Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.java.programmer > #26261 > unrolled thread

Java EE Architecture

Back to article view | Back to comp.lang.java.programmer

Contents

  Java EE Architecture art.marole@gmail.com - 2013-10-09 13:41 -0700
    Re: Java EE Architecture Lew <lewbloch@gmail.com> - 2013-10-09 14:45 -0700
      Re: Java EE Architecture Arne Vajhøj <arne@vajhoej.dk> - 2013-10-09 22:00 -0400
    Re: Java EE Architecture Arne Vajhøj <arne@vajhoej.dk> - 2013-10-09 22:18 -0400
      Re: Java EE Architecture art.marole@gmail.com - 2013-10-10 00:47 -0700
        Re: Java EE Architecture Arne Vajhøj <arne@vajhoej.dk> - 2013-10-16 20:37 -0400
    Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-10 13:26 +0100
      Re: Java EE Architecture 1connu <1connu999@gmail.com> - 2013-10-16 10:45 +0200
        Re: Java EE Architecture Donkey Hottie <donkey@fredriksson.dy.fi> - 2013-10-16 13:42 +0300
          Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-16 11:55 +0100
            Re: Java EE Architecture Donkey Hottie <donkey@fredriksson.dy.fi> - 2013-10-16 14:18 +0300
            Re: Java EE Architecture Donkey Hottie <donkey@fredriksson.dy.fi> - 2013-10-16 14:25 +0300
              Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-16 14:23 +0100
                Re: Java EE Architecture Donkey Hottie <donkey@fredriksson.dy.fi> - 2013-10-16 17:10 +0300
                  Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-16 16:44 +0100
                    Re: Java EE Architecture "L." <carisio@gmail.com> - 2013-10-16 10:15 -0700
                      Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-16 19:01 +0100
                        Re: Java EE Architecture "L." <carisio@gmail.com> - 2013-10-16 12:17 -0700
                          Re: Java EE Architecture Martin Gregorie <martin@address-in-sig.invalid> - 2013-10-16 21:01 +0000
                            Re: Java EE Architecture Arne Vajhøj <arne@vajhoej.dk> - 2013-10-16 20:40 -0400
                    Re: Java EE Architecture Donkey Hottie <donkey@fredriksson.dy.fi> - 2013-10-17 00:00 +0300
                Re: Java EE Architecture Wayne <nospam@all.invalid> - 2013-10-16 23:08 -0400
                  Re: Java EE Architecture Donkey Hottie <donkey@fredriksson.dy.fi> - 2013-10-19 17:05 +0300
                    Re: Java EE Architecture Lew <lewbloch@gmail.com> - 2013-10-19 14:48 -0700
                      Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-20 09:00 +0100
                        Re: Java EE Architecture "Qu0ll" <Qu0llSixFour@gmail.com> - 2013-10-20 19:40 +1100
                          Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-20 10:07 +0100
            Re: Java EE Architecture Arved Sandstrom <asandstrom2@eastlink.ca> - 2013-12-16 20:14 -0400
              Re: Java EE Architecture lipska the kat <"lipskathekat at yahoo dot co dot uk"> - 2013-12-17 08:47 +0000
          Re: Java EE Architecture Jeff Higgins <jeff@invalid.invalid> - 2013-10-16 20:19 -0400
            Re: Java EE Architecture Jeff Higgins <jeff@invalid.invalid> - 2013-10-16 21:24 -0400
        Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-16 11:49 +0100
          Re: Java EE Architecture "Qu0ll" <Qu0llSixFour@gmail.com> - 2013-10-16 22:24 +1100
            Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-16 13:39 +0100
              Re: Java EE Architecture Arne Vajhøj <arne@vajhoej.dk> - 2013-10-16 20:39 -0400
                Re: Java EE Architecture "Qu0ll" <Qu0llSixFour@gmail.com> - 2013-10-17 13:43 +1100
                Re: Java EE Architecture Jeff Higgins <jeff@invalid.invalid> - 2013-10-16 23:22 -0400
                  Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-17 08:53 +0100
                Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-17 08:57 +0100
                  Re: Java EE Architecture "Qu0ll" <Qu0llSixFour@gmail.com> - 2013-10-17 19:31 +1100
                    Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-17 09:56 +0100
                      Re: Java EE Architecture "Qu0ll" <Qu0llSixFour@gmail.com> - 2013-10-17 20:18 +1100
                        Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-17 10:56 +0100
                          Re: Java EE Architecture "Qu0ll" <Qu0llSixFour@gmail.com> - 2013-10-17 21:39 +1100
                            Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-17 12:08 +0100
                    Class Person [was: Java EE Architecture] "Chris Uppal" <chris.uppal@metagnostic.REMOVE-THIS.org> - 2013-10-18 08:19 +0100
                  Re: Java EE Architecture Arne Vajhøj <arne@vajhoej.dk> - 2013-10-17 21:00 -0400
                    Re: Java EE Architecture Arne Vajhøj <arne@vajhoej.dk> - 2013-10-17 21:02 -0400
                      Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-18 09:06 +0100
                    Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-18 09:04 +0100
                      Re: Java EE Architecture Arne Vajhøj <arne@vajhoej.dk> - 2013-10-19 22:43 -0400
                        Re: Java EE Architecture lipska the kat <"nospam at neversurrender dot co dot uk"> - 2013-10-20 09:05 +0100
                          Re: Java EE Architecture Arne Vajhøj <arne@vajhoej.dk> - 2013-10-20 09:24 -0400
        Re: Java EE Architecture Arne Vajhøj <arne@vajhoej.dk> - 2013-10-16 20:43 -0400

