Friday, February 21, 2020

rip larry tesler

Daring Fireball is eulogizing Larry Tesler, a giant in the field of UI, who codified terms like copy and paste, helped invent WiFi and Arm chips, and a host of other things. But this twitter thread DF links to makes me wonder:  -- specifically
We were content to literally put a string on the clipboard and insert it at the insertion point. As a 20-year-old college dropout tech writer I was not going to argue with Steve Capps. But Larry wanted it done better.
He wanted the spacing around words to adjust automatically.
I wonder if I consider that a good thing. When I do UX and UI, I often think in terms of "The Dao of Programming" - a program should aways do that which surprises the user least. Auto-adjusting spacing around copy and paste text... that is a flavor of "Guess What I Mean" in interface, vs "Do What I Say"

When it hits the sweet spot, it's great... but when then heuristic isn't perfect, it can be extremely frustrating to have to undo. Many editors today have similar problems, like when you copy a URL into a document that then gets turned into a clickable link, say... but then the editor has to decide if the letters you type right after are part of the link, or not. Usually not. But when you do want to change the link? Or when the editor has to figure out if you're changing the target URL or just the visual representation...there was a reason why many users back in the bad old days of Word Perfect 5.1 loved "reveal codes" so they could see that weird markup and get things precise.

Of course I'm not calling for a return to pre-WYSIWYG days, or arguing everyone should be writing in markdown or html or whatever... just saying that it wasn't just engineer laziness that would vote for Espinosa keeping paste as "just put the string into the insertation point" - there's a decent UX principle where it's better to risk making the user do a little more work than a LOT more work to undo a mis-guess, and to always be clear about what just happened.

(This entry lightly modified from me pontificating on the Lost in Mobile WhatsApp group - come join us if it sounds interesting!)

Thursday, February 20, 2020

css: sub-pixel borders

So in CSS, it looks like borders of less than one pixel width aren't really a thing, but you can achieve something similar with a CSS box shadow:
.thingy {
  0 0 0 0.5px white;

The first 0s are for the horizontal and vertical offset, we don't want a blur, and then we're trying for a 0.5 spread.

The result is something like this:
versus, regular border (with border-radius)
UPDATE: we want back to a simple 1 pixel border... the background image had a zoom effect on hover that would cause the badge to lose its "border" on the bottom, momentarily - very distracting.

Tuesday, February 18, 2020


Cult-of-Mac wondering if neumorphism could bring balance to the force of old-style real-world skeuomorphism and the flat-look that has been big ever since iOS 7:

Roughly speaking, everything looks like a paper cutout version of itself.

It makes me think about Win 3.1/Win95. In their own way they were an attempt at a kind of skeuomorphism, but restricted by the screen resolutions of the era...

Definitely meant to look like a physical button rising out of the screen, and the shadowing would flip when the button was in a pressed state...

Friday, February 14, 2020

note to self: disclosure panels in html5

I was forgetting the term "disclosures" - in html5 they are the summary/details pairs that let you make accordion-like things easily. (Technically accordions only show one thing at a time, but that's the name that sticks in my head.)

It's the tag <details>, but the  content is hidden except for the <summary> tag which acts like a clickable headline to show the rest.

   <summary>An Example</summary>
   Now that's what I'm talking about

An Example Now that's what I'm talking about

Glad that it's in html5 and we can stop building it over and over!

Tuesday, February 11, 2020

css - ellipsis after multiple lines vs after one line

CSS Quickie - So, there's the old trick of adding ellipsis like

text-overflow: ellipsis;     
overflow: hidden;     
white-space: nowrap;

which works pretty well for multiple line
not quite ready for primetime, "line clamping"
is more a less  a thing now:

display: -webkit-box;
-webkit-line-clamp: 3;
-webkit-box-orient: vertical;
overflow: hidden;

Browser support is still meh though.

One note, empirically it seems line-clamping tends to break on whitespace, while the old ellipsis will chop up parts of words.

Heh, I remember doing something like this in jQuery, I made a little open source jquery thing called TrunKit... it had automagic "show the full thing on hover" which was kind of fun...

Tuesday, February 4, 2020

the alternate worlds of a better apple

Like a lot of Apple fans, I pay attention to John Gruber's Daring Fireball blog and the associated podcast ("The Talk Show", which I still think is a corny name.) Recently on an episode with Ben Thompson - they're trying to figure out, on the iPad's 10th anniversary, "why did it stop breaking new ground"...

One thing Ben Thompson points out is the restrictions of the app store - specifically in not having a "reduced price for future upgrades" option - that it stopped companies from getting as serious about making full apps, from pushing the envelope of what was possible on a tablet. This restriction was mitigated a bit when Apple came up with subscription models, but it turns the game from "you own this software, pay us if you like it and want to keep with the current version" of paid upgrades to "keep renting this from us or lose access to your stuff" - which is like how every  damn thing in streaming land works, come to think of it.

And combine that both with "what app should cost more than buck? or $3-5 if extravagant" thinking from the iPhone store, thedevaluing of apps in general... It seems life would have been better if Apple had had a refund policy. Wouldn't it be cool if you could pay like $5-$20 for an app that really makes your life better, and nothing for all these garbage apps? But you weren't stuck paying more if the app couldn't pull its weight (I don't think so many apps are "one time use only", so that a deceitful user would get all the value of the app and then get a refund...)

Another thing Gruber writes about is : what was lost with iOS 7 minimalism - not just skeuomorphism, but a voluntary giving up of tons of other cues about focus and what not - it's funny how in both homes and software the drive for minimalism can cause people to shoot past healthy decluttering to unhelpful paring down

Getting back to the first point (which I think is more significant), the forced purchase models: whenever some is like "oh, these new phones are lame, where's the innovation these days??" I tend tochallenge that person to come up with at least one innovative idea themselves - ideally a brainstorm in the realm of feasibility - otherwise I kind of dismiss them as someone who might not be thinking about how all design is compromise, so often you have to give up something to get anything.... (or in the case of UI minimalism, maybe you give up too much...)

BUT... maybe I'm a hypocrite because I don't know what, exactly, we may have lost by not having a segment of the industry that could fund the development and sustaining of a higher tier of apps and soar above both the current race to the bottom, price-wise, the garbage-ad driven software, and even avoid "pay us forever or lose all your stuff" of subscriptions. It still feels like it's something.