NVAccess and the slow Erosion of trust: I still believe that NVDA is the best available screen reader, and I still donate monthly.
-
@prism @cachondo @NVAccess @MostlyBlindGamer Huh? NVAccess repeatedly complained that I was attacking them. As have you. That was my complaint. I never framed anything as an attack. Everyone else put on that framing, and it immediately killed all useful discussion. And that was my complaint. The fact that folks are so defensive about NVDA that you can’t have this discussion without putting words in my mouth means the entire thread is useless. Because this is a battle that you have to win by defeating me. So nothing can or will ever change, and nothing useful will happen, until that framing (that I never once placed or agreed to) is gone.
@fastfinge Again, the reason no discussion is happening is because you keep circling back to "no discussion can happen" after being pointed you to the discussion that happened, and being invited to ingage further on said discussions. You don't like that the magnifier is in core? Okay. I don't like that the magnifier is in core. So leave an issue comment. A response will be forethcoming. If the nature of that response is unsatisfactory, then you write the blog post.
I don't think you are being intentionally disingenuous, I'm sorry I said you made up the stuff in your post. I should not have. But imagine if someone, out of nowhere, wrote up a huge essay about something you did a year and a half ago, didn't ask you about it first, and signed off with "ultimately, the problem is trust." You would feel attacked by that, especially if it was your livelyhood. An open letter format comes off very aggressive to me, as an observer. I'm clearly not the only one who took it that way.
-
@fastfinge Again, the reason no discussion is happening is because you keep circling back to "no discussion can happen" after being pointed you to the discussion that happened, and being invited to ingage further on said discussions. You don't like that the magnifier is in core? Okay. I don't like that the magnifier is in core. So leave an issue comment. A response will be forethcoming. If the nature of that response is unsatisfactory, then you write the blog post.
I don't think you are being intentionally disingenuous, I'm sorry I said you made up the stuff in your post. I should not have. But imagine if someone, out of nowhere, wrote up a huge essay about something you did a year and a half ago, didn't ask you about it first, and signed off with "ultimately, the problem is trust." You would feel attacked by that, especially if it was your livelyhood. An open letter format comes off very aggressive to me, as an observer. I'm clearly not the only one who took it that way.
@prism @cachondo @NVAccess @MostlyBlindGamer And doing so misses the point entirely. Past decisions were made. The work is done, now. As I stated both in the introduction and the conclusion, the thing that concerns me is the pattern of decision making without strategic discussion I couldn’t find. When I point it out, I get pointed to implementation level discussion, IE the “how” and not the “why” or the “should we”. Re-litigating past decisions gets no closer to being more strategic about future decisions. There purpose is, once again, as I said in the article, to demonstrate the pattern I’ve noticed. -
@fastfinge Again, the reason no discussion is happening is because you keep circling back to "no discussion can happen" after being pointed you to the discussion that happened, and being invited to ingage further on said discussions. You don't like that the magnifier is in core? Okay. I don't like that the magnifier is in core. So leave an issue comment. A response will be forethcoming. If the nature of that response is unsatisfactory, then you write the blog post.
I don't think you are being intentionally disingenuous, I'm sorry I said you made up the stuff in your post. I should not have. But imagine if someone, out of nowhere, wrote up a huge essay about something you did a year and a half ago, didn't ask you about it first, and signed off with "ultimately, the problem is trust." You would feel attacked by that, especially if it was your livelyhood. An open letter format comes off very aggressive to me, as an observer. I'm clearly not the only one who took it that way.
@prism @cachondo @NVAccess @MostlyBlindGamer A practical example, because you’re trying to understand. If I propose a feature request to add a full fledged email client into core, I do all the work myself, and I create the pull request, does it get merged? What I’m hearing is that maybe it does, because someone else did all the work. This is my understanding of how both screen mag and on device image descriptions got into core. If it doesn’t, and obviously it shouldn’t, I can’t find the strategic direction around questions like “What belongs in NVDA at all?” That would make stopping this work quick and easy. Without this strong overarching strategic discussion, the discussion I pointed out I wasn’t finding in my article, and that is not just a pull request for a feature or change with no discussion of “why do this” or “should this be done at all”, only “how should we do this” discussion that takes the other two questions as assumed, eventually everything gets into core. And making changes and fixes gets harder and harder. At some point (maybe years from now, but eventually) everything just grinds to a halt. -
@prism @cachondo @NVAccess @MostlyBlindGamer A practical example, because you’re trying to understand. If I propose a feature request to add a full fledged email client into core, I do all the work myself, and I create the pull request, does it get merged? What I’m hearing is that maybe it does, because someone else did all the work. This is my understanding of how both screen mag and on device image descriptions got into core. If it doesn’t, and obviously it shouldn’t, I can’t find the strategic direction around questions like “What belongs in NVDA at all?” That would make stopping this work quick and easy. Without this strong overarching strategic discussion, the discussion I pointed out I wasn’t finding in my article, and that is not just a pull request for a feature or change with no discussion of “why do this” or “should this be done at all”, only “how should we do this” discussion that takes the other two questions as assumed, eventually everything gets into core. And making changes and fixes gets harder and harder. At some point (maybe years from now, but eventually) everything just grinds to a halt.
@fastfinge The "magnifier" (which mostly extends the windows magnifier and then calls it, as I understand) got into core because people asked for it. Image descriptions were taken out of core, because people didn't like them. All of which happened in public.
I do take your point about it being hard to predict the outcome of decisions made about what is or isn't accepted. But there is also the roadmap, product vision, and contributor guide, which seems like a reasonable place to frame any future discussions on what should or should not go into core, which you are of course free to start whenever you feel the need. 1/2
-
@fastfinge The "magnifier" (which mostly extends the windows magnifier and then calls it, as I understand) got into core because people asked for it. Image descriptions were taken out of core, because people didn't like them. All of which happened in public.
