The case against "It just works" or how to avoid #idiocracy
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Subtitle |
| |
| Alternative Title |
| |
| Title of Series | ||
| Number of Parts | 644 | |
| Author | ||
| License | CC Attribution 2.0 Belgium: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
| Identifiers | 10.5446/41655 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | |
| Genre |
FOSDEM 2018397 / 644
2
3
5
6
7
8
10
12
16
18
27
29
30
31
32
34
39
46
47
48
55
57
58
61
64
67
76
77
80
85
88
92
93
98
101
105
110
114
115
116
118
121
123
128
131
132
133
134
140
141
142
143
146
147
149
162
164
171
173
177
178
179
181
182
184
185
187
188
189
190
191
192
200
201
202
204
205
206
207
211
213
220
222
224
229
230
231
233
237
241
242
243
250
252
261
265
267
270
276
279
280
284
286
287
288
291
296
298
299
301
302
303
304
305
309
310
311
312
313
316
318
319
322
325
326
327
329
332
334
335
336
337
340
344
348
349
350
354
355
356
359
361
362
364
365
368
369
370
372
373
374
376
378
379
380
382
386
388
389
390
393
394
396
400
401
404
405
406
407
409
410
411
415
418
421
422
423
424
426
427
429
435
436
439
441
447
449
450
451
452
453
454
457
459
460
461
462
464
465
470
472
475
477
478
479
482
483
486
489
490
491
492
493
494
496
497
498
499
500
501
503
506
507
508
510
511
512
513
514
515
517
518
519
522
523
524
525
527
528
534
535
536
538
539
540
541
543
544
545
546
547
548
550
551
553
554
555
559
560
561
564
565
568
570
572
573
574
576
578
579
580
586
587
588
590
593
594
596
597
598
601
603
604
606
607
608
610
613
614
615
616
618
619
621
623
624
626
629
632
633
634
635
636
639
641
644
00:00
Gamma functionVacuumConfluence (abstract rewriting)Constructor (object-oriented programming)PlanningTrailInformation securityProcess (computing)ResultantFacebookVolume (thermodynamics)NeuroinformatikNumberTerm (mathematics)Open sourceUsabilityObservational studyInteractive televisionTwitterTask (computing)Operator (mathematics)SoftwareTextsystemMassIntegrated development environmentVirtual machineComputer configurationMessage passingMultiplication signDeclarative programmingAndroid (robot)BitConfiguration spaceProjective planeSet (mathematics)Constructivism (mathematics)Software developerLine (geometry)MereologyMobile appArithmetic meanSpacetimeExpert systemUniverse (mathematics)Element (mathematics)System callInterface (computing)TheoryBackdoor (computing)FamilyPhysical systemTable (information)Position operator1 (number)Point (geometry)Bookmark (World Wide Web)Medical imagingHoaxSound effectComplex (psychology)Ocean currentGoodness of fitMeasurementFlagInternet service providerGraphic designBlack boxUser interfaceDifferent (Kate Ryan album)CuboidNegative numberPresentation of a groupKeyboard shortcutObject-oriented programmingGame theoryOnline helpSoftware testingPRINCE2Moment (mathematics)Video gamePlastikkarteComputer animation
08:43
Sampling (statistics)Graph coloringWeb pageWordUniform resource locatorDifferent (Kate Ryan album)FrequencyDemosceneComputer programmingPasswordProcess (computing)Type theoryDigital photographyOpen sourceGoogolDirect numerical simulationKeyboard shortcutOrder (biology)Rule of inferenceInternetworkingGraphic designIP addressNumbering schemeGreatest elementPoint (geometry)PredictabilityNeuroinformatikRow (database)Source codeSoftware developerFocus (optics)Single-precision floating-point formatContrast (vision)Session Initiation ProtocolReal numberMereologyInteractive televisionUtility softwareAssembly languageCodeProgrammer (hardware)Link (knot theory)Inclusion mapDampingVisualization (computer graphics)SmoothingMemory managementVirtual machineControl flowMultiplication signTerm (mathematics)SoftwareLine (geometry)Goodness of fitEquivalence relationWechselseitige InformationPointer (computer programming)Text editorDigitizingRegular graphExecution unitOnline helpVideoconferencingCommitment schemeSpacetimeWaveFormal languageHost Identity ProtocolInformationGaussian eliminationComputer animation
14:40
DemosceneCodeComputer configurationLevel (video gaming)Game controllerNP-hardAssembly languageEncryptionMultiplication signRule of inferenceWeb pageConfiguration spaceBitComputer engineeringGroup actionNumbering schemeWordElectronic program guideRegular graphArithmetic meanSelf-organizationSoftware development kitPositional notationLine (geometry)EinsteckmodulProcess (computing)Computer animation
20:37
Computer configurationValidity (statistics)CASE <Informatik>Latent heatPoint (geometry)Multiplication signComputer animation
21:38
VideoconferencingTraffic reportingComputer animation
22:00
CollaborationismProgram flowchart
Transcript: English(auto-generated)
00:05
If you're ready to start, please welcome Michael. Thank you. Okay. Design is the creation of
00:23
a plan or convention for the construction of an object, system, or measurable human interaction. An open source is something that people can modify and share because its design is publicly accessible. So, open source design is
00:41
a design whose design is publicly accessible. I believe it's much more than that, and I'm here to convince you about that. I am Michael Dimitriou. I'm involved in both open source and design for a good 10 years now, and I believe that open source design should help users to exercise the freedom that open source provides.
