Alright, this is a general critique and gripe about MOST themes (and the direction of graphic arts in general).

How come everybody is still hung up with all this "Tiny" business ?

A few years ago Tiny became synonymous with cool in the graphics art and internet crowd ... and we were all inundated with these fancy flash websites where to go to the next page you had to mouse around all over the page and find some itty bitty 4 pixel by 4 pixel green arrow in the top left corner or something.

And we all agreed it was really cool ...
But come on ... havn't we grown up since then ?

This is a gimmick that is getting old.
And WE are getting old ... and our eyesight isn't as good as it used to be after spending 40 hours a week and about 5 years on the internet.

So I would like to see some nice big buttons if you ask me.

And there is no reason why big has to be "uncool".

We need to reach out and expand and come up with something new ... i mean ... tiny can't be cool forever.

anyways,

cherrio.

don't get bent out of shape i'm partly being tongue in cheek.

I will mos def agree that "tiny" can sometimes be considered synonomous with 'useless'. In some situations small can be useful, but not for something that you need to be able to read on a regular basis, or decipher what the graphic is.

I think the problem lies in the fact that most people who make themes, skins, whatever, come from a digital art/computer science backgroud. Yet user interface design should be based around ergonomics, behavioural and cognitive theory. These disciplines should be the primary focus. With that as the primary focus, graphic design is implicit.

Litestep should not be considered a tool for creation of interactive digital art, but a tool to help improve the ease of use of humanity's greatest creation to date.

But then again, if that's what you want to use it for, than more power to ya. Here's to circuit bending.

I should note that when I find a theme I like (with the reasonable size buttons, clear taskbar and systray), I tend to attempt to give the theme the ability to be reduced to a single button in one of the bottom corners. I find this useful when working in an app where screen-space is critical (ie: multitrack audio production... see: Cubase SX, Samplitude, etc.). This button usually allows direct access to a Tasks popup and a normal popup, as well as restoration to full-theme mode.

Tiny is fine, as long as it is in a corner and/or optional.
FI, a roll-up (AKA shade) feature is always handy, or buttons in corners, which really don't need to be big, as they just have to catch the slop from you flicking your wrist.

Otherwise I agree. Too many themes out there use itty bitty fonts and things (anti-aliased and too small to read popup fonts are a long-standing gripe I have), but oh well, that's for the themer to do. At least you can change it :).

Liquid Circuit: it already is a tool for creation of interactive ditigal art, and just like any other art, a lot of is crap, as many themers don't pay attention to detail. Sure, you can't pump out a theme a month, but you don't need to.

As far as tiny vs. large for me...it depends. Text should never be itty bitty, but anything else can be any size necessary...or as close as i can get, given the limitations of xLabel (and when xLabel is your limitation, you know you're asking too much :)).

I'll take "PPI" literally to mean pixels per inch. My running 1280x960 on a 17" monitor from 1999 CE results in around 105 PPI. Too bad the U.S. is still using archaic units of measure, huh.

How many pixels do you have per inch? It has not increased for the typical user, I don't think. Most people don't run a higher resolution to increase their PPI and then making everything on screen larger; they do it to show more at once, at the expense of everything on screen being smaller. Well, that's me at least. I'm aware of it. Sometimes I'll decrease the resolution if it bothers me enough. But once you increase your screen area be it an upgrade in monitor size or adding another monitor, you don't need to increase your PPI. It certainly feels that innovation in the hardware market has slowed down for a while now, to where more people are upgrading their screen real estate. And it seems people are fine with 'subpixel' font anti-aliasing... I certainly like ClearType, with its contrast tweaked to look quite brilliantly invisible on my CRT.

Sure, I'd love to have a densely hi-res screen, a print-quality display. For running high-res versions of moshi's clocks :D

Cerbie: A system shell is intended to be functional, not artistic. Just like fridge doors, toilet seats, and microwaves ovens are intended to be functional, or stereos, cd players, cars, jet engines, etc. Yes, through design you can create a beautiful piece of work, but function should really be the priority. What do you think the original developers of Litstep had in mind? I don't think they're biggest gripe with the Explorer shell was its lack of glitz and emotional impact.

I see a lot of themes inspired by such things as anime, video games, etc, with no improvement in useability when compared to Explorer. Entire themes designed to match a preexisting wallpaper. It all seems very strange and backwards from my perspective; see 'the wagon before the horse'.

And that is fine. I just like any interface I use, from my toilet to my midi sequencer to be as intuitive and poweful as possible. My toilet is quite powerful.

