Re: [bug-gtypist] [improvement] Fairest way to count errors in speed test.

[Felix Natter]

Alain DUCOULOMBIER <address@hidden> writes:

> Hello Gtypist team,

hello Alain,

Sorry for the very late reply - I was busy with my dayjob and another
open source project.

> I would like to submit a patch to change a little bit the way the errors 
> count is computed in speed test.
>
> I am a little bit puzzled each time I practice a speed test that despite I 
> fix all my mistakes before reaching the end
> of the exercise, my error rate is not zero percent (0.00%).
>
> As I would like to use the speed test to mesure my real speed in "normal" 
> conditions, where you have to fix all your
> mistakes before releasing the text (like an email composing for example). 
> This is a behavior close to the one offered
> by the www.typeracer.com website.
>
> So this patch aims at providing an errors rate calculation that counts only 
> the errors not fixed when the user reaches
> the end of the exercise.

I really like the idea, this makes sense :-)

To me the patch looks good, but before I apply it I'd like to ask Tim if
he wants this to take effect unconditionally (i.e. without a cli/config
file option).
@Tim: What do you think? (I think it's ok without an option).

> The current limitation is that you can not fix an error on a previous line as 
> the backspace key doesn't allow to go one
> line back.

I agree this should be fixed next :-)

> Thanks for your involvment in this great software, and let me know if this 
> patch is relevant.

Thanks. Unfortunately, GTypist development has slowed down a bit.

To quote from TODO (see
http://git.savannah.gnu.org/gitweb/?p=gtypist.git;a=blob_plain;f=TODO;hb=master):
  If you are interested in helping for [...], we will be glad to have you
  in the development team!

We also plan to include a way to change the lesson file (*.typ) from
within gtypist (using a <X>: command) for the 2.10 release
(which will include your patch one way or another).
We also look forward to your suggestions for 2.10!

Thanks, Cheers and Best Regards,
-- 
Felix Natter