@73, Nidsa07 already pointed out why people give simplistic advice before going and immediately thinking that its a problem with the program. I shall not reiterate it in this post.
Second, JAWS does not just rely on UIA and iAccessible2 for accessibility as NVDA does. JAWS relies on three components: the video intercept driver, UIA, and MSAA. (It may also rely on IAccessible2 and other APIs that I don't know of.) Either way, JAWS is capable of directly intercepting what is being drawn to the screen and where its being drawn. Scripts (a.k.a. hacks) can therefore be implemented to detect what's being drawn and, using that information, can deliver that information to the user. NVDA does not do this, for obvious reasons. I have no doubt that JAWS's codebase has surpassed the "unmaintainable" state and the only reason that its still maintained is because the developers know the code so well.
Now, before you come to the (understandable, but false) conclusion that JAWS is superior in this matter, it is not. In fact, the methods that I believe JAWS uses is in fact inferior to what NVDA does. JAWS relies on what the graphics card is drawing to get textual information. This means it has to do a lot more internal computation in order to figure out what exactly is important and what isn't. Add to that the UIA layer and you've got yourself a pretty big mess. Add MSAA and (optionally) IAccessible2 and you've got yourself... something indescribable. NVDA takes this concept of bloated accessibility and simplifies it. It relies on UIA, MSAA and IAccessible2 only when each technology is needed. If I'm not mistaken, NVDA prioritizes accessibility information: if UIA's information is better than IAccessible2, it runs with what UIA gives, and so on. This greatly simplifies not only the collection and maagement of accessibility information, but it also makes it much easier to hook into other accessibility APIs when its time to add them.
These are theories, however. I cannot attest that these claims are actually accurate, as I am not an NVDA or JAWS developer (thank god I'm not a JAWS dev, I'd hate that). However, its a pretty good assumption and would be the logical way to go about it.
You say that "many people are reporting this." Are they annotating other issues or creating new ones. If they're creating new issues, then no wonder NVAccess is so slow-- they have to sort through *every* issue, assign it, and close them and so on. A common git hub practice when reporting issues (I don't understand why people wouldn't do this) is to find the issue about the subject in question, read through it to se if there are suggestions you have not considered, and then, if you have exhausted all suggestions in the issue, you indicate that you, too, are suffering the issue. If people wouldn't open thousands of issues about the same issue and follow common git hub practices, I doubt NVAccess would be so slow-moving, nor do I think they'd have so many issues open.
@74, did I say that you were spamming them with issues? No, I did not. I said hat you most likely would. Please read my post more carefully instead of going on a tangent at the slightest movement. And I'll take the "stupid" part of your last sentence: none of us who said we weren't having these issues assumed (at least, I didn't) that you were stupid for 'having these issues'. We point out simplistic things as a diagnostic device. It is quite literally the first thing any organization or group of developers will tell you to do when you go to them with a problem. Do not assume, then, that that organization or group of developers thinks you are stupid. That is such an erroneous assumption as to be laughable. It is a basic diagnostic tool to ensure that the issue is not a user error and is in fact a code-related issue. Whether you take it as condescension is extremely weird, IMO.
"On two occasions I have been asked [by members of Parliament!]: 'Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out ?' I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question." — Charles Babbage.
My Github