What ergonomic design principles call for minimal contrast and reduced readability in user interfaces?

[debraspicher]

Just my own experience but based on general feedback of staff here it just seems the developer mindset reigns supreme versus a design mindset.  Could be borne out of necessity (platform needs, playing catch-up, etc...) or maybe designer input is being washed out of those talents just based on the amount of developers determining how things should be streamlined first according to their needs. Much of the programs have that feel and no doubt their developers have the talent, but that either testing is insufficient or there is not enough balance for a designers eye to come in and take over when it is most critical. The language around how things get done or why they haven't included a feature almost always end up being along the lines of some sort of technical limitation. It's not like they say we don't have the ability to discern what the user actually needs or can we ask you more questions about what you mean so we can get that to our designers and see why they made these decisions in the UI, just as an example? So I am wondering if they just don't have enough say at key points.

Of course the company could largely be code-oriented folk in culture. That would not help...

[Corgi]

2 hours ago, ChopperNova said:

In this case, accessibility/usability seems a distinction without a difference. The word disability itself has its own various contextual meanings, but word definitions miss the point: V2 is less pleasant to view for reasons given above. It also uses some non-standard contrasting on buttons, as @Granddaddy detailed quite well elsewhere.

...

I would guess that when confronted with the problems that users have been describing, laymen would say that they're having difficulty using the application, not accessing it. At least, in my country that would be the case.

Regardless, I agree completely that the distinction is unimportant. The interface problems exist whether they're described as 'usability' challenges or 'accessibility' challenges.

[deeds]

3 hours ago, ChopperNova said:

Computer science is a science

Whilst I agree with most almost everything you say... not this.

Yes, I can code. In about a dozen languages, including the lowest level stuff that's near no abstraction. Yes, I've designed and developed dozens of different types of applications and their UI and UX. And some hardware, too.

This field is art. Not engineering, not math, not science.

And, just in case you've forgotten, Fauci is The Science™, just ask him, he'll tell you.

There's more science in the art of propaganda and influence (advertising) design than there'll ever be in computing.

Whilst computers themselves require engineering and maths, industrial practices and material sciences to bring them into the world, the actual creation of software is almost entirely art with numbers.

And the amount of love towards the end user of software is THE determinant of the result. As seen here, for better ... or worse. 

 

 

[sfriedberg]

3 hours ago, deeds said:

This field is art. Not engineering, not math, not science.

As a PhD in Computer Science (1988 University of Rochester), whose day job was in OS internals for decades (file systems and network protocol stacks) and is now in complex data visualization including GUI work, I largely disagree with you.  There is a large element of craft.  But if you are not using mathematical principles and engineering methodology, you are doing it wrong.  Full stop.

I could entertain the argument that CS is not science, but only by the same arguments that one would argue that mathematics is not science, and that debate is not going to be settled in our lifetime.  (E.g., is mathematics a process of discovery or of invention?)

[Added in edit]  My grad school divided CS into Systems (i.e., operating systems, networks, databases), Theory of Computation, and Artificial Intelligence.  None of those divisions lack engineering, mathematics, or science.  (Well, maybe ToC didn't have much engineering at UR.) And UR had/has a strong cross-disciplinary Cognitive Science group, which absolutely involves science, in the traditional testable laboratory sense, as well as mathematics, philosophy, linguistics, and psychology.

You can work on a car without knowing a lot about engineering, and without knowing anything about logistics or factory production.  You can even customize your car by installing a fancy new sound system or shiny wheels.  But no amount of "art" will let you design and build a car from scratch, much less produce a model line of reliable cars.  Similarly, no amount of "art" will let you design and build a reliable, efficient, software system on time and on budget if you lack (or refuse to apply) the appropriate mathematical and engineering tools.

[deeds]

36 minutes ago, sfriedberg said:

As a PhD in Computer Science (1988 University of Rochester)

Conflict of interest, much

 

Computer "Science" is the dual acts of conceptualisation and subsequent creation of software, via computers. A computer is not designed and built to sit merely being a computer, it's built to run software, for people.

