So you really think Linux is better than Vista, do you?

[Stephan Rose]

On Tue, 27 Mar 2007 01:24:43 +0200, Stephan Rose


I found interesting stuff here:

http://en.wikipedia.org/wiki/File_Allocation_Table

Hmm interesting, thanks for that.

Learn something new everyday. =)

It's funny how no-one lit this fuse in XP, when it comes to malware.

I honestly don't think the fuse was nearly as easy to lit as Vista's. I mean
once you activate XP it's activated...no more issues. I suppose you might
get trouble when going to windows update or something, but I am not aware
of any built-in functionality that shuts the entire OS down Vista style in
XP. At least not without MS sneaking in such a DoS attack via auto update.
That I honestly don't know.

I think the fuse in Vista is significantly shorter and much easier to light
up.

A rootkit (in Windows-speak - you prolly have a clear idea of what it
means from the *NIX tradition) is malware that censors your view of
the system, e.g. so that a Dir /A doesn't show "protected" files,
Ctl+Alt+Del doesn't show "protected" processes, etc.

Ahh intesting, I actually didn't know about that. Haven't been using *nix
all that terribly long yet. =) I switched when MS released their latest
masterpiece.

Any malware can incorporate this functionality, so it's as senseless
to speak of "rootkits" as it is to speak of "viruses" or "worms", the
underlying fallacy being that one malware can't do any combination of
these behaviors and more.

However, the malware industry being as mature as it is, often you get
self-contained rootkits that are used to hide the malware vendor's
real code. It's like buying a reusable code library in "normal"
development, I guess.

By "open-use", I mean that Sony did not even try to ensure that only
their DRM commercial malware would be hidden by the rootkit. Nope,
anything the a wildcard match, e.g. BLAH*.*, would also be hidden - so
it wasn't long before traditional malware coders started to use this
to hide under Sony's "protection". Due dilligence? What's that?

I'd call it insanity!

If a private individual did that, they'd be jailed or at the very
least they'd be legally excluded from PC coding (a la Mitnick).

The courts didn't do that to "trust me, I'm a vendor" Sony, but we
can. I will not buy or resell any Sony goods, and I will be reluctant
to support these even on a pay-per-hour basis ("It's a Sony MP3
player. Call me back when you get something that doesn't suck")

Well I do like Sony's PS2, mostly because a lot of the Final Fantasy games
are on it which I really enjoy. =)

But beyond that..sony has nothing to offer to me I care in any way about.

But I need to see the info that Linux is acting on, right up there
with the filename - not in a "details" view, or "properties" click.

That actually is possible. If I switch my file view from thumbnails / icons
to a listview, then I can add columns to short all sorts of information.

Permissions being one, which would show little details such as if the
execute bit is set or not, read / write access, etc.

I can also turn on the MIME Type column to see as what type the OS
identifies the file.

Yup - but MS doesn't seem to "get" this.

There are a lot of things they don't seem to "get". =)

Ah! Can you "lock UI items"? Eudora can't, and it drives users
insane...

Well..can the API lock UI items? Yes.
Do I have a way to do it for the user implemented yet? No. =)

I am still too deep in development to really worry about that yet.

But yes, once development nears the end I will give the user ability to lock
the various panels once they are where they like.

"I'm going back to Outbreak, Eudora's too hard to use"
' Whaaat? Hard to use? '

Then I go over, and the damn thing's lost the mailboxes, toolbar, and
the least usefull UI element is huge and glued to everything else
while other things you want to use are bouncing around like loose
teeth in a skull. What a mess... of course it's "hard to use"!

Hahahaha! =)

There's no "safe" grey UI space to click anymore - every smudgy
attempt to just select something breaks it off or glues it onto
something else. What I call the "leprocy UI". Don't like.

Something like that is actually not very likely to happen in my case I don't
think, though I suppose there may be a user out there that will prove me
wrong. Time will tell.

There are constraints I can, and do, define how various panels can be
docked. So my toolbars basically stay at the top...the user can just
re-arrange them (actually due to a recent change in the UI there will only
be 1 toolbar left anyway...and I *might* even be able to eliminate that
one).

