The horrible aliasing on everything (at least in FF!) makes this a real no go. Everything just looks terrible, and the performance isn't all that good either.
(No IE11 supports at all, I'm hoping it at least runs well in Chrome / Webkit browsers?)
(I hate to be negative, but the aliasing really is quite extreme! Sorry!)
I saw the problem in FF too. I checked it out in Chrome though and it looks much better - the aliasing effect is much less noticeable and performance is better. I'd guess it isn't properly optimized for FF somehow.
> Performance is perfectly smooth for me even with the AA hack above on Windows.
Playing with it more, my performance complaints appear to actually be caused by individual elements losing my mouse when I am clicked down on them and then drag outside of their bounds. This makes it seem like I have to keep re-grabbing things, the end effect is a lot of stuttering as I release and reclick on an element to grab a hold of it ahead.
Edit: Pull request sent (just .coffee as requested).
Looks AMAZING in webkit browsers! Sadly in firefox I get "jaggies" on the diagonal edges due to the lack of antialiasing.
I noticed that adding a transparent outline (outline: 1px solid transparent;) to the class ".oridomi-mask" I smoothed these jaggies. Although they look better they still aren't quite as crisp as webkit.
I assume that is because when you add an outline, the whole thing becomes the texture and the texture filtering is better than no anti aliasing it is still worse than properly rendered polygons.
I've been away from professional web-dev for a while (kids!). My only real experience of IE10 is that I just handled an "IE10 on Win8 only" bug on a friends site so tested with IE10 and IE11 in a virtualised environment. I only have XP installs left of MS Windows.
Of course now that IE is supposed to work they won't let you use the conditional comment targeting method so I must - it seems - break it for IE11 by instructing all IE above 9 to fallback to IE9 compatibility mode.
Nothing else seems to fail with it and applying the hack (IE9 compat mode in a meta-tag) is problematic too.
Has IE changed that much?
That said the site is an application generated mess, it's painful. So maybe IE have really come around.
Modern IE is nice to use and has reasonable support. It will cost you as much (in dev time) to support as any other browser (or specifically, rendering engine), and its market share is comparable to Firefox.
The problem is old IE. fuck old IE and every bit of its ridiculous, antiquated and outdated self.
Author here. OriDomi actually does granular feature tests and has no awareness of what particular browser you're using. While the recent versions of IE do support CSS 3D transforms to a degree, they omit support for the "preserve-3d" transform style which allows the transformations of parent elements to affect their children. Without it, the visual tricks OriDomi does don't work, so the library will disable itself if it finds the client lacks support. Whenever IE begins to support "preserve-3d", OriDomi should start working without modification.
Funny coincidence, I just started a casual discussion about origami and programming with some friends yesterday:
So, my friend Torsten (who is really into papercraft and origami) explained to me the purity hierarchy within the papercraft community.
At the bottom is having to cut the paper and use glue in your creations. One level above that is if the form requires only cutting or tearing. Getting more and more pure is the form known as kirigami, which involves cuts but is made from a single sheet of paper as the base. At the top is origami - no glue, no cuts, only folds (although I'm not sure if there is such a thing as higher order origami), and ideally only one sheet of paper.
The comparison with functional programming ("no glue and cuts" and "no state") is so obvious I'm surprised I've never seen it in the wild - maybe because programming has more in common with algebra, whereas origami is geometric?
Although mostly unrelated to this library, we can conclude that some of the effects in OriDomi should be in a section called KiruDomi. ;)
Actually, from a usability perspective, I see a problem with using those effects, like "twist" and "fracture": once scrambled, it is not obvious from the shape in which way to drag to unfold again, unlike with an effect like "stairs"
Except the papercraft thing is meant to be an extra challenge, whereas functional programming has some practical advantages. Limiting to the bare minimum of just folding, not allowing things that would make it easier like glue or cutting, doing it without practical purpose but just to make impressive creations? I think the direct programming equivalent would be esoteric programming languages like brainfuck.