Page 2 of 3 — ← Prev page 1 [2] 3  Next page →

#26344

16.10.2013 18:44, lipska the kat kirjoitti:
> You say you are managing the personnel of a Company. What other types of
> personnel are there apart from Employees.

ExternalPerson. They are workers bought from a Vendor, external company.
Consultants, it you will.

-- 

Today is the last day of your life so far.

[toc] | [prev] | [next] | [standalone]

#26354

On 10/16/2013 9:23 AM, lipska the kat wrote:
> <snip>
> 
> You state
> 
> 'A Person may have more than one Identities' (sic)
> I'd prefer 'An Employee can have one or more Roles/Identities'
> A person with multiple identities doesn't really make sense does it?
> 
> How do you identify someone with multiple identities???
> 

That's easy: you name the class Sybil.

-- 
Wayne

[toc] | [prev] | [next] | [standalone]

#26382

17.10.2013 06:08, Wayne kirjoitti:
> On 10/16/2013 9:23 AM, lipska the kat wrote:
>> <snip>
>>
>> You state
>>
>> 'A Person may have more than one Identities' (sic)
>> I'd prefer 'An Employee can have one or more Roles/Identities'
>> A person with multiple identities doesn't really make sense does it?
>>
>> How do you identify someone with multiple identities???
>>
> 
> That's easy: you name the class Sybil.
> 

For example one Person: me.

I have multiple Identities. Donkey Hottie is one. There are others.

-- 

It is easy to find fault, if one has that disposition.  There was once a man
who, not being able to find any other fault with his coal, complained that
there were too many prehistoric toads in it.
		-- Mark Twain, "Pudd'nhead Wilson's Calendar"

[toc] | [prev] | [next] | [standalone]

#26384

Donkey Hottie wrote:
> For example one Person: me.
> 
> I have multiple Identities. Donkey Hottie is one. There are others.

This entire discussion has revolved around the definitions of "Person", 
"Identity", and other types as if they had a correct definition.

Whether a particular identifier is used in a program or not is a matter of art 
(where "art" is used in the technical sense, lest one need reminding here), 
and part of that is whether it conveys useful information to maintainers. In the 
restricted domain of discourse for any given system, for example a payroll 
system, a term like "Person" might well have a useful definition and scope.

To lay down a ukase like "One should never use a 'Person' type in a production system" 
is as pointless as it is self-serving. It has no bearing on any real issues of software 
development. Use terms that make sufficiently precise sense in the particular domain, 
and if that happens to include 'Person', so be it.

OTOH, one must indeed be cautious with such terms. Like every other engineering 
decision, there's a downside. The ease of interpretation of a term like "Person" 
comes with a concomitant ease of misinterpretation. To the degree that such 
misinterpretation is both likely and likely to damage system operation, yes, indeed, 
avoid those terms.

The problem lies not with the sentiment that "Person" is too ambiguous a term to use
safely, but the absolutism of the prohibition against a particular term without the context
and understanding to avoid cargo-cult programming.