Then there are 2 panes to the left, one is an explorer type view containing
the project structure and the second pane is the toolpane replacing the
need for a toolbar and popup dialog boxes to edit / create elements. It
basically contains a series of buttons that can be used to activate
commands and beneath the buttons, depending on the active command, various
controls can be added (it is basically space for an insertable child pane)
to edit or create an element.

That approach gives me the ability to never need a dialog box for over 95%
of all tasks and gives me the ability to allow the user to create drawing
elements via manual keyboard input for precise positioning if they so wish.

There also is a 3rd panel at the bottom for error information. Though that
will by default be turned off and normally only shown after an error check
is performed.

For the most part, what the user can customize, or will be able to rather,
is to turn on / off the different panels and change their positions. Either
dock them somewhere else or float them.

I'll have to make a recent screenshot of it and post it. =)

Not to mention low res for optically-challenged, and non-standard
large font sizes on LCDs where anything other than the carved-in-store
res gets blurry due to "text smoothing" effects.


You mean they can be *resized*? MS's UI folks are still struggling
with this new-fangled Win3.yuk feature :-(

Well it automatically sizes the dialog box correctly, the user can't
actually change it unless I wanted to allow the user to.

--
Stephan
2003 Yamaha R6

å›ã®ã“ã¨æ€ã„出ã™æ—¥ãªã‚“ã¦ãªã„ã®ã¯
å›ã®ã“ã¨å¿˜ã‚ŒãŸã¨ããŒãªã„ã‹ã‚‰

 

[cquirke (MVP Windows shell/user)]

//en.wikipedia.org/wiki/File_Allocation_Table[/url]

[/QUOTE]

Learn something new everyday. =)

Me2 - I thought I finally "knew" FATxx after all these years...

I honestly don't think the fuse was nearly as easy to lit as Vista's.

Del WPA.DBA or whatever the file is, and poof! "You must activate in
24 hours (XP Gold) or within 3 days (XP SP1/2)". Pretty easy, given
that the file always has the same name and is easy to locate.

I mean once you activate XP it's activated...no more issues.

See above... the payload's there, just no-one pulls the pin, even
though so many traditional malware coders are anti-MS. It's almost as
if folks were too scared to do this, which implies the ability of MS
to find them and take punitive action, thus posing the question as to
why they aren't already doing so, if they had that ability.

"Blue Suede Shoes" syndrome?

I am not aware of any built-in functionality that shuts the entire OS down

WPA (Windows Product Activation) has always had that power, long
before WGA(N). This subsystem isn't just a one-off activation
process, it's also a constant real-time (or at least, boot time)
monitoring of whether this is the "same" PC that was activated.

I think the fuse in Vista is significantly shorter and much easier to light

The trouble is, no-one's tracking the details on Vista's user-hostile
"validation" technologies, they way we did when XP introduced these
things. So we have no idea where we stand, in the minefield.

In XP, I can use Licenturion's tool to see exactly how many "lives"
(unchanged monitored components) I have left. In Vista, I have no way
to see this, plus I don't know whether the set of components (or their
relative weighting) has changed, or where the tripwire is.

I'm really hoping someone will prove me wrong and say "go to this
site, it's all been documented there", so in a sense I'm "trolling for
clue" (making a contentuous assertion, hoping to be proved wrong with
good information that I need and can use).

Ahh intesting, I actually didn't know about that. Haven't been using *nix
all that terribly long yet. =) I switched when MS released their latest
masterpiece.

Ah, OK. As you know, to us "MS fanbois", "root" is just a particular
directory on a disk volume's file system... NT's security model
peppers its NTFS permissions any way you like; parent/child
application to subtrees is often the result, but one isn't forced to
nest things in particular ways for these permissions to apply.

So when a *NIXer first hears about rootkits in Windows, the first
reaction is usually "huh? I thought you guys didn't have a 'root' "

It's not new, malware's often taken active steps to hide from the
infected OS and it's tools, via actively asserted behavior (as opposed
to static changes such as hidden attributes etc.). But when folks
were reminded what was possible, they coined a new name for what they
saw, and it's been a good "Mr Bogey Man" tern ever since.

