LinuxSTEP, GNUSTEP, GNU OpenStep, NeXTSTEP
DAVID L. JOHNSON
>
>How do you stop Bill Gates from taking over the world? Perhaps only
>Linus Torvalds plus Steve Jobs can do that.
>
I don't think Jobs' help is desireable, frankly.
>We need an OS which can rival Windwos NT in terms of functionality and
>availability. We then need a GUI which is sexier than Windows and as
>easily available.
Can you say X?
>
>Is plain Linux alone enough to do so? No, Linux hasn't got the sexy GUI
>Microsoft has.
Hardly true. There are several options, in X.
WINE - the Windows Emulator? Emulator can't be better
>than the original.
>
Huh? Wine is also not a gui.
>We need NeXTSTEP - the GUI which is sexier than Windows.
We need that like a fish needs a bicycle, to steal a phrase.
>Users of NeXTMail knew that email is perhaps the most important
>application in a GUI.
What?
--
David L. Johnson dl...@lehigh.edu or
Department of Mathematics dl...@chern.math.lehigh.edu
Lehigh University
14 E. Packer Avenue (610) 758-3759
Bethlehem, PA 18015-3174 (610) 828-3708
L S Ng
How do you stop Bill Gates from taking over the world? Perhaps only
Linus Torvalds plus Steve Jobs can do that.
We need an OS which can rival Windwos NT in terms of functionality and
availability. We then need a GUI which is sexier than Windows and as
easily available.
Is plain Linux alone enough to do so? No, Linux hasn't got the sexy GUI
Microsoft has. WINE - the Windows Emulator? Emulator can't be better
than the original.
We need NeXTSTEP - the GUI which is sexier than Windows. Can we have
NeXTSTEP on Linux? Yes, with GNUSTEP or GNU OpenStep we can. Find out
more from http://fvkma.tu-graz.ac.at/gnustep/gnustep.html. NeXTSTEP is
an alternative to Windows. GNU OpenStep is the NeXTSTEP for the masses.
A Chinese proverb goes 'to kill with a borrowed knife'. This is a game
between the GNU Generation vs. Microsoft. Linux is a knife borrowed from
the grand Unix tradition. NeXTSTEP, in the form of OpenStep, is another
knife borrowed from Steve Jobs.
Users of NeXTMail knew that email is perhaps the most important
application in a GUI. The Mail program in Windows 95 is a real killer
application. Only NeXTMail can be better. When Windows 95 is released,
expect to see the abandonment of plain old Unix Mail in favour of
Windows Mail. When this happen, other Unix users will realise how badly
they need something like NeXTMail.
Gisli Ottarsson PhD
Thank you, L S Ng, for this input and the informative http. NeXTSTEP
is indeed a fine system.
Please, no knee-jerk reaction from X lovers. Anyone intimately
familiar with NeXTSTEP will testify about its superiority over X.
Most of what L S Ng has to say is true (although I was never a fan of
NeXT mail). As much as I love my Linux/XFree86 system, I would
embarce Linux/GNU OpenStep in a minute (unless it turns out to be the
same resource hog as NeXTSTEP 486).
Good luck, GNU OpenStep developers.
Gisli Ottarsson
Darcy BROCKBANK
>In article <3kkaeg$c...@turner.ecs.soton.ac.uk>, ls...@ecs.soton.ac.uk (L S Ng) writes:
>>We then need a GUI which is sexier than Windows and as
>>easily available.
>Can you say X?
Can you say, "Ha ha ha ha ha ha ha...." ?
Great joke. I've not yet seen an X based interface which doesn't look
and feel, as well as behave like a 4 month student programming project.
- db
--
You smell of corduroy and lemon drops. -- Veruca Salt -- Baldric, you
wouldn't see a subtle plan if it painted itself purple and danced naked on
top of a harpsichord singing, "Subtle Plans Are Here Again" -- Atkinson --
The Lord loves a hanging, that's why he gave us necks! -- Hoek and Cat --
Robert Stephen Rodgers
L S Ng <ls...@ecs.soton.ac.uk> wrote:
>How do you stop Bill Gates from taking over the world? Perhaps only
>Linus Torvalds plus Steve Jobs can do that.
>
>We need an OS which can rival Windwos NT in terms of functionality and
>availability. We then need a GUI which is sexier than Windows and as
>easily available.
"We need apps."
[....]
>Users of NeXTMail knew that email is perhaps the most important
>application in a GUI. The Mail program in Windows 95 is a real killer
>application.
That's interesting. I think Exchange, as far as interface and functionality,
is pretty miserable. The only nice thing I can say about it isn't as bad as
IBM's Ultimail "Lite as a 50ton semi."
> Only NeXTMail can be better. When Windows 95 is released,
>expect to see the abandonment of plain old Unix Mail in favour of
>Windows Mail.
Not a chance.
>When this happen, other Unix users will realise how badly
>they need something like NeXTMail.
MIME.
novare
sam...@maggie.cs.mcgill.ca (Darcy BROCKBANK) writes:
> Great joke. I've not yet seen an X based interface which doesn't look
> and feel, as well as behave like a 4 month student programming project.
That might be a bit of an overstatement. Many commercial applications
are available under XWindows with duplicate interfaces.
I think that what is missing in Linux/UNIX is a good, free (money and
codewise) desktop metaphor (Finder, FileManager+ProgramManager,
what-have-you). Most of the way that NeXTSTEP looks has more to do with
many hours of design (visual design... art.) put into the interface
than with the internals underlying the display engines. X seems quite
fast, quite capable to me.
What Linux/UNIX/Mach/FSF needs is a new look and feel (Motif is OK, but
looks a little dated regardless. Perhaps a streamlining / redesign of
the Motif look along the lines of NeXTSTEP) that will give it a
distinct, fresh look separating it from other operating systems
combined with a ruthlessly effective system management/desktop metaphor
that borrows and builds on strong features from all current operating
systems.
The only thing that is really separating the common user from Linux is
the difficulties posed by the interface. I'm not saying take away
functionality, I'm saying add-on usability.
A bunch of good visual designers could do Free software a world of
good.
just my $0.02
Nathan Hand
: We need an OS which can rival Windwos NT in terms of functionality and
: availability. We then need a GUI which is sexier than Windows and as
: easily available.
A road-kill is sexier than windows.
: Is plain Linux alone enough to do so? No, Linux hasn't got the sexy GUI
: Microsoft has. WINE - the Windows Emulator? Emulator can't be better
: than the original.
Crash protection. Client-Server model. Free source. Coexist with
unix/X.
Oh yes... an emulator _can_ be better than the original.
--
"Ive never been so insulted in my life" +-----------------------
"Well, its early yet" +-----------------+ nat...@bin.anu.edu.au
----------------------+ I read the news reguarly -- sad, isnt it
Bogdan Urma
: the difficulties posed by the interface. I'm not saying take away
: functionality, I'm saying add-on usability.
I don't agree. The thing that's separating the common user from Linux
is Unix. The common user will neeed to learn Unix and how to administer
a Unix system, regardless of how pretty the GUI is. The common user does
not want to do this, which is understandable.
Bogdan
------
Bogdan Urma
Cornell University
Email: ba...@crux2.cit.cornell.edu
WWW: http://www.ruph.cornell.edu/burma/homepage.html
Dan Newcombe
>How do you stop Bill Gates from taking over the world? Perhaps only
>Linus Torvalds plus Steve Jobs can do that.
I don't think we need Steve to help :)
>We need an OS which can rival Windwos NT in terms of functionality and
>availability. We then need a GUI which is sexier than Windows and as
>easily available.
Sexier than Windows should be easier to do. I've shat things sexier
than windows. Freely availiable is a different story. I'm have on my
desk an add for Visual C++ 2.1 from MS. It uses the OS/2, er, uh, Windows95
Interface, and that one is starting to look nice.
Trying to develop a GUI for X will give you a few headaches. First, you have
to decide on a look and feel, and then implement it. This has been tried
several times (Andrew, XView/Openlook, Motif, Athena). No one seems to really
like what has been done. It's odd that a lot of apps are based on Athena,
even though the docs say it's only a sample. Motif costs, so we can rule
that one out. There was a project called Notif which was to be a free-Motif,
but it seems to have sizzled. Tcl/Tk, while good, is script based, and we
need something easily called from C/C++.
>Is plain Linux alone enough to do so? No, Linux hasn't got the sexy GUI
>Microsoft has. WINE - the Windows Emulator? Emulator can't be better
>than the original.
Wine will be better than Windows if for one reason only - being able to
run an app on one machine and have the output on another, using X. But alas,
it will probably never support the one Windows app I use : Cakewalk (for
MIDI).
>We need NeXTSTEP - the GUI which is sexier than Windows. Can we have
>NeXTSTEP on Linux? Yes, with GNUSTEP or GNU OpenStep we can. Find out
>more from http://fvkma.tu-graz.ac.at/gnustep/gnustep.html. NeXTSTEP is
>an alternative to Windows. GNU OpenStep is the NeXTSTEP for the masses.
I think right now it requires Motif to run. I read that it is in the plans
to get rid of that requirement, but that'll be a while, I'm sure.
>Users of NeXTMail knew that email is perhaps the most important
>application in a GUI. The Mail program in Windows 95 is a real killer
>application. Only NeXTMail can be better. When Windows 95 is released,
>expect to see the abandonment of plain old Unix Mail in favour of
>Windows Mail. When this happen, other Unix users will realise how badly
>they need something like NeXTMail.
That is the most off-the-wall idea I've ever heard. I don't think people
will flock to any OS just because the mail program kicks ass. What needs
to kick ass in this day and age is application interoperability.
-Dan
--
Dan Newcombe newc...@aa.csc.peachnet.edu
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
"And the man in the mirror has sad eyes." -Marillion
Ben A Lindstrom
: : The only thing that is really separating the common user from Linux is
: : the difficulties posed by the interface. I'm not saying take away
: : functionality, I'm saying add-on usability.
: I don't agree. The thing that's separating the common user from Linux
: is Unix. The common user will neeed to learn Unix and how to administer
: a Unix system, regardless of how pretty the GUI is. The common user does
: not want to do this, which is understandable.
Sorry, can we say Bull? There is no reason that a CORRECTLY managed UNIX
bo should need much administering...and if it's in an office type
enviroment the user should not NEED know any of the dirty details. I do
admit that do a lot of the system management from a "pretty GUI"
interface is a nicer, but not really needed. (Shell scripts and a bit of
spit and blood is just as effictive)
Doug DeJulio
Bogdan Urma <ba...@crux2.cit.cornell.edu> wrote:
>: The only thing that is really separating the common user from Linux is
>: the difficulties posed by the interface. I'm not saying take away
>: functionality, I'm saying add-on usability.
>
> I don't agree. The thing that's separating the common user from Linux
>is Unix. The common user will neeed to learn Unix and how to administer
>a Unix system, regardless of how pretty the GUI is. The common user does
>not want to do this, which is understandable.
In general, you're correct. I've owned a NeXTstep box though -- it's
the computer I used before switching to Linux.
I bought it because I wanted a Unix workstation, but when I got it, I
found out that the GUI was *so* good that even a braindead mehum could
administer the thing (in a one-user standalone environment with the OS
already installed, which was how it was shipped).
Slap the entire NeXTstep GUI on to Linux, and you've got something
that Joe and Jane Doe can use.
A lot of hackers will keep using X (not me), but for ordinary mehums
it's far too complex. For a hacker, a customizable configurable
interface is good, but for the genral public customizable is *BAD*.
Have a design professional who knows more about UI design than most
other people create a generic UI specification. Create a toolkit that
lets people build apps using that UI specification. Use a proprietary
windowing system, *don't* use a windowing system like X, or people
will install all sorts of little utilities and gidgets that screw up
your decreed-from-on-high consistency. While hackers may rebel at the
thought of such a system, it's better for non-hackers.
--
Doug DeJulio | R$+@$=W <-- sendmail.cf file
mailto:dd...@pitt.edu | {$/{{.+ <-- modem noise
http://www.pitt.edu/~ddj/ | !@#!@@! <-- Mr. Dithers swearing
Sergei Naumov
> Great joke. I've not yet seen an X based interface which doesn't look
> and feel, as well as behave like a 4 month student programming project.
That's not true. The look depends on a window manager. Get Fvwm and
configure it properly.
--
++++++++++++ http://sunsite.oit.unc.edu/sergei/Me/Serge.html ++++++++++
+ Sergei O. Naoumov se...@envy.astro.unc.edu tel: (919)962-3998 +
+Department of Physics & Astronomy, UNC-CH, Chapel Hill, NC 27599, USA+
++++++++++++++++++ !! 8 bit (Cyrillic) mail accepted !!++++++++++++++++
Michael Shandony
Darcy BROCKBANK <sam...@maggie.cs.mcgill.ca> wrote:
>Can you say, "Ha ha ha ha ha ha ha...." ?