-- 
Lew

[toc] | [prev] | [next] | [standalone]

#26389

On 19/10/13 22:48, Lew wrote:
> Donkey Hottie wrote:
>> For example one Person: me.
>>
>> I have multiple Identities. Donkey Hottie is one. There are others.
>
> This entire discussion has revolved around the definitions of "Person",
> "Identity", and other types as if they had a correct definition.
>
> Whether a particular identifier is used in a program or not is a matter of art
> (where "art" is used in the technical sense, lest one need reminding here),
> and part of that is whether it conveys useful information to maintainers. In the
> restricted domain of discourse for any given system, for example a payroll
> system, a term like "Person" might well have a useful definition and scope.
>
> To lay down a ukase like "One should never use a 'Person' type in a production system"
> is as pointless as it is self-serving.

What on earth are you babbling about now Bloch?
How is it 'self-serving'?

You yourself opined

'This entire discussion has revolved around the definitions of "Person",
"Identity", and other types as if they had a correct definition.'

Do you see that last bit? ...
The bit that says "... as if they had a correct definition"

Those few words describe perfectly to me your complete lack of 
understanding of the issue at hand, it's exactly *because* Person does 
not have a 'correct' (although I prefer exact) definition (in terms of a 
software system) that it should not be used.

Object Oriented software engineering is a state of mind Bloch.
You obviously don't have it (the required state of mind) so I suggest 
you consider your responses before you make a fool of yourself again

[snip more incoherent nonsense]

... to avoid cargo-cult programming.

Why use one word when 100 will suffice

jeez


-- 
Lipska the Kat©: Troll hunter, sandbox destroyer
and farscape dreamer of Aeryn Sun
GNU/Linux user #560883 - http://www.linuxcounter.net

[toc] | [prev] | [next] | [standalone]

#26391

"lipska the kat"  wrote in message 
news:fMmdnaMODK8GEP7PnZ2dnUVZ8nednZ2d@bt.com...

> What on earth are you babbling about now Bloch?

[snip]

> Object Oriented software engineering is a state of mind Bloch.
> You obviously don't have it (the required state of mind) so I suggest you 
> consider your responses before you make a fool of yourself again

[snip]

> [snip more incoherent nonsense]

You have put your hand up identifying yourself as a mindless troll again 
kitty.  Whenever someone disagrees with you, you resort to personal insult 
and steer clear of the technical argument.

> Lipska the Kat©: Troll hunter, sandbox destroyer

I have to laugh every time I see this sig.  The fact that you call yourself 
"Troll hunter" is as amusing as when countries put the word "democratic" in 
their name.

--
And loving it,

-Qu0ll (Rare, not extinct)
_________________________________________________
Qu0llSixFour@gmail.com
[Replace the "SixFour" with numbers to email me] 

[toc] | [prev] | [next] | [standalone]

#26392

On 20/10/13 09:40, Qu0ll wrote:
> "lipska the kat"  wrote in message
> news:fMmdnaMODK8GEP7PnZ2dnUVZ8nednZ2d@bt.com...
>
>> What on earth are you babbling about now Bloch?
>
> [snip]
>
>> Object Oriented software engineering is a state of mind Bloch.
>> You obviously don't have it (the required state of mind) so I suggest
>> you consider your responses before you make a fool of yourself again
>
> [snip]
>
>> [snip more incoherent nonsense]
>
> You have put your hand up identifying yourself as a mindless troll again
> kitty.  Whenever someone disagrees with you, you resort to personal
> insult and steer clear of the technical argument.

I think I've stated quite clearly my 'technical argument' it's not my 
problem if you don't get it.


-- 
Lipska the Kat©: Troll hunter, sandbox destroyer
and farscape dreamer of Aeryn Sun
GNU/Linux user #560883 - http://www.linuxcounter.net

[toc] | [prev] | [next] | [standalone]