I'd call it insanity!

Yup. The courts called it naughty and walloped Sony's writes, but the
fact that Sony still breathes in the DRM and software fronts implies
they were not walloped hard enough.

The arrogance didn't end there, either - Sony's first initial reaction
was "Who cares? Most users wouldn't know what a rootkit was, anyway!"
(IOW, because users are dumb enough to be exploited, we should have
the right to exploit them").

Short of physically walking up to you in the street and smearing feces
in your face, I don't know how they could have been more provocative.

Well I do like Sony's PS2, mostly because a lot of the Final Fantasy games
are on it which I really enjoy. =)

But beyond that..sony has nothing to offer to me I care in any way about.

Would you trust an insane DRM vandor to write the firmware of your DVD
writer or domestic player? How about your MP3 player? Do you like
having to use a proprietary storage format in your cameras and
camcorders? Do you trust a company that stays in denial for years
when its fire-hazard laptop batteries are in use all over the world,
including on airliners and in baggae holds?

Their track record speaks for itself.

That actually is possible. If I switch my file view from thumbnails / icons
to a listview, then I can add columns to short all sorts of information.

Nope, no cigar. If I see the file, I must see the risk level, without
having to switch to some mode that shows fewer files at a time without
scrolling. No "extra step". Always visible, always reliable.

MS OSs could still excel at that, but MS is screwing it up.

Tho I do like the features you describe.

There are a lot of things they don't seem to "get". =)

Yup - the challenge is to pick out the broad top level concepts from
which everything else flows, rather than carry around one's own
personal grab-bag of specific objections ;-)

As opposed to accidentally custimizing it the way they don't want it

Well..can the API lock UI items? Yes.
Do I have a way to do it for the user implemented yet? No. =)

This experienced user would like a safety catch that I can flip off
the one minute a year I decide to change the UI, vs. the 200+ times I
do so by accident (and the 2000+ times I get subconsciously stressed
from the expectation of screwing it up accidentally)

So yep, it would be a Good Thing To Add [TM]

I am still too deep in development to really worry about that yet.

Fair enough. I'd see it as vital polish before RTM, though, else it's
prolly going to come back in the top 10 beefs from tech support.

Something like that is actually not very likely to happen in my case I don't
think, though I suppose there may be a user out there that will prove me
wrong. Time will tell.

Grey UI space can be tricky at the best of times, but when it's
"live", it's really hairy - and I think this will get worse.

Normally, in the age of CRTs, a user with poor vision (e.g. anyone
much over 40, which is a lot of decision-making buyers) would drop to
a lower resolution to make things large enough to see. Those of us
with sprightly eyes would go for higher res so they can scroll less.

But today's LCDs only work properly (OK, "optimally") at a particular
resolution. Folks with poor vision have a choice between sharp and
clear but too small to see, or big enough to see but blurry because
the pixels no longer line up with the LCD's dots.

So instead, these users may keep the current resolution and opt for
larger fonts and icons - thus undermining dialog metric assumptions.

More to the point, the mouse movement thresholds are still scaled in
pixels, with as little as two pixels of movement signifying a desire
to drag-and-drop something. On a display that looks like 640x480b due
to larger UI elements but still running in a far higher resolution,
this translates to a verrrry twitchy mouse, and lots of slip-ups.

There are constraints I can, and do, define how various panels can be
docked. So my toolbars basically stay at the top...the user can just
re-arrange them (actually due to a recent change in the UI there will only
be 1 toolbar left anyway...and I *might* even be able to eliminate that
one).

A canny dodge may be to implimenmt your own pixel threshold, e.g.
movent that isn't over 32 pixels is ignored as a DnD attempt.

Well it automatically sizes the dialog box correctly, the user can't
actually change it unless I wanted to allow the user to.

If there's unbounded info within the dialog (i.e. anything that may
need a scroll bar, especially a horizontal scroll bar, if anything
more than a perfunctionary test data set is used) then IMO it should
always be resizable, and should ideally "remember" its size.

