Nick Hodges (CodeGear) wrote:
<snip>
 > Actually, Windows itself is Unicode based, so you are /now/ converting
 > from AnsiString to Unicode for every API call, and you won't have to do
 > that with UnicodeString.
<snip>

Ok.
Seamless converting is one thing
(even if it was being converted already people didn't know it -- that's a good thing. ),
requiring people to check and/or update all their code is another.

Why the heck is that so bad to require people
to review, update, or change change their sources
just because CodeGear decided to make the default string type
Unicode?
Think about it ...

Some things that pop into my head are:
(1)
Well some of us deal with "legacy code".
Do I really have to explain here that there are thousands of projects
out there that may have been put on the "back burner" for a few years
but may require an update later?
So now these applications will have to be "checked" or updated
in order to introduce changes?
Isn't that creating problems for other developers for no good reason?
( A potential answer to this problem:
So instead of upgrading these developers will probably JUST KEEP and use their old Delphi.)

(2)
Hey, what about all the code on the internet?
I'm not sure people realize the brain power required to comb-through all
sample code you download off the internet just because
it does "funny things" involving ansistrings.
So now I have to "check" other peoples' sources for compatibility problems?
( Hey, good luck finding those problems too. )

Yes, a compiler flag would've been very useful;
It would've solved a lot of issues I just mentioned.

UGGGGGGGGGGG!!!
This wasn't what Delphi was supposed to be about!!!
Old code was suppose to work with the new Delphi.
Now, it "may" or "may not".
So is that what it's about now?:
Making it easy for CodeGear to upgrade their software while at the same time
decreasing Delphi developers productivity in the real world?
I don't think so ...