>
>Great joke. I've not yet seen an X based interface which doesn't look
>and feel, as well as behave like a 4 month student programming project.
100% true. I have used X and Motif everyday for many years now and except
for Framemaker/Framebuilder every single program looks exactly as Darcy
described.
X based interfaces look unprofessional and sloppy.
=====================================================================
Mike Shandony | Telephone: (214) 684-7303
Bell-Northern Research, Inc. | BNR/NT Internal: (ESN) 444-7303
2201 Lakeside Blvd. MS D0307 | Fax: (214) 684-3748
Richardson, TX 75082-4399 | Internet: vanh...@bnr.ca
=====================================================================
cloister bell
>In article <SAMURAI.95...@maggie.cs.mcgill.ca>
>sam...@maggie.cs.mcgill.ca (Darcy BROCKBANK) writes:
>> Great joke. I've not yet seen an X based interface which doesn't look
>> and feel, as well as behave like a 4 month student programming project.
then you must not have tried out many serious x apps.
>That might be a bit of an overstatement. Many commercial applications
>are available under XWindows with duplicate interfaces.
e.g. mosaic and netscape.
>I think that what is missing in Linux/UNIX is a good, free (money and
>codewise) desktop metaphor (Finder, FileManager+ProgramManager,
>what-have-you). Most of the way that NeXTSTEP looks has more to do with
>many hours of design (visual design... art.) put into the interface
>than with the internals underlying the display engines. X seems quite
>fast, quite capable to me.
that sort of relates to what i think the underlying source of this debate is.
the mac os, windows, and nextstep have integrated the windowing system into the
os (ignore the whole debate about whether windows is or is not actually an os
for now, if you please). in doing so, they've made it pretty easy for users to
do things like changing fonts and colors, at the expense of a lot of
development time and complexity of the resulting os. however, in doing so
they've also imposed a lot of decisions on the users. a window on a mac looks
more or less one way in all apps and there isn't a damned thing the user can
really do about it, because it's part of the os.
the x guys decided they didn't want to make those kind of decisions for people,
so they have a system with lots of way powerful api's but none of the snazzy
user-oriented features like you have in windows, mac, nextstep. this means
that it's up to the window-manager writers to create that sort of
functionality. of course, that's a lot of work, so who's to blame for x's
difficulty of configuration? personally, i prefer the freedom to say "i hate
all the existing window managers, i'll go write my own", so my feeling is that
the blame lies with lazy window-manager writers (although i realize that it's a
hell of a lot of effort to write a window manager as it is).
it's a tradeoff that apple, microsoft, and next have made one way and x has
made the other way. however, i think that x has chosen the more flexible side
of the tradeoff.
(a note: i don't know if it's possible to replace the windowing system in
nextstep with one's own. if it is, i'd love to be corrected.)
Erik Fortune
In article <3kn131$p...@crchh78b.bnr.ca>, vanh...@bnr.ca writes:
> 100% true. I have used X and Motif everyday for many years now and except
> for Framemaker/Framebuilder every single program looks exactly as Darcy
> described.
>
> X based interfaces look unprofessional and sloppy.
I disagree. I'll take SGI's desktop tools over Windows any day.
-- Erik
+-------------------------------------+--------------------------------------+
| Erik Fortune | |
| Silicon Graphics International R&D | At Intel, Quality is Job 0.999897 |
| +1-415-390-1922 er...@sgi.com | |
+-------------------------------------+--------------------------------------+
All opinions are my own, so leave my employer out of this
Darcy BROCKBANK
>In article <SAMURAI.95...@maggie.cs.mcgill.ca>
>sam...@maggie.cs.mcgill.ca (Darcy BROCKBANK) writes:
>>Great joke. I've not yet seen an X based interface which doesn't look
>>and feel, as well as behave like a 4 month student programming project.
>That might be a bit of an overstatement. Many commercial applications
>are available under XWindows with duplicate interfaces.
OK, a 2 month student programming project :-).
cloister bell
> Tcl/Tk, while good, is script based, and we
>need something easily called from C/C++.
strictly speaking, this is incorrect. it is actually pretty easy to write a c
or c++ program that uses the tk widget set for its interface. yes, you still
write the interface as a tcl script, but you still get to do all the
interesting parts of the application in your language of choice. frankly,
having done one sample app this way, it was so easy to make a nice, powerful
interface that i'll probably never write another large-scale app in xlib
directly.
Ken-ichiro Aoki
Dan> se...@envy.astro.unc.edu (Sergei Naumov) wrote:
>> Once at a time Darcy BROCKBANK (sam...@maggie.cs.mcgill.ca) wrote:
>>> Great joke. I've not yet seen an X based interface which doesn't look
>>> and feel, as well as behave like a 4 month student programming project.
>>
>> That's not true. The look depends on a window manager. Get Fvwm and
>> configure it properly.
Dan> 99% incorrect. fvwm configured properly (whatever that means :) will
Dan> only control the window manager, the way borders and buttons on those
Dan> borders look and act. It doesn't matter how much tweaking I do to
Dan> fvwm, it still will not change the fact that Xaw is ugly.
....
Agreed; window managers don't change the way apps behave.
I use X (on sunos, HP-UX, linux) and Nextstep everyday. And I
have to say I like NS a lot. What I miss in the current
breed of X apps and X window managers is
1. drag and drop
2. gui
3. uniform interface (like print with every app.)
To elaborate:
1.
If I am using a drawing app, editing app and so on, it is very
intuitive to drop pictures, colors etc into the window. Also,
dropping files into applications is useful. i have
seen very few X apps that supports it in a natural way. In NS,
drag and drop is the norm.
2.
current breed of X apps usually don't have very nice GUI's. take
gnuplot for instance. it's nice and it's powerful, but to adjust
various styles (line widths, fonts, colors, etc.) it's just a
pain. In NS, all the plotting apps I know (free and shareware)
support GUI's for that (sliders, fontpanel, color panel,
etc). Things like colors and fonts etc: why do we have to adjust
that using a text editor!? It is something graphical and should
be adjusted graphically. (Like adjusting colors in fvwm, twm,
tvtwm, ugh.)
3.
It would be nice if people adhered to some interface guidelines
and also some minimal set of commands. In particular, being
able to print is very useful. xdvi, which is nice in many
aspects (nicer than NS's TeXView in some ways) doesn't support
print from the app, as far as I know. I realize some of this has
to do with NS's DPS. Sure, you can always open a terminal and
dvips or whatever, or invoke from emacs using acutex and use
compile or whatever (which is actually what I do) but this kinda
beats the purpose. It just uses X window as a terminal which can
open multiple windows, including graphics and just that. I
realize a lot of people use it that way, including myself, but I
expect something more from X wm. In NS, I do get a lot more.
----
After using NS for a few years, and getting used to it, my
expectations on window managers has gotten a lot higher.
I realize that X is a protocol so that in *principle* there is
nothing that prevents X apps/wm's interfaces being nicer. In
fact, possibly nicer than NS. However, considering how many
people use and develop X apps/wm's, it makes me wonder if there
is something inherently clumsy about the X protocol. I hope I am
wrong, since we seem to be stuck with it.
Though I love NS, I am resigned to the fact that I have to use X
at work; I really hope that X apps/wm's will get radically
better.
--
___Kenichiro Aoki (k...@th.phys.titech.ac.jp)
Dept.of Physics,Tokyo Inst. of Tech., Meguro-ku, Tokyo, Japan.
.... on the road oDo .......
Steve Dunham
: Thank you, L S Ng, for this input and the informative http. NeXTSTEP
: is indeed a fine system.
: Please, no knee-jerk reaction from X lovers. Anyone intimately
: familiar with NeXTSTEP will testify about its superiority over X.
: Most of what L S Ng has to say is true (although I was never a fan of
: NeXT mail). As much as I love my Linux/XFree86 system, I would
: embarce Linux/GNU OpenStep in a minute (unless it turns out to be the
: same resource hog as NeXTSTEP 486).
That would be Linux/XFree/GNUstep. GNUstep will use X as its
windowing system. (Don't worry - everything you think of when you
think X is actually the window manager and Xt/widget sets.)
Re: the superiority of OpenStep over X: you are comparing apples and
oranges here. X is a windowing system. OpenStep is a GUI. It seems
your statement is equivalent to: `A house is better than a cement
slab' - it's true, but you need the cement slab to build the house on;
I think that with its network transparency X is a good choice of
Windowing system to build a decent GUI on.
X is in need of a decent GUI - Maybe GNUstep will fill this rôle -
maybe it will collapse under its own gravity - we'll just have to wait
and see. (I hope it turns out well - I kinda like Objective C.)
An alternative to GNUstep could be Fresco - it will also be OO and
should be network transparent with the help of the CORBA stuff - so
decent inter-client communication will also be possible - the question
is whether anything will be built on this substrate.
Steve
dun...@gdl.msu.edu
Dan Newcombe
>Once at a time Darcy BROCKBANK (sam...@maggie.cs.mcgill.ca) wrote:
>> Great joke. I've not yet seen an X based interface which doesn't look
>> and feel, as well as behave like a 4 month student programming project.
>
>That's not true. The look depends on a window manager. Get Fvwm and
>configure it properly.
99% incorrect. fvwm configured properly (whatever that means :) will
only control the window manager, the way borders and buttons on those
borders look and act. It doesn't matter how much tweaking I do to
fvwm, it still will not change the fact that Xaw is ugly.
The point is that most of the freely availiable, and therefore the best
known X apps are based on top of Xaw or Xt, with a few on top of Xlib, for
those sadistic people :) Because most of them do not adhere to any
interface standard, it creates a headache for the user. While two apps may
have what looks to be a button, and these buttons could look exactly the
same, how you use or activate them can be radicaly different.
And because a lot of people use Xaw, which is only an example, they are
limited in what they can do.
Another problem is that people who design apps may not be that fluent in
what makes a good user interface. This is a complex topic, and I'm sure
IBM, MS, Apple, etc... have all spent tons of man hours hashing out how their
interfaces should look and act.
Also, I'm sure a lot of the people are more worried about getting the
underlying code that actually does stuff (drawing, morphing, etc...) working
than they are about making a pretty interface.
If you'd like to see some nice looking X apps, look at:
some of the Tcl/Tk stuff
xcoral
Andrew apps.
Orc
>Fundamentally I think that this the problem. The Linux community is so
>homegrown that most of its real users are computer science gurus who
>have trouble with thinking in a visually creative manner (Not saying
>that computer science gurus lack creativity). Apple, on the other hand,
>has the luxury of forcing their engineers to work closely with a bunch
>of visual people who may know very little about computers below the
>user level.
I disagree. It all comes down to "what you're most familiar
with" -- there are really very few people with the appropriate mix
of abilities and ambition who can push out new interfaces, and the
rest of us just stick to what we like. The Mac interface hasn't
made any drastic leaps in appearance since the first breadloaf came
out of Cupertino, but Apple is full of engineers who couldn't
imagine working with anything other than this interface.
If Linux is to have an interface that mimics some other machine,
I'd say that Windows is the way to go, just because of the user
base. I'd not object to a clone of the Mac user interface, but it
requires more underlying support (window layout) than a Windows-
like interface would.
____
david parsons \bi/ Me? I'll stick with a shell, thankyouverymuch.
\/
novare
newc...@aa.csc.peachnet.edu (Dan Newcombe) writes:
> Another problem is that people who design apps may not be that fluent in
> what makes a good user interface. This is a complex topic, and I'm sure
> IBM, MS, Apple, etc... have all spent tons of man hours hashing out how their
> interfaces should look and act.
Fundamentally I think that this the problem. The Linux community is so
homegrown that most of its real users are computer science gurus who
have trouble with thinking in a visually creative manner (Not saying
that computer science gurus lack creativity). Apple, on the other hand,
has the luxury of forcing their engineers to work closely with a bunch
of visual people who may know very little about computers below the
user level.
I think that a new interface for Linux could go a long way toward being
a killer app. I know that when old-school Mac users look over my
shoulder they say "dude. that screen looks cool with the little bevels
an' chit!" but turn tail as soon as I break out bash to do anything.
My suggestion would be to put togethor a team of visual people to work
with the various app teams (fvwm, xdm, etc.) and pretty the heck out of
em.
Project: Pretty Linux.
hmmm?
Dan Newcombe
>newc...@aa.csc.peachnet.edu (Dan Newcombe) writes:
>> Tcl/Tk, while good, is script based, and we
>>need something easily called from C/C++.
>strictly speaking, this is incorrect. it is actually pretty easy to write a c
>or c++ program that uses the tk widget set for its interface. yes, you still
>write the interface as a tcl script, but you still get to do all the
That is true...bit it still is not a call to a c function (or so I understand).
:)
Bora Akyol
and SUN X based interfaces, and I prefer the X based interfaces anyday.
