#PB_Any query

[Foz]

I know that you can specify the same id for a window and a gadget, however, if I use #PB_Any for both windows and gadgets, will the numbers generated ever overlap, or will they be unique?

[srod]

I am pretty sure they would be unique because I am reasonably sure that #PB_Any returns a pointer (for internal PB use).

[Foz]

I was also thinking on the same lines. I do hope so! Fred? Freak? Please pipe up

[netmaestro]

What difference does it make? Creating window 0 and gadget 0 and image 0 in the same program presents no problems - why would identical #PB_Any values for different object types be a concern?

[Foz]

Because I'm building a unique id list, and I want to know if I have to roll my own id generator, or if the #PB_Any will do it regardless.

[netmaestro]

afaik #PB_Any is a constant with a value of -1. To use the #PB_Any generator you'd have to actually create an object, no? What, then free it? Seems simpler to use a map. A map enforces uniqueness and it's fast.

[Foz]

You misunderstand, it's not for maps - it's for a hierarchical tree building, if I have a window and a gadget of the same id, then working out who the children are or who the parent is will get very awkward.

As I said - if using #PB_Any will get me a unique value regardless if it's assigned to a window or a gadget, then I'll just use that, otherwise, I'll use my own unique id generator. I just need to know if I am going to be doing that or not.

[Matt]

Nevermind

[nco2k]

you are wrong.

c ya,
nco2k

[Matt]

Can you correct me? Cause then I don't know

[Arctic Fox]

This post by freak may be of interest
http://www.purebasic.fr/english/viewtop ... 16#p284316

[Foz]

Ahhh perfect! Thanks Arctic Fox, the first line confirmed what I and srod suspected - the values are indeed memory addresses, therefore I can guarantee that they will be unique