Board index » delphi » Re: Late majority

2006-11-23 11:49:42 PM
delphi89
things like theming, AlignWithMargins, ActionMenus and thousands of other
small improvements!
 

"Dave Nottage [TeamB]" <XXXX@XXXXX.COM>skrev i meddelandet news:4564cdaa$XXXX@XXXXX.COM...
/ia
 

Markus Humm writes:
--
Nick Hodges
Delphi/C# Product Manager - CodeGear
blogs.borland.com/nickhodges
 

Peter writes:
to point out bad syntax.
--
Nick Hodges
Delphi/C# Product Manager - CodeGear
blogs.borland.com/nickhodges
 

Irakli writes:
VCL now has:
Margins
Padding
TFlowPanel
TGridPanel
New Memory Manager
Designer Guidelines
TCategoryButtons
TTrayIcon
Continued upgrades in the ActionBad area
TDockTabSet
Again, off the top of my head.
The VCL has been /far/ from "the same". We continue to add to and
enhance the VCL.
--
Nick Hodges
Delphi/C# Product Manager - CodeGear
blogs.borland.com/nickhodges
 

Nick Hodges (CodeGear) schrieb:
[snip]
(okay TrayIcon already was available with JEDI). Newer DB access drivers
(dbExpress).
Greetings
Markus
 

version of Delphi? I mean it is nice that you add these little
components, but rather you would do something about the whole
db-awarenes craziness that is been there since D1 and also about
all DataSet infrastructure. Does anyone on VCL team know the
word "interface"? Ever looked at other libraries? Swing maybe?
Where else do you need to inherit from such complex object as
DataSet just to get data binding to controls? Only in Delphi.
Not updated in 10 years - and that, mind you, is probably
*the most* important part of VCL.
Everything else can be done by any component vendor, but the
fundamental things cannot, because it is connected to the whole
infrastructure which you designed. Also, if you are improving
things with visual controls, why not do styling and theming
(and make it generic, so that it can be attached to any control)?
Why not update Grid/DBGrid controls? (compare to .Net grid)
I'm not even asking for Report designer... that is also been
10 years of bad choices...
This all doesn't matter that much anymore. WPF just blows your mind...
It's a huge step in the other direction and changes the whole
paradigm of how we design controls and interfaces.
The big question is: where have you been for 5 years? Were you
doing WPF type of visual library? You weren't? Well, then you're
done. it is like developing in Turbo Vision when Windows 3.1 arrived...
Microsoft has done a cruel thing by further attaching development
platform to the OS. They now have strategical advantage in all
areas. The only way to complete is to do anything/everything better.
Make programmers significantly more productive. That takes innovation.
That's why I am saying, stop following MS, think of what you
can do to beat them. Find a better direction. Invent a new language
Delphi++, make it so we can design applications faster better
with more visual aids...
Sorry, it just hurts to see how bad things are and on top of that
hearing this nice things like "we've added TrayIcon component"...
what a great achievement! in 2006... (I wrote my first one I think
in ~1998). This is just ridiculous... (You did make some progress
on compiler and added some features, I give you that ;)).
All you are saying in this thread is that generally you've been
spending resource on the *wrong* and obsolete things. This is really
too bad.
Cause I don't want to leave Delphi, and I will buy Delphi 2007 if it
is good enough to work with, but realistically, you must do
something to bring language and VCL to the today standard of
quality and features.
Michael
 

 

Michael Baytalsky writes:
The major thing the VCL needs is unicode. I personally think this one
enhancement would get a number of companies upgrading to the latest
version. Just think if all of those cool DevExpress/TMS (insert 3rd
party here) components supported unicode.
The problem is that Borland has been too conservative about approaching
the unicode subject. They seem to be looking for the perfect solution
which isn't always possible without a little pain.
Having something (that works for 50% of users), is sometimes better
than having nothing.
Also you need to remember that VCL.NET also gets these improvements and
with VCL.NET being the framework of choice for the Compact Framework
this is important. While many don't care about VCL.NET - some folks do.
Some also might ask why the vcl still doesn't have native support for
PNG images, why come controls still don't paint with a themed look and
feel under XP.
Most people that bag the VCL don't seem able (or willing) to respond to
questions such as - what would you like added to the VCL? What
enhancements do you look forward to in upcoming releases? What QC
entries have you made to try and influence the direction of the VCL.
Maybe your QC entries are under a different name because searching on
your surname reveals a donut of entries.
Also don't give me the throw away line that QC doesn't work - because I
assure you it does. Just because report XXX and report YYY aren't
opened doesn't mean that it doesn't work. All you have to look at is
reports AAA to WWW that have been opened/closed (as fixed).
--
Compact Framework for Delphi 2006: www.jed-software.com/cf.htm
QualityCentral Windows Client: www.jed-software.com/qc.htm
Visual Forms IDE Add In: www.jed-software.com/vf.htm
 

