Re: academics and employment
Eloquence is the only synth I will use for myself.
From: firstname.lastname@example.org <email@example.com> On Behalf Of JM Casey
Sent: Tuesday, August 21, 2018 11:41 AM
Subject: Re: [nvda] academics and employment
I found that once I switched back to Eloquence, JAWS crashed a lot less. It’s been performing the best I’ve ever seen since I dropped Vocalizer as my primary synth. I got tricked into liking the more human-sounding voice, but the thing is obviously a little resource-intensive and seems to cause other problems as well. Maybe people who experience a lot of JAWS crashes should consider using a less intensive synth. I have 8 GB of ram and a fairly fast multi-core processor and yet Vocalizer was clearly still causing problems for whatever reason.
Jaws only seems to crash a lot for me on web pages, while a page is loading mostly, and it seems only with IE, but I use IE more, so it is hard to be sure about that.
I think Jaws may crash a lot more than people realize, because now it comes back on its own, so a page will be loading, and I hear nothing for a couple minutes, and then I hear it start up with its usual start up phrase.
Thank you for sharing your insights on this! They were a pleasure to read. I like to read and write a lot. I do not mind reading long messages.
When I generalized about JAWS, I was being informal. I should have asked the group not to take my advice as solid proof. Overall, it sounds like what we really need to do is have a large group of people reproduce the exact steps that cause JAWS to crash, and if it is consistent on that large number of machines, we need to use a debugger and log files to step through the app or site JAWS is crashing on when it happens, document it, and send that report to Freedom Scientific.
The conclusion I drew from what you wrote, overall, is that nothing is perfect. I use the tools that will solve the problem along the way, and suggest improvements as I go.
This message is rather long. You may or may not want to read it all. I hope you find it interesting and useful.
Others may be able to give you an answer or some sort of guide about a good sample size. I don't have any formal answer. Intermittent or not, if there is a performance problem on only one machine that is known, it can't be generalized to be a problem that affects a lot of or most machines without more information. I never say the person isn't having the problem unless I know the cause and it isn't what is being claimed but off and on, when someone generalizes about a performance problem because they have it, I say that it hasn't been demonstrated that it is a general problem.
Recently, a member said they had the same problem on two or three computers. Ssince the problem occurred in a very popular application and no one had reported it on any list I follow except him, I surmised and said, not as a definite conclusion, but as, what I consider to be a likely state of affairs, that the person had done the same thing on all those machines, installed or uninstalled the same program or something, whatever it might be that caused the problem on all machines. I use the nature of the problem, if I've seen it reported by others, and other factors, if they seem relevant, but my main point is that I see people generalize about a product by saying how terrible it is because of a performance problem on their machine and generalization can't be done.
As far as the announcement of visited links is concerned, I believe that is true, that it can't be turned on or off. But not to use a screen-reader because of one tiny behavior like that is, in my opinion, absurd.
If someone is that bothered by visited link announcement, delete browsing history. That's what visited links are determined by. The information is stored in browsing history and turning it off will cause all links to be shown as unvisited.
Second, if you tab through links, or use the letter k to move just by links, you will hear the link spoken and visited spoken afterword. This is another example of where NVDA should be user customizable. You should be able to turn off visited as an announcement but further, you should be able to set NVDA to announce visited before or after a link and it should be consistent no matter how you move through links.
As I said yesterday, it might be a good idea to have a small group convene and consider and get input from users about how NvDA should be made more customizable. In this respect, you also can't adjust when other controls are heard, do you want to hear button or combo box or check box, for example, before or after you hear the text announced.
I haven't used System Access enough for a long time to know how it may have changed in more recent years. Regardless, it simply is more limited than NVDA on the Internet and in general. It wasn't developed to be a powerful screen-reader, it was intended to be powerful enough to meet less demanding users' needs at a time when there were no or no decent considerably lower price alternatives to JAWS and Window-eyes whereas NVDA is intended to be a much more powerful screen-reader
Take one example, the simulated mouse in System Access is much less capable. there are times when you must use the simulated mouse to activate a link or control. But you often can't do it using the System Access mouse, at least that is my recollection. And talk about annoyances, the simulated mouse in System Access is programmed to make a horrible nerve-grating noise when you click it. What were they thinking? In my opinion, that noise is much more disgusting and annoying than hearing the word "visited" when I don't want to.
On the other hand, years ago, I used one or two web sites that couldn't be used to perform their main function when other screen-readers were used. I could use a link or control to perform an essential action that wouldn't work with another screen-reader. Also, on this or that page, System Access would read changes that occurred on the page, not where I was working when I would take an action and that was essential because I wouldn't have known that the page had changed otherwise even if I could have read the new material. And it would have been ridiculously inefficient to look around the page every time I guessed that it might have changed somewhere.
There is no one ideal screen-reader and I strongly feel that not using a screen-reader because of this or that small or tiny behavior is absurd. Such behaviors may be annoying but it's a classic example of the cliché about a mountain out of a molehill.
Gene----- Original Message ---------- Original Message -----
I believe in using whatever resources are at my fingertips, in any given situation. If I come across an inaccessible app at work, I am not going to have the time to program out a fix for it, plus do the programming I do as part of my job, and meet the deadline.
The reason my friend uses System Access and not NVDA is that he claims with NVDA, you cannot turn off the announcement of visited and unvisited links.
Also, Gene, you were talking about sample size and testing software earlier. What is an accurate sample size for testing an intermittent problem? When JAWS crashes on the machines I used, it is intermittent. I wish they taught us more about troubleshooting intermittent problems in school. It is a skill that I have much room for growth in.
System Access isn't nearly as powerful as NVDA but it works with web pages differently. While this generally doesn't matter in terms of accessibility of web pages, at times it does. So it's good to have as a resource. But you don't have to install it. Try System Access to Go. As a once in awhile resource, it may be useful.
You used to have to use System Access to Go with Internet Explorer. I'm not sure now, though I believe that in order to use the talking feature that speaks when you open the web site, you still may.
----- Original Message -----
I had a problem with System Access where it totally locked up my computer when I installed it. I may have to retry. My friend keeps wanting me to try it.