#26957

On 10/16/2013 07:55 AM, lipska the kat wrote:
> On 16/10/13 11:42, Donkey Hottie wrote:
>> 16.10.2013 11:45, 1connu kirjoitti:
>>> lipska the kat a écrit :
>>>> On 09/10/13 21:41, art.marole@gmail.com wrote:
>>>>> Hi all,
>>>>>
>>>>> I would like to prepare typical Java EE solution (EJB, JSF, JPA) now,
>>>>> and I have quite stupid problem related with domain objects. I want
>>>>> to create Person class as persistence class, but I don't know in
>>>>> which module I shoud put this class: Business Layer (ejb module) or
>>>>> in Integration Layer (jar module).
>>>>>
>>>>> Person class is more related with logic, so I should put it in ejb
>>>>> module, but in this case I will need to create dependency relation
>>>>> from Integration Layer to Business Layer as Integration Layer need to
>>>>> know what object should store in DB (ex. void store(Person
>>>>> object){};).
>>>>> Put Person class in Integration Layer will remove this unwanted
>>>>> dependency, but I feel that Person class is more Business class then
>>>>> Integration class.
>>>>>
>>>>> What do you think about it?
>>>>
>>>>
>>>> A Person class ... A PERSON class., please, please, *please* don't
>>>> have a Person class
>>>>
>>>> For the sake of whatever omnipotent deity you pray to, do not have a
>>>> Person class
>>>>
>>>> *People* do not belong in a computer system.
>>>>
>>>> Why?
>>>>
>>>> Because it's a poor abstraction.
>>>>
>>>> How are you going to model consciousness, sentience, the 'mind', the
>>>> 'soul'? let alone arms legs, eyes and teeth :-)
>>>>
>>>> You *do not need* a Person class.
>>>>
>>>> Let's put it another way
>>>>
>>>> The very fact that you are having problems deciding where your Person
>>>> belongs should alert you to the fact that it is a poor abstraction.
>>>>
>>>> Try Account, or Record, or if you absolutely *must* User but not
>>>> Person, you will end up regretting it.
>>>>
>>>> lipska
>>>>
>>>>
>>> Why not ?
>>> You can have an abstract class Person with minimum field (name, date
>>> birth) and sub class depending of role of theses persons (ex User
>>> (login/password) ) in your model.
>>
>> I see no problem with Person.
>>
>> I have interface IPerson, and implementations Person and ExternalPerson.
>> I'm a lead programmer on an Identity Management solution.
>
> You *cannot* be serious ... IPerson groan :-(
>
> If you are developing an Identity Management system surely Identity
> might be a better abstraction ... just a thought.

Must admit, IPerson made me cringe too. In any context I can think of.

I'm not as antagonistic to the Person concept as an implementation. 
Despite what you said, all a coder needs to do with that concept is 
model just what they need. If I was going to model a pedestrian mobility 
study using some random algorithms, I might well call my discrete 
construct a Person.

AHS
-- 
When a true genius appears, you can know him by this sign:
that all the dunces are in a confederacy against him.
-- Jonathan Swift

[toc] | [prev] | [next] | [standalone]

#26964

On 17/12/13 00:14, Arved Sandstrom wrote:
> On 10/16/2013 07:55 AM, lipska the kat wrote:
>> On 16/10/13 11:42, Donkey Hottie wrote:
>>> 16.10.2013 11:45, 1connu kirjoitti:
>>>> lipska the kat a écrit :
>>>>> On 09/10/13 21:41, art.marole@gmail.com wrote:
>>>>>> Hi all,
>>>>>>
>>>>>> I would like to prepare typical Java EE solution (EJB, JSF, JPA) now,
>>>>>> and I have quite stupid problem related with domain objects. I want
>>>>>> to create Person class as persistence class, but I don't know in
>>>>>> which module I shoud put this class: Business Layer (ejb module) or
>>>>>> in Integration Layer (jar module).
>>>>>>
>>>>>> Person class is more related with logic, so I should put it in ejb
>>>>>> module, but in this case I will need to create dependency relation
>>>>>> from Integration Layer to Business Layer as Integration Layer need to
>>>>>> know what object should store in DB (ex. void store(Person
>>>>>> object){};).
>>>>>> Put Person class in Integration Layer will remove this unwanted
>>>>>> dependency, but I feel that Person class is more Business class then
>>>>>> Integration class.
>>>>>>
>>>>>> What do you think about it?
>>>>>
>>>>>
>>>>> A Person class ... A PERSON class., please, please, *please* don't
>>>>> have a Person class
>>>>>
>>>>> For the sake of whatever omnipotent deity you pray to, do not have a
>>>>> Person class
>>>>>
>>>>> *People* do not belong in a computer system.
>>>>>
>>>>> Why?
>>>>>
>>>>> Because it's a poor abstraction.
>>>>>
>>>>> How are you going to model consciousness, sentience, the 'mind', the
>>>>> 'soul'? let alone arms legs, eyes and teeth :-)
>>>>>
>>>>> You *do not need* a Person class.
>>>>>
>>>>> Let's put it another way
>>>>>
>>>>> The very fact that you are having problems deciding where your Person
>>>>> belongs should alert you to the fact that it is a poor abstraction.
>>>>>
>>>>> Try Account, or Record, or if you absolutely *must* User but not
>>>>> Person, you will end up regretting it.
>>>>>
>>>>> lipska
>>>>>
>>>>>
>>>> Why not ?
>>>> You can have an abstract class Person with minimum field (name, date
>>>> birth) and sub class depending of role of theses persons (ex User
>>>> (login/password) ) in your model.
>>>
>>> I see no problem with Person.
>>>
>>> I have interface IPerson, and implementations Person and ExternalPerson.
>>> I'm a lead programmer on an Identity Management solution.
>>
>> You *cannot* be serious ... IPerson groan :-(
>>
>> If you are developing an Identity Management system surely Identity
>> might be a better abstraction ... just a thought.
>
> Must admit, IPerson made me cringe too. In any context I can think of.
>
> I'm not as antagonistic to the Person concept as an implementation.
> Despite what you said, all a coder needs to do with that concept is
> model just what they need. If I was going to model a pedestrian mobility
> study using some random algorithms, I might well call my discrete
> construct a Person.