Craig writes:
correct, you're still likely writing {*word*99}py buggy code, by definition <g>.
--
Kristofer
 

 

Nick Hodges (CodeGear) writes:
 

at all. There simply must be a strategy on how to work with 3rd party
vendors to make sure that components are available and that there's
competition between vendors and everybody is happy ;).
Most of the users can work it out perfectly with Ansi and save trouble
of thinking how many characters string contains (remember, once you go
unicode, unless you do UCS4 you don't know the number of chars in string
unless you iterate it). Unicode is not an easy thing to support and
there are a lot more problems with VCL then this one - which is important,
but I doubt that it will result in more sales for CG.
Frankly, I'd develop a brand new VCL from scratch (almost) instead
of adding features to existing VCL.
is that Borland/CG fail to deliver "sense of innovation". The users don't
feel that they're working with cutting edge stuff. Microsoft has pushed
that boundary far away with .Net.
My feeling here is that everyone in this NG will cry out "release Unicode
and Win64", but once it is released, there will be very few sales still.
Why? I guess because there are still very few people here to make the
difference. In order to push sales you need hype. In order to produce
hype you need to invent something "cool" and promote it loudly (something
like "super-duper new rad approach to database development - create
your whole application automatically from database schema in 3 mouse clicks").
This is what we really need more then any particular feature.
On top of that we need IDE that is superior in speed & quality to VS.
Is that too much to ask? ;))
do .Net, you use C#. Choosing VCL.Net is a very strange approach to .Net.
I, personally, don't know anybody who even think about going this
route, while almost everybody now do something with .Net.
There's been so much criticism on how Delphi.Net sucks compared to
C# that I don't even want to go into it. I would rather simply use C#
or Chrome. Producing VCL.Net is like creating Turbo Vision for Windows...
imho.
Also, PNG is not a problem with free PNG libraries available for years.
But ImageList editor sucks big time and hasn't really been updated... ever.
reimplement db-awareness using interfaces" - this is just absurd.
Either CG will figure it out (cause it is quite obvious) or they
don't [have resource]. QC will not help there, I believe. They should
make their own decisions. QC is good for asking to add a property
X to class Y, or report a bug. IMHO.
My sort of emotional response to Nick was because of the "improvements"
he mentioned. They just sounded really pathetic to me ;(. I mean
it's sad. If this is what we are proud of... then we are just dead.
Just look at how many things MS has released over the last 5 years.
Tons of innovative technologies! What did Borland do? If not for
MS, we would have never even see generics in Delphi - and we didn't even
see them yet... Delphi lacks features tremendously. If they don't
see it, no QC will help. If they need QC to see how much IDE sucks,
then we should probably just run away - there's no hope in guiding
blind people with QC through the realms of software development.
The point of what I was saying is that VCL should really be
redesigned to today's standard. Any particular feature doesn't
bother me - I can always find replacement components that do it.
Michael
 

Michael Baytalsky writes:
very important to make unicode vcl, because business requires it and
it's already de facto standard, today everybody uses unicode in
non-english countries and if codegear wants to enlarge business then
they have to do it at any cost, for example in my country there is a
tend to leave delphi and move to VS.NET, why? one major reason for this
is lack of unicode with all other useful things already implemented by
competitors in their tools.
 

but I am not sure it is the one that could make sales go up... I might be
wrong, though. Also, this will require major overhaul of 3rd party components
so in many cases even though you have Unicode VCL, you won't be able to
use it because of lack of 3rd party components that support it.
I myself, mostly reside in Ukraine, but I do a lot of work for US customers
and I never experienced the pressure. In Ukraine we also have 3 codepages
(Eng, Rus, Ukr) so it could be tough without Unicode, but so far
people manage to get by with Ansi.
Also, if you really need Unicode, what stops you from using 3rd party
components to implement it? As far as I know, there are quite a few
such suites on the market. Am I wrong? I don't really have much experience
with Unicode, I must admit ;).
Michael