Linux-Misc Digest #312, Volume #25 Wed, 2 Aug 00 01:13:02 EDT
Contents:
Re: MP3's skip : How I solved it (Stewart Honsberger)
Re: Which IDE linux C programers use? (David M. Cook)
Re: Partition Problem (Dennis Lee Bieber)
Re: Carnivore and Privacy: An Oxymoron? (root)
Re: Learn Unix on which Unix Flavour ? (Grant Edwards)
Re: GUI IDE (WYSIWYG) (Prasanth A. Kumar)
Re: Unexpected daily disk activity... (Mary P)
Re: Building a Linux Server from scratch: Experiences? ("D. C. & M. V. Sessions")
Re: Goosing the Mouse (blowfish)
Re: Come to my Linux's Website! (blowfish)
Re: Which IDE linux C programers use? (Grant Edwards)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Stewart Honsberger)
Crossposted-To: alt.os.linux.mandrake
Subject: Re: MP3's skip : How I solved it
Reply-To: [EMAIL PROTECTED]
Date: Wed, 02 Aug 2000 04:07:33 GMT
On Tue, 01 Aug 2000 18:37:49 -0400, Gordon Gilbert wrote:
>> Most drives work well with these features,
>> but a few drive/controller combinations are not 100% com-
>> patible. Filesystem corruption may result. Backup every-
>> thing before experimenting!
>
>I guarantee it has nothing to do with the drive and controller (i.e.
>Ultra 33 and Ultra 66 work fine in Windows). It has everything to do
>with Linux being an average of 2-3 years behind Windows in terms of
>hardware support.
Attack #1.
>It simply doesn't support DMA or UDMA transfers
>using a Promise Ultra 66 controller (apparently).
Interesting how you could come to such a conclusion after such a singular,
uneducated, unresearched test.
>Mandrake 7.1 added support for the Promise Ultra 66 controller, but
>I guess it was only PIO support, which defeats the entire purpose of
>having an Ultra66 controller. Of course that'll change, eventually.
>But for now, I'm stuck with 3MB/sec when I Should be getting more
>like 20. Heck, my new Maxtor drive (that Win98 is using) is rated
>at 43MB/sec! How fast does Linux access it? You guessed it.
>3MB/sec!
That's too bad. It's also too bad that you don't realize that, for one
thing, the kernel has some to do with DMA support. Of course you knew
that - right? Otherwise you wouldn't be making juvenile attacks at those
who are offering the only advice that could be offered, as well as
attacking Linux.
BTW - I get 46MB/sec and 49MB/sec out of my two Ultra ATA 33 drives.
That couldn't be because I have DMA support properly compiled into
my kernel - could it?
>> As you can see, they told you explicitly that there is no guarantee,
>> and that you should backup your data before fiddling. What you were
>
>What does this have to do with getting my KDM login working again?
"If you drive your car whilst wearing a blindfold, we make no guarantees
you will arrive safely at your destination."
What do you make of that statement? That General Motors should give you
a new paint job and replace your transmission?
>I see a lot of "you should have backed up" responses, but nothing
>helpful in fixing my problem. Yes, you can rub that salt in all you
>want, but it accomplishes nothing for the problem at hand.
No, it accomplishes the task of teaching you to backup before fiddling
with your hardware settings.
Are we speaking Swahilli here?
>Yes, and the lesson learned is that Linux sucks when it comes to
>modern HD support.
Attack #2.
>You either put up with the lousy 3MB/sec
>transfers or you try to improve them by changing a parameter in
>hdparm. There is no other choice.
Unfounded statement.
>> first place, know any caveats of your hardware, and know the consequences
>> before you go playing with something you obviously don't understand.
>
>What don't I understand Master Yoda? I don't need to understand the
>source code of hdparm to *use* it. I read the man page. It wasn't
>all that helpful (like most man pages). I took the risk. I hate
>slow hard drive performance. Linux screwed my hard drive over a
>bit. End of story. I still have slow hard drive performance and
>always will until I decide to buy a machine that is completely Linux
>friendly (i.e. 3 year old hardware) or until Linux catches up (at
>which time I'll probably have even newer hardware).
My 6G Ultra ATA 33 drive is quite less than 3 years old, thank you. You
might also be interested to know that Linux beat Microsoft in supporting
Ultra ATA 100 drives. But, of course, your mind appears set. You don't
want to listen to reason. Rather than take it as a lesson learned, you've
decided to trash Linux.
