Re: How is verbosity decided
Hi all,toggle quoted messageShow quoted text
Surveys were done to find out what people want from NVDA (look at calls by NV Access when the last survey came out a while ago).
As for the issue at hand: I think there is an unspoken, deeper issue that's giving folks headaches: perfectionism. We strive for a perfect software, a perfect productivity tool, a perfect screen reader, or a perfect "cure" for access issues to a point where we demand reality match ideals. Several people gave testimonials as to why that's not really the case, and I provided one pragmatic way of resolving verbosity issues, specifically element announcements.
One thing folks should think about is that machines are directed by humans. Technological progress such as artificial intelligence, augmented reality, and even screen reading may present an idea that machines are perfect and can do whatever they want. But behind bits, Bluetooth radios, and accessibility event handlers is humanity. World history has shown that, despite brilliant plans and inventions, humans are not perfect: sometimes making mistakes, or even leading others to suffer because of the way individuals interpret reality for their benefit. Humans are not perfect, therefore things designed by them carry this marker, and this applies to digital technology as well.
What I'm about to say is my own thought: several weeks ago I heard a comment (either from this list or somewhere) that NVDA community is saturated with perfectionism. I must say I agree with this analysis: we have become so dependent on NVDA to a point where we forget about the humanity of the community. We demand a more perfect screen reader and have forgotten to instill reflection of reality and pragmatism to our posterity. NVDA has become a god and a superhero, a troubling sign of imminent crises. Our devotion to NVDA gave birth to a "cult of productivity" that will come back to haunt the community later.
The biggest crisis the NVDA community will face and must confront in 2020's will be perfectionism and forgetting about stakeholders and their humanity. Marketing literature emphasize NVDA's benefits for users, especially first time computer users. But users are not the only stakeholders: developers, industry watchers, accessibility specs authors and implementers, and even folks outside of this community who are watching us carefully (this includes organizations such as various consumer advocacy groups, Microsoft, Mozilla, Google, and countless others). I heard anecdotes that criticized this community for focusing too much on the ideal, rather than available paths given the reality we are facing. I think this is a sign of perfectionism, specifically expecting NVDA to be the savior we've been waiting for. We must get rid of the mindset that NVDA is the perfect productivity companion. I intentionally used "the" to highlight the seriousness of this problem.
The first step to solving the overwhelming saturation of perfectionism within NVDA community is listing stakeholders and thinking about their humanity and relationships amongst them. What sustains a community is not the product, but level of trust among its members, and the first step in improving trust and relationships is recognizing who these stakeholders are. Users are not the only source of feedback for developers; they also watch market trends, and increasingly, industry specifications such as ARIA (Accessible Rich Internet Applications). But developers do not stop at reading specs and respond to user feedback: they also think about how different people and organizations interpret documentation. What makes developers both happy and angry is differing interpretations and implementations, which, at the philosophical level, shows that human beings are imperfect.
The second step, specifically for fellow developers and add-on authors, is to not put everything in version 1.0 of something. Do not put a fancy user interface in version 1 of an add-on. Release the basics in version 1, and then refine code based on subsequent feedback (this is perhaps my biggest headache I get when reviewing add-ons: attempts to satisfy the community and oneself by putting everything in version 1). In addition to recognizing other stakeholders besides users and understanding their humanity, developers must recognize that they are not gods - we are humans, therefore we are imperfect. The best we can do is listen to and learn from people.
The third step, specifically for users, is to broaden our assumptions and thought process. Users are not the only star in the universe. Just as many stars make the night sky bright, users should consider many things. As it is, one of the causes of saturated perfectionism is the "assumption cage" we locked ourselves into - trying to just apply old assumptions into new reality. It is important to learn lessons from the past and reflect them here. However, it is even more important to get away from the idea that we can just apply what worked in the past to the reality we are facing. I do know that I'm aiming the last bit at certain portion of the community who can't get rid of golden days of old screen readers and wish NVDA to emulate them. As much as old screen readers and their golden eras are great history lessons and we can learn many things from those times, NVDA is a different thing and faces its own issues. Simply put, do not apply your idea of a perfect screen reader to another entity without first checking out what you are facing; or to use terms I coined: do not become desensitized due to information blackout (that is, do not limit yourself to assumptions you already have i.e. be pragmatic at times and think beyond what you know and interact with).
I wrote the above statement both as a critique and to stir some reflection and dialogue. I know that I am an influencer, more so within the NVDA community. Although I do sing praises about NVDA, I believe it is my duty as a community member to offer suggestions and criticisms. I do know that my words do have an effect of swinging topics, therefore I weigh my words carefully. But at times, I believe honest and direct statements are better in terms of helping people confront the issues we are facing and will face in the coming years. And the whole discussion of element verbosity, and the deeper issues surrounding community saturation of perfectionism, is one of those moments where direct confrontation and philosophical pondering could be the best path forward. Please think, think carefully, and think carefully and critically again.
Hope this helps.
From: firstname.lastname@example.org <email@example.com> On Behalf Of Gene
Sent: Thursday, December 17, 2020 7:28 AM
Subject: Re: [nvda] How is verbosity decided
We've heard from a few people saying they like this or that setting as it is. I think it might be of real benefit to take a survey to determine what most people, at least if enough respond, want. It has been traditional since the very early days when all this verbosity became possible for screen-readers to announce it. I haven't heard of one survey of users done to find out what they want. Whether I am right about what people want or not, I think my point is valid. We don't know if the amount of verbossity is what people want or not.
From: Brian Vogel
Sent: Thursday, December 17, 2020 9:17 AM
Subject: Re: [nvda] How is verbosity decided
On Thu, Dec 17, 2020 at 02:57 AM, Gene wrote:
If ;people think I'm wrong, is there some way to survey typical users?- Gene, even if there were a survey, you just will not accept the fact that what you, or I, or any given individual prefers does not mean that many others will prefer it.
You have a good point in terms of a completely new product, or feature, sometimes, but once something has been in wide release like NVDA has, changing what have been defaults for features of longstanding becomes way more trouble than it's worth.
There is absolutely something to your point about figures announcement in that it should be able to be turned on/off at will. As to the rest of your position, not so much. You cannot seem to take on the information that multiple users have presented here that your opinion, and desires, do not match theirs. As a result, what you see fit to have turned on/off by default is incongruent with what they would have turned on/off by default.
In the end, and not just for screen readers, it is the absolute obligation of the end user to seek assistance in customizing things to their liking.
No one at any software development house can ever create something that makes everybody happy, and particularly as far as what setting ship as default out of the box. The tools exist to allow users to create their own best experience and if they're concerned with doing that then they need to explore them, with whatever assistance is necessary, or alone if none whatsoever is available. This isn't a blindness-related thing in any way.
It simply is, and has always been, for any piece of software. The more complex the software the more true it is if you're looking to get as close to exactly what you prefer.
If you can't, or won't, acknowledge the absolute truth of the preceding paragraph then you cannot be reasoned with. No excuses about "lack of training" or similar change it.
Brian - Windows 10 Pro, 64-Bit, Version 20H2, Build 19042
[Regarding the Supreme Court refusing to hear the case brought by Texas to overturn the votes certified by 4 states:] Pleased with the SCOTUS ruling, but also immediately slightly terrified of where this crazy train goes next.
We should know by now there’s a bottomless supply of crazy.
~ Brendan Buck, former adviser to Speakers of the House Paul Ryan and John Boehner