> Do you have the right metric here?
Well, there is probably no right metric, do you know one which fits better?
Consider that it should be possible to extrapolate the current quality level
without having perfect information, e.g. you only know something about
the discovered bugs, not about the hidden ones.
Volant wrote:
Do you have the right metric here?
Imagine that an error cuts every identifier after 2 characters - not acceptable,
or after 20 characters - bad,
or after 200 character - no one will even notice.
And this is exactly ONE (imaginary) bug (of the same kind).
Concerning "lines of code" see the text books about software metrics,
especially the requirements for "lines of code" meaning something.