Re: Demo: Modelling Cost of Travel Paths Between Towns

On 16 Nov 2004 22:57:04 -0800, Neo wrote:

>Summarization

(snip)

Hi Neo,

I'm sorry, but your point is completely invalid. The actions needed to correctly change John's name are different from the actions needed to change Paula's preference - in any model.

   (Ref by name)                    (Ref by pointer)
   Mary likes John           1: Mary         ->1 likes ->3
   Paula likes John          2: Paula        ->2 likes ->3
                             3: John

1. John's name changes to Johnny.
   Update each occurence     Update each occurence
   of 'John' to 'Johnny'     of 'John' to 'Johnny'
         yields:                      yields:
   Mary likes Johnny         1: Mary         ->1 likes ->3
   Paula likes Johnny        2: Paula        ->2 likes ->3
                             3: Johnny

2. Paula changes preference - she now prefers Mary.
   Update the 'likes'        Update the 'likes'
   relationship for Paula    relationship for ->2
         yields:                      yields:
   Mary likes Johnny         1: Mary         ->1 likes ->3
   Paula likes Mary          2: Paula        ->2 likes ->1
                             3: Johnny

If the wrong update strategy is chosen in either the "ref-by-name" model or the "ref-by-pointer" model, the data becomes corrup. If the correct update strategy is chosen, the data doesn't become corrupt. If a user accidentally chooses the wrong update strategy, the data will become corrupt. That won't change until computers become equipped with a mind-reading device.

In case you still don't believe me, reread the message with message-ID <picqa01mpenqaq3ftiqif7nhiclvpohjm8_at_4ax.com> (or if you prefer google: http://www.google.com/groups?q=g:thl2447151731d&dq=&hl=en&lr=&c2coff=1&selm=picqa01mpenqaq3ftiqif7nhiclvpohjm8%404ax.com). In this message, I try (as prompted by you) to enter a person and a dog, both named 'john', in XDb1. It fails. In your reply to this message (messageID <4b45d3ad.0405202106.4d079b08_at_posting.google.com>, or http://www.google.com/groups?q=g:thl2447151731d&dq=&hl=en&lr=&c2coff=1&selm=4b45d3ad.0405202106.4d079b08%40posting.google.com), you indicate that I chose the wrong data modification strategy, or (to quote you) "The steps you took to verify it were inappropriate." Which proves that chosing the incorrect data modification strategy can cause corrupt data in ANY database.

>> > In your example, you stored john (a fact) twice.
>>
>> You call 'john' a fact?????
>
>Yes, 'john' is a fact. In fact, a, b, c ... are facts.
>You may want to read OT "A Normalization Question".
>Feel free to provide arguments to the contrary.

All my arguments to the contrary are here: http://dictionary.reference.com/search?q=fact

Best, Hugo

-- 

(Remove _NO_ and _SPAM_ to get my e-mail address)