01:07
Current design trends do not promote our freedom, probably the opposite. The things around us are designed to keep us out in the name of ease of use, security, or good looks. But the user that uses something most easily,
01:23
the most secure user, the happiest user is the one that knows how anything works. So, making something automagically work will cause a few moments of initial delight, but the effects in the long-term are exactly the opposite.
01:40
I realized this by showing the incredible advances in deep learning and neural image manipulation to friends and co-workers during the past few years. People that don't understand the complexity of the system involved are the least impressed. Technology is magic for them and can do anything.
02:00
There's a story I read when I was little which described the family dinner during which one of the adults starts floating over the table. Despite what one would think, the children are the one that are the least amazed. They are used to seeing amazing new things all the time. Adults, on the other hand, know that floating over the table is
02:21
impossible and are genuinely surprised. Trying to make technology look like magic, helps keep our users child-like, which is good at making that sound, it's not. You might wonder why I mentioned clickbait and fake news during the notes of the presentation.
02:41
It's because I believe that fake news calls for universal backdoors and conspiracy theories all have one thing in common. They stem from the inability of the crowd to distinguish between what is impossible and what's possible or improbable. The ability to do that improves with a firmer understanding of how things work.
03:01
Unless everything is reduced to a black box, only the elusive expert understands how things work. In fact, nobody does. All the things are just some of black boxes and any given expert can only know what one of those parts does. So, nobody understands how the whole works.
03:24
But let's go back a little bit to the design movements of yesteryear. Let's go back to other flaws in modernism and the declaration on ornament is crime. Or to the Russian constructivists that thought all art was political and that building the USSR was a great art project. To the both of us modernists that thought that
03:41
art should adapt to the new era of the machine. We can also go to today's emotionally durable design that aim to save the environment by designing things that we will love more as time passes. All those guys and many, many more in the history of the art and design thought that they could change the world with their creations. They thought they could make the lives of people better.
04:03
Some by promoting social status, other by diminishing it, some by promoting technology, and others by promoting mass consumption. I believe that today's design trends try to promote an easier life but fail, actually promoting blissful ignorance.
04:22
Do you know what the biggest search term by volume is on Google? Facebook. Why? Because people can't distinguish a search term from URL. And number three is Google. I think this is our fault and we have to fix it. It has huge security implications and it's clearly a
04:42
failure of our design. We are trapped in our bubble and we've lost track of the real results of our work. We think that anyone knows basic things about technology because we do, but the reality is that many don't. And since we managed to get rid of manuals through
05:00
intuitive design, now it's our job to treat them. There are many more examples of how current design fails, but I think the most spectacular manifestation is Facebook's fake news problem. They try to reduce the unlimited complexity of a thing called news to the presence or not of this.
05:24
A small red flag. Of course, they managed to do the opposite of what they intended. It's deeply ingrained in silicon volume mentality. Simplify anything down to a single click. And that brings me to why it fails.
05:40
It fails because our world's complex. Complexity is there and it is inevitable. The universe is complex. Our solutions to the problems caused by complexity usually involve adding even more complexity. In our effort to simplify things as designers, we end up just hiding complexity behind user interfaces that make assumptions. By driving over the complexity with a road-paving roller
06:03
machine of a UI, we destroy all kinds of nuances people love to love. Just visit any kind of community that loves a certain something, be it coffee, tea, audio equipment, cars, whiskey, gaming, keyboards, jeans, desktop environments, whatever, and you will find that the discussion always gets down to the slight nuances that make
06:21
each one's favorite thing unique to the point sometimes of absurdity. So let's go to typesetting. It's an example of the results of UI oversimplification. Latech might be intimidating, but it requires you to invest the time to learn everything about typesetting before you produce a very nice document.