I do take your point about it being hard to predict the outcome of decisions made about what is or isn't accepted. But there is also the roadmap, product vision, and contributor guide, which seems like a reasonable place to frame any future discussions on what should or should not go into core, which you are of course free to start whenever you feel the need. 1/2
@prism @cachondo @NVAccess @MostlyBlindGamer I read those, in the course of writing up a feature request for PGP signing of addons. Then I scrapped the feature, because I realized I have no idea what a feature request needs to go ahead, how priority is decided, etc. My fear was that it would happen if someone decided to code it, and not happen if nobody wanted to code it. And that’s just not how to build complicated things. So I decided it’d be better not to waste even more developer time by submitting the feature request at all and risking someone just does it and things get even more bogged down. -
@prism @cachondo @NVAccess @MostlyBlindGamer A practical example, because you’re trying to understand. If I propose a feature request to add a full fledged email client into core, I do all the work myself, and I create the pull request, does it get merged? What I’m hearing is that maybe it does, because someone else did all the work. This is my understanding of how both screen mag and on device image descriptions got into core. If it doesn’t, and obviously it shouldn’t, I can’t find the strategic direction around questions like “What belongs in NVDA at all?” That would make stopping this work quick and easy. Without this strong overarching strategic discussion, the discussion I pointed out I wasn’t finding in my article, and that is not just a pull request for a feature or change with no discussion of “why do this” or “should this be done at all”, only “how should we do this” discussion that takes the other two questions as assumed, eventually everything gets into core. And making changes and fixes gets harder and harder. At some point (maybe years from now, but eventually) everything just grinds to a halt.
@fastfinge How-ever, if I understand correctly, you want them to be more proactive about solicitting community feedback. Frankly there are pros and cons to this approach: it slows down development and leads to fiascos like when they tried adding a sound pack, everyone faffed around for months without settling on something they liked, and ultimately it got closed with no action taken. Maybe you liked that outcome because it's better suited for an add-on. In any case.
As I said earlier, public communications can be a job in itself. Considering the size of the team and the various tasks they have before them, what level of reach-out do you consider appropriate?
-
@fastfinge How-ever, if I understand correctly, you want them to be more proactive about solicitting community feedback. Frankly there are pros and cons to this approach: it slows down development and leads to fiascos like when they tried adding a sound pack, everyone faffed around for months without settling on something they liked, and ultimately it got closed with no action taken. Maybe you liked that outcome because it's better suited for an add-on. In any case.
As I said earlier, public communications can be a job in itself. Considering the size of the team and the various tasks they have before them, what level of reach-out do you consider appropriate?
@prism @cachondo @NVAccess @MostlyBlindGamer Hmm, no. I want them to be more proactive about strategic direction. Getting a feature or pull request, and deciding each time “does this belong in NVDA” is going to lead to problems. The questions like what should addons do? What should be in core? What should addons not do? Why is NVDA deciding to do what it’s doing this year? Should all be clearly understood by everyone. When the strategic direction is set, we can all make decisions about what to work on, what to request, etc, based on that direction. Without it, things just…kind of happen. And that’s not sustainable. -
@prism @cachondo @NVAccess @MostlyBlindGamer Hmm, no. I want them to be more proactive about strategic direction. Getting a feature or pull request, and deciding each time “does this belong in NVDA” is going to lead to problems. The questions like what should addons do? What should be in core? What should addons not do? Why is NVDA deciding to do what it’s doing this year? Should all be clearly understood by everyone. When the strategic direction is set, we can all make decisions about what to work on, what to request, etc, based on that direction. Without it, things just…kind of happen. And that’s not sustainable.@prism @cachondo @NVAccess @MostlyBlindGamer Another example of the lack in overall direction. NVDA seems to have no policy that I can find about AI generated code. But looking at the open pull requests, nobody has submitted any fully AI generated code. Waiting until it happens and then deciding is the kind of reactive stance that leads to bad, poorly thought out decisions. And this is strategic direction that needs to be provided by NV Access, not me, some random dude online.
-
@prism @cachondo @NVAccess @MostlyBlindGamer Hmm, no. I want them to be more proactive about strategic direction. Getting a feature or pull request, and deciding each time “does this belong in NVDA” is going to lead to problems. The questions like what should addons do? What should be in core? What should addons not do? Why is NVDA deciding to do what it’s doing this year? Should all be clearly understood by everyone. When the strategic direction is set, we can all make decisions about what to work on, what to request, etc, based on that direction. Without it, things just…kind of happen. And that’s not sustainable.
@fastfinge NVDA publishes in-process updates every month and a roadmap at least once a year, describing their plans for future work. In fact they just posted the most recent one this month https://www.nvaccess.org/post/nvda-roadmap/
-
@fastfinge NVDA publishes in-process updates every month and a roadmap at least once a year, describing their plans for future work. In fact they just posted the most recent one this month https://www.nvaccess.org/post/nvda-roadmap/
@prism @cachondo @NVAccess @MostlyBlindGamer Yes, I subscribe it and read every issue. And again, that documents how they’re doing the work, and the progress of that work. Not why it was decided to do that work at all. The soundpack thing is actually a good example. More community feedback is fine. But eventually, there needs to be a decider. Who can say “Doing X is most in line with NVDA’s mission, vision and strategy because of Y. So we’ve decided that we’re doing X.” I might disagree with either X or Y. But the reasoning is there. The strategy is there. Without that overall vision and strategy, what happened, happens. Everyone argues for months and nobody does anything. Or someone just decides and does something for reasons only clear to them, and outside of any framework or overall goal for the project.