Perhaps you should give up on Linux. It sounds to me as if you don't want
control over your computer, but are happy with the reverse.
>> For the record, BTW, Linux systems aren't the only ones that can fail
>> to activate DMA on 'capable' equipment and bomb miserably when it's forced.
>
>The difference is that Windows enables it when it's supported. This
>is because the manufacturer includes drivers for Windows.
Or, could it be the fact that the Windows kernel is a one-size-fits-all
solution, with no specific optimizations for the hardware the OS sits on?
Could it have anything to do with the fact that your Linux kernel just
might not have support enabled for your specific IDE controller in order
to enable DMA?
Have you thought of posting your boot messages (man dmesg) to demonstrate
what was displayed on boot-up? Have you considered asking if your specific
IDE controller is supported, and where to look to go about getting it
supported? Did you think to ask what would happen if you enabled DMA on
your system without knowing how to go about it? Did you have anything
running at the time?
No. You saw somebody who posted their fix to the group and decided that it
was a fix that would work for you, too. Sorry to tell you - but it didn't.
You broke your OS, and now it's time to learn from that experience and move
on.
>Linux doesn't enable it half the time when it IS supported. So, you can't
>trust it. You end up trying hdparm anyways.
Back that up with some facts, please. You sound like an idiot right now.
Cool off for a while. Your postings look an awful lot like reactionary,
rather than thought out discourse. It's never good to respond in anger,
especially when discussing a subject you're clearly not an expert in.
>The man pages for hdparm are hopelessly outdated (there's no *current*
>list of supported hard drive controllers in the man pages).
Oh? Are you offering to keep it current? Do you understand that Open Source
Software authors often have things like careers and other projects on the
go that would prevent them from doing such Earth-shatteringly important
things as updating a man page with a list of tested hardware? No. You want
what's best for you and screw the world. That's the wrong attitude to have,
my friend.
>The funny thing is that Promise did write a driver for Linux. I
>guess it didn't include DMA support either.
Are you USING Promise's driver for Linux, or did you just hear about it
and assume it was implemented?
>> I've done it to my share of Windoze systems in past, and they sure didn't
>> like the results to the point where a format was in order.
>
>Oh, well you clearly didn't understand what you were doing, then, I
>guess. :P
What's to understand? The help pages indicate that if you want to enable
DMA, you enable it. If it doesn't work, disable it.
Windows is re-installed on a very regular basis anyways, so it wasn't very
painful for me to do.
>I already said I had previously backed up all my downloads and my
>home directory. There are configuration files scattered about the
>system though (i.e. to get Glide to work, to get other apps global
>configs to work, to set up global configs for window managers, etc.
>etc.)
Perhaps if you don't back them up, you'll learn another valuable lesson;
being able to configure things repeatedly. You may even find that you do
it better the second and third time around.
>> Then you have 6.5 times more backup storage space than I have. You can
>> make backups of static data to CDR's, and backups of other data to CDRW's.
>
>I wouldn't put static data on CDRs for the reason that most linux
>software isn't really static. Some software gets updated as often
>as once a day. That would waste CDrs.
650 megs for ~$1.80, and you're worried about wasted CDR's? Man, you
really must be short on cash.
>And before I can put backups
>on my CDRW drive, I have to get my CDRW drive working properly under
>Linux. I guess I'll have to play with CDRecord because the
>frontends for it suck.
All of them suck? There isn't a single frontend out there worth your
precious time, or is it the fact that they don't resemble the Windows
counter-parts enough, or that you have to perform manual input in order
to make them function?
> But that's the problem with Linux.
Attack #3. Attacking the OS for software written for it. Tsk.
>You end
>up doing most things in the shell because there are no better/easier
>tools to do it with. Sure people can tout how POWERFUL the shell
>is. That may be true, but most of the time that's just offered to
>cover up the fact that most Linux software isn't user friendly.
Isn't idiot friendly, my friend. All the software I need is covered for
both GUI and CLI. I'm not sure what your problem is. Perhaps you're
either too fussy, or you can't live without the Windows API.
>Where I can just copy files over to the CD-RW drive using explorer
>in Win98, I have to use CDRecord from the command line in Linux.
Have you researched this? I know I've seen mumblings about mounting a
CDR(w) disk as a volume and copying files to it like normal. Of course
not. You prefer the attack first, look like a heel later approach to
solving problems.
