Re: pfASD last level not shown

New Message Reply Date view Thread view Subject view Author view

From: Marco Tartaglia (m.tartaglia++at++acsys.it)
Date: 04/27/2001 03:10:01


Thank you Yair for your very quick answer.

But what you said doesn't explain why if i remove the last paging files
(finest level) we have no warning messages, no crashs and we have the
same behavior with or without these files; as their aren't really used.
For our application the allocation memory is critical (we need to load
many pfASD node) and having a large amount of memory allocated and not
used, limits us a lot, in fact the allocation of level N: [4 ^N quads]
is much more than the SUM of all the previous N-1 levels: [((4^N) - 1)
/ 3 quads] .
Do you think that you can help us more ?
Is there any work around to avoid the allocation of 'apparently' unused
memory ?
Could be any problem in our use of ASDGen functions ?

Thank you again

Marco

Yair Kurzion wrote:

> Hello Marco !
>
> I believe the highest level (say - level N) contains only vertices. These
> vertices lie on edges of triangles at level (N-1). pfASD doesn't display
> these vertices but uses them to match neighboring triangles at level (N-1).
> Two triangles of level (N-1) are considered neighbors if they share a reference
> vertex (pfTerrainFace::refvert) on one of their edges.
>
> -yair
>
>
>> we are using pfASD node to represent a terrain from GTOPO30 files; we
>> are using ASDGen tool to create paging asd and we experienced a problem:
>> the last (finest) level was allocated but not displayed.
>> We think that it is allocated because the fmem[level][x][y] printing
>> show up to the last level number.
>> Unfortunaly this level is not displayed (we aren't able to see it) and
>> it is not even loaded because aif we remove the lst level pages nothing
>> changes.
>

-- 
              Marco Tartaglia

ACS - Advanced Computer System SpA via della Bufalotta, 378 Sc. M - 00139 Roma

Tel: +39687090311 Fax: +39687201460 email: m.tartaglia++at++acsys.it web: www.acsys.it


New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Fri Apr 27 2001 - 03:15:53 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.