I think HN has one of the web’s most successful communities, but my interests skew more towards the front-end and I’ve wanted to create an HN for webdesign for awhile now. At the moment I’m the only one posting links and designs but I’m hoping the site could eventually gain enough traction to truly thrive on its own.
A couple of potentially interesting design decisions:
* Points for both links and comments are hidden until after you vote
* Uses oAuth and OpenID exclusively (even though none of your personal info is displayed and you can still choose your username independently this may be a big turnoff -- but I also think it’s potential for keeping out spam could outweigh its potential drawbacks)
* Was contemplating creating a sense of scarcity by limiting signups to 10/20/whatever per day. Of course at the moment it wouldn’t matter either way, but could be an interesting way to drum up interest
Anyway, it’s just a side-side-project and may not appeal much to this crowd, but feedback is appreciated.
On first open, I immediately thought "whoa, I'm lost."
Then I thought... "way too much effort to parse this information."
Of course its almost surely because I am so used to the HN simplicity and directness. But it is indeed hard to parse for me. Lists are very very friendly. I find myself being able to parse HN very fast, starting from the top left, reading each listed line until it loses my interest, and then proceeding. Very hard to do on your site since much of my effort is ignoring the design aspect and also the fact that I have to scan so wide horizontally does not help.
Much easier for my mind to know that each line is its own concept, and a newline refreshes the concept. With your box model, my mind has to consciously give meaning to the structure and boxes. Same thing for hierarchy. There is an assumed, implied, and very natural order with a list. Not so with your box model.
Perhaps I'm just thinking too much but for usability, your site gets a "too much effort." and I'd much prefer HN.
Of course as you may have guessed, I'm not a design guy.
I can appreciate and love good design, but my take is that design should complement the goal in not only an aesthetic but utilitarian manner.
I feel there is just too much benefit in using the HN list model to go with anything else.
I have to agree. The large bold type in short lines is hard to read. As with HN, you should be able to scan quickly. I like larger type than most people so that's good, but longer line lengths are essential. Another subtle thing about HN and is the color. The black (or almost black) and white is too jarring for quick scanning.
If you want to keep the box design with images check out http://www.notcot.org/. Also, this article: http://www.brandonwalkin.com/blog/2009/08/10/managing-ui-com... specifically the part where he mocks up the Aperture 2.0 filter window with iLife-style controls (in regards to your double border on the boxes).
Despite all of this, though, it is a good idea and a good start. I hope it works out.
You also have some errors in your html ( http://vldtr.com/?key=webdesignerd.com ). Especially the double use of id="header" could lead to some unexpected behaviour at some point. And don't forget to encode & as &
Thanks a lot, apsurd, I agree with a lot of that and if that's the overwhelming feedback I'll reconsider.
I just dislike how images look to the left of the text on Digg, Reddit, etc. so I wanted to try and break out of that a bit. But if it's unreadable it's unreadable.
I'm sorry, it's unreadable. It was in fact the first thing that struck me as I entered the site.
Anyways, I highly recommend some books on cognitive ergonomics - it's a bit odd that we have a bunch of UX guys running around who don't even know what the term means.
It's generally understood that, for western countries anyways, people read content from top to bottom, left to right. When trying to present something in a skimmable, easily parsable way, you must stick with this format.
The somewhat flaky nature of your thumbnailing is also distracting. When presenting a list, the list must be consistent - and right now it is anything but. Behaviour and location of information radically differs between thumbnailed vs. non-thumbnailed entries.
No offense, but the move away from the top-down list seems driven more by an attempt to be different rather than any perceived gain in usability.
No offense was intended - I was more referring to my experience with UX guys who don't know the basics of cognitive ergonomics - I did not mean to imply that you did not yourself.
Nonetheless, the main page is a pretty gigantic violation of basic cognitive ergonomics tenets, and IMHO trades way too much usability for what I perceive as mere prettiness.
Also, I'm very skeptical of the usefulness of the thumbnails altogether - the thumbnail communicates nothing (usually it's a screengrab of a wall of text) and doesn't tell me anything about what I'm about to visit.
> IMHO trades way too much usability for what I perceive as mere prettiness
I don't disagree with you. I'm going to try and find the time to add a toggle b/n list view and grid view so the user can save their preference. Then if it seems that most use the list view, I might just make that the default.
> Also, I'm very skeptical of the usefulness of the thumbnails altogether
Thumbnails of designs are a pretty standard convention with "design gallery"-type sites (http://cssremix.com, http://www.unmatchedstyle.com, etc). It's meant to do nothing more than give you a small idea of the site you're about to visit. The only alternative would be listing the site's title or URL and that certainly wouldn't be any more helpful.