What about ... oh I don't know, Pedestrian?
It kind of says what it is, doesn't it?

What about guide dogs?

You say

"all a coder needs to do with that [Person] concept is model just what 
they need"

Well this is the point isn't it?

To me, saying that you just model what you need implies that the 
component you are modeling could be made less complex by reducing the 
concept to a more limiting name. Person could mean anything, Pedestrian 
means Pedestrian, nothing more.

Contrary to what some respondents to this thread think, this is not a 
dogmatic or idealogical view on my part. I have proved to myself and 
others (so far) by practical application that all a Person class does is 
confuse the issue at hand.

It's all about focusing our imperfect minds on complex problems. Giving 
things suitably limiting names helps us to discard issues irrelevant to 
the current problem space.

Good to chat with you again

Lipska

-- 
Lipska the Kat©: Troll hunter, sandbox destroyer,
treacherous feline and farscape dreamer of Aeryn Sun
GNU/Linux user #560883 - http://www.linuxcounter.net

[toc] | [prev] | [next] | [standalone]

#26347

On 10/16/2013 06:42 AM, Donkey Hottie wrote:
> 16.10.2013 11:45, 1connu kirjoitti:
>> lipska the kat a écrit :
>>> On 09/10/13 21:41, art.marole@gmail.com wrote:
>>>> Hi all,
>>>>
>>>> I would like to prepare typical Java EE solution (EJB, JSF, JPA) now,
>>>> and I have quite stupid problem related with domain objects. I want
>>>> to create Person class as persistence class, but I don't know in
>>>> which module I shoud put this class: Business Layer (ejb module) or
>>>> in Integration Layer (jar module).
>>>>
>>>> Person class is more related with logic, so I should put it in ejb
>>>> module, but in this case I will need to create dependency relation
>>>> from Integration Layer to Business Layer as Integration Layer need to
>>>> know what object should store in DB (ex. void store(Person object){};).
>>>> Put Person class in Integration Layer will remove this unwanted
>>>> dependency, but I feel that Person class is more Business class then
>>>> Integration class.
>>>>
>>>> What do you think about it?
>>>
>>>
>>> A Person class ... A PERSON class., please, please, *please* don't
>>> have a Person class
>>>
>>> For the sake of whatever omnipotent deity you pray to, do not have a
>>> Person class
>>>
>>> *People* do not belong in a computer system.
>>>
>>> Why?
>>>
>>> Because it's a poor abstraction.
>>>
>>> How are you going to model consciousness, sentience, the 'mind', the
>>> 'soul'? let alone arms legs, eyes and teeth :-)
>>>
>>> You *do not need* a Person class.
>>>
>>> Let's put it another way
>>>
>>> The very fact that you are having problems deciding where your Person
>>> belongs should alert you to the fact that it is a poor abstraction.
>>>
>>> Try Account, or Record, or if you absolutely *must* User but not
>>> Person, you will end up regretting it.
>>>
>>> lipska
>>>
>>>
>> Why not ?
>> You can have an abstract class Person with minimum field (name, date
>> birth) and sub class depending of role of theses persons (ex User
>> (login/password) ) in your model.
>
> I see no problem with Person.
>
> I have interface IPerson, and implementations Person and ExternalPerson.
> I'm a lead programmer on an Identity Management solution.
>
>
>

Higgins 2.0 :)
<http://wiki.eclipse.org/Persona_Data_Model_2.0#Person_entities.2C_attributes.2C_links_and_contexts>

[toc] | [prev] | [next] | [standalone]

#26352

On 10/16/2013 08:19 PM, Jeff Higgins wrote:

> Higgins 2.0 :)
> <http://wiki.eclipse.org/Persona_Data_Model_2.0#Person_entities.2C_attributes.2C_links_and_contexts>
>
When owl:sameAs isn’t the Same: An Analysis of Identity Links on the 
Semantic Web
<http://events.linkeddata.org/ldow2010/papers/ldow2010_paper09.pdf>

[toc] | [prev] | [next] | [standalone]

#26328

On 16/10/13 09:45, 1connu wrote:
> lipska the kat a écrit :
>> On 09/10/13 21:41, art.marole@gmail.com wrote:
>>> Hi all,
>>>
>>> I would like to prepare typical Java EE solution (EJB, JSF, JPA) now,
>>> and I have quite stupid problem related with domain objects. I want
>>> to create Person class as persistence class, but I don't know in
>>> which module I shoud put this class: Business Layer (ejb module) or
>>> in Integration Layer (jar module).
>>>
>>> Person class is more related with logic, so I should put it in ejb
>>> module, but in this case I will need to create dependency relation
>>> from Integration Layer to Business Layer as Integration Layer need to
>>> know what object should store in DB (ex. void store(Person object){};).
>>> Put Person class in Integration Layer will remove this unwanted
>>> dependency, but I feel that Person class is more Business class then
>>> Integration class.
>>>
>>> What do you think about it?
>>
>>
>> A Person class ... A PERSON class., please, please, *please* don't
>> have a Person class
>>
>> For the sake of whatever omnipotent deity you pray to, do not have a
>> Person class
>>
>> *People* do not belong in a computer system.
>>
>> Why?
>>
>> Because it's a poor abstraction.
>>
>> How are you going to model consciousness, sentience, the 'mind', the
>> 'soul'? let alone arms legs, eyes and teeth :-)
>>
>> You *do not need* a Person class.
>>
>> Let's put it another way
>>
>> The very fact that you are having problems deciding where your Person
>> belongs should alert you to the fact that it is a poor abstraction.
>>
>> Try Account, or Record, or if you absolutely *must* User but not
>> Person, you will end up regretting it.
>>
>> lipska
>>
>>
> Why not ?
> You can have an abstract class Person with minimum field (name, date
> birth) and sub class depending of role of theses persons (ex User
> (login/password) ) in your model.

Well let's see.

A User is not a subclass of Person really is it?

A User uses the system, at the very least a user can be a human being or 
another computer so if you want to go down that road then User should be 
your base class. A User has credentials, whatever they may be.

Given the above, what does a Person class give you that (say) Account 
doesn't ... A User *has an* Account.

User says what it is, a User. It certainly *doesn't* imply that the User 
is a Person. (see above)
Account says what it is, a record of a Users interaction with the system 
for example.

When we think about a system, certainly when we are starting to design 
it, we need to use words that help us to partition the system into 
meaningful abstractions. User, Account, Session, Address etc etc.

Introducing Person into the model clouds our thinking. A Person is not a 
cohesive abstraction of some part of a system. It is a word we use to 
describe the other similar sentient entities we encounter in our version 
of reality. The word Person comes with a whole load of built in baggage 
that does nothing to assist us in deciding on the components of our system.

That is why the class Person is not a good abstraction and doesn't 
belong in our model.

-- 
Lipska the Kat©: Troll hunter, sandbox destroyer
and farscape dreamer of Aeryn Sun
GNU/Linux user #560883 - http://www.linuxcounter.net

[toc] | [prev] | [next] | [standalone]

#26333

"lipska the kat"  wrote in message 
news:Qsydne-OrL7R8sPPnZ2dnUVZ7sidnZ2d@bt.com...

> A Person class ... A PERSON class., please, please, *please* don't
> have a Person class

Gosh, this has to be the biggest load of bollocks I have seen in a long, 
long time!

Please tell me Oh Feline One, if your system deals with "people", how are 
you going to model them?

SentientBeingOTheHumanVariety? FamiliarPrimate?  HomoSpeciesPostErectus?

Anything but PERSON!

--
And loving it,

-Qu0ll (Rare, not extinct)
_________________________________________________
Qu0llSixFour@gmail.com
[Replace the "SixFour" with numbers to email me] 

[toc] | [prev] | [next] | [standalone]

#26335

On 16/10/13 12:24, Qu0ll wrote:
> "lipska the kat"  wrote in message
> news:Qsydne-OrL7R8sPPnZ2dnUVZ7sidnZ2d@bt.com...
>
>> A Person class ... A PERSON class., please, please, *please* don't
>> have a Person class
>
> Gosh, this has to be the biggest load of bollocks I have seen in a long,
> long time!