From computers themselves all the way through to making the most impossibly absorbing video game... all of which, from top to bottom, all of it is art. Calling it a science is a horrible misnomer because it's permitted excusing the focus on diabolical analysis of humans with these magical boxes rather than empowering them to create.

As Eugene Jarvis said: the only legitimate use of a computer is to play games

and the more we see of what they can do in the hands of the technocrats, the more true this statement becomes, every single day.

 

[sfriedberg]

14 hours ago, deeds said:

Conflict of interest, much

Well informed opinion, rather.

14 hours ago, deeds said:

Computer "Science" is the dual acts of conceptualisation and subsequent creation of software, via computers.

It's much more than that.  In fact, I would classify most software creation (which is what I have been doing for a living for 40 years, starting in the early 1980's and full-time in industry since 1992) as software engineering, guided by CS but by no means encompassing CS.  A great deal of software creation is also guided by non-CS disciplines, such as art/design, human factors engineering, and various business or artistic disciplines (the "why we are writing a program in the first place" motivation).

I certainly will not argue that purposeful software development needs more than CS.  But just as there is a difference between a do-it-yourself craft project and a professional artwork or construction, there is a difference between DIY hacking around and professionally developed software.  And a large part of that difference comes from knowing the professional techniques.  And here I am not talking about this year's trendy libraries or languages, but core CS material.

One thing I find non-engineers do not understand very well, if at all, is that engineering (of any specialty, but specifically software engineering in this discussion) is a highly creative process.  If technical technique is "craft", then the creative aspect is "art".  (Of course, one can say the same thing about mathematics.)

14 hours ago, deeds said:

the only legitimate use of a computer is to play games

Hahahahah.  A fine piece of rhetoric, but a literal piece of nonsense.  Do you realize how many 100's of millions of people would be without electricity, running water or fuel if all the non-gaming computers in the world just stopped working?  How many companies would collapse without a means to manage logistics, schedule activities, even inventory the contents of their warehouses?  How many people would thereby be thrown out of work and impoverished?  How many people would starve or freeze to death?  There are many, many more computers in the world than the familiar midtower in the spare bedroom or the laptop or tablet in your bag, and many of them perform essential services, even life-critical services.

18 hours ago, deeds said:

the actual creation of software is almost entirely art with numbers.

Oh dear God, no. Numbers have almost nothing to do with software at all!  That wasn't even true back in heyday of FORTRAN which is a relentlessly numerical language to write software in.

The crucial art in software development is "art with structure" or "art with algorithms".  The actual creation of most software is 75% boilerplate and package integration, with a lot of struggle around integration.  And in "art with structure", you can read that as either "data structure" or "program structure" and be correct either way.

-------------------------

Okay, this discussion has gotten way off course, and I have "helped".  Let me help direct it back to the original topic.  Yes, fewer lessons from human factors engineering regarding legibility and ease-of-access seem to have been applied to the V2 UI than the V1.  And apparently this was deliberate.

[François R]

On 11/20/2022 at 9:05 PM, Granddaddy said:

Affinity 2 user interfaces seem less readable than in Affinity 1. Difficulties with discerning what is on the screen have been described in numerous posts in several threads in these forums.

I'm willing to attribute some of my difficulties to my aging eyes, though dozens of other applications do not cause me problems. I have also noticed in recent years that it is fashionable to make printed publications less readable by using dark gray or dark blue text on black backgrounds and even worse. I also find the controls on the dashboard of my 2014 Honda Accord to be far less usable than the controls on my 1998 Accord. The 1998 controls could be set almost entirely by touch. The 2014 controls, which consist of rows of identical buttons, require actually looking at the dash to first find the needed control and then to make the desired adjustment.

This leads me to ask three questions in regard to user interface design:

1.) What principles of ergonomic design suggest that reducing visibility/contrast/distinctiveness produces a better user experience?

2.) What advantages accrue to the user through the use of very low contrast and very low distinctness in the user interface? 

3.) Do reduced contrast and easily-confused icons reduce the possibilities for user error?

Perhaps those in the graphical design business could educate me about this.