Coming to why X based GUI looks clumsy, I think that's because most
sofware is put together by people that are enthusiasts and do not
have the time to make software look "pretty". I have been running
LINUX for 3 weeks now, and would not give it back in exchange of anything.
NEXTStep was a good interface, but again it might hog resources
which I belive are quite important unless you have at least 16 Megs
of RAM>
I believe that Mosaic for example is a good example of a nice Xbased
application ditto for Netscape. Anyway functionality of UNIX and X can not
be found on any other operating system in the world now.
Of course these are all my own humble opinions.
Bora Akyol
Stanford University
Grad Student.
Daniel Louis Potter
> That's not true. The look depends on a window manager. Get Fvwm and
> configure it properly.
Yes, you can configure the way the window borders and the underlying
interface works, but beacuse of the organiziation of X, you can't really
change the look and feel of each program. A lot of them were developed
using some really primitive looking Xt pieces (look at things like Xmag
and bitmap). I've seen NeXTStep and although it might be a pain to keep
running, the interface is incredibly nice. I think they could do away
with the file browser completely and adopt a mac/os2/win95 system, but it
was still nice. Personally, though, I believe that people want GNUStep
for the development tools. If it really does fill in with Application
Builder, it will be one of the great triumphs of free software.
BTW thought i'd mention, I liked next's interface so much after using it
a bit, I configured fvwm's "goodstuff" module to display a nice bar along
the right side, and a clock on the bottom. Once we got the NeXT.xpm icon
from the X distribution and wrote an XView program that said it was an
about box for next and put it at the top. The reaction by a few people
was quite fun =)
--------------------------------------------------------------------------
Daniel Potter dlp...@jove.acs.unt.edu
UNT Box 6909 Denton, TX 76203
--------------------------------------------------------------------------
Paul F. Kunz
> I don't agree. The thing that's separating the common user from
> Linux is Unix. The common user will neeed to learn Unix and how to
> administer a Unix system, regardless of how pretty the GUI is. The
> common user does not want to do this, which is understandable.
Well I know a lot of common users of NeXTSTEP (which is a UNIX OS)
that don't even know they are using UNIX. They even have to be
taught that that icon that represents their home directory is a 'home'
and not a 'house'. The only UNIX command that some of these users
know is 'tn3270' to get to the mainframe, yet they can successfully do
UNIX things like set permissions, copy files, create symbolic links,
navigate an NFS file system, etc. I think a good GUI goes a long
ways. These users are much happier with this GUI+UNIX then they
would be on platforms thought to be for the common user (DOS, Windows,
MacOS).
--
Paul F. Kunz Paul...@slac.stanford.edu (NeXT mail ok)
Stanford Linear Accelerator Center, Stanford University
Voice: (415) 926-2884 (NeXT) Fax: (415) 926-3587
Paul F. Kunz
> In article <SAMURAI.95...@maggie.cs.mcgill.ca>, Darcy
> BROCKBANK <sam...@maggie.cs.mcgill.ca> wrote:
>> Can you say, "Ha ha ha ha ha ha ha...." ?
>>
>> Great joke. I've not yet seen an X based interface which doesn't
>> look and feel, as well as behave like a 4 month student programming
>> project.
> 100% true. I have used X and Motif everyday for many years now and
> except for Framemaker/Framebuilder every single program looks
> exactly as Darcy described.
> X based interfaces look unprofessional and sloppy.
How true. Take for example, file viewers/browsers. They are
terrible on most X-based platforms. Yet, it is not the fault of X
windows. As a conter example, take the FileViewer in the examples of
the objcX distribution (the beginnings of GNUStep). It is a X based
application that has an interface like the FileViewer under the
NeXTSTEP Workspace manager. Another example is a file selection
panel. The one that comes with Motif is terrrible. While the one in
objcX is like the NeXTSTEP OpenPanel and SavePanel. Both are X based,
but what a difference in easy of use.
Linux plus GNUStep can be a great and professionally looking
combination and yet be based on X Windows.
Kristian Köhntopp
> That might be a bit of an overstatement. Many commercial
> applications are available under XWindows with duplicate
> interfaces.
... and they don't integrate.
Your commercial X application with nice, flashy interface does
not match in look and feel and in communication capabilities
with the apps delivered with the systems nor does it match
other applications from other vendors. At my DEC Alpha I had to
explain to my users why there are two types of scroll bars
(Motif and Athena) that are handled completely different. Same
goes for the closing of windows (Window close button within the
app window and hitting close in the Motif title bar terminates
the app vs. window close button in the title bar minimizes or
hides the window) and several other things (for example
filename dialog boxes...)
No dropping of images or text documents from newsreaders into
file managers, no standard format for editable formatted text
with several fonts, no standard consumer scale access to
scaleable fonts, no useable defaults editor (in an environment
where defaults are *everything*), and yes, in some applications
not even cut and paste of simple plain text: in short no
nothing at all.
Then look at Nextstep again.
I know of several "no keyboards please, I'm managment" types
that are actually using Nextstep for several years now (and
yes, they do real work with it) and they have yet to see a
terminal emulator window.
> I think that what is missing in Linux/UNIX is a good, free
> (money and codewise) desktop metaphor (Finder,
> FileManager+ProgramManager, what-have-you).
This is of course missing, too.
> Perhaps a
> streamlining / redesign of the Motif look along the lines of
> NeXTSTEP) that will give it a distinct, fresh look separating
> it from other operating systems combined with a ruthlessly
> effective system management/desktop metaphor that borrows and
> builds on strong features from all current operating systems.
But what use is this new and fresh look, when every other
application comes back with Athena scroll bars and Buttons?
Kristian
--
Kristian Köhntopp, Harmsstraße 98, 24114 Kiel, +49 431 676689
"Anspruchsvolle 'Poesie' auf CD: Blusen Wunder
Windowsoberfläche supereinfach mit Cheftaste und Lupenfunktion!
Die Serie für Kenner zum Sammeln. Kein Platz auf der Festplatte notwendig."
-- ESCOM-Katalog EXTRA 3.5
Doug DeJulio
Paul F. Kunz <Paul...@slac.stanford.edu> wrote:
> How true. Take for example, file viewers/browsers. They are
>terrible on most X-based platforms. Yet, it is not the fault of X
>windows. As a conter example, take the FileViewer in the examples of
>the objcX distribution (the beginnings of GNUStep).
I can't. The current objcX GNUstep stuff requies Motif to compile,
and my school's site license doesn't cover a student's home system.
I'm not sure I can justify spending something like $200 for Motif if
all I'm going to do with it is temporarily help debug objcX, when
they're planning on removing the Motif requirement at some point
(when?).
Could someone compile up the existing GNUstep applications, statically
linked to objcX and Motif, and stick 'em on an FTP site? I'd like to
look at 'em.
(Also, if there's a vendor out there who'll sell Motif libraries to a
student dirt cheap, contact me...)
Matthew Harrison
>
> 100% true. I have used X and Motif everyday for many years now and except
> for Framemaker/Framebuilder every single program looks exactly as Darcy
> described.
So that's 95% true, then :?)
_ \\ Matthew Paul Harrison, BioRad Microscience UK
\> \\ Ph (+44)(0)442 232 552, Fax 234 434
/\/\-\ \\ Email matt...@bio-rad.demon.co.uk
L S Ng
>In article <3kkaeg$c...@turner.ecs.soton.ac.uk>, ls...@ecs.soton.ac.uk (L S Ng) writes:
>>
>>How do you stop Bill Gates from taking over the world? Perhaps only
>>Linus Torvalds plus Steve Jobs can do that.
>>
>I don't think Jobs' help is desireable, frankly.
I am speaking metaphorically. Steve Jobs lend his NeXTSTEP to the public
in the form of OpenStep. So GNU OpenStep in a sense is Steve Jobs giving
a hand to the other Unix communities to stop Bill Gates. Linux plus
OpenStep is what I mean by 'Linus Torvalds plus Steve Jobs'.
>>We need an OS which can rival Windwos NT in terms of functionality and
>>availability. We then need a GUI which is sexier than Windows and as
>>easily available.
>Can you say X?
If you look into the http page for GNU OpenStep (repeated here again,
http://fvkma.tu-graz.ac.at/gnustep/gnustep.html), you will find that the
GNUStep is implemented on top of Xlib. It is still X. But not the X with
the plain and rather ugly *wm.
This is not a war between X and GNUStep. I rather think of GNUStep as a
replacement of Motif.
Tom Gall
|>
|> In article <3kn131$p...@crchh78b.bnr.ca>, vanh...@bnr.ca writes:
|> > 100% true. I have used X and Motif everyday for many years now and except
|> > for Framemaker/Framebuilder every single program looks exactly as Darcy
|> > described.
|> >
|> > X based interfaces look unprofessional and sloppy.
|>
|> I disagree. I'll take SGI's desktop tools over Windows any day.
SGI's desktop I think is probably the one exception when it comes to a really
NICE implementation of X/Motif.
Beyound that, the CDE desktops do seem to give Motif a bit of a facelift. Still
it's no where as nice as NeXTSTEP.
|> -- Erik
+-------------------------------------+--------------------------------------+
|> | Erik Fortune |
|> | Silicon Graphics International R&D | At Intel, Quality is Job 0.999897
--
Hakuna Matata!
Tom
#include <std-disclaimer.h>
|o| Tom Gall "Where's the ka-boom? There was supposed to be |o|
|o| IBM Rochester an earth shattering ka-boom!" -Marvin Martian |o|
|o| tom_...@vnet.ibm.com (NeXTMail -- Sure!) |o|
Martin Michlmayr
> I would embarce Linux/GNU OpenStep in a minute (unless it turns out to be
> the same resource hog as NeXTSTEP 486).
Well, I think so.... With the DPS (Display PostScript) version (*) and color,
you will come about the same requirement.
(*) not out yet.
> Good luck, GNU OpenStep developers.
Thank you very much! We will need it!
--
Martin Michlmayr | t...@tci002.uibk.ac.at | t...@fvkma.tu-graz.ac.at
GNU OpenStep Development Team, Manager of the Documentation Department
http://fvkma.tu-graz.ac.at/gnustep/gnustep.html
Darcy BROCKBANK
>Once at a time Darcy BROCKBANK (sam...@maggie.cs.mcgill.ca) wrote:
>>Great joke. I've not yet seen an X based interface which doesn't look
>>and feel, as well as behave like a 4 month student programming project.
>That's not true. The look depends on a window manager. Get Fvwm and
>configure it properly.
What's not true? My subjective statement/opinion? It's very nature,
being subjective, means that you can't just say it's true or false. You
just have to agree or disagree :-).
Yes, I've used pro X implementations, andwindow managers, etc., and this
is what I'm talking about, as well as the general design and
implementation of apps for X based stuff. This is why I said "X based
interfaces" and not X.
I've tried quite a few, and my statement holds.
Dan Newcombe
>In article <3kn5qm$j...@news-read-1.peachnet.edu>,
>Dan Newcombe <newc...@aa.csc.peachnet.edu> wrote:
>>that one out. There was a project called Notif which was to be a free-Motif,
>>but it seems to have sizzled.
>I think you mean "fizzled?" Had it "sizzled," I think that would typically
>mean that it would have gotten released...
Eek!!! What a typo. Yes...that is what I meant.
>>Tcl/Tk, while good, is script based, and we
>>need something easily called from C/C++.
>a) You *can* call Tcl/Tk from C/C++. The last issue of Linux Journal had
>an article on this very thing.
As I have said in other articles, it's not a true C interface, but what
seemed to be a semi-awkward interface...though that's from reading the
article. Others have nicely told me to get a book on Tcl/Tk as it can
be amazing. I guess I'm sorta hesitant, as almost every Tcl/Tk app I've
ever gotten off the net has not worked in one way or another. And I'm
dealing with a simple slackware setup...so it's not like I compiled and
installed it myself.
>We've got Tcl/Tk, Perl/Tk, Scheme/Tk. Why *not* C/Tk? It can currently
>be implemented by pipelining out to Tcl, but C/Tk may already exist, and
>if not, it would certainly leverage the work already done on Tk.
Rexx/Tk...now that would be nice!!!
Larry Hastings
>application. [...] When Windows 95 is released,
>expect to see the abandonment of plain old Unix Mail in favour of
>Windows Mail.
You don't need to worry about Microsoft Exchange superceding
Unix mail. Microsoft Exchange, as far as mail goes, is just
a client; it speaks Microsoft Mail, CompuServe Mail, presumably
The Microsoft Network Mail, and Internet Mail. The latter
is standard old email, which can parse both UUENCODEd and
MIMEd attachments.
Whether Windows 95 eats everyone else's lunch remains to be
seen; but, at least in this category, you don't have to worry
about it establishing another bizarre proprietary standard.