Limiting to the bare minimum of just folding, not allowing things that would make it easier like glue or cutting, doing it without practical purpose but just to make impressive creations?
My first experience with functional programming was being quizzed to solve certain maths problems with Haskell, and not being allowed to save intermediate results in a variable like an integer was really frustrating due to existing habits. Yes, yes, accumulators and recursion - point is our instructor gave us zero hints about how to think differently, and just kept saying "no, Haskell has no variables."
It might say more about how poorly guided my first encounter with functional programming was, but still, I'd say that is a lot like what you just described.
Author here. Twist and fracture are among the more unconventional, dramatic methods that I left in for those looking for cut-up effects over folding. As you found out, you can even target the body element and the OriDomi will stretch to the window dimensions. That middle portion in the last screenshot looks like a stuck hover state probably due to the fact that you have nested OriDomi instances on the page. Thanks for trying it out.
If you're in to origami, check out the Can you fold it-challenge from Tetra Pak: http://canyoufoldit.com/
You get €5,000 for passing the qualification phase. Are you up to the challenge?
Edit: clarification, 10 teams will be selected from the qualification phase for the design phase and awarded €5,000. That's at least how I understand it :) (I am not involved in the competition in any way.)
You get 5,000 if you complete the design phase, not the qualification phase. For that you need to complete the qualification phase AND be selected to go there.
Lovely work on the annotated source code ... it explains a lot of the tricks and techniques, and makes the otherwise perhaps abstruse math accessible: http://oridomi.com/docs/oridomi.html
Looks awesome, but it seems like all the examples on the page incorporate a weird "wiggling" effect in addition to the folding. Is it possible to turn this off?
Yes. That wiggling is the "ripple" mode I added recently which adds a cascading delay. It's disabled by default when creating new OriDomi instances, but I enabled it for the demos on this page. You can also programmatically turn it on and off:
Author of OriDomi here. Really excited to see this at the top of HN.
I recently pushed out version 1.0 with some new features that make OriDomi easier to use.
It now has an internal queueing system so you can chain up effect sequences instead of dealing with callback handlers. [1] What looks like a chain of synchronous calls will now be deferred and sequenced automatically, e.g.:
Secondly, OriDomi elements are now fully responsive, meaning they adapt to changing dimensions caused by media queries, window resizing, etc. You can see this behavior with the demos by resizing your browser, or on the current iteration of my personal site. [2]
Finally, there are two new methods that make it easy to do custom folding behavior and content manipulation, .map() [3] and .modifyContent() [4]. The former lets you pass your own function to define folding based on input angle and panel index.
It's still relatively lightweight with zero dependencies and 7.6KB gzipped and minified. More than anything, a testament to what you can do with the modern browser API.
Looking forward to seeing what else people do with it. Check out the annotated source [5] for a look at how it works.
52 comments
[ 3.4 ms ] story [ 104 ms ] thread(No IE11 supports at all, I'm hoping it at least runs well in Chrome / Webkit browsers?)
(I hate to be negative, but the aliasing really is quite extreme! Sorry!)
It's a hack but it does give nice anti-aliased edges.
Performance is perfectly smooth for me even with the AA hack above on Windows.
Playing with it more, my performance complaints appear to actually be caused by individual elements losing my mouse when I am clicked down on them and then drag outside of their bounds. This makes it seem like I have to keep re-grabbing things, the end effect is a lot of stuttering as I release and reclick on an element to grab a hold of it ahead.
Purely perceptual, but still nasty.
Looks AMAZING in webkit browsers! Sadly in firefox I get "jaggies" on the diagonal edges due to the lack of antialiasing.
I noticed that adding a transparent outline (outline: 1px solid transparent;) to the class ".oridomi-mask" I smoothed these jaggies. Although they look better they still aren't quite as crisp as webkit.
