[nfbcs] JAWS JAWS JAWS

[Mike Gorse]

That's an advantage of NVDA, and open source software in general--if there 
is a bug that a developer hasn't fixed yet, then you can fix it, at least 
in theory. That being said, there is a significant learning curve in terms 
of understanding how a piece of code works, understanding the APIs that it 
uses, and so on, even if you know the programming language, so digging 
down and fixing a bug probably won't be something that most users are 
realistically going to do.

On Mon, 4 Feb 2019, Gary Wunder via nfbcs wrote:

> Hello, Doug. You will probably get some flak for this post, but as a
> programmer for thirty years I can validate all of your points, if not the
> specific numbers that go along with them. I think we have every right to
> complain when a product does not meet our expectations, but everyone will
> find in his or her personal experience the need to prioritize and the
> reality that not everything we want to do will get done and certainly not in
> the timeframe we wish. Thank you for the post.
>
>
> -----Original Message-----
> From: nfbcs [mailto:nfbcs-bounces at nfbnet.org] On Behalf Of Doug Lee via
> nfbcs
> Sent: Monday, February 04, 2019 11:25 AM
> To: NFB in Computer Science Mailing List
> Cc: Doug Lee
> Subject: Re: [nfbcs] JAWS JAWS JAWS
>
> An Economics 101 story problem, where numbers are fictitious since I don't
> have real ones:
>
> A company has 50 employees and 50,000 customers. In an average month, 1000
> work requests from customers come in.  Each employee has approximately five
> hours out of each day in a five-day work week
> directly for handling customer work requests. The average time required to
> fill each of the 1,000 submitted requests is estimated at ten hours, because
> of the need to test carefully after each one to make sure that the changes
> it produces do not upset something else that could affect all 50,000
> customers.
>
> The company's CEO does the math and figures out that
> - One employee needs an average of two days to fill one request.
> - If all 50 employees work solidly, this means 25 requests per day will be
> filled on average.
> - Under ideal conditions, this translates to 40 business days of work to
> fill all pending requests from one month.
> - In an average month, accounting for holidays, there are about 20 business
> days.
> - This means that current request-handling capacity is only half of what is
> needed to fill all requests.
>
> You are the CEO. Which of the following is your plan for handling this
> problem?
>
> 1. Hire more employees, thus forcing the price of the product to increase.
>
> 2. Filter the work requests down so that only half actually require work.
> (Keep in mind, this filtering process itself will also take some of those
> employee hours, besides disappointing a lot of customers.)
>
> 3. Force employees to fill twice as many requests in the same amount of
> time, thus risking quality control problems.
>
> 4. A solution that escaped my attention. This is entirely possible; I'm not
> a CEO. :-)
>
> Disclaimers:
> - I bet the customer count is higher for Vispero. Not sure about employee
> count.
> - I also think 20 days is a high estimate for number of business days in a
> month averaged over a year for the average private-sector company.
> - Ten hours per request is an average and is ment to account for a whole
> host of things, including occasional major issue resolution requests,
> required research, testing on multiple systems, Windows versions, etc.
> - Five hours per day for customer requests is probably above reality for
> Vispero; after all, even if you ignore meetings and other normal company
> hum-drum, innovation and independent research and experimentation are part
> of
>  the job of a screen reader developer/maintainer.
>
> On Mon, Feb 04, 2019 at 11:42:41AM -0500, NFBCS mailing list wrote:
> Don't you hate when a company refuses to patch certain bugs?