DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma;
	h=message-id:date:from:user-agent:mime-version:to:cc:subject
	:references:in-reply-to:content-type:content-transfer-encoding;
	b=drCAF1zhnV1refXKBqUqTgP21jNTuLL+H+8uJ7QkeT5I3Qfrxnc4mmRAlgLz+ATR5N
	lzPTRG2PWnbm3PHF3EMX+QK82Do7UnP9SpJf12Dm+z51fn16tHPekKonozPymkATJrgr
	jaAj1Tk9pBotWt0UQAfS2wDzAozt3JEXy0Lqs=
Message-ID: <4BBD8873.5040003@gmail.com>
Date: Thu, 08 Apr 2010 09:40:35 +0200
From: Yeb Havinga <yebhavinga@gmail.com>
User-Agent: Thunderbird 2.0.0.24 (Windows/20100228)
MIME-Version: 1.0
To: Achilleas Mantzios <achill@matrix.gatewaynet.com>
CC: pgsql-sql@postgresql.org
Subject: Re: Table Design for Hierarchical Data
References: <n2i5ab13581004061033j4ed25734xd0504cb95a1fc092@mail.gmail.com>
	<201004071139.42140.achill@matrix.gatewaynet.com>
	<4BBCEC03.3050100@gmail.com>
	<201004080956.06446.achill@matrix.gatewaynet.com>
In-Reply-To: <201004080956.06446.achill@matrix.gatewaynet.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit

Achilleas Mantzios wrote:
> Στις Wednesday 07 April 2010 23:33:07 ο/η Yeb Havinga έγραψε:
>   
>> Achilleas Mantzios wrote:
>>     
>>> Στις Wednesday 07 April 2010 11:06:44 ο/η Yeb Havinga έγραψε:
>>>   
>>>       
>>>> Achilleas Mantzios wrote:
>>>>     
>>>>         
>>>>> You could also consider the genealogical approach, e.g.
>>>>>
>>>>>
>>>>> The parents of any node to the root, i.e. the path of any node to the root are depicted as
>>>>> parents[0] : immediate parent
>>>>> parents[1] : immediate parent of the above parent
>>>>>   
>>>>>       
>>>>>           
>>>> What I have more than one parent?
>>>>     
>>>>         
>>> Then it is no longer neither a tree, nor a hierarchical structure, but rather a graph.
>>> This a totally different problem.
>>>   
>>>       
>> My question was actually an attempt to point at the inability of what 
>> you call the 'genealogical approach' database design to store 
>> information of more than one parent.
>>     
>
>
> Are you suggesting that we should change our definition of trees ADT, just because it does not
> fit the mere detail that humans have two parents?
> Or are you just suggesting that the "genealogical" term is inaccurate?
>   
The latter, but rethinking it, why would genealogical be a bad word when 
applied to graph algorithm 'stuff' when words like parent, child, 
ancestor, sibling are common use. When I read 'genealogical' I had only 
the connotation 'family relations' in mind. I suspect that if looking at 
the definition of the word genealogy alone, it could very well include 
the study of single parent transitive relationships. However, not 
exclusively, so yes, IMHO something called the genealogical approach 
should not preclude polyhierarchies.

regards
Yeb Havinga