--
larry hastings, the galactic funkster, funk...@hyperion.com
"SORCERER Hours: 10-12 2-4" --Gumby, "Goo & The Witch part 2"
<a href="http://www.hyperion.com/~funkster">My WWW homepage</a>
Des Herriott
>
> [ ... ]
> There was a project called Notif which was to be a free-Motif,
>but it seems to have sizzled.
Yes, but a new project, LessTif, has started. Check out this URL:
http://www.cs.uidaho.edu:8000/hungry/microshaft/
It's still very much a work in progress according to the authors, but it
seems that a lot of the work has already been done. So we may have a
free Motif-replacement pretty soon now, which is great news.
--
Des Herriott, Micro Focus, Newbury, UK / "Fashion is something so ugly it
d...@mfltd.co.uk / has to be changed every 15 minutes"
http://www.mfltd.co.uk/~dnh / -- Senser
Terry Wilcox
Erik Fortune <er...@westworld.engr.sgi.com> wrote:
>
>In article <3kn131$p...@crchh78b.bnr.ca>, vanh...@bnr.ca writes:
>> 100% true. I have used X and Motif everyday for many years now and except
>> for Framemaker/Framebuilder every single program looks exactly as Darcy
>> described.
>>
>> X based interfaces look unprofessional and sloppy.
>
>I disagree. I'll take SGI's desktop tools over Windows any day.
>
That's not saying much. I'm constantly surprised that people put
up with the Windows interface. That Program/File Manager thing
is unusable to most people. Every day I watch people drop back
to DOS to copy files.
Of course we all use X-Terms for everything in X. It would be
nice if the X-using workstation manufacturers could at least agree
on some common tools.
SGI might have some nice tools, but they're not part of X. Until
X has some standard and usable tools, it'll be too complicated.
Terry Wilcox
--
Terry Wilcox
Arcane Systems Ltd.
te...@arcane.com
Terry Wilcox
Ben A Lindstrom <mou...@netnet.net> wrote:
>
>Sorry, can we say Bull? There is no reason that a CORRECTLY managed UNIX
>bo should need much administering...and if it's in an office type
>enviroment the user should not NEED know any of the dirty details. I do
Who does the correct managing (aka administration) for this box that
doesn't need much administration? Somebody has to set it up to begin with.
I've got a couple Unix boxes at home, I administer a bunch at work.
They require a lot of administration at the beginning and a lot if
you want to make changes. Somebody has to do that work.
Our users do have to learn a lot of Unix because using X requires it.
They end up having to know about Window Managers and terminal types.
It never ends. The advantage is they learn something.
Michael J. Suzio
>: The only thing that is really separating the common user from Linux is
>: the difficulties posed by the interface. I'm not saying take away
>: functionality, I'm saying add-on usability.
> I don't agree. The thing that's separating the common user from Linux
>is Unix. The common user will neeed to learn Unix and how to administer
>a Unix system, regardless of how pretty the GUI is. The common user does
>not want to do this, which is understandable.
Why? As many people have pointed out in the past, MS Windows is fully as
complex to administrate as Unix is. I myself was rendered helpless this
weekend trying to get Windows sound drivers working for a friend's father
who had swapped out sound cards, but evidently still had bits of junk in
his .INI files that were still trying to use the old sound card
configuration. Ick, I ended up having to tell him "Reinstall Windows, I
cannot figure out where the error is, and there isn't an "uninstall"
option."
Now, this is just an example - I don't want some MS Solutions person
telling me what the correct answer was. The point is, I would have had to
spend some time learning more about Windows than just the basic interface
in order to solve this problem. In a similar amount of time, I am dead
sure I could teach a competant user how to administrate a Unix box for
personal use. It's not that hard, really.
Let's look at the issues:
1) Command-line interface. Dead simple. How is "ls" any harder than
"dir"? There are more options available, and you need to understand how
Unix wildcarding is different from DOS, but that's about it. So you give
the person a two page summary of Unix commands and maybe a page on how the
shell works and it's interaction with other programs. Scripting is another
issue, but that adds maybe two more typed pages.
2) GUI. X is a little less feature-rich than MS-Win, and I would not
argue with the assertion that the setup is more tedious for X (although
MS-Win setup is not piece of cake, either, esp. with video drivers).
However, it looks like this feature is improving, I understand better
XConfig programs are out now that can do most of this setup for you. Once
it's setup, X is easy to use and to customize.
3) User administration - simple. Either you give them a GUI tool to do
adding and deleting, *or* a simple set of scripts (every admin has at one
point in their life written one). After that, what else needs be done? I
have a pal who needs to use my computer, I give him an account and he's all
set. What beyond this needs to be done for personal use?
4) Machine tweaking. OK, this is beyond autoexec's and config.sys's. It's
at about the same level as .INI's however :-). Really, any system setup
files just need to be well-documented and mostly hidden from the user
needing to grovel through them. Even as an admin, I *hardly ever* mucked
with rc files or XDefaults. What I had worked, I didn't worry about the
rest. The biggest offender in the "obscure setup file" area is, of course,
sendmail. I have no answer for that one, but maybe we need a nice GUI for
that that would setup reasonable default settings. I have certainly seen
good templates within my company where the admin just needs to edit 5-6
lines of a "stock" sendmail.cf and they are all set.
Anyway... Unix *is* accessible, you just need to overcome people's fear of
the unknown and maybe concentrate a little on some nice tools to make life
simpler (if a little less flexible).
- Mike
PS: But Linus is right - apps first, admin tools second
L S Ng
>L S Ng <ls...@ecs.soton.ac.uk> wrote:
>>How do you stop Bill Gates from taking over the world? Perhaps only
>>Linus Torvalds plus Steve Jobs can do that.
>>
>>availability. We then need a GUI which is sexier than Windows and as
>>easily available.
In <3kl73v$9...@mocha.eng.umd.edu> rsro...@Glue.umd.edu (Robert Stephen Rodgers) writes:
>"We need apps."
Apps comes after a sexy GUI. When you have NeXTSTEP, making apps is a
snap!
If you have an ugly GUI, the apps will look ugly too.
Perhaps you should note that XMosaic was developed originally on
NeXTSTEP. This is a perfect example how GUI will affect apps
development.
Christopher B. Browne
>to decide on a look and feel, and then implement it. This has been tried
>several times (Andrew, XView/Openlook, Motif, Athena). No one seems to really
>like what has been done. It's odd that a lot of apps are based on Athena,
>even though the docs say it's only a sample. Motif costs, so we can rule
>that one out. There was a project called Notif which was to be a free-Motif,
>but it seems to have sizzled.
I think you mean "fizzled?" Had it "sizzled," I think that would typically
mean that it would have gotten released...
>Tcl/Tk, while good, is script based, and we
>need something easily called from C/C++.
Can I disagree gently from two directions:
a) You *can* call Tcl/Tk from C/C++. The last issue of Linux Journal had
an article on this very thing.
b) Not only can you call Tcl/Tk from C/C++, you can also call C/C++ from
Tcl/Tk. For an application where the interactions between application and
GUI are not all-pervasive it probably makes sense to split the app. into
independent pieces and integrate it using Tcl/Tk. Or use the UNIX shell
if that's more appropriate.
Putting this another way, applications that are operated basically through
menu structures (opening files, saving files, changing options, that sort
of thing, "run some process on the data") might be made more flexible and
robust by doing the GUI essentially independently of the application.
I'll admit that significant pieces of a WYSIWYG word processor have so many
co-dependencies between application and GUI that this would not be a great
approach in that case. (Other highly interactive GUI applications might
be similar.)
If the interface is fairly nice (which it does seem to be), and is fairly
popular (which it is), then I'd argue that it should be used as the basis
for further work. Or at least cloned for C/C++ purposes.
We've got Tcl/Tk, Perl/Tk, Scheme/Tk. Why *not* C/Tk? It can currently
be implemented by pipelining out to Tcl, but C/Tk may already exist, and
if not, it would certainly leverage the work already done on Tk.
--
Christopher Browne - cbb...@io.org
Fatal Error: Found [MS-Windows] System -> Repartitioning Disk for Linux...
Michael Shandony
Ken-ichiro Aoki <ao...@madonna.physics.ucla.edu> wrote:
>I realize that X is a protocol so that in *principle* there is
>nothing that prevents X apps/wm's interfaces being nicer. In
>fact, possibly nicer than NS. However, considering how many
>people use and develop X apps/wm's, it makes me wonder if there
>is something inherently clumsy about the X protocol. I hope I am
>wrong, since we seem to be stuck with it.
As a person who has done several projects for both school and work using X
Windows and Motif, this are the reasons that I feel that it is difficult to
make an X-based GUI look good.
X lets you configure every little stinkin' thing. It gives you almost
total control over what each GUI element will look and act like. While
this level of freedom is really nice in some ways, it is bad in others.
For example, because of this flexibility, it is very easy for a designer
to make his/her GUI look and act very differently from any else's or just
from project to project. Do I really care that a pushbutton can be armed
and/or activated? I only ever care when it's activated. I find choosing
colors to be very difficult, using GCs to be unintuitive, and placement of
widgets almost a black magic (even with a form widget). I could go on and
on, but I think you get the idea.
(BTW, my GUIs do look good. I spend a good deal of time of them, but they
never look the same from project to project.)
To the person who wrote that SGI's GUI looks good, I agree. I have had
the pleasure of programming GUIs on SGI's machines before and their
interface is nice AND very easy to program. It does what I mean, rather
than what I said. :-) What's even stranger is that their GUI is built on
top of X windows and I like it!
NEXTSTEP still has the best GUI by a long shot. This is in comparison to
everything else that I have seen.
=====================================================================
Mike Shandony | Telephone: (214) 684-7303
Bell-Northern Research, Inc. | BNR/NT Internal: (ESN) 444-7303
2201 Lakeside Blvd. MS D0307 | Fax: (214) 684-3748
Richardson, TX 75082-4399 | Internet: vanh...@bnr.ca
=====================================================================
Darcy BROCKBANK
>I believe that Mosaic for example is a good example of a nice Xbased
>application ditto for Netscape. Anyway functionality of UNIX and X can not
>be found on any other operating system in the world now.
>Of course these are all my own humble opinions.
I wonder if the fact that Mosaic's interface is appealing is that it was
designed and prototyped on NEXTSTEP, by a NEXTSTEP user/programmer?
Terry Wilcox
Sergei Naumov <se...@envy.astro.unc.edu> wrote:
>Once at a time Darcy BROCKBANK (sam...@maggie.cs.mcgill.ca) wrote:
>> Great joke. I've not yet seen an X based interface which doesn't look
>> and feel, as well as behave like a 4 month student programming project.
>
>That's not true. The look depends on a window manager. Get Fvwm and
>configure it properly.
I'm using Motif as I write this. The window frames all look alike, but
similarity ends there. Each application looks completely different.
None of the applications have a good feel to them. Unlike most
other systems, I don't feel comfortable in unfamiliar applications.
Learning the interface for one application doesn't help with other
applications.
The window manager doesn't make the application interfaces any better.
Darcy BROCKBANK
>In article <3kn131$p...@crchh78b.bnr.ca>, vanh...@bnr.ca writes:
>>100% true. I have used X and Motif everyday for many years now and except
>>for Framemaker/Framebuilder every single program looks exactly as Darcy
>>described.
>>
>>X based interfaces look unprofessional and sloppy.
>I disagree. I'll take SGI's desktop tools over Windows any day.
In dealing with two crappy things, it's appropriate to choose the lesser
of two evils :-).
Anyway, NEXTSTEP's UI beats all. I don't even bother arguing it anymore,
since only the Truly Religious seem to disagree. No sense in trying to
convert the Truly Religious...
- db
(A serious NEXTSTEP, Windows, Windows NT, SunOS, X (yes, and SGI on
occasion), vt100, and DOS user...)
Kennel
> In article <SAMURAI.95...@maggie.cs.mcgill.ca>,
> Darcy BROCKBANK <sam...@maggie.cs.mcgill.ca> wrote:
> >Can you say, "Ha ha ha ha ha ha ha...." ?
> >
> >and feel, as well as behave like a 4 month student programming project.
> 100% true. I have used X and Motif everyday for many years now and except
> for Framemaker/Framebuilder every single program looks exactly as Darcy
> described.
> X based interfaces look unprofessional and sloppy.
"Mechanism not Policy"
{the alpha and omega of X}
is
A Disaster.
In war, there is no substitute for victory.
In software, there is no substitute for quality.
Kennel
> Coming to why X based GUI looks clumsy, I think that's because most
> sofware is put together by people that are enthusiasts and do not
> have the time to make software look "pretty".
People programming Nextstep have even less time to make the software
look pretty but it happens anyway, somehow.
Michael Mellinger
Hi,
I'm a naive Windows programmer with a lot of money who wants to
program in X Windows. Could someone tell me how to write X apps? I'm
considering some these solutions, along with any others that anyone
else can offer.