Really?, I find that hard to believe given the amount of bollocks 
appearing in this group :-)

It is a tricky concept to grasp but in the end most people who are 
capable of thinking rationally come around to my point of view.

> Please tell me Oh Feline One, if your system deals with "people", how
> are you going to model them?

Sigh ... my systems never *deal with people*, that's the point. You seem 
to have missed entirely. It is *never* the case that a Person class is 
required.

'People' and other actors interact with the system. I have never found a 
need to include a Person class in any system I have ever had anything to 
do with.

It's a fact, it is possible and it pays dividends. Open your mind, think 
about it ... it really does make sense.

-- 
Lipska the Kat©: Troll hunter, sandbox destroyer
and farscape dreamer of Aeryn Sun
GNU/Linux user #560883 - http://www.linuxcounter.net

[toc] | [prev] | [next] | [standalone]

#26349

On 10/16/2013 8:39 AM, lipska the kat wrote:
> On 16/10/13 12:24, Qu0ll wrote:
>> "lipska the kat"  wrote in message
>> news:Qsydne-OrL7R8sPPnZ2dnUVZ7sidnZ2d@bt.com...
>>
>>> A Person class ... A PERSON class., please, please, *please* don't
>>> have a Person class
>>
>> Gosh, this has to be the biggest load of bollocks I have seen in a long,
>> long time!
>
> Really?, I find that hard to believe given the amount of bollocks
> appearing in this group :-)
>
> It is a tricky concept to grasp but in the end most people who are
> capable of thinking rationally come around to my point of view.
>
>> Please tell me Oh Feline One, if your system deals with "people", how
>> are you going to model them?
>
> Sigh ... my systems never *deal with people*, that's the point. You seem
> to have missed entirely. It is *never* the case that a Person class is
> required.
>
> 'People' and other actors interact with the system. I have never found a
> need to include a Person class in any system I have ever had anything to
> do with.
>
> It's a fact, it is possible and it pays dividends. Open your mind, think
> about it ... it really does make sense.

I think that depends on the system in question.

And the fact that you have never had the need for a Person class does
not imply that no case exist where it is needed.

Example: in Denmark we have something called Central Person Register for
all danes. I find it very likely that they have a Person class, because
they deal with persons.

Arne

[toc] | [prev] | [next] | [standalone]

#26353

"Arne Vajhøj"  wrote in message 
news:525f31cc$0$297$14726298@news.sunsite.dk...

> Example: in Denmark we have something called Central Person Register for
> all danes. I find it very likely that they have a Person class, because
> they deal with persons.

Nah, I am sure it's called IntelligentBipedalMammal or something similar ;-)

--
And loving it,

-Qu0ll (Rare, not extinct)
_________________________________________________
Qu0llSixFour@gmail.com
[Replace the "SixFour" with numbers to email me] 

[toc] | [prev] | [next] | [standalone]

#26355

On 10/16/2013 08:39 PM, Arne Vajhøj wrote:
> On 10/16/2013 8:39 AM, lipska the kat wrote:
>> On 16/10/13 12:24, Qu0ll wrote:
>>> "lipska the kat"  wrote in message
>>> news:Qsydne-OrL7R8sPPnZ2dnUVZ7sidnZ2d@bt.com...
>>>
>>>> A Person class ... A PERSON class., please, please, *please* don't
>>>> have a Person class
>>>
>>> Gosh, this has to be the biggest load of bollocks I have seen in a long,
>>> long time!
>>
>> Really?, I find that hard to believe given the amount of bollocks
>> appearing in this group :-)
>>
>> It is a tricky concept to grasp but in the end most people who are
>> capable of thinking rationally come around to my point of view.
>>
>>> Please tell me Oh Feline One, if your system deals with "people", how
>>> are you going to model them?
>>
>> Sigh ... my systems never *deal with people*, that's the point. You seem
>> to have missed entirely. It is *never* the case that a Person class is
>> required.
>>
>> 'People' and other actors interact with the system. I have never found a
>> need to include a Person class in any system I have ever had anything to
>> do with.
>>
>> It's a fact, it is possible and it pays dividends. Open your mind, think
>> about it ... it really does make sense.
>
> I think that depends on the system in question.
>
> And the fact that you have never had the need for a Person class does
> not imply that no case exist where it is needed.
>
> Example: in Denmark we have something called Central Person Register for
> all danes. I find it very likely that they have a Person class, because
> they deal with persons.
>
> Arne
>
>
"When I use a word, it means just what I choose it to mean - neither 
more, nor less."

