Re: [racket-users] Debugging racket applications

2018-03-05 Thread 'Paulo Matos' via Racket Users
On 05/03/18 17:42, Greg Hendershott wrote: >> OK, doing racket -l errortrace -t as suggested in >> http://docs.racket-lang.org/errortrace/quick-instructions.html helps. > > The equivalent in racket-mode is to set `racket-error-context` to `'high`: > > > https://github.com/greghendershott/ra

Re: [racket-users] Debugging racket applications

2018-03-05 Thread 'Paulo Matos' via Racket Users
On 05/03/18 16:20, Matthias Felleisen wrote: > > All debugging aids impose a performance penalty. Always. > > Well, maybe I didn't clarify my comment. However, setting -g on a gcc build does not impose performance penalties. Simply adds debug information (making compile-time slower and execu

Re: [racket-users] Debugging racket applications

2018-03-05 Thread Greg Hendershott
When I started programming, debuggers seemed like magic. Eventually I realized that a debugger "just" rewrites your program into a debug-able (e.g. trace-able and/or step-able) version of your program.[^1] Which means a more complicated -- and without hardware assistance, significantly slower --

Re: [racket-users] Debugging racket applications

2018-03-05 Thread Greg Hendershott
> OK, doing racket -l errortrace -t as suggested in > http://docs.racket-lang.org/errortrace/quick-instructions.html helps. The equivalent in racket-mode is to set `racket-error-context` to `'high`: https://github.com/greghendershott/racket-mode/blob/master/Reference.md#racket-error-context

Re: [racket-users] Debugging racket applications

2018-03-05 Thread Matthias Felleisen
All debugging aids impose a performance penalty. Always. > On Mar 5, 2018, at 10:07 AM, 'Paulo Matos' via Racket Users > wrote: > > > > On 05/03/18 15:28, 'Paulo Matos' via Racket Users wrote: >> >> Is there anything I can try to improve backtrace generation? >> > > OK, doing racket -l

Re: [racket-users] Debugging racket applications

2018-03-05 Thread 'Paulo Matos' via Racket Users
On 05/03/18 15:28, 'Paulo Matos' via Racket Users wrote: > > Is there anything I can try to improve backtrace generation? > OK, doing racket -l errortrace -t as suggested in http://docs.racket-lang.org/errortrace/quick-instructions.html helps. Does errortrace introduce a performance penalty

[racket-users] Debugging racket applications

2018-03-05 Thread 'Paulo Matos' via Racket Users
Hi, I have been using Greg's racket mode and racket on the command line to start my application. One of the most frustrating/infuriating things about full time racket programming at the moment is debugging. For example, here's a run of my program: $ racket /var/tmp/lt15198411601519841160715/0/dri