>area. Even if I liked doing it that way, I guarantee most of the
>general public won't. And if you ever want mainstream software
>support for Linux, driver support, etc., then Linux has to be
>acceptable to the mainstream, like it or not.
To tell you the truth, I really couldn't care less. Mainstream support
tends to draw users like you towards Linux, and that's not my vision of
Linux's future.
>> Or, you could use an automated script to put everything into a tarball
>> in the middle of the night and back it up once a week.
>
>I don't leave my system running 24/7. Cron isn't much use when you
>aren't running Linux at regular intervals.
Have you read the documentation for cron? It's not limited to the wee
hours of the night when you're curled up with your teddy bear.
>If Linux is ever to be a successful home operating system,
There you go again; making your own projections as to what Linux has to
do to become "sucessful". Believe it or not, Linux wasn't designed to be
a LUSER platform. If you want that, you've already got plenty of products
from that company in Redmond you could pay out the ear for. They sound
more up your alley.
You're trying to make Linux into Windows, and that's just not going to
happen.
>Automated scripts are also one of those user unfriendly processes.
Why, because you have to type, rather than point and grunt, er, click?
>> It took me all of 20 lines, comments included, to have my /home directory
>> backed up and a reminder mailed to me each week. Every time you find
>> something else you need/want backed up, add a line or two.
>
>A graphical frontend could make this much simpler.
Good for it! Write one!
Oh, wait, a graphical front end that can't make use of the literally
millions of possible functions of the Bash scripting language without
becomming a several megabyte monster of an application. A graphical
front end that would have to be updated every time the Bash scripting
language were ever to be changed/improved.
Gosh.. I think we're at a stalemate here. Here I am with my scripting
language with which I could run an entire BBS system (been written
before - I'll track down the URL if you'd like it), or perform such
simple tasks as backing up a directory tree, or running a Seti@Home
client with certain specific options.
I'm sorry that Bash isn't idiot friendly. Truly, I am.
>> The Bash scripting language is extremely powerful. Learn it. Love it.
>
>Ah, now we get to the punch. Linux will never become mainstream
>with this kind of attitude.
What, you mean the attitude whereby you have to actually tell your computer
what you want it to do? Vis: Do some reading in order to make your computer
function?
No, you're thinking of that Redmond-esque product again. It's the one that
tells its users where they want to go, and for how much.
>Sorry; it just won't. I shouldn't have to learn a scripting language to
>backup a few files.
I shouldn't have to learn all those annoying by-laws just to drive a few
blocks to get groceries but damnit, doesn't life just suck that way?
>This is not to say *I* won't, but rather to say that the mainstream
>public definitely will NOT.
Aww! Does that mean that Linux WON'T get more idiot users? {sob} Damn..
>When Linux learns this lesson once and for all, it might become mainstream.
There's that "mainstream" word again. I'm not sure if you even know how
to properly apply it to this situation.
>> Then you're not looking at it in the right light. What you're looking for
>> appears to be a response of "You took fate into your hands, screwed up,
>> but if you type xxx, everything will be back to normal". Sorry, but you're
>> not going to get that. Instead, you got a lesson telling you not to play
>
>Is that because you don't have a clue how to fix it? Just say so.
Alright, I have no idea how to fix a system which has been bunged by a
gung-ho user who's changed hardware parameters in order to follheartedly
make it 'wurk bettr'. But then again, that's part of what I do for a
living. I fix systems for people who think they're gurus. Y'know how I
fix 'em half the time? I backup whatever I can find that's valuable and
I re-install whatever smoking hulk of an OS remains. Often with pleas
of "Don't tinker with it if you don't understand it" relayed to the
customer afterwards.
But they come back... They always come back.. {smile}
>Stop telling me what I should have done. I can't timetravel back and
>tell myself, "Hey that's going to corrupt KDM. Don't do it!"
I could also time travel back to the first time I ever stuck my tongue
in a light socket. But then I would never have learned that tounge +
light socket = ouch!
>Geeze. It's like talking to a brick wall.
I know. But, with time, you might get better.
>> Stop thinking of Linux as a toy might be a good start.
>
>Give me a break. I said I was using Linux as a learning tool to
>learn Linux. Is that making it a TOY? (shrug) I'm done wasting my
>time here.
We'll miss you!
--
Stewart Honsberger (AKA Blackdeath) @ http://blackdeath.tinys.cx
[EMAIL PROTECTED] (Remove 'thirteen' to reply privately)
Humming along under SuSE 6.4, Linux 2.4.0-test4
------------------------------
From: [EMAIL PROTECTED] (David M. Cook)
Subject: Re: Which IDE linux C programers use?
