Hacker News Reader

Comment by GuiA | Hacker News Reader

As a heavy HN user, I often skim the front page to see what's happening. The point/comment ratio often tells a lot about an article.

First thing I don't like: the Y meter. Most HN readers are nerds, and nerds like exact values (same reason I have the percentage on for my devices's battery)

Second thing I don't like, which breaks the design completely for me: my eyes have to travel a huge distance to see how many comments stories have. It makes skimming the front page much, much worse.

HN's design is about extreme conciseness. There are a few things that could be done better, but if you redesign it that has to be your #1 goal.

View on HN

I agree about the point/comment ratio. The Y-meter is useful, though, at a glance when I'm looking at 25 stories on a big screen. So... just put the points inside the Y-meter!

Why represent the data twice? That's not minimalism. Besides, if you were the OP on an article, you couldn't tell the addition of even say, 5 points without the number itself, making the Y-meter useless. Plus, all data outside 0-400 is completely lost. Something with 401 votes looks exactly the same as 2000 votes.

Maybe minimalism isn't the real goal. The data might be represented twice, but functionally the representations are not redundant. The number provides maximum granularity, but the Y meter facilitates instantaneous comparisons of many stories simultaneously.

I'm not sold on the design overall but there is nothing in principle wrong with representing data in both a imprecise and precise form. A lot of user interfaces do this. Some have three layers. [#######---] 73% 9441/12890

" Plus, all data outside 0-400 is completely lost. Something with 401 votes looks exactly the same as 2000 votes. "

I'm not sure that's a bad thing.

I disagree, the Y meter is the best part. I rarely ever look at the number of points until after I look at the article. I just click on what looks interesting on the front-page. Having an easy up-front visible way to discern new vs established articles would be much more helpful than an exact number buried under too much information.

A point made is that the Y meter obscures an otherwise instantly understandable metric with a visualization that must be deciphered. If you don't care about that number, sure, it's less information and thus perhaps less clutter. For myself, it's chart cruft.

> the Y meter obscures an otherwise instantly understandable

I'm not sure I agree with that, because it takes me a noticeable amount of time and effort to look at the scores for two different stories, located at two different parts of the page, and to compare them. It's far easier to get an instant notion of how a story is performing compared to other stories using the Y meter, for me at least.

My recommendation for the Y meter is dead simple: just also show the number.

Nice design, but I'm not sure I like the Y-meter myself. I don't like when websites/apps change to hide information or remove detail. My eyes already do a pretty good job discerning big numbers from small numbers. Now I have to understand a 70% full square versus an 90% full square?

It bothers me a little that (<)100 is half-way around the meter while (<)400 is all the way around. I don't usually quibble about small details, but if you're going to make a meter, just make it sensible.

Why not have a continuous (or at least less discrete) meter?

perhaps faint separators between stories - a small, light line, or alternating colors, would be helpful ... that way it's much easier to tie the comment-info to the story.

I'm like you: I also insist on the % showing, not just the icon (though I like the icon, too).

I wonder if it would be useful to show the ratio of comments to article points.

Just a little more whitespace would help.

I agree about the faint separators. I'll take under consideration in my follow up post - Stay tuned.

Cheers, Hila

> (same reason I have the percentage on for my devices's battery)

Do you have the wireless signal strength as well?

Duh, no.

I want to know exactly how much percentage I have before my battery dies, i.e. measuring the remainder of a finite resource AND estimating rate of depletion.

I don't care if my signal is -65 vs. -66 dbM, because that's largely out of my control and polled every 5-10 seconds, so even if I did choose to use that to try and find a better signal, the response granularity isn't enough to help me make a decision before my call is dropped.

Reply on news.ycombinator.com