Universal TV remotes are another example of where style tends to have priority over ease of use. But then again, this is a very old discussion, and every one mos def SHOULD have their own opinion.

Oh, and about the small buttons in the corner: anything small is hard to notice. When you first begin using a theme, if a feature is hidden (ie: incredibly small) it is unlikely the enduser will even notice it is there, nor is it likely that they will realize that it does have a function. If it is small, at least make it bright. Humans have very poor spatial acuity in they're peripheral field, so small buttons will go unnoticed more often, being used less. Humans rely primarily on they're vision initially when learning a new task. Through practice they begin to rely less on vision, and (in this example) begin to rely more heavily on motor memory. The more obvious such a button is, the more the user will remember to use that feature. The more they are reminded that feature is there, the more likely they are to practice using it. Once they become acustomed to flicking their wrist to the corner of the screen for these tasks, THEN the size/contrast of the button does not really matter.

Wow, long post... sorry guys! lol Anyway, they point is, before you begin designing your theme, decide who the intended audience is. If you alone are going to be using it, then by all means do what you want. But if you want to make a 'revolutionary', or well respected theme, 4x4 pixel buttons in each of the corners aren't going to get you anywhere. Nor is an *INSERT CONSOLE GAME HERE*-inspired bar theme. It all depends on your audience. The best themes are the ones that a lot of people find useful... in my eyes they are the ones that can cater to everyone. Hooray pop culture.

Functional: yes, the shell is supposed to be functional. So let's go with this analogy. The shell is the electrical and plumbing. The modules are the fridge, stove, microwave, cupboards, etc., and the theme you make is the fully decked out kitchen, complete with a coffee maker and some catfood on the floor. Art does not need to have any glitz or emotional impact.

From good old Webster:
1 : skill acquired by experience, study, or observation
2 a : a branch of learning: (1) : one of the humanities (2) plural : LIBERAL ARTS b archaic : LEARNING, SCHOLARSHIP
3 : an occupation requiring knowledge or skill
4 a : the conscious use of skill and creative imagination especially in the production of aesthetic objects; also : works so produced b (1) : FINE ARTS (2) : one of the fine arts (3) : a graphic art
5 a archaic : a skillful plan b : the quality or state of being artful
6 : decorative or illustrative elements in printed matter

All but #6 can apply to theming LS. In particular, an art is practically anything learned from experience and observation that is used to create something, as opposed to a science, that is used more to understand something. They overlap, of course. The execution of the art makes a work of art.

X-inspired themes: previous UIs and nature are where good inspiration comes. And boredom--always good for epiphanies :). Or as Poe had it, "epochs of most intense tranquility".

Universal remotes can do/be both, those that do just cost an arm, a leg, and a first-born child :).

Small things: everyone should at least skim the readme of a theme, and often with such things (in my case especially), the whole point is to be out of the way, yet easy to get to--and everything is optional. Reading docs, even if not thruroughly, is very important with LS, as you could have a very useful theme with only LSXCommand, jKey, mzScript and a VWM module (I used to have such an animal, for when I broke what I was working on). Even I don't want some of my own stuff in the way sometimes. No static spot can be reached faster than corners--so that's a good spot for misc. crap (except the top-right--or top-left if you're using Mac windowblinds skin). I have it visible so as not to have what I consider an annoying fact of hotspots: they are there but you can't see them at all.

Audience: the audience is myself, and anyone else who has similar thinking (according to my MBTI, that's only 1% or so :)). If you try merely to please others (as a group, not counting contract work and friends), you will fail every time.

Cater to everyone: You cannot cater to everyone's wants. Someone who's used a PC as a GUI-based tool from the start won't find my themes anywhere near as useful as I do, I don't think. I started with a CLI, and still keep at least two handy (LSH (cmd to look/act like BASH, but familiar as cmd) and LSXCommand). Yet look at all the themes out there w/o LSXCommand in them now.

My thoughts: It must take a lot of work to search google and amazon, use a calculator, and start up most apps!
I use a GUI for what it is better at, and nothing else, unless I can't help it (the systray, FI, is a necessary evil). A GUI can do many things well, and I want to maximize the use of those features, but I want nothing gratuitous(sp), and nothing I can do faster and easier with a command line (including getting to it).

tnl: MS is supposed to be working on this for Longhorn, and I hope Linux folks get the notion, as well. That decorations should be based on pixels (window borders and such), but actual work area should be measured in real-world units, independent of the almighty pixel.