Xlib
Xt
Tk
Galaxy
TeleUSE
XVT
Athena
Suit
InterViews
Fresco
Geez, there are sure a lot of ways to program in Xwindows. It's too
bad that 90% of all commerical software is written for the Windows
API(Win32 or MFC). Gosh, wouldn't it simply be easier to make these
APIs freely available under X so that we naive Windows programmers
could effortlessly port our many thousand applications to to run under
X Windows on those really neat Sun, SGI, HP, IBM, and Linux boxes.
-Mike
cloister bell
>cloi...@u.washington.edu (cloister bell) wrote:
>>newc...@aa.csc.peachnet.edu (Dan Newcombe) writes:
>>> Tcl/Tk, while good, is script based, and we
>>>need something easily called from C/C++.
>>strictly speaking, this is incorrect. it is actually pretty easy to write a c
>>or c++ program that uses the tk widget set for its interface. yes, you still
>>write the interface as a tcl script, but you still get to do all the
>That is true...bit it still is not a call to a c function (or so I understand).
>:)
i'm not quite sure what you mean by that, but i think it is. the way it works
is that you instantiate a tcl interpreter in your C application. then you
write a bunch of functions in your app and register them with the interpreter.
having done so, you're allowed to call your functions from the tk script that
runs your interface. so if you're writing a raytracer with a nice interface,
for example, you probably have a function in your app that recalculates the
image. you can hook that up to a button in the interface so that the user can
just click to call the function. on the other side, it's also possible to
invoke/evaluate tcl code directly from your c program by just passing it off to
the interpreter. i'd suggest getting a copy of ousterhout's tcl/tk book, as it
has a whole lot of stuff in it related to doing this.
--
+-------------------------------------------------+---------------------------+
|tactical nuclear sdi stealth nsafood signature. | cloi...@u.washington.edu |
+-------------------------------------------------+---------------------------+
Alain Knaff
: Hi,
: I'm a naive Windows programmer with a lot of money who wants to
^^^^^^^^^^^^^^^^^^^
: program in X Windows. Could someone tell me how to write X apps? I'm
: considering some these solutions, along with any others that anyone
: else can offer.
Motif :-)
Alain
Robert Rodgers
ls...@ecs.soton.ac.uk (L S Ng) wrote:
>In <3kl73v$9...@mocha.eng.umd.edu> rsro...@Glue.umd.edu (Robert Stephen Rodgers) writes:
>>"We need apps."
>
>Apps comes after a sexy GUI. When you have NeXTSTEP, making apps is a
>snap!
You are mistaken. If apps are going to come, they're going to come
whatever the interface is like. Even DOS had some good apps -- and it
didn't have a non-command interface *at all*. For NeXTSTEP, the apps
are simply not going to come.
>If you have an ugly GUI, the apps will look ugly too.
There's looking ugly (MS Windows 3.x) and there's being cruddy,
unarguably hideous and poorly designed (most X apps).
>Perhaps you should note that XMosaic was developed originally on
>NeXTSTEP. This is a perfect example how GUI will affect apps
>development.
This is an absolutely empty example.
Ken Stagg
[snip]
> In war, there is no substitute for victory.
>
> In software, there is no substitute for quality.
Sure there is: Marketing. Just ask Microsoft. They don't know anything
about quality, but...
-Ken
John M Dow
look nice that the system and applications were insipid and bug filled? A case of all
polish and no furniture.
john
Jeff Dege
: >>Tcl/Tk, while good, is script based, and we
: >>need something easily called from C/C++.
: >a) You *can* call Tcl/Tk from C/C++. The last issue of Linux Journal had
: >an article on this very thing.
: As I have said in other articles, it's not a true C interface, but what
: seemed to be a semi-awkward interface...though that's from reading the
: article. Others have nicely told me to get a book on Tcl/Tk as it can
: be amazing. I guess I'm sorta hesitant, as almost every Tcl/Tk app I've
: ever gotten off the net has not worked in one way or another. And I'm
: dealing with a simple slackware setup...so it's not like I compiled and
: installed it myself.
The recent article in The Linux Journal on calling TCL/TK from C/C++
was interesting, in a way, but it was showing a very non-standard way of
accessing TCL for use in very odd situations. The entire TCL interpreter
is a C library, and it was originally meant to be bound into C programs.
TCL's original purpose was to provide a common scripting language across
multiple applications. That TCL is succeeding as a pure scripting
language is only possible because it has been bound into a command-line
shell. Try reading Part III (Writing TCL Applications in C) and Part IV
(TK's C Interfaces), in Ousterhout's book ``TCL and the TK Toolkit''.
--
,sig under construction
Vassili Leonov
On 22 Mar 1995, Michael Mellinger wrote:
>
> Hi,
> I'm a naive Windows programmer with a lot of money who wants to
Good. I don't know though what 'lot of money' means :-) I assume you
got all them programming for Windo$e... Well - some people earn for
living in even more perverse ways... :-) :-)
> program in X Windows. Could someone tell me how to write X apps? I'm
It's easy - requirements, design, implementation :-) If you write them
under GPL then testing is not your problem :-)
> considering some these solutions, along with any others that anyone
> else can offer.
>
> Xlib
This is the low level access to X protocol. Mostly able to get messages
from the mouse and draw lines in the window.
> Xt
This is the library to build windows objects with (widgets).
Then, the next would be
-Motif
A library of these widgets ready. Rather separate issue is that there is
a window manager coming with it (mwm).
> Tk
This is some sort of a scripting, easy expandable language, with bunch
of widgets in it.
> Galaxy
??
> TeleUSE
??
> XVT
Is this Xview? This is same as Motif, only worse. Sexy - but somehow you're
not exicited :-)
> Athena
Oh well - this is the same as Motif, only much smaller and not sexy. But
people are fighting agains obscenity on Internet - right?
> Suit
??
> InterViews
This is like Motif, but also with some tools to actually build your
user interface. Was here befor Visual Basic and Visual Xyz - but the goal
and idea is the same.
> Fresco
Yeah - Motif is nice, but expen$ive - so this one is like it - only
better and free - but not finished yet.
>
> Geez, there are sure a lot of ways to program in Xwindows. It's too
> bad that 90% of all commerical software is written for the Windows
> API(Win32 or MFC). Gosh, wouldn't it simply be easier to make these
Let the Windows rot in sin... It's too sexy...
> APIs freely available under X so that we naive Windows programmers
OK, the API at the C level is orders of magnitude better with X then it's
in MSWin. It's nowhere near 600 function calls. Documentation is very
good - just buy volumes 0-8 of O'Railly X-programming series (maybe
volume numbers are sligtly different):
0 - Protocol
1 - Xlib prog.man
2 - Xlib ref.man
3 - Xt prog man
4 - Xt ref man
5 - Motif prog
6 - Motif ref
7 - X users guide
8 - X admin guide
Easch book is no more then 700-1100 pages (but no less also :-)
and you're ALL set. And you know - there is NOTHING hidden, nothing
undocumented and gonna stay here forever.
Vassili.
mmalcolm Crawford
> was spent making it look nice that the system and applications were
> insipid and bug filled? A case of all polish and no furniture.
>
Have fun,
mmalc.
Dan Newcombe
gone]
funk...@spud.Hyperion.COM (Larry Hastings) wrote:
>ls...@ecs.soton.ac.uk (L S Ng) writes:
>>The Mail program in Windows 95 is a real killer
>>application. [...] When Windows 95 is released,
>>expect to see the abandonment of plain old Unix Mail in favour of
>>Windows Mail.
>Unix mail. Microsoft Exchange, as far as mail goes, is just
>a client; it speaks Microsoft Mail, CompuServe Mail, presumably
>The Microsoft Network Mail, and Internet Mail. The latter
>is standard old email, which can parse both UUENCODEd and
>MIMEd attachments.
This reminds me of something posted in alt.fan.bill-gates a bit ago.
Someone was trying to prove that Windows (or NT or Win95) was so much
better because it supported 32-bit mail.
Meaning of course that the mail program was a 32bit executable, but
that is not the way it came out sounding.
-Dan, must get an alpha for 64-bit mail
Paul F. Kunz
> Perhaps you should note that XMosaic was developed originally on
> NeXTSTEP. This is a perfect example how GUI will affect apps
> development.
I believe this is not correct. The first GUI web browser was
developed on NeXTSTEP, but XMosaic did not copy that GUI. In fact,
XMosaic had a better GUI then the original NeXTSTEP browser. A
fairly rare case where the NeXTSTEP application was not a nice as the
equivalent X one.
--
Paul F. Kunz Paul...@slac.stanford.edu (NeXT mail ok)
Stanford Linear Accelerator Center, Stanford University
Voice: (415) 926-2884 (NeXT) Fax: (415) 926-3587
David Alan Gilbert
>In article <SAMURAI.95...@maggie.cs.mcgill.ca>
>sam...@maggie.cs.mcgill.ca (Darcy BROCKBANK) writes:
>I think that what is missing in Linux/UNIX is a good, free (money and
>codewise) desktop metaphor (Finder, FileManager+ProgramManager,
>what-have-you). Most of the way that NeXTSTEP looks has more to do with
>many hours of design (visual design... art.) put into the interface
>than with the internals underlying the display engines. X seems quite
>fast, quite capable to me.
>What Linux/UNIX/Mach/FSF needs is a new look and feel (Motif is OK, but
>looks a little dated regardless. Perhaps a streamlining / redesign of
>the Motif look along the lines of NeXTSTEP) that will give it a
No - thats not enough. The look of the system is important - but things
like Motif and InterViews look nice; integrated file manager etc. etc.
is important - but its still not enough.
You need at least the following:
1) True integration of file managment - i.e. programs don't have 'load'
or 'open' options any more - you just double click/drag the file onto
them (ditto with import); I'm not aware of any WIMP to do that nicely
other than Acorn's Risc OS; most do double click to load but thats it.
2) Applications sharing services etc. - the example which I was shown on a
Next many years ago went something like, say a number of apps use an
editable text pane, you buy a spell checker - suddenly all the text panes
have spell check facilities; this demonstration was about 4 years ago
as I remember; and I don't think the text editing pane was replaced,
I think it was more of the spell checker offering some type of service to
it.
Making X look pretty isn't the issue - that just takes a little effort
on the part of the artistic guys; making it powerful is a whole different
kettle of fish.
Dave (gilb...@cs.man.ac.uk)
--
-------------------------------------------------------------------------------
- David Alan Gilbert - gilb...@cs.man.ac.uk - G7FHJ@GB7BEV -
------------ (University of Manchester - AMULET Group) ------------------------
Felix Sebastian Gallo
>Black text on colored background is much easier to read allowing one
>to keep the font size small.
Colored text on a black background is IMHO superior because white is a
very radiant color which induces shimmer and fuzziness at lower refresh
rates or in the presence of interlacing. Being particularly sensitive
to this, I can't read a screen of black text on white on most monitors.
The good reason to use a white background is that paper is (usually)
white or close to white, so if paper is a final output medium, you'll
get a closer WYSIWYG feel by mimicking it on screen.
Ideally, the user gets to select all colors, of course. White might
not even be available on some output device you don't know about.
>Paul F. Kunz Paul...@slac.stanford.edu (NeXT mail ok)
>Stanford Linear Accelerator Center, Stanford University
>Voice: (415) 926-2884 (NeXT) Fax: (415) 926-3587
--
Felix Gallo as...@io.com
"stabbing someone is a direct result of several factors." - Kevin Lord
Tim Smith
>Once at a time Darcy BROCKBANK (sam...@maggie.cs.mcgill.ca) wrote:
>> and feel, as well as behave like a 4 month student programming project.
>
How long have you had these telepathic powers?
--Tim Smith
Kennel
For them, yes. For us, the statement stands.
> -Ken
Lewis
>No - thats not enough. The look of the system is important - but things
>like Motif and InterViews look nice; integrated file manager etc. etc.
>is important - but its still not enough.
>You need at least the following:
> 1) True integration of file managment - i.e. programs don't have 'load'
>or 'open' options any more - you just double click/drag the file onto
>them (ditto with import); I'm not aware of any WIMP to do that nicely
>other than Acorn's Risc OS; most do double click to load but thats it.
Gods no, don't get rid of them! I spend most of my time _in_ the
application, I don't want to have to go back to the OS to open another
file. Add the draggability as an option (like many Macintosh based
packages do), but keep my open option, please.
--
Lewis Tanzos - le...@ds9.lesn.lehigh.edu - lj...@lehigh.edu
Erik Fortune
In article <PFKEB.95M...@kaon.SLAC.Stanford.EDU>, pf...@kaon.SLAC.Stanford.EDU writes:
> Black text on colored background is much easier to read allowing one
> easier to read. Result is trim, crisp windows that look pleasant
> while still being X/Motif.