[toc] | [prev] | [next] | [standalone]

#26356

On 17/10/13 04:22, Jeff Higgins wrote:
> On 10/16/2013 08:39 PM, Arne Vajhøj wrote:
>> On 10/16/2013 8:39 AM, lipska the kat wrote:
>>> On 16/10/13 12:24, Qu0ll wrote:
>>>> "lipska the kat"  wrote in message
>>>> news:Qsydne-OrL7R8sPPnZ2dnUVZ7sidnZ2d@bt.com...
>>>>
>>>>> A Person class ... A PERSON class., please, please, *please* don't
>>>>> have a Person class
>>>>
>>>> Gosh, this has to be the biggest load of bollocks I have seen in a
>>>> long,
>>>> long time!
>>>
>>> Really?, I find that hard to believe given the amount of bollocks
>>> appearing in this group :-)
>>>
>>> It is a tricky concept to grasp but in the end most people who are
>>> capable of thinking rationally come around to my point of view.
>>>
>>>> Please tell me Oh Feline One, if your system deals with "people", how
>>>> are you going to model them?
>>>
>>> Sigh ... my systems never *deal with people*, that's the point. You seem
>>> to have missed entirely. It is *never* the case that a Person class is
>>> required.
>>>
>>> 'People' and other actors interact with the system. I have never found a
>>> need to include a Person class in any system I have ever had anything to
>>> do with.
>>>
>>> It's a fact, it is possible and it pays dividends. Open your mind, think
>>> about it ... it really does make sense.
>>
>> I think that depends on the system in question.
>>
>> And the fact that you have never had the need for a Person class does
>> not imply that no case exist where it is needed.
>>
>> Example: in Denmark we have something called Central Person Register for
>> all danes. I find it very likely that they have a Person class, because
>> they deal with persons.
>>
>> Arne
>>
>>
> "When I use a word, it means just what I choose it to mean - neither
> more, nor less."

I couldn't have put it better myself
And this is *exactly* my point

What *you* choose it to mean may not be what someone else chooses it to 
mean, this is why we need to step down a level and think about what 
exactly we *do* mean when we name a concept.

Anyone who thinks that the names we give our concepts are not important 
is ... deluded.

What does 'Employee' say to you?
Now what does 'Person' say to you?

Think about what these words mean to someone who may not be in your head 
but has to maintain your system.

IMHO the names we give things is the most important part of the system 
documentation


-- 
Lipska the Kat©: Troll hunter, sandbox destroyer
and farscape dreamer of Aeryn Sun
GNU/Linux user #560883 - http://www.linuxcounter.net

[toc] | [prev] | [next] | [standalone]

#26357

On 17/10/13 01:39, Arne Vajhøj wrote:
> On 10/16/2013 8:39 AM, lipska the kat wrote:
>> On 16/10/13 12:24, Qu0ll wrote:
>>> "lipska the kat"  wrote in message

<snip>

> I think that depends on the system in question.
>
> And the fact that you have never had the need for a Person class does
> not imply that no case exist where it is needed.

I'd bet the farm that I could come up with a better name for *any* 
concept, in any model.

> Example: in Denmark we have something called Central Person Register for
> all danes. I find it very likely that they have a Person class, because
> they deal with persons.

Good grief I hope not

Citizen, Resident, Fish Pie, *anything* but Person :-)


-- 
Lipska the Kat©: Troll hunter, sandbox destroyer
and farscape dreamer of Aeryn Sun
GNU/Linux user #560883 - http://www.linuxcounter.net

[toc] | [prev] | [next] | [standalone]

#26358

"lipska the kat"  wrote in message 
news:r7udnfKuf9XNBcLPnZ2dnUVZ7s-dnZ2d@bt.com...

> Good grief I hope not
>
> Citizen, Resident, Fish Pie, *anything* but Person :-)

Let's face it, there are no "people" in the world.  The world is inhabited 
by citizens, residents, fish pies, accounts, users and lots and lots of 
animals.

There is not one "person" on Earth!

--
And loving it,

-Qu0ll (Rare, not extinct)
_________________________________________________
Qu0llSixFour@gmail.com
[Replace the "SixFour" with numbers to email me] 

[toc] | [prev] | [next] | [standalone]

Page 2 of 3 — ← Prev page 1 [2] 3  Next page →

Back to top | Article view | comp.lang.java.programmer

csiph-web