👆📖

🙂

[deeds]

5 hours ago, sfriedberg said:

How many companies would collapse without a means to manage logistics, schedule activities, even inventory the contents of their warehouses?

How did they do it before computers?

 

[fde101]

On 11/28/2022 at 3:25 AM, deeds said:

Computer "Science" is the dual acts of conceptualisation and subsequent creation of software, via computers. A computer is not designed and built to sit merely being a computer, it's built to run software, for people.

Computer science deals with programming and information technologies and is definitely a mathematics-based practice, and firmly on the science/technology side of things. Computer science, in the strictest sense, is not an art form.

 

Software development involves many practices, however.

 

First you need a reason to develop a particular piece of software - this may come from an actual need (which is more on the "science" side of things - I have some goal I am trying to reach, so how do I do it?), or simply from a desire to create (which is more on the "art" side of things).

Then the software needs to be designed.  For a modern graphical application, this virtually always involves BOTH art and science practices: there is the design of the user interface and the determination of how the user interacts with the software, which falls more on the "art" side of things (but that art is backed by science), and there is the engineering of how the software will present that interface and how the interface will be made to perform the task(s) it is intended to accomplish (which falls firmly on the "science" side of things).

The UI design and the like, which are more "art", are technically not in the domain of Computer Science - it is a somewhat distinct practice which is part of software development, but is not a true computer science task.  The engineering of how the UI is presented, and of how the underlying work is performed - that is computer science.

There is also the creation of documentation for the produced software, and writing is generally considered an art form.

 

Granted that there is some degree of artistic involvement in some forms of engineering - engineers do need to be creative at times, and sometimes there is an "art" behind the science, just as art itself has science behind it (why does mixing certain colors of paint together produce some other color?  That is science, enabling the art; how did anyone ever come up with the idea of genetic algorithms?  That was a creative/artistic method of developing concepts which are applied scientifically, as an engineering practice...)  so there are not firm lines here, but it is definitely not accurate to say that computer science is an art.

[sfriedberg]

7 hours ago, deeds said:

How did they do it before computers?

Seriously?  Go watch an old movie with office or factory scenes, set in the 1940s or 1950s, and ideally filmed before the 1970s.  Look at the acres of office workers sitting in grids of desks, processing stacks of paperwork.  Then realize that the largest such office that could be operated cannot handle more than a tiny fraction of the work that a commercial software suite can handle.  And that "human attention" was absolutely the last thing that those masses of humans were intended to supply.

You ever work in a good-sized library?  Ask a librarian about mis-shelving books on the stacks.  Or a warehouse worker about mis-placing goods on the racks.  Ask about quarterly or annual inventory checks, and how long those took before computers stored and indexed the data.  Ask about the impact of barcodes and RFID tags.

I could go on and on.  The short answer is "with massive mind-numbing and soul-crushing human labor, endless seas of paperwork, frequent losses and inaccuracies, and yet at a vastly smaller scale, and with a much slower response time".

[deeds]

11 minutes ago, sfriedberg said:

Seriously?  Go watch an old movie with office or factory scenes, set in the 1940s or 1950s, and ideally filmed before the 1970s.  Look at the acres of office workers sitting in grids of desks, processing stacks of paperwork. 

Full employment, guaranteed. 

"The only legitimate use of a computer is to play games."

[Granddaddy]

New threads keep springing up concerning reduced readability/usability of the user interface in Affinity 2 when compared to Affinity 1.

On Monday we were alerted to the invisible arrowheads in the font selection drop down menu.
https://forum.affinity.serif.com/index.php?/topic/174868-serif-fire-your-affinity-20-interface-designers/

On Tuesday the unreadability of layer names was addressed
https://forum.affinity.serif.com/index.php?/topic/174876-layer-names-barely-readable-ui-prefs-dont-help-can-barely-use-v2/

On Wednesday morning we are alerted to the fact that "large" no longer means large when it comes to layer thumbnails. "Large" thumbnails in Affinity 2 appear similar in size to the medium thumbnails in Affinity 1, making them more difficult to decipher. How does this reduced thumbnail size improve the usability of Affinity 2? 
https://forum.affinity.serif.com/index.php?/topic/175008-large-layer-thumbnails/

[GuyMiklos]

User Testing is the best way to discover the greatest number and most serious usability problems, although there are other less expensive methods that pick up many usability issues.

