Updating Apps Sucks

In part because god damn UI designers like fucking with design and typically make things worse.

See the Twitter App:

I mean WHY MAKE THE NEW TWEET BUTTON A FLOATING CONTENT OBSCURING MONSTROSITY FROM HELL?! Esp since it seems to have moved from the upper right chrome to make room for the sparkle button of “home”…which takes you to some bullshit selection of tweets.

Honesty, it’s hard to imagine that they’d do worse if they were dying of some illness whose only cure spittle from my sputtering reaction.

This ain’t rocket science and it ain’t art. Just leave things alone for the most part once you have a reasonable design. Floaty things BAD!


Joe Clark on iPhones

Catch up isn’t happening but at least I’m not falling behind.

Joe Clark is almost always worth reading. He’s a bit polemical but that’s ok. He has an inspired discussion on the usability problems with iPhones:

Very advanced, very tuned-in people learn about, and learn how to use, new Apple features by watching them being demonstrated onstage during Apple keynote events.

Then there’s everybody else.

Don Norman told us several times several years ago () that iPhones hide their functions.

(His list of example encounters hits home hard.)

The scary thing is that iOS is fairly useable as these things go. A lot of the problem is the whole “flat design” ethos which kills chrome in favor of content. There are fewer cues and more “you need to know the gesture or to tap”. Invisible features suck.

My example is how horrifically ugly and unusable the music app has become. I can barely use it. I had to provide phone tech to Zoe overseas to turn off single song repeat (it’s really hidden!).

So Apple, shape up! Make features discoverable.

Two Text UI Todo Managers

I’ve been studying Text UI (TUI) frameworks in Python for a while now. In my class, we use the built in argparse module to mange command line argument handling. I fantasise pushing up to REPLs and then widget based full screen console apps. Ideally, the TUI widget framework would have GUI and Web based backends but alas none do. Also, I have a couple of grade and exam management tools with argument handling that I’m looking to add a better front end. A TUI is appealing because it’s lightweight, portable, and easier to security audit.

In this investigation, I stumbled across a couple of TUI task and todo managers. There are, of course, dozens and based on editors, file formats, etc etc. These accidentally juxtaposed by one of them being in an open tab from Hacker News and the other is built on one of the frameworks I’m playing with. The first is a node.js app called Taskbook which has a kind of Trello mentality. The other is a Python/urwid toolkit based app called todotxt-machine and is designed around the todo.txt file format.

(I haven’t fully investigated but it certainly seems that todo.txt is sufficient to capture Taskbook’s data model.)

So, with minor differences in list layout, they have very similar functionality but very different interaction designs. Taskbook is entirely a shell app and all interaction is via arguments from the shell. This is great for integration with scripts and thus larger workflows but can get get when manually manipulating lots of items. It’s primary mode is batch.

Contrariwise, todotxt-machine is a full screen, interactive console app with a bit of menuing, mouse support, and so on. It’s virtues and vices are the reverse of Taskbook.

Implementationwise, though I don’t have actual stats or anything, it’s much easier to add the command line processing than the console UI. Adding the Taskbook argument handling should be a doddle (given feature parity).

In a similar summery, Taskbook’s JSON is more ready out of the box for processing whereas todo.txt is much easier for people to manipulate.

Keypad Research

Just some tab cleanup!

Phone dialers are a fascinating part of our lives. I remember moving from rotary dialers to a keypad and it was amazing. It was so much faster and easier on the hand! But the real game changer was the way it engaged muscle memory. (This is all fading as contact lists make remembering or even dealing numbers increasingly less necessary.) Most of us only experienced rotary or dialpad so it may not be clear just how designed the dialpad is. But Bell Labs experimented with a ton of layouts:

And Sajid Saiyed is revisiting this design space today! Dialing on a touch screen with a hand held, modern smartphone is very thumb oriented, but screen sizes make the pads less thumb friendly esp with the “dial” button pushing everything up a row. You can participate in the study (which looks for issues across demographics groups, which is interesting…I didn’t see across tab on screen size, though) using an iPhone app.

The current leading concept is very simple and obvious…move the seldom uses buttons (* and #) up top and slot the call button beside the 0. This brings every thing down and makes the call button easier to hit. Neat!