Agreed, but I prefer AvantGarde to Helvetica -- it's still crisp but it
has a little more style.
-- Erik
+-------------------------------------+--------------------------------------+
| Erik Fortune | |
| Silicon Graphics International R&D | At Intel, Quality is Job 0.999897 |
| +1-415-390-1922 er...@sgi.com | |
+-------------------------------------+--------------------------------------+
All opinions are my own, so leave my employer out of this
Pete Hardie
Michael Shandony <vanh...@bnr.ca> wrote:
>X lets you configure every little stinkin' thing. It gives you almost
>total control over what each GUI element will look and act like. While
>this level of freedom is really nice in some ways, it is bad in others.
>For example, because of this flexibility, it is very easy for a designer
>to make his/her GUI look and act very differently from any else's or just
>from project to project. Do I really care that a pushbutton can be armed
>and/or activated? I only ever care when it's activated. I find choosing
>colors to be very difficult, using GCs to be unintuitive, and placement of
>widgets almost a black magic (even with a form widget). I could go on and
>on, but I think you get the idea.
IMHO, the solution is to find the setup you like, and save it for later
use. I'm doing just that for my current company - we decide the colors
for bg/fg, selection of buyyons, etc, and those are put into macros
(or functions) and applied to everything.
--
Pete Hardie, nanobrewer, amateur philosopher
pe...@nyet.atl.ga.us
Jerry Shekhel
>>
>> Sure there is: Marketing. Just ask Microsoft. They don't know anything
>> about quality, but...
>
>
>Ken, this is a popular comment---but what exactly is there to back it up?
>What proof do you have that MS is lacking in quality controlled compared
>to others in the computer industry?
>
>I think this is mostly a myth, based on misinterpreting microsofts
>fledgling products, which had severe time pressures on their
>original development.
>
>Personally, if I had to pick a company to deliver a major new product
>of high quality, I would pick MS---they can _afford_ the luxury
>of thorough design and quality control at this point.
>
You'll probably get lots of angry replies to this, but I agree 100%. Nobody's
products are under nearly as much scrutiny or used by nearly as many people as
Microsoft's. So of course you hear more about Microsoft bugs than anyone
else's. I've worked in the UNIX world for over 6 years, so I know what a mess
most commercial UNIX products are, not to mention the fact that they are
usually enormous (most are statically linked), slow, and have terrible and
inconsistent user interfaces. MS productivity applications are light years
ahead of anything available for UNIX, in functionality, user interface, speed,
etc. To someone who's never used anything but UNIX, seeing something like
Excel 5 running on a machine with a mere 16MB RAM, dragging objects to and
from other applications, is a revelation. And although I know nothing about
its internals, I must say that NT is the most impressive piece of
system software I've ever seen, except of course for a select few Linux
kernels :-)
On the other hand, nothing in the mainstream PC world can touch UNIX for power
from the command line. The problem is that the power comes at a cost that 95%
of the computer-using public can't afford. I'm a developer, though, and
that's why there'll always be room on my hard drive for Linux, which is still
my favorite OS.
>
>Barry Merriman
>
Jerry
j...@ix.netcom.com
Paul F. Kunz
> As a person who has done several projects for both school and work
> using X Windows and Motif, this are the reasons that I feel that it
> is difficult to make an X-based GUI look good.
I have a somewhat different opinion on why X/Motif applications
don't look nice. It seems on many systems the default is white
letters on a colored background. The default font is typewritter
like. It's hard to read. So the fix seems to be make the font
larger. Then the buttons are too small for the text, so make them
larger. Then the Windows are too small so make them larger. The end
result is that your 1K by 1K screen becomes cluttered with ugly,
chunky windows.
In developing the objcX library, I initially tried to stay with the
X/Motif defaults. But lately, I switched to following NeXTSTEP style.
Black text on colored background is much easier to read allowing one
to keep the font size small. Use of Adobe Helvetica font which is
easier to read. Result is trim, crisp windows that look pleasant
while still being X/Motif.
--
Hallvard Paulsen
> Great joke. I've not yet seen an X based interface which doesn't look
> and feel, as well as behave like a 4 month student programming project.
>
>
> - db
If you are talking about individual application interfaces there might
be a grain of truth in what you are saying. However I must say that
windows applications with lots of "sub-windows" are a big pain to
work with. Using Fvwm I presently have 16 logical X screens on my
computer. Each of which may be compared to the total windows screen.
Also I can freely move any single window from one screen to another.
(But since no windows user seems to be able to imagine how it is
possible to run more than one application at the time this kind
of functionality is probably nothing you guys think you need.)
I think the "desk-top" metaphor has gone to far. My (real) desktop
is usually very messy. I * do not * want the same mess on my computer
screen. What is the point of having a computer if you are supposed to
do the work in the same way as you do with paper and pencil? I thought
computers were supposed to *improve* the way we do our work, not just
cost a lot of money and enable us to do the work in the same way we
always did it!
Yes, I believe something is fundamentally wrong in the minds
of some important people in the industry. (Not to mention the
computer press, populated by people that do not produce anything
but self fulfilling prophesies!)
puhhh
--
Hallvard Paulsen
Robert Rodgers
vanh...@bnr.ca (Michael Shandony) wrote:
>In article <3l9fpv$g...@nntp3.u.washington.edu>,
>Mike Kenney <mi...@wavelet.apl.washington.edu> wrote:
>>Why did people continue to prefer the DOS CLI interface
>>to the Mac? The Mac was available long before Windoze.
>
>Ignorance, cost, and IBM's backing of Microsoft.
128k.
--
"Amiga is IBM-compatible, too. A simple piece of software teaches
Amiga to emulate the IBM operating system, so you can run most IBM
programs. You'll have instant access to the largest library of
business software in the world..." [Ad for Commodore Amiga, Dec 1985]
Bogdan Urma
: In article <KSTAGG.95M...@eskimo.com> kst...@eskimo.com (Ken Stagg)
: writes:
: > >
: > > In software, there is no substitute for quality.
: >
: > Sure there is: Marketing. Just ask Microsoft. They don't know anything
: > about quality, but...
: >
: > -Ken
: Ken, this is a popular comment---but what exactly is there to back it up?
: What proof do you have that MS is lacking in quality controlled compared
: to others in the computer industry?
You are joking, right?
1. Windows 2.x, 3.x
2. MS-DOS
The above two are so unstable and buggy that it's not even worth
discussing with someone, if s/he cannot realize that already.
: I think this is mostly a myth, based on misinterpreting microsofts
: fledgling products, which had severe time pressures on their
: original development.
Now it's a reality based on TODAY's MS products.
: Personally, if I had to pick a company to deliver a major new product
: of high quality, I would pick MS---they can _afford_ the luxury
: of thorough design and quality control at this point.
MS can afford to put out buggy software that doesn't get fixed since
people are already used to accepting that. That's what MS can afford.
Bogdan
------
Bogdan Urma
Cornell University
Email: ba...@crux2.cit.cornell.edu
WWW: http://www.ruph.cornell.edu/burma/homepage.html
Jim A. Fetters
direction, but X isn't the problem. When you consider closed systems
such as Macintosh --- whose GUI is 99% identical from the 1984 version,
that is the step in the WRONG direction. Several issues are at work
here:
1. you are really talking about bad programming design of the user
interface
2. a desktop metaphor for dealing with program/data "objects"...
The desktop metaphor is not one that I choose to subscribe to.
I've used the SGI and find myself going to the x-terms to load and
find programs, etc. I was once asked by my instructor how he could
set up an icon for PhotoShop on the SGI so that the students could
just click on the icon. I can't understand that approach because
in my opinion what could be more easier than typing "photoshop"
at the shell prompt?
I am 100% for GUI development within the program level, but as far
as the desktop metaphor goes I am against it. Look at Microsoft's
"Bob" (appologies to the Church of the SubGenius)... it takes that
desktop concept to hideous new levels. The desktop is an "abstraction"
--- too much of an abstraction for me. GUI tools may be developed
but I will err on the side of freedom of choice and risk "bad GUI"
than sacrifice the freedom and versatility of X/UNIX.
The whole desktop thing is over-rated. Think about it. You spend
most of your time *in* an application. You want to *do* something.
When I use the Macintosh, I am so frustrated when using applications
and I keep seeing the underlying desktop, or if i slip the mouse pointer
between the edge of the window and the bottom of the screen and
click by mistake, I'll be transformed back into the desktop/Finder.
Like the original PARC scientists used to say, "Don't Mode Me In"..
a desktop creates an unwanted mode - much too "modal" for my tastes.
Within an application I want all the fancy features, but outside of
that I still want my shell.
Jim
Barry Merriman
>
> Sure there is: Marketing. Just ask Microsoft. They don't know anything
> about quality, but...
>
> -Ken
Ken, this is a popular comment---but what exactly is there to back it up?
What proof do you have that MS is lacking in quality controlled compared
to others in the computer industry?
I think this is mostly a myth, based on misinterpreting microsofts
fledgling products, which had severe time pressures on their
original development.
Personally, if I had to pick a company to deliver a major new product
of high quality, I would pick MS---they can _afford_ the luxury
of thorough design and quality control at this point.
--
Barry Merriman
UCSD Fusion Energy Research Center
UCLA Dept. of Math
bmer...@fusion.ucsd.edu (Internet; NeXTMail is welcome)
Michael Pizolato
:pf...@kaon.SLAC.Stanford.EDU (Paul F. Kunz) writes:
::Black text on colored background is much easier to read allowing
::one to keep the font size small.
:
:Colored text on a black background is IMHO superior because white
:is a very radiant color which induces shimmer and fuzziness at
:lower refresh rates or in the presence of interlacing. Being
:particularly sensitive to this, I can't read a screen of black
:text on white on most monitors.
:
:The good reason to use a white background is that paper is (usually)
:white or close to white, so if paper is a final output medium,
:you'll get a closer WYSIWYG feel by mimicking it on screen.
:
:Ideally, the user gets to select all colors, of course. White
:might not even be available on some output device you don't know
:about.
I prefer the black on white myself, but of course one's sensitivity
to things makes all the difference. Allowing the user to select
the colors solves the problem, and I doubt there are enough mono
machines left to be a consideration in this area. Of course, I'm
writing this on a mono NeXTstation, so it doesn't apply to me :-).
The best text/background color combination I've seen was the default
colors in Cello, a Windoze web browser. They used black on a sort
of off-white (too yellow to be tan, too brown to be yellow, too
white to be either). I've always preferred paper of that color
for reading print, and on screen that combination is extremely
readable, even with very small fonts (which I prefer, so I can
squeeze more stuff onto my screen). I'd love to be able to use
that combination, but even if I had a color machine I wouldn't be
able to set the background color for text areas in NEXTSTEP :-(.
I agree with NeXT's general principle that color should be used
only where it's truly effective, and that allowing pink windows
with purple borders is fluff rather than good customizibility (is
that a word?), but in the case of text foreground/background colors
it's far too important an issue for people with differing
sensitivities,
given the amount of time users spend reading text on screen.
-m
--
Anderson Financial Systems Inc. Michael_...@afs.com
Suite 106 Vox: 215/653-0911
909 Sumneytown Pike
Springhouse, PA 19477
Martin Michlmayr
[GNUStep http: "http://fvkma.tu-graz.ac.at/gnustep/gnustep.html"];
/ I think right now it requires Motif to run.
This is correct.
/ I read that it is in the plans to get rid of that requirement,
/but that'll be a while, I'm sure.
You are correct again. We will have another version, that does not need
Motif.
Well, The DPS (Display PostScript) version will take a while, but
there is already a library, libNeXT, that uses Xlib.
However, the problem is that we can`t release this library under the
terms of the GPL, because the developers need money to give it away.
We will soon show a version of it with objcX and hope that people interested
in this progress may find the money.
--
Martin Michlmayr | t...@tci002.uibk.ac.at | t...@fvkma.tu-graz.ac.at
GNU OpenStep Development Team, Manager of the Documentation Department
http://fvkma.tu-graz.ac.at/gnustep/gnustep.html
Steven Miale
Dan Newcombe <newc...@aa.csc.peachnet.edu> wrote:
>Trying to develop a GUI for X will give you a few headaches. First, you have
>to decide on a look and feel, and then implement it. This has been tried
>several times (Andrew, XView/Openlook, Motif, Athena). No one seems to really
>like what has been done. It's odd that a lot of apps are based on Athena,
>even though the docs say it's only a sample. Motif costs, so we can rule
>that one out. There was a project called Notif which was to be a free-Motif,
>but it seems to have sizzled. Tcl/Tk, while good, is script based, and we
>need something easily called from C/C++.
May I suggest Python? It's a wonderful language; object-oriented, absurdly
easy to extend, incredibly powerful (it's been used as the basis for everything
from a web browser to a virtual reality research project), and has modules
that support X and Tk, as well as a few platform-independent toolkits,
like SUIT and stdwin.
