New Icon Set...[echo]...

[David Nielsen]

man, 07 08 2006 kl. 14:09 -0400, skrev Diana Fong:
> I believe there might be a few things confusing the matter...hopefully 
> this will help to clear it up a bit.
> 
> [Bluecurve, Tango, Echo...etc]
> For those that like Bluecurve, don't fear it will still be included in 
> our next releases.  However, this should not stop us from trying to 
> creating a new icon set.  Perhaps Echo might not be THE icon set ... but 
> maybe by exploring this together we can come up with something better 
> than it's starting point. By opening its development early on, it is 
> hoped that through the input of various people, the new set could 
> somehow be better. 

It just feels like months ago the answer was get Tango upstream and it
will be our default. Then suddenly BAM.. here's Echo with the words New
and Theme used by people with @redhat.com in their emailaddies.
Understand my anger with this, it does not exactly help with the feeling
that Fedora is indeed a COMMUNITY project. This is exactly the kind of
thing that leads to ideas like Fedora is nothing more than ungrateful
betatesting for RHEL.

> By saying that you just don't like it...is valid but perhaps by creating 
> a few examples with modifications, the rest of us can follow your 
> example to creating something better.  If you say further that it's just 
> not an improvement over the others and efforts should not be spent on 
> this...perhaps you could suggest better alternatives...even if you've 
> got a great idea in a completely different look, it would also be 
> great...and this start would have been all worth it...for echo to serve 
> as a starting point to something more usable.  At this point no icon set 
> is without faults, echo is an icon project intended at creating, what I 
> had hoped to be a better style, but more importantly to really develop 
> an icon theme that is from the collaboration of the community.  If small 
> sized icons are what's dear to you, please do help with the creation of 
> them.  By getting involved, you can influence the direction of things 
> rather than ...perhaps feeling a bit antagonized by what other people 
> create and throw at you.  If you believe it is beyond help 
> then...well...there are other icon sets, not everyone has to use the 
> same one.  However, I have seen some useful suggestions even in the many 
> emails this weekend and ask that perhaps you can think of it as it's own 
> project.  Bluecurve, Tango and others aside, how can we improve echo.

I opened up with the fact that I have no SVG creation skills, I also
have no intention to learn Inkscape. I can make comments on what does
not work and what does, that is the extend to which I can go. Don't
start this put your code where your mouth it. If that was the standard
for all Free Software then we would never have any users. We have to be
open to comments without giving the standard reply "so where is your
fix". It is a perk when users fix your issues in addition to giving you
input not a requirement.

> [smaller icons]
> Some people's overall dislike of the set was a bit confused with the 
> issue of it looking bad in smaller sizes.  My question is, therefore, at 
> it's current size...do you like it?  At this point...I'm looking for 
> feedback on the look and feel of the icons.  The angle, the colors, the 
> shadow.  Previous messages of the shadow being a bit too dark was useful 
> as well as someone's current mention of the "grave cross" "add" icon.  
> With the shadow, this icon set is no longer a simple copy and paste of 
> icons into a square, accommodations have to be made for the shadow, and 
> therefore, with different shapes, icons such as the "pause" might seem 
> larger than the "skip forward" and whatnot it's placed next to.  But 
> these are exactly the kinds of consideration that are useful when 
> brought up.

Do I like it at it's current size, no I think the prespective is a fancy
way of trying to look shiny without really adding much to my experience,
it just looks wrong to me. Rather than being a sutle effect that looks
nice it seems invasive and in your face, a bit like overdoing a good
thing. 

One study you do deserve full credit for though, the facing and
prespective study you did on Bluecurve, that pointed out some
interesting issues. I promptly put on my whiteboard of things to
remember in my review toolbox, I knew Bluecurve was kinda bad in that
respect but the range of issues even on a sigle toolbar is scary. 

I don't like the color scheme much either, I was never a big fan of that
light blue glass look, it honestly reminds me of the month I spend in
the hospital: cold, clinical and industrial (and everything you hear
about hospital food is absolutely true btw. it's not fit for human
consumption).. it's a bit erie feeling to me.
 
The battery icons especially just seem overdone. It's one of those
things that's hard to put your finger on. Then again I am a big fan of
fairly simple icons, when looking at the level of detail I tend to favor
the less is more thinking. I also like repeating images to underline
functionality. For additional detail read the reply I wrote for the
remote folder icon.

> Each smaller sizes will have to be either cleaned up at the .png or the 
> vector stage to make it more readable.  Currently, smaller sizes have 
> not been created.  Help is greatly appreciated in this area.  The 
> package created by Leon was a quick test of the icons and should not be 
> criticized.  Even with this package, we learn that the icons cannot 
> simply be scaled down, work has to be put in to clean up the smaller 
> versions...but maybe we can see which icons did work by simply shrinking 
> and thus lessen the amount of work. At the smaller size I do realize 
> that the current perspective will cause trouble and thus the suggestion 
> for them to be head on and definitely simplified. 

There is just no possible way to do prespective below at least 24x24,
the ratio between right and left in terms of pixels to convey
information with would be laughable. It makes something small even
smaller and thus makes the job of getting information across harder. I
think we might need to think of that kind of thing as a luxury reserved
for very big icons to give them a swanky feel. It looks great on my
desktop (if I had icons there)  not so much most other places. 

