[
Lists Home |
Date Index |
Thread Index
]
Bullard, Claude L (Len) wrote:
> If you define evolution as the accretion of new features, sure. If
> you mean adding new
> instances of those features, no. In any layered system, one must be
> careful to specify
> which layer is evolving vs one that is merely accreting.
>
> If you define evolution as 'change' one *might* say this is
> evolution. Amazon added categories
> so the addition of the categorical function is evolution (a new
> feature) but adding categories
> is extensibility.
>
> In a biological system, evolution is a feature that is inheritable.
> Comparing that to Amazon,
> those are not evolution unless adding categories or search to Amazon
> adds it to any
> descendant of Amazon or any system derived from Amazon. If Amazon
> cannot have
> decendants, it cannot evolve. In this sense, evolution is an
> observable process of
> populations as they adapt to their environment, shape their
> environment, and then
> adapt to those changes (mediated feedback). So my question to you
> is, what
> is the equivalent of Amazon.com genes? Adding categories could mean the
> information is evolving, but Amazon evolved only when the categorical
> function
> was added, and then it accretes categories.
>
> My position would be that XML evolution is in the addition or deletion
> of features
> of a schema (at any metalevel one cares to work) because it defines
> the system.
>
> So one might look at how instances acquire new elements and attributes
> that are
> then added to their schemas as definitions. For example,
> aggregation: why would
> one relax constraints such that composite documents become homogenous?
> One might do that if working on one product in isolation and fits HTML
> elements into
> say SVG applications.
>
> Is XML an evolution of SGML or simply an adaptation (it lost features, so
> maybe it is devolution)?
depends doesn't it - was losing most of our furry/hairy covering
evolution or devolution? sometimes evolution means losing things in a
tradeoff for an overall better position. one of the great lessons of
chess and warfare (although today warfare must progress without loss,
but the debate continues). there is any number of birds that have lost
the ability to fly in exchange for an advantage on the ground or in the
water.
as a counter to the "more categories isn't evolution" you could also
consider the giraffe - is the long neck evolutionary or not.
topographically it's the same as other mammal necks, 7 bones, etc - it's
just very long. evolution or extension ;) ?
i think they're both the same. one way or another the system is changing
to be better in a given situation. it may lose features, gain features,
or change them. it's different because for some reason it needs to be.
rick
>
> len
>
>
> *From:* Roger L. Costello [mailto:costello@mitre.org]
>
> Hi Folks,
>
> I am interested in hearing your thoughts on the differences
> between an information system that is extensible versus an
> information system that is evolvable.
>
> For example, suppose that Amazon.com gives users the ability to do
> a keyword search. Further, suppose that Amazon empowers its users
> to create new keywords (and map the new keywords to information at
> the Web site). Is this support for new keywords an example of
> extensibility or evolution? By adding new keywords has
> Amazon merely been extended, or has it evolved?
>
> Let's take another example. Suppose that Amazon gives users the
> ability to search by book category (e.g., Fiction, Non-Fiction,
> etc.) Further, suppose that Amazon empowers its users to create
> new categories (and map the categories to information at the Web
> site). Is this support for new categories an example of
> extensibility or evolution? By adding new categories has Amazon
> merely been extended, or has it evolved?
>
> If adding new keywords and adding new categories are merely
> examples of extension, then can you give an example
> of evolution? /Roger
>
begin:vcard
fn:Rick Marshall
n:Marshall;Rick
email;internet:rjm@zenucom.com
tel;cell:+61 411 287 530
x-mozilla-html:TRUE
version:2.1
end:vcard
|
