office-collab — archive

[Svante Schubert]

Hello John, On 21.09.2012 23:43, John Haug wrote: > Perhaps I haven't absorbed all this sufficiently yet, but I've had a reservation about MCT I've never understood enough to fully prove. Doesn't all this relative positional referencing cause difficulty for an implementation? No, quite the opposite. By using relative positioning the implementations are able to bring changes of multiple users into relation. In addition we are able to separate the changes from the content, allowing to sent changes without the content (and vice versa). > > As Svante noted early in the thread: >> As result to this the numbers of the arguments might be increased (or decreased) when the operation is being moved up/down, as components earlier counted have or have not existed before. > So each time an arbitrary change is rejected or the document content is altered (change-tracked or not), the numerical references of some/all the change tracking markup may be invalidated and must be recalculated by the application, right? > > That seems like a significant and unavoidable burden. I was a hardware guy in school, not a software/algorithms guy, but that has to be unpleasant big-O behavior! Or am I (hopefully!) wrong in my understanding? Moving an operation through the list is similar to a sort, which naturally depends on the algorithms being used and in the worst scenarios of Quicksort (as algorithm) is O(n log n), though it is being said that Quicksort is often faster in practice than other O(n log n) algorithm. If I am not mistaken the additional adoption of the position is merely a constant factor being eliminated in the complexity. No complexity that my mobile wouldn't be able to handle.. Hopefully I could distract some of your uncertainty and doubt, John. Svante > > John > > >