06:42
Word processors promise to make this easy, but in reality, they never allow the masses to produce work comparable to that of professional typesetters. Why is that? Because typesetting is hard work. You have to know all about kerning and hinting and widows and orphans and positive and negative space, and no amount of buttons and nice interface elements is
07:04
going to make that easier. It just provides the false sense of competence, so all the CVs I receive say, excellent use of Microsoft Word in the same paragraph in which they use spaces to write a line of sentence. We make the effort to build a million different options in
07:23
our software, and then we go and hide them behind this. This essentially says, hey, I assume you're too stupid to change these settings. Please don't. This doesn't mean that your app should have a 500 typesetting dialogue, but why hide things about config or Android developer settings so hard?
07:41
And then there are the options that aren't even there. On one hand, we try to promote open source software as something that provides freedoms, and then we turn around and take those freedoms away by hiding them behind compiler flags as we know that our users don't
08:01
know how to recompile our software. If this isn't two-faced, I don't know what is. So I think it's our duty as open source design to have a political agenda, to actually do try and change the world, to be something more than automated operators of AB studies and usability testing, to care not only whether our
08:20
user will do the task quickly and effortlessly, not only if she's going to engage more with our item, but also if she will leave the interaction happier, wiser, and why not a better person? It's not a coincidence that Deepa Napton found that the computer skills of undergraduates dropped off
08:40
between 1996 and 2006. It's the period of Steve Jobs' return to Apple and the same period, so an almost religious purge of anything that resembled the command line or programming for consumer computers. They come without any hint of programmability.
09:02
They are for primary purpose. To be fair, this gave birth to excellent visual computer interactions and helped computers reach millions of people. I have nothing against that, but we took it too far.
09:22
So what do I propose? To force everybody to code in assembler like real men? No, of course not. Technology must be accessible to everyone. It must have an easy on-ramp and not be intimidating. Is this possible? You talked about typesetting a few minutes ago.
09:40
On one hand, you have a latter which is steep and intimidating. On the other hand, you have a Microsoft Word which just promotes ignorance. I'm sorry. I have to take a sip.
10:02
Is there a middle ground? I think stuck edit and markdown is a very good example of a middle ground. It doesn't look hieroglyphics to an untrained eye. It is easy to learn. And a single row of buttons allows you to do most of the things you'd want to do in a document, while at the same time, you see exactly what the computer sees. After the few presses of the B button in a markdown
10:23
editor, if you're slightly interested, you'll understand that it all does is surround the word you are typing with double asterisks. And you'll be able to do it, saving yourself a trip to the toolbar. AutoCAD is another nice example of a closed source. It says point is click as MS Paint, but there's a nice
10:42
little command line on the bottom that repeats every command you do, even if you click on it with a mouse. After a few times, you start learning what the commands are, and you can type them yourself. There's also a nice, predictable shortcut scheme that
11:01
helps you work with shortcuts without having to go to the manual. So let's break it down to some simple rules. One, teach. You know those troubleshooters that never work? Why not, instead of trying to fix the problem, you present your user with the steps he would take in order to understand why his printer or his internet access
11:22
doesn't work? Teach them the process of the elimination, and while you are doing this, explain to them what DNS is and why this is probably the problem when you can reach the IP address and not the google.com. Two, ask politely to show more information.
11:40
I'm pretty sure that there are millions of people that are curious what happens between pressing their keyboard in the URL bar and seeing the webpage they intended to see. So ask them politely. Did you ever wonder what happens behind the scenes when you type here? Not only you get to show all the fancy things you do on
12:02
every keystroke, but also explains why you ask if you want to enable search suggestions and why the answer shouldn't be. Of course, why do you even ask? By the way, I've seen people pasting passwords on the URL bar to see if they are correct. Three, use the jargon. Stop replacing intimidating jargon with softer words.
12:23
It's confusing to those who know. It's different in every brand and makes for an awful, awkward situation when you are trying to give or get help. These are the words that Samsung and LG use instead of color subsampling. Don't be patronizing.
12:42
This is not a good 404 page. I like the heap language, but what this page says is frankly bullshit. There are no palumpas in the computer searching for your mistyped web page. Check the spelling, or go to a wayback machine. And because design isn't all new software, five, make repairs easy. Include a repair manual.
