[Chapter-delegates] Preview on ISOC website developments

[Alan Levin]

Hi,

In response to both these 'issues'...

On 05 Dec 2007, at 12:42 PM, Patrick Vande Walle wrote:
> Khaled KOUBAA wrote:
>> Alan,
>> Good you raised this argument.
>> And both situation are logic, but in my point of view a membership
>> system should provide at any time a reliable statistic and  
>> information
>> about the real number of member.
>> As I told if I have the right to be member three time each time with
>> different data or more being member twice using the same data,  
>> this is
>> not reliable.
> In my perception, each record in a database should have a unambiguous,
> globally unique identifier. An email address does not fit that  
> criteria.
> It can change, be shared, can become invalid, be recycled for another
> person, etc. Thus it cannot be used to distinguish one record from
> another. In the present case, the membership number should be the one
> distinguishing one member from another.
>
> I agree with Alan there are situations where several people share a
> single email address. I have 2 such members in my chapter, one of them
> being virtually inexistent for ISOC. I also agree some members may  
> want
> to receive communications from ISOC on several e-mail addresses.

So we decided after much debate that the 'unique identifier' was not  
a number but the unique combination of name/surname/email address.

As far as a reliable 'statistic and information about the real number  
of members', we did some tests and calculations over time. There were  
almost always the same number of 'shared' email addresses as there  
were individuals with multiple addresses subscribed. So it wasn't  
entirely accurate but effectively the number of email addresses on  
the list is the number of members of the organisation. We agreed that  
an accurate number members was not needed nor practical until we had  
a proper membership system.

hth

aLan

-- 
Alan Levin
Tel: +27 21 409-7997