I'm currently putting together a Python package. For now, you can get it
from ftp.cwi.nl. Version 1.2 should be out in a week or so, but you can
get the beta from /pub/guido. It compiles right out of the box.
You can also check out comp.lang.python.
Steve
--
Steven Miale <http://www.cs.indiana.edu/hyplan/smiale.html>
Chris Bitmead
>Hi,
> I'm a naive Windows programmer with a lot of money who wants to
>program in X Windows. Could someone tell me how to write X apps? I'm
>considering some these solutions, along with any others that anyone
>else can offer.
>
>Xlib
>Xt
>Tk
>Galaxy
>TeleUSE
>XVT
>Athena
>Suit
>InterViews
>Fresco
>
>Geez, there are sure a lot of ways to program in Xwindows. It's too
>bad that 90% of all commerical software is written for the Windows
>API(Win32 or MFC). Gosh, wouldn't it simply be easier to make these
>APIs freely available under X so that we naive Windows programmers
>could effortlessly port our many thousand applications to to run under
>X Windows on those really neat Sun, SGI, HP, IBM, and Linux boxes.
You can't program in X without at least linking with Xlib. You must then
choose a widget library to use. Many are Xt libraries.
These days most commercial apps use Motif, so this would be the most
sensible choice.
If you want a free widget library, you could use either Athena (ugly),
Athena 3-D (better), or xview. If you want to use C++ then interviews is
very nice.
But Motif is the best choice if you want to sell commercially.
--
Chris Bitmead
chr...@ind.tansu.com.au
Daniel Louis Potter
> Colored text on a black background is IMHO superior because white is a
> very radiant color which induces shimmer and fuzziness at lower refresh
> rates or in the presence of interlacing. Being particularly sensitive
> to this, I can't read a screen of black text on white on most monitors.
I have the opposite problem. On my computer I run X, and I have
configured my Xterms to be white on black. This is because A) I like this
color set for shells and B) minicom works better that way. Now my
computer runs in an interlaced 1024x768 mode and doesn't look all that
great. That text usually burns you up before you're done. One the other
hand, the black on white text is usually easier to read. If it bothers
you, just turn down the brightness. (That may sound silly, but hey..) I
think that what Mr. Kunz said is probably the easiest to read all around,
black text on a COLORED background (ie, some shade or derivitive of
grey). That always seemed to look nicest on my yucky interlaced screen,
and my roomate's good computer.
> Ideally, the user gets to select all colors, of course. White might
> not even be available on some output device you don't know about.
Ideally, we're all running NextStep ported to Linux. =)
--------------------------------------------------------------------------
Daniel Potter dlp...@jove.acs.unt.edu
UNT Box 6909 Denton, TX 76203
--------------------------------------------------------------------------
Terry Wilcox
Jim A. Fetters <fet...@enuxsa.eas.asu.edu> wrote:
>I agree that a consistent GUI would be a step in the right
>direction, but X isn't the problem. When you consider closed systems
>such as Macintosh --- whose GUI is 99% identical from the 1984 version,
>that is the step in the WRONG direction. Several issues are at work
>here:
Yeah, the Mac's so bad that even it's crappy apps look and work better
than X based apps. We needed a low end drawing app for X. Somebody
recommended Xfig, saying it was great. What a load of crap. It
wasn't half as good as 1984 Mac software, but it's good for an
X app.
>The desktop metaphor is not one that I choose to subscribe to.
>I've used the SGI and find myself going to the x-terms to load and
>find programs, etc. I was once asked by my instructor how he could
>set up an icon for PhotoShop on the SGI so that the students could
>just click on the icon. I can't understand that approach because
>in my opinion what could be more easier than typing "photoshop"
>at the shell prompt?
That's exactly the problem with X. It's easier to do everything from
a terminal window. Much easier. That's fine for those of us who
know Unix. Unfortunately, I've got dozens of users who don't know
Unix. They use Unix boxes to get work done, not learn about shells
and paths and X and window managers etc. If we could find a consistent
and usable interface (including some kind of decent file browser) for
X, life would be much better. That product just doesn't exist.
>The whole desktop thing is over-rated. Think about it. You spend
>most of your time *in* an application. You want to *do* something.
I spend most of my time in multiple applications. I want them all to have
the same look and feel. I want them to to communicate transparently.
In X I spend all of my time in terminal windows. I have to waste
my time learning new interfaces with every new app, so I avoid using
new software if I can. The command line tools are just easier to use
than most X apps.
That's not a good thing.
Terry Wilcox
--
Terry Wilcox
Arcane Systems Ltd.
te...@arcane.com
Gregory Propf
appealing to me. I have to say I agree about the general clunkiness
of X apps. However, as I read the posts I feel that something is
neglected in this discussion. The strength of both X and UNIX is not
that they are neccessarily very user friendly or very 'nice' to
people. Their strength is standardization. I can write a POSIX
compliant X app on my Linux box that you can probably compile on your
SGI, Sun or RS/6000 with only a few changes to the makefile. I cannot
write an app for the Windows API that will even begin to port to OS/2
or any other system. History shows that in the long run standardized
products win out over non-standard products.
If GNUStep is an X-extension it may work. If it turns out to
introduce all kinds of incompatibilities we will be back at something
like the Windoze API (i.e. 600 functions that all do something that
doesn't work on any other system). I would like a better interface to
X but you really can't turn UNIX into a pure point-and-click (System 7
style) system without turning it into something that is fundamentally
not UNIX anymore. Standards are the strength of Linux/X.
===========================================================================
| Greg Propf, Towson State Univ | "First things first, not necessarily |
| Towson, MD (physics '96) | in that order" - Dr. Who |
| pr...@zeus.towson.edu | |
===========================================================================
Gregory Propf
: In article <3kvdtd$1...@news.umbc.edu>,
: Gregory Propf <pr...@zeus.towson.edu> wrote:
: >If GNUStep is an X-extension it may work.
: NeXT published a standard called OpenStep. You write a program using
: the OpenStep standard, just like writing a program using ANSI C and
: the Xlib library calls. Then you can compile that program on any
: OpenStep compliant system.
OK - good.
: Right now, NeXTstep is an OpenStep compliant system. Sun is
: apparently teamed up with NeXT to put out a version of OpenStep for
: SunOS. And these folks are working on GNU OpenStep, a free
: implementationof the standard.
Alright, sounds good.
: So, you grab the source code to *any* OpenStep application, whether
: written for a NeXT, a sun, or anything else, and compile it under
: *any* OpenStep environment, and it should work.
: Satisfied?
Yes, actually I am. As long as we are setting up a standard and not
reinventing the wheel the thing sounds good. I wasn't being critical
- just cautious. I'd hate to see Linux turned into another one of
those - 'you need to rewrite all your code to make it work here'
systems. I attempted to learn programming for Windows once and was
very unhappy. The compiler created 12 source files for a simple
'hello world' program. I couldn't follow the logic even using the
debugger! Then I realized I needed to join a sort of developer's
secret society (at the cost of $500/yr) to get the really useful info
about the system. I gave up. I just wanted to make sure that GNUStep
was NOT a step (bad pun intended) in that direction. So far GNUStep
sounds really great! - GP
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Gregory Propf, Towson State Univ. (physics) pr...@zeus.towson.edu %
%-----------------------------------------------------------------------------%
% "At roughly 600,000 employees (including accountants, auditors and soldiers)%
% the IRS is currently the largest terrorist organization in the world. Not %
% even the strongest squadron can hope to stand against them if they decide %
% to make an example of you. They'll either assault you themselves, or hire %
% the work out" - Strike Commander Technical Manual (Sudden Death magazine) %
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Garance A. Drosehn
> TO ALL: Let me say first of all that the idea of a better Linux
> GUI is appealing to me. I have to say I agree about the general
> clunkiness of X apps. However, as I read the posts I feel that
> something is neglected in this discussion. The strength of both
> X and UNIX is not that they are neccessarily very user friendly
> or very 'nice' to people. Their strength is standardization.
This is one of the great bullshit claims of Unix.
> I can write a POSIX compliant X app on my Linux box that you can
> probably compile on your SGI, Sun or RS/6000 with only a few
> changes to the makefile. I cannot write an app for the Windows
> API that will even begin to port to OS/2 or any other system.
That app which is horribly-tied to MS-Windows will run on a few
million computers (all of which are running MS-Windows, or are
running MS-Windows emulators).
To write a POSIX-compliant X app means "to do nothing". If you're
going to write an application that *does* anything, you're almost
certainly going to call some routines which are not defined in POSIX.
> History shows that in the long run standardized products win out
> over non-standard products.
This may be true, as I don't feel like trying to think up some
counter-example.
However, the "standard" system right now is MS-DOS and MS-Windows.
Those are the most common operating systems in the computer industry
today. They own about 70%-90% of all computers sold (depending on
who's doing the estimate of sales, personally I don't think it's
really quite that high). While the Unix community was beating it's
chest with standardization efforts, Microsoft just went out and
got most of the world to buy their product. As such, they are
becoming the defacto standard, and a defacto standard has a lot
more significance than some fringe element claiming that they have
all agreed to some other "standard" interface.
Note that I'm not against the idea of standardized systems, where
no single company has total control over the system. However, back
here in the real world, Microsoft is the most standard thing around.
--
Garance Alistair Drosehn = g...@eclipse.its.rpi.edu
ITS Systems Programmer (handles NeXT-type mail)
Rensselaer Polytechnic Institute; Troy NY USA
Douglas McClure
folks know exactly the score on some items.
In article <3kvh95$1...@news.umbc.edu> pr...@zeus.towson.edu (Gregory
Propf) writes:
> Doug DeJulio (dd...@pitt.edu) wrote:
>
> : NeXT published a standard called OpenStep. You write a program using
> : the OpenStep standard, just like writing a program using ANSI C and
> : the Xlib library calls. Then you can compile that program on any
> : OpenStep compliant system.
>
> OK - good.
You can compile it on any system with OpenStep in as much you NEVER call
anything outside of the OpenStep library. If you do, then you will have
to be doing #ifdef's around code that is going to be dependent on the
system you are using.
> : Right now, NeXTstep is an OpenStep compliant system. Sun is
> : apparently teamed up with NeXT to put out a version of OpenStep for
> : SunOS. And these folks are working on GNU OpenStep, a free
> : implementationof the standard.
>
> Alright, sounds good.
NEXTSTEP is NOT OpenStep compliant. It never has been. OpenStep is a
standard based on what NEXTSTEP 4.0 is going to look like. When 4.0 comes
out, THAT will be OpenStep compliant, and not until then. There is no
OpenStep system currently available at all.
-d
Mark van Hoeij
>In article <D5xxt...@ennews.eas.asu.edu>,
>>The desktop metaphor is not one that I choose to subscribe to.
>>I've used the SGI and find myself going to the x-terms to load and
>>find programs, etc. I was once asked by my instructor how he could
>>set up an icon for PhotoShop on the SGI so that the students could
>>just click on the icon. I can't understand that approach because
>>in my opinion what could be more easier than typing "photoshop"
>>at the shell prompt?
>That's exactly the problem with X. It's easier to do everything from
>a terminal window. Much easier. That's fine for those of us who
>know Unix.
You are saying here that someone needs to know Unix in order to be able to
type photoshop.
Besides, you haven't answered the question. So the question remains: what could
be easier to learn how to start up photoshop than by typing photoshop?
Why do some people think that CLI is always more difficult than a GUI, even
in this example where typing photoshop in a CLI is clearly the most easiest
thing to remember.
>Unfortunately, I've got dozens of users who don't know
>Unix. They use Unix boxes to get work done, not learn about shells
>and paths and X and window managers etc.
Users don't need to know about paths and windows managers. They only need
to know the name of the app. Type that name and the app starts. Simple, right?
>If we could find a consistent
>and usable interface (including some kind of decent file browser) for
>X, life would be much better. That product just doesn't exist.
It does exists. Slackware came with some kind of program manager and file
manager. I forgot the names, but you can find them.
Mark van hoeij
Doug DeJulio
=?ISO-8859-1?Q?Kristian_K=F6hntopp?= <kr...@black.schulung.netuse.de> wrote:
>> Perhaps a
>> streamlining / redesign of the Motif look along the lines of
>> it from other operating systems combined with a ruthlessly
>> effective system management/desktop metaphor that borrows and
>> builds on strong features from all current operating systems.
>
>But what use is this new and fresh look, when every other
>application comes back with Athena scroll bars and Buttons?
My understanding is that GNU OpenStep is going to have several
versions. At *first* it'll run on top of X11, but then it'll have the
option to run on its own windowing system, something like Display
GhostScript, as well. Please correct me if I'm wrong.
