cleo loses its settings at depth > 7
Nik
My cleo settings on nodes at depth > 7 are not permenant.
Each time after restarting LEO cleo loses my settings on these nodes!!
I'm using LEO 4.4.8 beta 3.
Regards
Nik
Terry Brown
Nik <nitr...@googlemail.com> wrote:
> I have the following strange problem with cleo plugin.
> My cleo settings on nodes at depth > 7 are not permenant.
Strange indeed, I'll look into it, thanks.
Cheers -Terry
Terry Brown
Nik <nitr...@googlemail.com> wrote:
Hmmm, can't duplicate that on the trunk. I assume you're not using
thyrus's speedups? Not that I'm in anyway suggesting they're to blame,
just that they do have something to do with recursion, but I can't see
why depth of a simple outline should matter.
Ha - I bet I know what the problem is. I need to update the cleo docs.
to explain this. Mainly for historical reasons, cleo stores its
attributes on vnodes. So for this tree:
A
B
C
D
B
C
D
(who remembers the dark ages when we couldn't use indentation :-)
Assuming that the Bs are clones of each other, C and D are in fact the
same vnodes (i.e. both Cs are the same vnode, and both Ds are the same
vnode). So they will always have the same cleo attributes.
Hmm, I guess this isn't seeming so much like an explanation as I
thought.
Do you have a simple test case?
Cheers -Terry
Nik
outline tree
which led to the observation mentioned im my first posting!
I have now reorganized my outline tree and still have a similar
problem.
My outline tree has now the following structure:
@chapters
@chapter AAA
BBB
CCC
DDD
aaa
bbb
ccc
....
All nodes under DDD lose their cleo settings after a LEO restart.
I have also tried the following tree:
@chapters
@chapter AAA
BBB
CCC
aaa
bbb
ccc
....
The same problem with the children of CCC!
It seems that it does not depend on the depth of nodes.
Regards
Nik
On Apr 2, 8:17 pm, Terry Brown <terry_n_br...@yahoo.com> wrote:
> On Wed, 2 Apr 2008 01:47:44 -0700 (PDT)
>
Nik
Only the children of thin nodes suffer from cleo amnesia!
It seems that the problem is related to thin nodes (at least in my
case!)
Regards
Nik
Terry Brown
Nik <nitr...@googlemail.com> wrote:
> I have only organizing and thin nodes in my outline tree.
> Only the children of thin nodes suffer from cleo amnesia!
> It seems that the problem is related to thin nodes (at least in my
> case!)
Ah, @thin nodes, I should have thought of that. Leo doesn't save
attributes on vnodes which descend from @thin nodes.
I'm sure there's a reason for that although I don't know what it is,
Edward?
Cleo could switch to storing its attributes on tnodes, but it's not
clear that that's always desirable, sometimes it's nice to have cleo
"markup" on a node in one context and not in another - talking about
clones here of course.
I'll wait to see what Edward says about vnodes which descend from @thin
nodes before thinking more about possible solutions.
Cheers -Terry
Edward K. Ream
Ah, @thin nodes, I should have thought of that. Leo doesn't save
On Thu, 3 Apr 2008 03:05:27 -0700 (PDT)
Nik <nitr...@googlemail.com> wrote:
> I have only organizing and thin nodes in my outline tree.
> Only the children of thin nodes suffer from cleo amnesia!
> It seems that the problem is related to thin nodes (at least in my
> case!)
attributes on vnodes which descend from @thin nodes.
I'm sure there's a reason for that although I don't know what it is,
Edward?
The reason is simple: vnodes do not exist in @thin trees. gnx's in files derived from @thin refer to tnodes, not vnode. Thus, Leo creates "anonymous" vnodes when reading thin derived files.
This can never change, the @thin read code is robust precisely because it does *not* save vnode info.
Cleo could switch to storing its attributes on tnodes, but it's not
clear that that's always desirable, sometimes it's nice to have cleo
"markup" on a node in one context and not in another - talking about
clones here of course.
I'll wait to see what Edward says about vnodes which descend from @thin
nodes before thinking more about possible solutions.
It's not an easy problem. You would like the flexibility of per-vnode attributes, but that won't work for @thin files.
But whatever you do, please wait until after 4.4.8 final to do it :-)
Edward
Nik
possible problems with tree depth. It was an unfortunate
coincidence that all thin nodes in my outline tree were
located at depth 7! After some simple experiements with
my outline tree I found that cleo 'problem' is related to
thin nodes. I hope that we will have a solution in near furure :-)
Regards
Nik
On Apr 3, 12:54 pm, "Edward K. Ream" <edream...@gmail.com> wrote:
Edward K. Ream
> But whatever you do, please wait until after 4.4.8 final to do it :-)
branch. I should have said,
"whatever you do, please wait until after 4.4.8 final to commit it to
the trunk."
Edward
Edward K. Ream
> I apologize for my first misleading posting regarding
> possible problems with tree depth.
expected to understand the bug in detail :-)
Edward