[r-t] Method extension

[Andrew Johnson]

> From: Philip Saddleton <pabs at cantab.net>
> Date: 01/05/2015 07:33
> 
> I don't think that there should be a requirement to name a method that 
> happens to be an extension the same as the parent, unless that extension 

> path has already been chosen by methods at different stages being given 
> the same name.
> 
> PABS
> 
That's sensible. If M8 extends to Ma10, Mb10, Mc10, but Ma10 somehow is 
the 'best' extension, then ringing Mc10 before Ma10 shouldn't force the 
extension path to be Mc=M.

If a method has been rung at two different stages then how much choice is 
there with extension paths?

> That could be accommodated by allowing a line to have length zero. But
> I don't think it is desirable - you could then add a totally unrelated
> piece of work, e.g. a wrong place section to a right place method.
I agree with this, though I'm happy with a method 'contracting' to another 
method with the loss of a piece of work.

Where it gets slightly tricky is internal versus external places. If you 
have a parent method on stage n with a place at the back at n then if you 
extend the method and keep the place at n then you create some new work 
above. You aren't inventing new place notation, but the effect is 
different.

E.g. PB Minimus -> Little Bob Minor.

Andrew Johnson
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 
741598. 
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU