent.

I really dislike shortened names in programming and I'm not sure what I'm going to do about Bootstrap using col, btn, or lg. Probably switch to Foundation.

Those extra bytes add up every day. ;)

More that you have to memorise what the word is and what letters they're removing from it.

("Is it Lg, or Lrg? Err, hold on, I'll check...")

This, exactly this.

well, that's kind of funny, because I made this https://gist.github.com/zalew/6227522 for my work with Foundation. yeah, you should use Foundation, but because it's a better product, not because of class names.

Still no official Datepicker I guess, unless I missed it.

Using <input type='date'> should bring up a datepicker in modern browsers.

I wouldn't recommend it: http://caniuse.com/input-datetime

You can always use JavaScript to check for date input support and make a JS fallback if necessary.

and what about the WTF video ? Am I the only one to be surprised ? I find it rather funny =).

In my day a framework was a major piece of generic software used to deploy databases, large scale data flow and service millions of reequests per minute.

These days you write a fucking preprocessed CSS file, a few JS helper functions and it's a framework.

In my day, a framework was a number of interconnected standards, ledgers and transoms. Made from steel and the sweat of man.

These days, throw a bunch of fucking zeroes and ones together and it's a framework.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact