All I meant was that we should use a more refined notion of "depth limit"
which does a good job of cutting off the output before it grows too large.
That might mean limiting not only the depth, but the width too, for example.
Cheers,
Simon
pepe wrote:
To clarify, this is only when doing ':show bindings', not when doing
':print'.
Ensuring that the output given never gets too large would mean
arbitrarily cutting it to make it fit, or giving the user some control
over these settings. I remember talking to you about this in ICFP, and I
got the impression you thought it was overkill. Did I misunderstand ?
For now we have a shallow (fixed) depth limit in the term reconstruction
step which helps to limit the size of terms, but unfortunately not the
size of the output given by the pretty printer. I plan to experiment
with Pretty.fullRender with well-chosen settings, that should be a big
win. Any other suggestion will be very appreciated.
pepe
On 14/11/2007, at 10:17, Simon Marlow wrote:
Pepe Iborra wrote:
Tue Nov 13 09:45:39 PST 2007 Pepe Iborra <[EMAIL PROTECTED]>
* GHCi debugger: added a new flag, -fno-print-binding-contents
The contents of bindings show at breakpoints and by :show bindings
is rendered using the same printer that :print uses.
But sometimes the output it gives spans over too many lines and the
user may want to be able to disable it.
Then wouldn't it be better to ensure that it never gets too large? Or
is that too hard?
Cheers,
Simon
_______________________________________________
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mailman/listinfo/cvs-ghc