You may have to limit how small something can be sized, to preserve UI
safety against tactics like "This is safe.txt NOT!.pif"
though for that particular example, a floating tooltip is best and a
status bar expansion less best (as used has to look away to see it)

--------------- ----- ---- --- -- - - -

IRC is just multiplayer notepad.

 

[cquirke (MVP Windows shell/user)]

cquirke (MVP Windows shell/user) wrote:

Well the machine I am doing my work on always has an identical copy to what
is on the server. So if the server goes down, my local copy is still there.

Cool, I like that. The installations I generally support are not
server-centric and are more concerned about data survival than
blocking internal access, so I do a similar thing from the server
side, but it's more granular; the backups get done and pulled from the
systems once a night, with 5 day's depth kept on the source system.

The only thing, without backups, that I could loose is the change history in
the event of the server going dead. The current data will always be
preserved as long as I maintain my working copy on my local machine.

OK, same as mine but in reverse. If all PCs except one are stolen,
one set of typically one day old data is available for all PCs. If
the PC holding the data is still alive, there are 5 day's depth of
backups. Not as robust as yours, but still better than default (which
in XP, was "why would home users need Backup?")

It has the ability to backup the SQL Data of course, but that's then just
another step I have to do.

Is the extra step shutting down the engine so data files aren't in
use? That's why I have to exclude SQL data (and make special
arrangements for it) in my systems.

I like!

It's less funny (for me) because it's true... :-/
....but in many other ways, I'm luckier

Tell me more. =)

Google( Bart PE ), then (if interested) un-mung and email me, but
you'd need to use a really clear subject line as I won't have you
white-listed out of the spam-bucket yet.

"Bart programming" would prolly catch my eye

There are two things I'd code if I could:
- .INI-driven generic UI front-end for CLI-driven scanners
- .INI-driven log file processor, to extract and combine data

I already have a wizard that runs a sequence of 8 av scanners from a
non-HD boot, which precludes malware interception. Of those scanners,
about 5 are CLI-driven; one has a GUI front-end, the others I wrap in
a crude batch file front end.

But the real pain is having to wade through and correlate 8 logs, plus
the 3 from the antispyware scanners that follow.

Referring to my cross-platform development or the visual studio thing above?

Referring to the strong Vista-readiness (i.e. if older Windows, no go)

When the time comes, other vendors will still have to support
compromise OK-on-XP developments, and that's going to hurt.

-------------------- ----- ---- --- -- - - - -

"If I'd known it was harmless, I'd have
killed it myself" (PKD)

 

[Stephan Rose]

//en.wikipedia.org/wiki/File_Allocation_Table[/url]

Learn something new everyday. =)

Me2 - I thought I finally "knew" FATxx after all these years...

I honestly don't think the fuse was nearly as easy to lit as Vista's.

Del WPA.DBA or whatever the file is, and poof! "You must activate in
24 hours (XP Gold) or within 3 days (XP SP1/2)". Pretty easy, given
that the file always has the same name and is easy to locate.[/QUOTE]

Hmm now there is something I actually didn't know. I realize the activation
info has to be stored on the computer somewhere, but I didn't expect MS to
do it in such a blatantly stupid way.

See above... the payload's there, just no-one pulls the pin, even
though so many traditional malware coders are anti-MS. It's almost as
if folks were too scared to do this, which implies the ability of MS
to find them and take punitive action, thus posing the question as to
why they aren't already doing so, if they had that ability.

Good point, though I do wonder how many people actually know about it. I
suppose an experienced malware writer would...but how common knowledge is
that file? This is the first time I've ever even heard of it in all my
years of using XP.

"Blue Suede Shoes" syndrome?


WPA (Windows Product Activation) has always had that power, long
before WGA(N). This subsystem isn't just a one-off activation
process, it's also a constant real-time (or at least, boot time)
monitoring of whether this is the "same" PC that was activated.

Fair enough but it is far more trigger happy, and therefore in my mind
unstable, in Vista than XP I'd say. I suppose in XP the only thing one
would have to worry about is the existance of that file.

