Board index » kylix » Re: How increase Kylix sales

2003-10-06 03:08:15 AM
kylix2
"JQP"wrote:
between systems/desktops goes beyond a combobox. And I think very few people
won't agree with you that the Windows interface is a good interface. It's
what is behind the interface and the company behind that. Is .NET really
better than WIN32? Only for Microsoft developers, we from Borland are used
to a good framework. But we _must_ port our applications to .NET in the next
years. And the users only notice the change because they have to buy a
faster computer again. Something like that is not going to happen soon on an
opensource system.
Peter
 

"Peter Agricola" < XXXX@XXXXX.COM >wrote in message
better or worse than the other ... just different. There is no good reason
not to pick one and make it the standard. The fact that debate continues
over this just amazes me.
The Win32 API will be around for a long time, just like the DOS API is still
around and available under Windows.
 

PA>Only when you go to automatic transmission. Not the other way.
PA>I can name very good reasons why my car has a hand-operated gearbox,
PA>but it is OT. People have also good reasons to choose a particular
PA>Desktop.
PA>There are many different cars on the road and you are used to that.
PA>I'm sure you don't want that changed.
But i never buy a car with driver place on the right side (in Russia all
drivers seats at left side, while law does not enforce this for imported
cars), or a car with manual gear-box that use different scheme of positions
that all other, or different order of pedals. IOW a car where driver actions
based mainly on reflexes differ from other majority.
PA>systems.
Of course not seen if you use only one action with it - MouseClick.
Examples from memory.
Edit box:
Action: Select all text in the edit box
Konqueror: Ctrl+A
Mozilla: Ctrl+A moves cursor to the beginning of text
GIMP: Ctrl+A moves cursor to the beginning of text
Action: Delete text from cursor to the beginning of text
Konqueror: Ctrl+W
Mozilla: Ctrl+W
GIMP: Ctrl+W deletes one "word" from the left of cursor
Action: Move cursor to the end of text
Konqueror: Ctrl+E, End
Mozilla: Ctrl+E, End
GIMP: End
Action: Move cursor through words
All use same keybinding but Mozilla's understanding what "word" is mean is
differ from Konqueror's and GIMP's. IDLE under IceWM also has it's own
understanding what is "word" different from others.
Just spend some time analyzing each of widgets and you will find that they
differ in many places.
---
Andrew V. Fionik, Papillon Systems, Unix Programmers Group
For reply use "ender" instead of "fionika" in e-mail.
 

{smallsort}

PA>[...] Is .NET really better than WIN32?
PA>Only for Microsoft developers, we from Borland are used to a good
PA>framework. But we _must_ port our applications to .NET in the next
PA>years.
PA>And the users only notice the change because they have to buy a
PA>faster computer again. Something like that is not going to happen soon
PA>on an opensource system.
I not very much familiar with NET (just started learning), but as i
understand it offer many good features for programmer. Also... who said that
you _must_ port your applications to .NET. I'm not going to port any in near
future. Win32 will be around few years so it is good chance to make
refactoring for your applications and port it on the NET only when
necessary.
 

"JQP" < XXXX@XXXXX.COM >wrote in news: XXXX@XXXXX.COM :
can be round or a square. This does not change anything in the behavior
of the checkbox. If it contains a check-mark or a cross does not change
anything and depends on the taste of the user.
I have not noticed any significant difference between Linux-desktops and
Windows. The only thing I once noticed was, that under a specific
configuration, I had to place a window manually (the window did not just
pop up, but I just got a frame before it showed up and I had to place it
where I wanted before I could see it). This was quite unusual for me, but
I noticed improvements in a few cases (e.g. when I opnened the 3rd shell
and I did not want to have it placed over the others), and disadvantages
in other cases (e.g. if a application had to open a window many times).
It seems that the UIs get closer and closer with the default settings.
That is IMO a good development. But it is also fine that a user can
configure it in a completely different way if he /chooses/ to do it.
disadvantages in some details.
In the Win3.x days we had these flat white windows and buttons were just
rectangles. Borland had it's own 3d-look back then. M$ shows up with it's
ctrl3d later. So MS thought that it had to "improve" the UI. We did not
have standard-tree-views, but some applications had their own. MS
implemented "standard" trees later, because they improved the UI in some
cases.
If we really wanted to pick one standard, MS will never be allowed to
change it's GUIs because it would violate the standard. Or do you think
that MS is the standard and everybody else have to follow this standard,
even if there are better ways to handle a specific problem? Don't make it
better as long MS does not realize that it can do it better? Not a good
strategy.
I see no reason to do something different /just/ to do it differently,
but I see good reasons to /improve/ a GUI if it is neccessary. It's fine
to stick with the "standard" as long as possible, but it should not be a
religion.
We had multiple virtual desktops under Linux since the beginning, and MS
tries to sell it as innovation if it also starts doing it now.
Is it "unusual" for the average Windows-user not have tons of
applications on the same desktop? May be. But why should people who do
not want this mess not use the advantage of multiple desktops?
And no, I do not like GUIs which are completely different just to be
different. But I like it if they carefully improve the GUIs.
 

