4 ms·
Hundreds of NASA photos. Arranged in a new and unique way. Never before you had to wait so many seconds watching sliding animations to view the next picture. Am
by comboy 8y ago
Hundreds of NASA photos. Arranged in a new and unique way. Never before you had to wait so many seconds watching sliding animations to view the next picture. Am I missing something?
Random link shows just a gray background under firefox then I have to wait 30 seconds to watch the initial animation again (ok, I can skip it down to 10 seconds).
I don't know what did they use machine learning for. What's good about having small bubbles as links instead of a clear list of albums?
I'll drink coffee just in case but I think that's not the problem here.
- whatshisface 8y agoThis is what happens when the designer works on it for too long - imagine flipping the pages back and forth but with only lorem ipsum on them, you would optimize for the experience of sitting there and flipping the pages back and forth instead of the experience of browsing content. One big thing you have to watch out for in design is the difference in experiences between you while activating the transition animation for the 1,000,000th time versus the user seeing it for the first time.
- laumars 8y ago> One big thing you have to watch out for in design is the difference in experiences between you while activating the transition animation for the 1,000,000th time versus the user seeing it for the first time. This. 100% this!! It's sometimes hard to not want to make pretty user interfaces - in fact why shouldn't we have pretty user interfaces?! But the problem I find some designers suffer from is they're so focused on an initial short term wow that they completely forget (or simply don't care?) about any longer term usage. They say first impressions count but so does long term exposure.
- pault 8y agoGood designers work from the content out, not the page in.
- Angostura 8y ago> What's good about having small bubbles as links instead of clear list of albums? The idea is to spatially cluster related categories of image together, I believe.
- mrhappyunhappy 8y agoIronically, they had a caption “scared of heights? Imagine going for a spacewalk” and when you click that you see a bunch of images of people in suits. Why get creative leaving me to believe I will look at actual spacewalks then just show astronauts. Sure they are about to go out but it’s not the same. Also, art sex experiments? Am I the only one seeing that? Nice choice of urls. Over designed, otherwise nice images. Loved the auroras - probably the coolest thing.
- spectramax 8y agoThere is a fundamental lack of understanding what user want. Just like those crappy catalogue viewers with flipping animations - just let me fucking download the pdf file. This is not good design no matter how you put it. Imagine if this page was built using simple HTML without JavaScript. Just a list of albums as you said, that load in less than 100ms and you scroll through using your mouse wheel. The world is full of superfluous display of design in the form of decoration and a middle finger to the user. It’s sad and I don’t know what these designers are taught in their colleges. It is infuriating. Of course there are exceptions in good design but boy it makes me sad that the world is filled with vaneers on shitboxes.
- espeed 8y agoThese are Art Experiments -- it's a playground for design -- it's what artists do. Playing with design and ideas is how you think new thoughts and jump the curve -- discover subtle aspects -- elements that work that you can combine and overtime as it comes together the bigger picture emerges into something new.
- spectramax 8y agoIt’s about the content and if the designer wants to add flair - that takes away from the content by slowing down my browser to a crawl. I can understand if this is a UX/UI art experiment. But it’s not. It’s about NASA images that are collected and grouped together using machine learning. Edit: I’m not against the experiment. I’m against how it’s presented in the browser window.
- espeed 8y agoSo you understand the essence of art, and you're saying experimenting with machine learning as an artistic tool should play no part -- it's out of scope? Have you read the book Hackers & Painters [1]? http://www.paulgraham.com/hp.html http://www.paulgraham.com/hp.html Do any of these other Arts & Culture Experiments speak to you? https://experiments.withgoogle.com/collection/arts-culture https://experiments.withgoogle.com/collection/arts-culture [1] https://www.amazon.com/gp/product/0596006624 https://www.amazon.com/gp/product/0596006624