This would mean, once GNU OpenStep gets to the point where you can get
all your work done on it, you can just punt on X11 totally. This
prevents people from running standard X11 apps without being experts
and jumping through hoops, and this prevents the problem you're
talking about.
I know *I* plan to do this if things ever get to that point...
--
Doug DeJulio | R$+@$=W <-- sendmail.cf file
mailto:dd...@pitt.edu | {$/{{.+ <-- modem noise
http://www.pitt.edu/~ddj/ | !@#!@@! <-- Mr. Dithers swearing
David Simmons
Daniel Louis Potter <dlp...@jove.acs.unt.edu> wrote:
>Felix Sebastian Gallo (as...@pentagon.io.com) wrote:
>
>> Colored text on a black background is IMHO superior because white is a
>> very radiant color which induces shimmer and fuzziness at lower refresh
>> rates or in the presence of interlacing. Being particularly sensitive
>> to this, I can't read a screen of black text on white on most monitors.
>
>hand, the black on white text is usually easier to read. If it bothers
>you, just turn down the brightness. (That may sound silly, but hey..) I
White (or light-colored) text on a dark background is easiest for the brain
to extract information from because if the background is white, then the
background is carrying the visual energy, and the black text is merely a
lack of visual energy. The brain must perform a negative of the image
in order to obtain the information. (commentary on information and
energy equivilance postponed for now... :)
David
--
David Simmons, System Administrator sim...@ee.msstate.edu
Mississippi State University Electrical and Computer Engineering
Visit my home page! http://www.ee.msstate.edu/~simmons
Doug DeJulio
>> : the OpenStep standard, just like writing a program using ANSI C and
>> : the Xlib library calls. Then you can compile that program on any
>> : OpenStep compliant system.
>You can compile it on any system with OpenStep in as much you NEVER call
>anything outside of the OpenStep library. If you do, then you will have
>to be doing #ifdef's around code that is going to be dependent on the
>system you are using.
Right. Just like ANSI C. Stick in a single non-ANSI library call,
and all bets are off. Stay 100% within the spec though, and you've
got real portability.
>> : Right now, NeXTstep is an OpenStep compliant system.
[...]
>NEXTSTEP is NOT OpenStep compliant. It never has been. OpenStep is a
>standard based on what NEXTSTEP 4.0 is going to look like....
Oops. Thanks for correcting me, I was misinformed on this one.
Maureen Lecuona
>The point is that one does not need to know what to type. Not only did
>you need to know the name of the program, you also needed its proper
>spelling, the executable name, its capitalization, and the optional/
>manditory command line arguments. When it is an icon, you just point and
>click. That is why your instructor wanted the icon. He wanted to teach
>about the program, not how to get it running.
That may be so, but when a desktop gets full of icons, they start to get
confusing and I for one tend to forget which one is for what product.
I believe that the desktop metaphor is nice for many folks, but the
command line is also good for others. Why should everyone have to fit
into the same mold? I thought the idea is to allow for peoples preferences,
not for programmer dictates. Many times I find as a user that I get
extremely annoyed at the inflexibility of icon based products, and so,
I end up not using them.
You can't dictate preferences and end up with something that pleases
all users...
Just my $.02
M. Lecuona
--
|| maureen || For this world that men have made, none of us is bad enough.
|| || For the world that made us, none is good enough. -Edward Abbey
|| lecuona || Those who dream of the joys of living in a space colony should
|| || live in a space colony. -Edward Abbey
Doug DeJulio
Sounds like you misunderstand what GNUStep is.
NeXT published a standard called OpenStep. You write a program using
the OpenStep standard, just like writing a program using ANSI C and
the Xlib library calls. Then you can compile that program on any
OpenStep compliant system.
Right now, NeXTstep is an OpenStep compliant system. Sun is
apparently teamed up with NeXT to put out a version of OpenStep for
SunOS. And these folks are working on GNU OpenStep, a free
implementationof the standard.
So, you grab the source code to *any* OpenStep application, whether
written for a NeXT, a sun, or anything else, and compile it under
*any* OpenStep environment, and it should work.
Satisfied?
Michael Shandony
>I've used the SGI and find myself going to the x-terms to load and
>find programs, etc. I was once asked by my instructor how he could
>set up an icon for PhotoShop on the SGI so that the students could
>just click on the icon. I can't understand that approach because
>in my opinion what could be more easier than typing "photoshop"
>at the shell prompt?
The point is that one does not need to know what to type. Not only did
you need to know the name of the program, you also needed its proper
spelling, the executable name, its capitalization, and the optional/
manditory command line arguments. When it is an icon, you just point and
click. That is why your instructor wanted the icon. He wanted to teach
about the program, not how to get it running.
=====================================================================
Mike Shandony | Telephone: (214) 684-7303
Bell-Northern Research, Inc. | BNR/NT Internal: (ESN) 444-7303
2201 Lakeside Blvd. MS D0307 | Fax: (214) 684-3748
Richardson, TX 75082-4399 | Internet: vanh...@bnr.ca
=====================================================================
Aleksandar Milivojevic
: that one out. There was a project called Notif which was to be a free-Motif,
: but it seems to have sizzled. Tcl/Tk, while good, is script based, and we
: need something easily called from C/C++.
There is also project called LessTif. You can have an early look into
sources also. URL is:
http://www.cs.uidaho.edu:8000/hungry/microshaft/lesstif.html
Don't expect too much from it, they just started with project.
--
CAHN'S AXIOM: | Aleksandar Milivojevic | mail: Jakova Gotovca 5
When all else fails, | ---------------------- | 54520 Slatina
read the instructions. | e-mail: al...@srce.hr | Croatia
Terry Wilcox
>In <3kuj30$4...@pentagon.io.com> te...@pentagon.io.com (Terry Wilcox) writes:
>
>>That's exactly the problem with X. It's easier to do everything from
>>a terminal window. Much easier. That's fine for those of us who
>>know Unix.
>
>You are saying here that someone needs to know Unix in order to be able to
>type photoshop.
>Besides, you haven't answered the question. So the question remains: what could
>be easier to learn how to start up photoshop than by typing photoshop?
>
>Why do some people think that CLI is always more difficult than a GUI, even
>
>manager. I forgot the names, but you can find them.
But is that Photoshop, photoshop, or PhotoShop? If a non-Unix
user forgets the name, how does he find it?
What's easier? Ask any Mac or Windows user. Double-click the
Photoshop icon. It starts. Two clicks, no chance of a typo
or a wrong name.
If you're going to have tools that normal people can use,
why not make getting to those tools easy? Why only go
part of the way?
Pohl Longsine
: Like the original PARC scientists used to say, "Don't Mode Me In"..
: a desktop creates an unwanted mode - much too "modal" for my tastes.
You misunderstood what they meant. They were implying that the user
interface should behave the same from anywhere, i.e. that it shouldn't
put you into a mode where you have to switch to a different interface
paradigm just to do something simple. They accepted that small
modal sessions were necessary. (i.e. There are unsaved changes, quit
anyway?)
You complain that the Macintosh's interface will go to the desktop menu
when you click on the desktop. This is not a problem with GUIs in
general. This was a mistake by the Mac design team. You really
should try a decent GUI, like NeXTstep. It may change your perspective.
"Don't Mode Me In" was their mantra.
Character-based interfaces will mode you in like nobody's business.
--
____/| | Pohl Longsine, OpenStep Software Developer
\ o.O| GPF! | "I don't do Windows."
=(_)= CTLALTDLT! | plo...@inetnebr.com (Internet Nebraska)
U (Bill Gates, The Cat) | NeXT & MIME mail formats accepted.
DAVID L. JOHNSON
complained about Motif app's looking bad because they don't like the default
colors. Well, you can either sit on a tack, or you can change the colors,
fonts, etc., to please _you_. On the old NeXTs, or on modest hardware in
NeXTSTEP, you can have any color you want, as long as you want grey.
Also, there indeed are several desktop gui things out there. Many of them have
D&D, configurable icons, and many options to choose from. It may not come that
way, any you may have to edit the resource file to get it to look the way you
want, but you're supposed to be able to read&write -- and it's just not that
hard to do.
A partial list: GREAT (binary only, widely available), mfm (browser),
xapplaunch (no D&D), xfm, xfilemanager. The last two are available with source,
and don't need Motif. Use the 3d Xaw widgets to get it prettier if you want.
I trimmed the list of newsgroups this goes to. Perhaps the cross-posting to
the NeXT advocacy group was another reason this thread seems misinformed.
--
David L. Johnson dl...@lehigh.edu or
Department of Mathematics dl...@chern.math.lehigh.edu
Lehigh University
14 E. Packer Avenue (610) 758-3759
Bethlehem, PA 18015-3174 (610) 828-3708
Scot W. Stevenson
: Great joke. I've not yet seen an X based interface which doesn't look
: and feel, as well as behave like a 4 month student programming project.
This is more a question of taste than anything else. Some of us like
a desktop to be a desktop, that is, clean, and not cluttered with
all of those icons stuffed in every available corner ("A clean desktop is
a clean..."). The only good interface I have ever seen, period, was on
the Atari ST. Every thing else, with the possible exception of Apple on
good days, has been painful.
Y, Scot
--
Scot W. Stevenson sc...@catzen.gun.de Essen, Germany
"How do you know I'm mad?" said Alice. "You must be," said the Cat, "or you
would not have come here." --- Alice in Wonderland
Sergei Naumov
> >That's not true. The look depends on a window manager. Get Fvwm and
> >configure it properly.
> I'm using Motif as I write this. The window frames all look alike, but
> similarity ends there. Each application looks completely different.
I agree but then you have to use something like OI.
--
++++++++++++ http://sunsite.oit.unc.edu/sergei/Me/Serge.html ++++++++++
+ Sergei O. Naoumov se...@envy.astro.unc.edu tel: (919)962-3998 +
+Department of Physics & Astronomy, UNC-CH, Chapel Hill, NC 27599, USA+
++++++++++++++++++ !! 8 bit (Cyrillic) mail accepted !!++++++++++++++++
Christopher B. Browne
>direction, but X isn't the problem. When you consider closed systems
>such as Macintosh --- whose GUI is 99% identical from the 1984 version,
>that is the step in the WRONG direction. Several issues are at work
>here:
>
>interface
>2. a desktop metaphor for dealing with program/data "objects"...
>
>The desktop metaphor is not one that I choose to subscribe to.
I'd agree that there's something basically wrong with the "desktop."
Being critical of it isn't quite enough. I haven't seen the
"advanced GUI research" into other sorts of metaphors.
- Do they exist?
- Is there a metaphor that *is* superior to the desktop?
- Or is GUI basically flawed by a lack of good metaphors?
I'm often quite critical of GUIs. ("Often" may be understating
things.) If what I'm really being critical of is the metaphor
of a desktop with "shifting papers" then there may actually be
something I could like. Is that alternative out there? Better
still, is there an X windows manager that demonstrates it?
--
Christopher Browne - cbb...@io.org
Fatal Error: Found [MS-Windows] System -> Repartitioning Disk for Linux...
Steve Dunham
: Paul F. Kunz (pf...@kaon.SLAC.Stanford.EDU) wrote:
: : >>>>> On 22 Mar 1995 10:02:18 -0600, vanh...@bnr.ca (Michael Shandony) said:
: : > As a person who has done several projects for both school and work
: : > using X Windows and Motif, this are the reasons that I feel that it
: : > is difficult to make an X-based GUI look good.
: The reason is the default Motif (1.2 and earlier) resource settings
: are so bad they make you puke. The default athena settings are slo
: bad, so are the defaults in Tk. And since most X programmers are
: hackers, they don't give a damn. So bad defaults + hackers make X apps
: look the way they are.
Or the X programmers have a .Xdefaults file that fixes the default
Athena settings.
(Currently I use:
*Command.background: tan
*MenuButton.background: tan
*Label.background: LightGray
*Dialog.background: LightGray
*Form.background: LightGray
*Box.background: CadetBlue
*SimpleMenu.background: gray95
But I still have some fine tuning to do.)
Jagadeesh Krishnamurthy Venugopal
Aleksandar> Dan Newcombe (newc...@aa.csc.peachnet.edu) wrote: : that one out.
Aleksandar> There was a project called Notif which was to be a free-Motif, :
Aleksandar> but it seems to have sizzled. Tcl/Tk, while good, is script
Aleksandar> based, and we : need something easily called from C/C++.
Tcl/Tk is very easily callable from C and C++. There are many ways of doing
this but my personal favorite is the excellent idea of Matt Welsh in Linux
Journal. In essence what you do is start off a "wish" interpreter, and attach
a file handle to it's stdin and stdout. Then communicate with it using those
two filehandles or file pointers.
With some knowledge of Unix you can easily support the event driven model of
interface handling with this approach.
You still have to know tcl and tk to do this but it is a very easy and
flexible way of calling it from C or C++.
Jagadeesh