[Date Prev][Date Next][Thread Prev][Thread Next][Author Index][Date Index][Thread Index]
Re: [zzdev] Re: [zzdev] Re: [zzdev] gzigzag/Java ZZTreeRaster.java (fwd)
- To: Antti-Juhani Kaijanaho <gaia@xxxxxx>
- Subject: Re: [zzdev] Re: [zzdev] Re: [zzdev] gzigzag/Java ZZTreeRaster.java (fwd)
- From: "Tuomas J. Lukka" <lukka@xxxxxx>
- Date: Fri, 14 Jul 2000 06:57:15 +0300 (EEST)
- Cc: zzdev@xxxxxxxxxx
- In-reply-to: <20000714135129.F7245@xxxxxxxxxxx>
On Fri, 14 Jul 2000, Antti-Juhani Kaijanaho wrote:
> On Fri, Jul 14, 2000 at 01:43:15PM +0300, Antti-Juhani Kaijanaho wrote:
> > > The way I expected it would be to put the accursed cell in the middle: show
> > > its parent and siblings and children and grandchildren and
> > > greatgrandchildren etc.
> >
> > Hm, I'll have to think about that. Actually, that's probably quite easy.
>
> I need on input on this one: if the substructure being shown is not
> a tree, how should this raster cope with it? My current method is
> (I think...) to show a spanning tree of the subnetwork that consists of
> all cells reachable from the "root" poswards on the two dimensions shown.
> However, this method needs a "root" cell, and finding it from the accursed
> cell is easy only if the substructure is a true tree.
Don't rely on it being a root - the parent can have siblings too. Think
relative, and think of being in a big tree.
If cells really lie outside the tree, show a connection and a very small
version of the cell (for example).
Actually, if you're deep in a tree, showing the whole path outwards
might be nice: just show parent's parent etc.
Tuomas