"Andreas Prucha" < XXXX@XXXXX.COM >wrote in message
Apple has a standard GUI, MS has a standard GUI, Linux doesn't. It
desperately needs one if it is ever to make a dent in the desktop. The
Linux standard can be different from the others, it doesn't matter.
What matters is consistency and compatibility. Any Linux based computer
should have the standard GUI available. Any new version of the standard
should be thoroughly tested to insure backward compatibility. As it stands,
the Open Source community doesn't waste much time on compatibility testing.
 

On 10/06/03 03:22 +0900, JQP wrote:
while combo boxes and that particular ilk of object is pretty
consistent, sliders and scroll bars can look and perform
differently than, say, a Windows user might expect. A few minutes
playing with the standard X implementation of Emacs proves to be
very entertaining, if not enlightening, when it comes to how
/interesting/ scroll bars can be.
JQP's arguments can be summed up as:
* Linux sucks rhino dung because of the dearth of
standardization, apps and quality development tools
End of that enlightening story. ;^)
One thing that I see in all this is that people like to shred
Linux usability based on the clueless user. Well, clueless users
(Hi, Mom!) have trouble with Windows and Macs, too. Put a clueful
user in front of pretty much ANY user interface and they'll
figure it out.
How long did it take you guys to learn KDE or Gnome well enough
to execute and close apps? Ten- or fif{*word*249} seconds? With an
average office user executing 2-to-5 apps -- period -- to earn
their living, how significant is this business of widgets and
whatnot anyway? Smart users will recognize that the widgets look
different but work much the same, and won't care about it. Dumb
users will simply remember that to do X they have to click Y when
Z happens. They won't recognize that the widgets are different
because they don't even grasp the concept of widgets.
In my experience, consistency within an application or suite of
applications counts for much more than whether a button looks
like a Motif button or a Windows button. Keep your OK, Cancel and
Apply buttons in the right order in a dialog -- this is a really
common APPLICATION issue -- to ensure your user doesn't click
Cancel out of locational habit when they really wanted to click OK.
And if you're really in the dark about what constitutes a
worthwhile user interface under Linux, Kylix developers should
feel free to borrow heavily from the Microsoft UI guidelines. The
document is well thought out and really is an intelligent
discussion on how to give your user a consistent look-and-feel
experience. Developers dealing with l10n and i18n issues would do
well to read Kano's "Developing International Software for
Windows 95 and Windows NT" -- dated but still very relevant in
its theory. Finally, Alan Cooper wrote a thought-provoking book:
"About Face: The Essentials of User Interface Design." You might
not agree with what he says in the book, but I guarantee he'll
give you food for thought.
trane
--
//------------------------------------------------------------
// Trane Francks XXXX@XXXXX.COM Tokyo, Japan
// Practice random kindness and senseless acts of beauty.
// mp3.com/trane_francks/
 

On 10/06/03 06:12 +0900, JQP wrote:
failed. RedHat has its own vision of how things should be.
UnitedLinux might have been a good idea, but RH wouldn't play
and, of course, SCO is involved. Frankly, I couldn't suggest
abiding by ANYthing SCO thinks is right. UnitedLinux is destined
to die a cold, lonely death.
Currently, RH is the de facto standard of Linux distributions.
They have the largest market presence and they promote Gnome. If
you want a current "standard," code your apps according to Gnome
coding conventions. Your Kylix apps won't look the same, but they
can be coded to behave in a very similar fashion. That is what
counts the most.
trane
--
//------------------------------------------------------------
// Trane Francks XXXX@XXXXX.COM Tokyo, Japan
// Practice random kindness and senseless acts of beauty.
// mp3.com/trane_francks/
 

Trane Francks < XXXX@XXXXX.COM >wrote in news: XXXX@XXXXX.COM :
TurboLinux, also UnitedLinux-partner seems to be quite pouplar in the
Asian region.
large as in the US.
they prefere it. And I prefere Gnome, for example. However, I do not see
any big difference between them. And in the latest release RH tried to
make both appear almost identically.
it's a better idea to avoid desktop-dependancies.
 