> Something I found interesting/useful was the small icon 
> discussion...when does the icon start to break down?  I had originally 
> thought of only the 16x16 as requiring the simplified version and was on 
> the fence for 24x24...but from the feedback, it seems that the icons 
> need to be simplified even before that size.  So, instead of saying that 
> it's just horrible and unusable at smaller sizes, if we can figure out 
> at what point the icons need to be fixed, then that would be 
> constructive.  I know in the Leon's package, there included 22 and 24 in 
> addition to 48...and the 22 was probably horrendous...but there's also 
> 32 and 36 versions that are used as well...how do those look?

I'll have a look at it once I figure out why Evolution keeps crashing on
me. <elmer fudd>be wwery quiet, I'm hunting bugs</elmer fudd>. Right now
that is higher on my list of things to do because well.. without
Evolution giving feedback is kinda hard. It might be helpful if we could
get a few of the icons done like so:

One prespective all the way (pure scaling) and one flat (also pure
scaling) right under it with the same motive so we can see where the
breaking point is. We don't need to consider information overload for
that test, just to see where the breaking point is, I would wager it
sits in at 32x32 or 36x36 depending on the complexity of the icon. 

> I've also posted some of the early icons in the fc5 screen setting 
> [1]...those are actually cleaned up smaller icons.  Is that _still_ 
> bad...or maybe better?  A few pixels cleaned here and there can really 
> make a difference. =)

In the very small sizes you really need a defined outline to be as
viewable as possible it really makes all the difference (and once you
have it on smaller icons it looks silly to not have it on larger ones as
they will look very different). 

> [svg file size]
> Yes, they are large...horribly so.  Nicu's example of slimming it down 
> is awesome and I really hope someone could perhaps run through them.  I 
> wish my original wasn't so large to begin with but at this point I am 
> trying to create as many as possible to provide for better coverage and 
> to also get to the smaller icons in time for release.  Maybe it's 
> wishful thinking...I dunno, but must be hopeful.  Attached is an example 
> from Andy Fitzsimon...when he first saw my svgs he laughed at the 
> ridiculous sizes...but instead of criticizing them...he went ahead and 
> showed what could actually able to get it down to 1.9k!  If you want to 
> check it out...fyi. it came with the warning of...

Strawman... If you can't take criticism and insist on only taking
feedback from people who can code you are building an elitist system.
While I agree that it is a meritocracy there is something to be said for
the fact that out of the many users we have only about 1% can actually
do SVG design to the degree of being useful. That doesn't hopefully make
us useless in your eyes, some of us have no direct skills but have
intimate knowledge of what is needed to hit certain users and we can
provide testing. I think of criticing artwork as filing crasher bugs, I
can't code C to the degree where I could fix Evolution fast enough to
keep up but I can point out when it just doesn't work.

> "Warning: it uses an svg filter for the drop shadow ( gnome's librsvg
> displays this fine but Inkscape will only display it on our next release )"

That sounds odd, technical issues aside the ligmus test is putting a
release out there for testing on the actual desktop not what it looks
like in a designers tool.

> So while he might not be interested in populating yet another icon set, 
> he is interested in optimizing svg files...and that's awesome.  It does 
> make me hopeful to see this and really contributes to the evolution of 
> the icon set.  But this is what I mean...with the feedback and talents 
> of various people the set can evolve into something more. 

It still seems like taking a well thought out design guideline and a
very complete and well tested icon set and throwing it out. I have yet
to see any argument against Tango that didn't boil down to Fong or Duffy
plain not liking it. While that's fine I would like to see some
justification for reinventing the wheel rather than adapting Tango to
our needs. And don't say branding because that is perfectly possible
within Tango, it worked for Ubuntu why not for Fedora?

> There is, of course, always personal preference so while it might be 
> better for some, it might not be for others. 
> We all have different needs and are entitled to our opinions...let's 
> just try to focus and make what we write here useful to this project.

I still think that is the wrong attitude, before discarding an idea you
have debate. I don't want to wake up 3 months from now to another
brilliant idea from behind the RedHat walls because when that happens we
get stuff like Echo that from ground up was not designed with the idea
in mind to cater to users that do not have 20/20 eyesight. You can't
think of every issue on your own that's why we need the community, and
we need to make an RFC on an idea because starting a ton of work and
labelling it as "the new theme". It makes it seem like a decision was
taken away from the community and is being stuffed down my throat.

There is personal opinion and then there are areas where you really need
some serious testing and design. These would be things like usability
and accessability, but also things like cultural bias, the obvious one
here is money not being green many places outside the US, we need to
think about these things. One thing I'm very set on is testing of icons,
like Jimmacs blur scripts, simulating red/green color blindness, etc,
there are things that I don't think about either. No one person is
perfect so it would be good to work hand in hand with the usability team
and do real testing to see if we hit common pitfalls.

I would love to help set up an icon test squad, I can take care of my
own handicap and with a bit of wiggling I can send screenshots or my
laptop around to my former collegues and get some real feedback from
special needs users. But we need more, if anyone knows other special
needs users we should welcome them here so that we can get the best
compromise between a cool look and a usable look. 

> Sorry about the long email, was just trying to address the many points 
> brought up over the weekend.

Long email.. long reply.

- David