In the 1990s, Jakob Nielsen conducted some excellent research into the efficacy of various methods in discovering usability problems, comparing the various methods with one another. User Testing is the golden standard against which all other methods are compared. In user testing, if a user is unable to complete a representative task that the software is expected to support within a reasonable time limit, then the interface design is awarded "Task Failure", and the designers have to go back to providing affordances in the workflow that will carry the user through to the conclusion of their task within reasonable time.

I think the Affinity interfaces are excellent in many ways. One of those ways is that they is far more intuitive than the convoluted structures of Adobe products. But providing an interface with greater usability than Adobe's labyrinth of widgets is fairly easy. 

Sorry. I didn't mean to moan about Adobe. But it took me about a month of using Affinity to achieve greater fluidity in design work than after ten tortuous years with Adobe Illustrator and Photoshop. So well done Serif! 

I'm curious - does the Affinity team apply user testing? And if so, how do you select test subjects and the tasks they are given?

[GuyMiklos]

Add to the list of poor visibility, the disclosure triangles in the Text Styles panel, which are small

light grey arrowheads on a slightly lighter grey background . This slick, new, ever-so-fashionable look is very hard to discern without a good long squint. This adds between about 1 to 3 seconds to my search time, and interrupts my focus and work flow, every time I use the panel.

Better would be a larger icon

Better would be strong contrast

Best would be a larger icon AND strong contrast

but no, the person responsible approved minimal icons with minimal contrast. Has this person any background at all in GUI usability? 

I would guess (perhaps naively) that this would only take an hour or so to fix the contrast at least, so

please, please, please fix this soon.

[deeds]

2 hours ago, GuyMiklos said:

User Testing is the best way to discover the greatest number and most serious usability problems, although there are other less expensive methods that pick up many usability issues.

In the 1990s, Jakob Nielsen conducted some excellent research into the efficacy of various methods in discovering usability problems, comparing the various methods with one another. User Testing is the golden standard against which all other methods are compared. In user testing, if a user is unable to complete a representative task that the software is expected to support within a reasonable time limit, then the interface design is awarded "Task Failure", and the designers have to go back to providing affordances in the workflow that will carry the user through to the conclusion of their task within reasonable time.

I think the Affinity interfaces are excellent in many ways. One of those ways is that they is far more intuitive than the convoluted structures of Adobe products. But providing an interface with greater usability than Adobe's labyrinth of widgets is fairly easy. 

Sorry. I didn't mean to moan about Adobe. But it took me about a month of using Affinity to achieve greater fluidity in design work than after ten tortuous years with Adobe Illustrator and Photoshop. So well done Serif! 

I'm curious - does the Affinity team apply user testing? And if so, how do you select test subjects and the tasks they are given?

There's a vastly better and faster way, hire a principled, considerate designer, that also happens to love end users and digital design, and empower them to make wholesale decisions on UI ---- AND ---- UX

User research is only necessary if you don't do this, and only works if you're willing to hire an expert in both asking questions AND understanding the responses AND is empowered to make AND correct wholesale decisions that should have been made correctly at the UI and UX design phase.

None of this is going to happen. Just take a look at the rate of UI and UX fixes and improvements in version 1, then slow that down by a considerable margin because they're now concurrently working on 3x apps on 3x platforms. 

[Old Bruce]

This looks like a good place to put some of the screenshots I have been collecting.

Is there a disclosure arrow here?

The contrast is far too low for these two as well. Very hard to find.

Then there is this bloody awful result of the latest craze in designing interfaces

Which is set?

[GuyMiklos]

On 11/30/2022 at 9:30 PM, GuyMiklos said:

fix this soon.

Can everybody clearly see the disclosure triangle next to Body and next to Sub-heading1 ?

[François R]