Date: 2 Aug 2000 04:14:41 GMT
On Tue, 1 Aug 2000 19:50:17 -0500, Jerry L Kreps <[EMAIL PROTECTED]> wrote:
>widget placements look? You don't use a wysiwyg design tool? Do you make
>pencil notation of the estimate of the number of pixels you need to move
>something? Or, do you hand code the x,y, dx, dy values determined by a
>manually designed screen pattern?
Most Linux graphical toolkits don't use absolute placement. Gtk, for
instance, uses horizontal and vertical boxes for placement, and also allows
you to use the Tk packer model (which is just a generalization of the box
model) or a table (grid) type of layout. Absolute placement is allowed for
as well. IMO, it's a lot easier to implement your UI layout concept with a
packer model.
Starting with a sketch on paper is a good idea, though, whether you're going
to use a design tool or not.
That said, glade is a really cool graphical layout and code generation tool.
It will even pump out a skeleton diretory for a typical project and populate
it with the standard files. It uses XML for its layout files, and these can
be loaded dynamically, for instance, by Python programs. Together with a
good programming editor (e.g. XEmacs), this makes for a fairly complete
environment.
http://glade.pn.org/
I'm sure KDE fans can speak to the tools available in that case.
Dave Cook
------------------------------
From: Dennis Lee Bieber <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux.mandrake,comp.os.linux.setup
Subject: Re: Partition Problem
Date: Tue, 01 Aug 2000 21:14:05 -0700
On Tue, 01 Aug 2000 20:34:13 GMT, Martin Racette
<[EMAIL PROTECTED]> declaimed the following in
alt.os.linux.mandrake:
>
> Partition Magic tell me this error (this is the exact text from
> PM3.05, and PM 4.0 has the same):
> "The logical drive chain is incompatible. DOS, OS/2, Windows 95, and
> Windows NT require that logical partitions be chained together in
> ascending order. This error occurs when one or more of the logical
> partitions are chained together in the wrong order. Other operating
> systems may not have this requirement. Some versions of Linux fdisk
> chain logical partitions together in the order created. This situation
> is very dangerous and can cause loss of one or more partitions when
> using DOS's FDISK."
>
>
> How can I correct this situation ???
>
Try using Partition Magic 5.0, which understands ext2fs and
linux swap partitions.
--
> ============================================================== <
> [EMAIL PROTECTED] | Wulfraed Dennis Lee Bieber KD6MOG <
> [EMAIL PROTECTED] | Bestiaria Support Staff <
> ============================================================== <
> Bestiaria Home Page: http://www.beastie.dm.net/ <
> Home Page: http://www.dm.net/~wulfraed/ <
------------------------------
From: root <[EMAIL PROTECTED]>
Subject: Re: Carnivore and Privacy: An Oxymoron?
Date: Wed, 02 Aug 2000 04:21:16 GMT
[EMAIL PROTECTED] wrote:
>
> Check out <a
> href="http://www.linuxsecurity.com/feature_stories/feature_story-63.html">thi
> s</a>
>
> Here is a quote from the article:
> "When one really thinks about it, the main reason for computer security
> is data privacy. People protect their systems so that unwanted people
> can't see data they're not authorized to see. Well, what if there was no
> way to protect your privacy because all incoming and outgoing data was
> being viewed by a third party. This is the potential power that the FBI
> wields."
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
At least the FBI doesn't spam! There's a lot of truth
in that article but to be frank I'm much more concerned
with corporate spying and snooping.
Just yesterday I saw an *uncommanded traffic event* put
to one of the mailing lists (forget which one) for
expertise.
------------------------------
From: [EMAIL PROTECTED] (Grant Edwards)
Crossposted-To: alt.os.linux,alt.solaris.x86,comp.os.linux.advocacy
Subject: Re: Learn Unix on which Unix Flavour ?
Date: Wed, 02 Aug 2000 04:31:16 GMT
In article <8m7uju$a43$[EMAIL PROTECTED]>, Tim Prince wrote:
>
>At the risk of getting blown off the planet, I'll mention that
>cygwin/W2K is improving faster, much more attractively priced, and gives
>more opportunities for GPL software and portability to invade the
>Windows world.
>
>> > After they changed their name to Interix, I got a demo CD of
>> > their Open NT (or whatever the name was of their Posix/Gnu/NT)
>> > thing. I never got a chance to install it before it expired,
>> > so I don't know how well it worked. It looked like an
>> > impressive piece of work based on the literature.