When creating a username I was slightly thrown off by the inconsistency of calling it a 'username' versus a 'display name'. At first (w/o reading helptext) I thought clicking 'Create display name' was going to take me to another page to create a new name. Maybe think about going with one or the other, I would prefer username.
Also, when choosing a username there is no way to cancel the action and return to the main page (ie clicking off the dialog or maybe having a cancel button.)
That said, the signup process was really smooth and painless. Congrats!
Yeah, I was trying to add as many textual clues as possible to the fact that your Gmail username, etc. wouldn't ever be shown, but instead your "public" or "display" name. But, you're right, "display name" is inherently confusing.
> when choosing a username there is no way to cancel the action
That's actually intentional so that users can't post or comment without a username. But I should add helper text to explain that.
The design is confusing. Not because there's too much information (there's not) but because there's no hierarchy. You haven't told me where to look. There's not an element on the page that has any more visual importance than any other. Comparing the site to HN makes me think these are going to be ranked/ordered, but we haven't ranked anything.
Below, jlees suggested splitting out the design and the news sections. I think that'd help. The grid layout you currently have for the whole site would work fine for the design section, and then I'd do news the same way HN does (why re-invent the wheel?). If you look at the HN homepage, it doesn't take more than 1/2 the width of the browser (or in the cases it does, links could be truncated/wrapped). You could have the news/design sections side by side on the homepage. The split alone would help me decide where I should spend my time.
Before when I was the only one posting, I turned ranking/ordering off. Now that there are a few other users I've turned ordering back on and there's now a /newest page where everything's ranked only by freshness.
I think part of the confusion might stem from the fact that points are hidden until after you vote, so I'll have to think on that one.
I didn't mind the layout as much as some people here, but I think that having a list version wouldn't be bad. Would it be possible to allow the user to choose?
As for the idea - great! I already found a few points of interest and that's just from scanning the front page.
My first impression was that it is very difficult to quickly scan which articles I want to look at, but as I look at it, it seems that it's more usable because I can see what's behind the link. Maybe a combination of the two strategies might help?
I like how you follow HN's lead and make read content headlines a lighter gray. But there's no corresponding effect for the screenshot slides. That definitely reduces the usefulness of the read state.
Yup. You can set opacity for elements in all modern browsers. IE6, too.
On some sites, I've used it as a subtle hover state for image links, so they "pop" a little. (Set the default opacity to 92% and have it go to 100% on hover.)
My plate is so full these days that I won't use a new site / service unless I can consume via RSS. Didn't seem to find a feed on the page, which makes it effectively inaccessible for me.
The idea there is to feature more than just articles, so the user can submit designs, too. I'm not sure that'll stay or not, it's just an attempt to appeal to the visual person a bit more.
Re: overwhelming - Yeah, I agree, I've tried to keep everything toned done as much as possible, but there's probably still too much visual noise.
Oh so there is an option to submit a design.. is that for a live site? That's a pretty cool idea.
Hopefully you can attract a healthy designer community there - one of the things that seems like a recurring theme around here is that there are very few designers in this crowd. I can think of maybe 2 or 3.
I would split out design based voting and leave interesting news as text, HN-style. The way it's currently presented just doesn't incite me to really click or interact, and seems weird - reminds me a lot of Evernote, actually. I may be an outlier, though.
I'm gonna go ahead and suggest the opposite: keep the news as is, but reduce the textual noise. That is, remove the tld, "upvote," submitter, etc. and only show them on mouseover. These are all distractions from the important content, the article names. That's why the NY Times skimmer works; all the metadata is absent.
Since your audience are front-end designers and developers, they like to look at pretty things. The Hacker News crowd strives for functionality.
which I personally find pretty readable. But the appeal of that experiment is probably due to the fact that it's less busy than Webdesignerd, so I'll work on that.
A couple of potentially interesting design decisions:
* Points for both links and comments are hidden until after you vote
* Uses oAuth and OpenID exclusively (even though none of your personal info is displayed and you can still choose your username independently this may be a big turnoff -- but I also think it’s potential for keeping out spam could outweigh its potential drawbacks)
* Was contemplating creating a sense of scarcity by limiting signups to 10/20/whatever per day. Of course at the moment it wouldn’t matter either way, but could be an interesting way to drum up interest
Anyway, it’s just a side-side-project and may not appeal much to this crowd, but feedback is appreciated.