Of course now that IE is supposed to work they won't let you use the conditional comment targeting method so I must - it seems - break it for IE11 by instructing all IE above 9 to fallback to IE9 compatibility mode.
Nothing else seems to fail with it and applying the hack (IE9 compat mode in a meta-tag) is problematic too.
Has IE changed that much?
That said the site is an application generated mess, it's painful. So maybe IE have really come around.
The problem is old IE. fuck old IE and every bit of its ridiculous, antiquated and outdated self.
http://caniuse.com/#feat=transforms3d
http://stackoverflow.com/questions/19483742/does-ie11-suppor...
So, my friend Torsten (who is really into papercraft and origami) explained to me the purity hierarchy within the papercraft community.
At the bottom is having to cut the paper and use glue in your creations. One level above that is if the form requires only cutting or tearing. Getting more and more pure is the form known as kirigami, which involves cuts but is made from a single sheet of paper as the base. At the top is origami - no glue, no cuts, only folds (although I'm not sure if there is such a thing as higher order origami), and ideally only one sheet of paper.
The comparison with functional programming ("no glue and cuts" and "no state") is so obvious I'm surprised I've never seen it in the wild - maybe because programming has more in common with algebra, whereas origami is geometric?
Although mostly unrelated to this library, we can conclude that some of the effects in OriDomi should be in a section called KiruDomi. ;)
Actually, from a usability perspective, I see a problem with using those effects, like "twist" and "fracture": once scrambled, it is not obvious from the shape in which way to drag to unfold again, unlike with an effect like "stairs"
Actually, it does occur in the wild ;):
http://www.cs.ox.ac.uk/jeremy.gibbons/publications/origami.p...
http://www.google.com/search?q=algebraic%20origami
Thanks!
Er... what? What does "glue and cuts" mean in functional programming?
And functional programming has state, like most other programming paradigms (see the ST and State monads).
My first experience with functional programming was being quizzed to solve certain maths problems with Haskell, and not being allowed to save intermediate results in a variable like an integer was really frustrating due to existing habits. Yes, yes, accumulators and recursion - point is our instructor gave us zero hints about how to think differently, and just kept saying "no, Haskell has no variables."
It might say more about how poorly guided my first encounter with functional programming was, but still, I'd say that is a lot like what you just described.
How would multiple sheets of paper really differ from cuts?
You get €5,000 for passing the qualification phase. Are you up to the challenge?
Edit: clarification, 10 teams will be selected from the qualification phase for the design phase and awarded €5,000. That's at least how I understand it :) (I am not involved in the competition in any way.)
Update: Correction, didn't work the first time I went to the link but the second time was fine.
For those interested in going beyond what the docs cover, the annotated source page is a great place to understand the inner workings.
http://oridomi.com/#ripple
I recently pushed out version 1.0 with some new features that make OriDomi easier to use.
It now has an internal queueing system so you can chain up effect sequences instead of dealing with callback handlers. [1] What looks like a chain of synchronous calls will now be deferred and sequenced automatically, e.g.:
folded.curl(50).collapse().setSpeed(2000).stairs(-29).foldUp().unfold();
Secondly, OriDomi elements are now fully responsive, meaning they adapt to changing dimensions caused by media queries, window resizing, etc. You can see this behavior with the demos by resizing your browser, or on the current iteration of my personal site. [2]
Finally, there are two new methods that make it easy to do custom folding behavior and content manipulation, .map() [3] and .modifyContent() [4]. The former lets you pass your own function to define folding based on input angle and panel index.
It's still relatively lightweight with zero dependencies and 7.6KB gzipped and minified. More than anything, a testament to what you can do with the modern browser API.
Looking forward to seeing what else people do with it. Check out the annotated source [5] for a look at how it works.
[1] http://oridomi.com/#queueing
[2] http://oxism.com
[3] http://oridomi.com/#custom
[4] http://oridomi.com/#content-manipulation
[5] http://oridomi.com/docs/oridomi.html