>>
>> Most products do look impressive, in the literature.
I just noticed that in the Linux Journal 2000 buyers guide, Microsoft has an
ad for Interix: $99 for the whole thing -- it used to be more like $400.
--
Grant Edwards grante Yow! WHOA!! Ken and
at Barbie are having TOO
visi.com MUCH FUN!! It must be the
NEGATIVE IONS!!
------------------------------
Subject: Re: GUI IDE (WYSIWYG)
From: [EMAIL PROTECTED] (Prasanth A. Kumar)
Date: Wed, 02 Aug 2000 04:32:06 GMT
"Michael Westerman" <[EMAIL PROTECTED]> writes:
> What is avaliable
> i can prog in delphi, java, vb, pascal
>
> i am ready to start in linux but other than jbuilder i don't know what tools
> are avaliable for this.
>
> i am willing to learn a new language eg i have done some c (turbo c++from
> dos )
>
> links, documents and general pointers what ever.
<snip>
Borland is going to bring something similar to Delphi for Linux called
Kylix.
--
Prasanth Kumar
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Mary P)
Subject: Re: Unexpected daily disk activity...
Date: 2 Aug 2000 04:36:36 GMT
Reply-To: [EMAIL PROTECTED]
>Martin Brown wrote:
>>
>> It's nice to know that I am not the only one who is up regularly at
>> 0400. :)
>> --
Sure. The baseline is flatter then :-)
Thanks for the informative replies in this thread.
MP
--
.--- ..- ... - ..-. --- .-. ..-. ..- -.
When a person lives a long time, and then they die
while they're eating a sandwich, then they're eating that
sandwich for ever.
-anonymous second-grader
_
. .
V
// \\
// \\
(W W)
------------------------------
From: "D. C. & M. V. Sessions" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.hardware
Subject: Re: Building a Linux Server from scratch: Experiences?
Date: Tue, 01 Aug 2000 21:37:18 -0700
John Broadhead wrote:
>
> > ' Actually don't get mad about this if it's a UDMA-66 cable. For some
> > ' strange reason that's how they're supposed to be.
> >
> > You're kidding me! Aren't you? The replacement cables that I bought
> > did not have such a cut. Or is the cut done differently?
>
> No, really I'm serious. There are two types of IDE cables. The original
> 40 conductor type, which are good up to UDMA-33, and the new 80
> conductor, yet 40 pin ones. Don't ask me why it has 80 conductors if it
> only has 40 pins, since I don't know.
Ground. To go at UDMA-66 speeds, the system has to have clean signaling
on the cable. Which means that line impedance has to be well-controlled
(to reduce reflections and settling time) and the crosstalk has to be
held down (which means bleeding off the capacitive coupling between
lines in the cable.) Putting ground on every other wire is the easiest
way to do this (and SCSI did it from Day 1) but the original IDE saved
connector pins by skipping nonessentials like that.
Now they have to make up for the original crummy design with funny
connectors and fine-pitch cable. The ground that they get isn't very
good because the connector doesn't provide a low-impedance ground path,
but it's the best that can be done with backward compatibility.
--
| Bogus as it might seem, people, this really is a deliverable |
| e-mail address. Of course, there isn't REALLY a lumber cartel. |
| There isn't really a tooth fairy, but whois toothfairy.com works. |
+----------- D. C. & M. V. Sessions <[EMAIL PROTECTED]> ----------+
------------------------------
From: blowfish <[EMAIL PROTECTED]>
Reply-To: ..
Subject: Re: Goosing the Mouse
Date: Tue, 01 Aug 2000 21:38:57 -0700
"D. C. & M. V. Sessions" wrote:
>
> I hate mouse acceleration. The nonlinearity throws me
> off every time. What I really want is just a mouse with
> adjustable sensitivity, so that moderate mouse movements
> map to a full screen movement.
>
> Yeah, gpm can be set to report a multiplier, but that just
> means that you skip every other pixel or whatever. What I
> want is a more sensitive mouse. Rumor has it that some mice
> can actually be told to increase their sensitivity.
>
> Any suggestions?
>
I don't like mouse. It's the MOST useless thingy ever invented for
computer.
I use a Wacom digitizing tablet, with a pen pointer/drawer.
It might or might not be what you want. But won't hurt to check it out.
- Alex / blowfish.
