From: Dimitre Novatchev <dnovatchev@gm...>  20050618 03:36:00

On 6/18/05, Andre Cusson <ac@...> wrote: > Hi, > > We seem to be on different planets. Definitely. > > > >Be informed that there exist pure XSLT solutions to the problem of > >finding a path between two nodes of a multigraph  I have my own > >solutions and wouldn't be surprised if other people have developed > >theirs. > > > >I would be glad to demonstrate such a solution on a sample graph, > >provided by anyone interested. > > I have no problem finding the paths between nodes, this is trivial, witho= ut > assign. > > It would be nice if instead you had a solution for the problem at stake, No problem at stake has been defined; where is:  the source xml to be transformed.  the wanted properties of the transformation.  the wanted result If in addition the problem is that a pure XSLT solution is "inefficient", then the alleged more efficient solution must be provided so that it would be possible to compare it to pure XSLT solutions people will put forward. > but I have not seen any. I have not seen any problem defined as described above. > > Maybe you do not understand, or maybe I was not clear enough. > > But when you write: > > What I just want is that people are not misinformed about the alleged > > impossibility (or inefficiency) of solving this problem in a pure > > functional (without modification of values of variables) way. > I understand that you are accusing me of misinforming people Until now people have seen only some unsupported statements that a problem exists. > and my feeling > now is that you do not have a solution and you do not want anyone to know= . Andre, I asked you more than once to provide a definition of the problem  a concrete source xml, the desired properties of the transformation and the desired result; also the "efficient" solution which uses saxon:assign. You obviously are not able provide such problem definition that will prove your claims. Then it is logically to conclude that this problem simply does not exist. It is high time to either stop the rhetoric or provide concrete example of your claims. I am deeply disappointed that what looked promising to provide a new challenging problem turned out to be just words. Anyway, if you ever really happen to have a definition of a problem that you want to be solved in pure XSLT, I would be happy to give it a try. Thanks, Dimitre 