13:01
Link to disassembly videos. Don't use glue. Embrace wear and tear. Stop cherishing smoothness and newness. Scratches and marks show use and utility. Use that in your design. Six, make it replaceable, interchangeable, show its inner workings. RevLab 3D printers are a very good example of what I'm talking about.
13:20
You can see and learn exactly how they work. You can infinitely repair them. And you can upgrade them with parts you can buy off the shelf. And then you can go and build your own. Contrast that to regular printers, which are almost always cheaper to replace than to repair.
13:41
Check sometimes it's cheaper to replace than buy new inks. And because we know nothing about them, they seem to never work. We have a certain focus on the tools nowadays and not the crafts.
14:00
We have fixated on the digital tools that allow us to do crafts. And now people say that they know AutoCAD instead of architectural drafting. They say that they know Photoshop instead of graphic design or digital photo development. And this here is both the craft and the equivalent open source software, making them look inferior. If you know how to dig, the brand of the shovel
14:21
shouldn't matter. So I think that what I described is what open source design should mean. But design is always controversial, as it should be. And so in order to avoid hijacking the term, I would like to call this approach transparencies.
14:42
I invite you all to join me into creating examples of transparent design and put these ideas into even more concrete rules and approaches and influence the future of design. I have created a GitHub organization, which is still
15:01
empty, but I'd like to fill it with examples, rules, and other things that would help future designers follow this approach. Thank you very much.
15:31
And if you have any questions, why, we should.
15:50
We shouldn't do, you make it easier for users.
16:06
And you just say, don't. They have to learn. I'm OK to learn, but sometimes it's just too difficult, because there's a large diversity of users, young, old people, not computer engineers.
16:25
So where do you place the level? I don't think it's about the level. That's why I said we don't need to make everybody code an assembler. Assembly is hard.
16:41
We don't have to make things hard. We have to make them easy, but in the process, show the people what's happening behind the scenes. So when something goes wrong, which it will go, and we all know it, instead of just believing that printers came from hell, we understand that it's the roller thing in
17:02
the printer that it's for, and then we can go to the store and buy another and put it in there. And to be able to do that, it first means that the whole community requires the manufacturer to be like that.
17:20
So there was a 3D printer manufacturer who started putting filament into cartridges, like the regular printer guys do. And they didn't manage to keep that approach not even a year. Because everybody else uses interoperable filament, and you
17:42
can buy filament from everywhere and put it in your printer, so nobody could accept buying the proprietary thing. That's what I'd like to see, a design that makes users understand what's behind, and then they want to learn.
18:06
And because they learn, they are empowered to do things and not be controlled by whatever scheme is devised so that manufacturers, for example, can sell more printers.
18:22
I don't know if I answered the question. Yes, sir? Once you learn the examples, I don't do this. There's a lot of preference in there if they are done wrong, if the user changes them in the wrong way,
18:41
that they could change security. Or if users start relying on these preferences and somebody wants to remove them, it's a lot of effort to kind of make that transition. So for users that just kind of see some guide and say, oh, I'm going to change the preference, I can do this specific thing, and they don't want to
19:00
carry the whole, why is this happening? How would you manage that using this process? So the question in a few words was why I was against the about config warning and what about the users that don't want to learn. Actually, that page was put there because I anticipated this kind of question.
19:20
I wanted to talk a little bit about that. I think that instead of having that page and then a bunch of options that are in JavaScript notation, with dots, there should be no warning, and then every option should have two lines of text.
19:41
I'm sure that this is very short time needed, compared to the time needed to implement an option in the code. That explains what is it and why it should or it should not be clicked and maybe for some especially dangerous options,
20:05
there should be a warning near the specific option. Because in about config, until recently, it was the control tab switch between the latest tabs option, and other options that were very dangerous and allow disabling encryption and everything else.
20:23
So bundling them all in the same basket and then telling the user, okay, go learn and then come back, is much worse than describing what everything does and why he should or he should not click on it. But if you go into describing every line,
20:42
then you're going to say, the user is going to say, oh, I see, I understand this, to support every, that's the point behind this. Okay, that's a valid point, yes. Is there options for a specific use case to rely on?
21:03
Yes, that is a valid use case, and I understand that there might be backlash if you go and remove something that people use. Yeah, that's something that maybe you should be discussed more or maybe you're just right.
21:22
Anything else or do I have more time? I don't know. Okay. Okay, I will do that. Yeah, I'm sure it's a lot.
21:42
I'll do that, but it's just the credits for the except the GitHub repo, all the others are just the credits for the videos and the artwork I used, but I will put them into Petaborf. Thanks.