> --
> | Bogus as it might seem, people, this really is a deliverable |
> | e-mail address. Of course, there isn't REALLY a lumber cartel. |
> | There isn't really a tooth fairy, but whois toothfairy.com works. |
> +----------- D. C. & M. V. Sessions <[EMAIL PROTECTED]> ----------+
--
- If Vi is God's editor. Then, God must have too much free time on his
hands,
lives a very boring and unproductive life; so he needs Vi to waste his
time.
Simplicity rules. That's why I use Easy Edit (ee).
------------------------------
From: blowfish <[EMAIL PROTECTED]>
Reply-To: ..
Crossposted-To: comp.os.linux.networking,comp.os.linux.powerpc,comp.os.linux.security
Subject: Re: Come to my Linux's Website!
Date: Tue, 01 Aug 2000 21:45:29 -0700
WORLOK wrote:
>
> I hate to be critical, but not a very original site. Looks like a
> grammar school project. ...but he gets a B for effort I guess.
>
> STOP using that WINDOWS crap, Sam Tang!
>
Oh mi gosh! He uses M$ Front Page to do a Linux site! :-|
And all the animated .gifs... :-|
-Alex / blowfish.
> -------------------------------------------------------------------
>
> In article <8m4lfr$2hk8$[EMAIL PROTECTED]>,
> "Maarten W.G. Andriessen" <[EMAIL PROTECTED]>
> wrote:
> >
> > "Sam Tang" <[EMAIL PROTECTED]> wrote in message
> > news:8m2ktk$27o$[EMAIL PROTECTED]...
> > > Come to my Linux's Website!
> > > http://www.asia-comp.net/linuxcity/
> >
> > ....he said while posting in Windows with Microsoft Outlook Express
> ;-)))
> >
> > X-Newsreader: Microsoft Outlook Express 4.72.3110.5
> > X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3
> >
> > *busted*
> >
> > Maarten
> >
> >
>
> --
> ================================
> Viva Linux!! Viva La Revoluti�n!
> ================================
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
--
- If Vi is God's editor. Then, God must have too much free time on his
hands,
lives a very boring and unproductive life; so he needs Vi to waste his
time.
Simplicity rules. That's why I use Easy Edit (ee).
------------------------------
From: [EMAIL PROTECTED] (Grant Edwards)
Subject: Re: Which IDE linux C programers use?
Date: Wed, 02 Aug 2000 04:48:43 GMT
In article <[EMAIL PROTECTED]>, Jerry L Kreps wrote:
>>I prefer emacs (+etags +make +gdb) roughly on the console.
>>Especially for code we produced by our self.
>>
>>RH> *I* have found that using a point-and-click programming interfaces tend
>>RH> to slow down code production -- the point-and-click interface has too
>>RH> much interface overhead (lots of excess hand movement, distracting 'eye
>>RH> candy', etc.).
>
>This begs a question...
Not really. "Begging a question" refers to a logical fallacy where somebody
is arguing in support of X, but part or all of their argument is based on
the premise that X is true. [Sorry -- I couldn't resist. Erroneous use of
that phrase is a pet peeve.] ;)
>When you design a graphical screen (form, dialog, or what ever you call it,
>do you have to compile and run just to see how your widget placements look?
Sometimes.
>You don't use a wysiwyg design tool?
Never.
>Do you make pencil notation of the estimate of the number of pixels you need
>to move something?
Pixels? Who worries about pixels? Do you plan where each bit goes when you
write a program?
Calculating things in pixels is useless. It's why things running under MS
Windows have mucked up dialog boxes where things are hidden, overlap, get
clipped by the window, and can't resize properly.
You've got no idea how big a pixel is on your user's screen, you don't know
what fonts are being used in what sizes, and if you're good, you don't even
know how long the strings are.
All decent UI's let let the user configure their own font families and
sizes. You have no way of knowing what size or style font looks good on my
screen. If your program is internationalized, you don't even know what set
of text strings is being used.
>Or, do you hand code the x,y, dx, dy values determined by a manually
>designed screen pattern?
You use geometry-based layout. You groups things into hierarchies. You
specify the relative locations of things with respect to each other. You
specify which things are allowed to grow and shrink in which dimensions.
You let the widget toolkit worry about pixels. I've got better things to do
than count pixels or squint at the screen trying to line up buttons.
--
Grant Edwards grante Yow! I would like to
at urinate in an OVULAR,
visi.com porcelain pool --
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.misc) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Misc Digest
******************************
