[
Lists Home |
Date Index |
Thread Index
]
On Fri, 3 Jun 2005 7:52 pm, Didier PH Martin wrote:
> but at least programming and software development is so much better now..
>
> Didier:
> What????
We now have CVS and Google. They make it much better than before..
> Can you expand on this? Frankly some pieces (like server side) are
> indeed in better shape. On the other hand, the client side is still on the
> darker side....and in shape like a coach potato.
Depends who you talk to I guess.. I would question how much call
there is for traditional "servers" and traditional "clients".
It's quite insulting for Small/Medium businesses to always be lumped with
the "cheap clients", and the big business be fed an "expensive
server". That's only to impress the boss of the "enterprise". I personally
think people in the smaller businesses need to be given more respect.
These days they are often the clever ones who don't want to put up
with all the crap that goes along with working for a big company.
> We do not have yet a common object serialization. We have
> just a remote procedure call mechanism with SOAP but no ways to move
> objects. Moreover, if these objects are to be used in three different
> environments it is even less obvious.
> ..
> So, yes in all cases it's not easy to do. The question is now: Is it easier
> to do with off the shelf XML technologies than with something else?
I guess it would be. Off the shelf is usually the easiest.
> Do we really need the second tier? Can we have more responsive
> applications? People found the new acronym "AJAX" for more responsive
> applications. Did anyone played with google map? Is the experience better
> then with "a page at a time" kind of application? ( ref
> http://map.google.com ). Did anybody experimented with SQL/XML? Is the
> possibility to create objects from yemplates submitted by a client and
> filled by a server a way to resolve the impedance mismatch (at least for
> dynamic languages).
So many of questions... how to answer..
I think though the best answer is to just try them out for yourself.
Our software "rules" aren't really rules, they are conventions. Everybody
is free for themselves to try out different or new paradigms if they think
that it might help.
> Any real progress now is to occur on the client side.
Quite possibly.
> a) is it possible to move a collection of objects from server to clients
> instead of an object at a time. Is it possible to "check out" these objects
> and work on them even offline. Is XML useful to encode such objects?
Yes.
> b) is it possible to build a less complex and still robust architecture on
> a two tier system. On the one hand clients running applications on the
> other hand data/object/service sources.
Yes.
> c) can both a and b be viable with a low cost of ownership?
Yes.
> d) Can we improve the experience of users and have at least responsive
> applications like we got in a not so far past with visual basic,
> powerbuilder and other component based development/run-time environments?
Possibly. Just install more RAM and a faster hard-disk. :-)
> e) can we leverage the new generation of machine having more power than
> several mainframe of a not do far away past. Is using them as simple HTML
> decoder the best usage of so much power. Are these machine incapable to run
> applications? This is not what I see with more and more powerful games.
Yes.
> f) is there any alternative to rich environments like XAML? How do we
> expect a user to react in front of a web page and in front of a highly
> interactive XAML application?
Don't know. I've never used XAML.
> And so on and so forth....And have a good week end.
ok. you too.
David
--
Computergrid : The ones with the most connections win.
|