AP>We had multiple virtual desktops under Linux since the beginning,
AP>and MS tries to sell it as innovation
AP>if it also starts doing it now.
Where MS says that multiple desktops is innovation?
AP>Is it "unusual" for the average Windows-user not have tons of
AP>applications on the same desktop? May be.
No. It is unusual to have tons of apps on the same desktop. Why?
1. Windows apps tends to be multifunctional. So there are less different app
windows at same point of time.
2. Windows apps (especially heavy and multifunctional) start and shutdown
significantly faster than it's Linux counterparts, so there is no reason to
keep app in memory when it's needed. Just close it. On linux user usually
switch to other desktop and load other app in hope that switching to old app
will be faster than full load.
3. Windows graphics faster than linux'es, desktop redrawing take less time
and user rarely see screen flickering so common for Linux desktop. Same for
window resizing - for linux better to open window on full screen and switch
between desktops instead of windows. Less flickering.
AP>But why should people who do not want this mess not use the advantage
AP>of multiple desktops?
That so called "advantage" is purely solution of bunch of problems that was
created by lack of speed of some OS subsystems and way of thinking of
certain developers.
 

TF>Currently, RH is the de facto standard of Linux distributions.
TF>They have the largest market presence and they promote Gnome. If
TF>you want a current "standard," code your apps according to Gnome
TF>coding conventions. Your Kylix apps won't look the same, but they
TF>can be coded to behave in a very similar fashion. That is what
TF>counts the most.
How interesting. In Russia RedHat pretty rare. There are Linux Mandrake or
local linux distributors like ALT Linux or ASP Linux. I heard that SuSe is
very popular in the Europe. So i cannot agree that RH "is the de facto
standard".
 

On 10/06/03 13:57 +0900, Ender wrote:
developer who codes on Win2k and seldom has fewer than 30
windows/apps open at any given time. That's an extreme, to be
sure, but it shows that for any statistic, an opposite and
equally valid statistic can be quoted. :)
fairly common for applications such as Office to run as a stub in
the startup folder or system tray. The application startup time
remains similar; it's just split between system startup/login and
the time that the user clicks on the application button. If half
the app is already in memory, of course the apparent startup will
appear faster to the user.
purposes. Besides, with a gig of RAM, I don't need to worry about
"wasting" memory. (FWIW, I installed that much memory to
facilitate dropout-free audio recording with Cakewalk under
Windows. I just happen to put it to good use under Linux, too.)
enjoy the multi-desktop paradigm. Even MS recognized that;
multi-desktop capabilities were brought to Windows NT as part of
the NT Resource Kit. It worked reasonably well, too.
trane
--
//------------------------------------------------------------
// Trane Francks XXXX@XXXXX.COM Tokyo, Japan
// Practice random kindness and senseless acts of beauty.
// mp3.com/trane_francks/
 

On 10/06/03 15:21 +0900, Trane Francks wrote:
so badly managed that it's often not possible to run more than 3
heavy-footprint applications regardless of how much physical
memory that system may have installed.
trane
--
//------------------------------------------------------------
// Trane Francks XXXX@XXXXX.COM Tokyo, Japan
// Practice random kindness and senseless acts of beauty.
// mp3.com/trane_francks/
 

"Ender" < XXXX@XXXXX.COM >wrote in
Ok, what's running at the moment.
-Mozilla(3 times)
-Delphi
-Help&Manual
-VNC
-Some other strange remote-control
-Notepad (3 times)
-XNews
-Eudora
-Cmd (2 times)
-IBConsole
I think that's all.
 

Trane Francks wrote:
immediately switched to it and never rolled back. So all my words about
NT-class os'es.