Re: Optimizations for pure-height approximations. (issue1817045)

[Joe Neeman]

On Wed, Jul 21, 2010 at 3:09 PM, <address@hidden> wrote:

On 2010/07/21 21:04:38, Neil Puttock wrote:

Ah, don't worry, I think I see what's going on.  It's creating a new

hash table

whenever the method's called:

+ if (!to_boolean (scm_hash_table_p (pure_height_cache)))

This is always true.

I think it should be

if (scm_hash_table_p (pure_height_cache) != SCM_BOOL_T)

OK, this is complete bollocks, of course. :)

It seemed right since I'm no longer getting the excessive memory usage.

Wait, you mean that you're running it with 1000 and not having problems now?