Making generalized Trie type in D

[Dmitry Olshansky]

On 11.06.2012 22:30, Roman D. Boiko wrote:
> On Monday, 11 June 2012 at 18:27:58 UTC, Roman D. Boiko wrote:
>> On Monday, 11 June 2012 at 18:23:39 UTC, Roman D. Boiko wrote:
>>> On Monday, 11 June 2012 at 18:16:45 UTC, Dmitry Olshansky wrote:
>>>> I meant an operation pseudo-XOR x P^ y where x is part of snapshot
>>>> and y is part of diff page.
>>>>
>>>> x P^ y == x when y == T.init
>>>> x P^ y == y when y != T.init
>>> I understood, just mentioned some alternatives (maybe not reasonable
>>> given your trie implementation).
>> Or probably not understood :)
>>
>> Why pseudo?
0 ^ 0 = 1 while T.init P^ T.init == T.init :)
other are almot the same
but
1 ^ 1 = 0 while x P^ y == y
So it's not symmetric when x,y != T.init for starters ;)

> Is this article somehow related, or just uses the same term?
> http://iwct.sjtu.edu.cn/Personal/mxtao/paper/liujc_icc11_PID1075548.pdf
>
looks like it might be but I can't penetrate this scientish somehow ( I 
usually can) in any case I believe it's more of coincidence.

-- 
Dmitry Olshansky