When a tree is viewed from the search/browse web page, the link causes the latest build of Phylowidget to launch from the phylowidget.org website, e.g.:
http://www.phylowidget.org/full/?tree='http://www.treebase.org/treebase-web/tree_for_phylowidget/TB2:Tr1000'
This results in an instance of a fully-feathured Phylowidget, and usually performs well. By contrast, when a tree is viewed from /phylows/, it causes a special (but older) copy of Phylowidget to launch, (a) which has fewer features, but (b) also has features for saving modified trees back to the database. For example:
http://purl.org/phylo/treebase/phylows/tree/TB2:Tr1000?format=html
This is not ideal because this older version of Phylowidget is buggy, and the added "Save to database" might be a security hole. Better to have it behave like the other link, or alternatively, use some other viewer altogether (e.g. jsPhyloSVG).