Vista on the other hand...

The trouble is, no-one's tracking the details on Vista's user-hostile
"validation" technologies, they way we did when XP introduced these
things. So we have no idea where we stand, in the minefield.

Hence why I prefer to keep Vista off my machine and watch the minefield from
a distance. =)

In XP, I can use Licenturion's tool to see exactly how many "lives"
(unchanged monitored components) I have left. In Vista, I have no way
to see this, plus I don't know whether the set of components (or their
relative weighting) has changed, or where the tripwire is.

Well apparently the trip wire is attached to the foot with about 1 mm of
slack...

I'm really hoping someone will prove me wrong and say "go to this
site, it's all been documented there", so in a sense I'm "trolling for
clue" (making a contentuous assertion, hoping to be proved wrong with
good information that I need and can use).

You and me both. I'd love to see that site...but I won't hold my breath!

Ah, OK. As you know, to us "MS fanbois", "root" is just a particular
directory on a disk volume's file system... NT's security model
peppers its NTFS permissions any way you like; parent/child
application to subtrees is often the result, but one isn't forced to
nest things in particular ways for these permissions to apply.

So when a *NIXer first hears about rootkits in Windows, the first
reaction is usually "huh? I thought you guys didn't have a 'root' "

It's not new, malware's often taken active steps to hide from the
infected OS and it's tools, via actively asserted behavior (as opposed
to static changes such as hidden attributes etc.). But when folks
were reminded what was possible, they coined a new name for what they
saw, and it's been a good "Mr Bogey Man" tern ever since.

Thanks for the explanation. =)

Yup. The courts called it naughty and walloped Sony's writes, but the
fact that Sony still breathes in the DRM and software fronts implies
they were not walloped hard enough.

The arrogance didn't end there, either - Sony's first initial reaction
was "Who cares? Most users wouldn't know what a rootkit was, anyway!"
(IOW, because users are dumb enough to be exploited, we should have
the right to exploit them").

Short of physically walking up to you in the street and smearing feces
in your face, I don't know how they could have been more provocative.



Would you trust an insane DRM vandor to write the firmware of your DVD
writer or domestic player? How about your MP3 player? Do you like
having to use a proprietary storage format in your cameras and
camcorders? Do you trust a company that stays in denial for years
when its fire-hazard laptop batteries are in use all over the world,
including on airliners and in baggae holds?

I don't trust anything DRM. Matter of fact, I despise anything DRM. Even the
whole HD DVD / Blue Ray DRM stuff...what's the point? I mean, what's there
to prevent a user from reading the raw data from the respective disk, strip
the DRM..and watch happily ever after?

Their track record speaks for itself.


Nope, no cigar. If I see the file, I must see the risk level, without
having to switch to some mode that shows fewer files at a time without
scrolling. No "extra step". Always visible, always reliable.

Actaully the list view shows more files usually. File icons are relatively
large in thumbnail mode. And you can of course also set your view to list
with the columns of information you like permanently so you don't need to
switch.

MS OSs could still excel at that, but MS is screwing it up.

Tho I do like the features you describe.

Thanks =)

As opposed to accidentally custimizing it the way they don't want it

Good point

Well..can the API lock UI items? Yes.
Do I have a way to do it for the user implemented yet? No. =)

This experienced user would like a safety catch that I can flip off
the one minute a year I decide to change the UI, vs. the 200+ times I
do so by accident (and the 2000+ times I get subconsciously stressed
from the expectation of screwing it up accidentally)

So yep, it would be a Good Thing To Add [TM]

Thanks for the suggestion, consider it done. =)

Fair enough. I'd see it as vital polish before RTM, though, else it's
prolly going to come back in the top 10 beefs from tech support.
Agreed!



If there's unbounded info within the dialog (i.e. anything that may
need a scroll bar, especially a horizontal scroll bar, if anything
more than a perfunctionary test data set is used) then IMO it should
always be resizable, and should ideally "remember" its size.

I avoid creating dialog boxes that need to be resized. If there is any
reason to be able to resize it, I do allow it. But in general, I try to
design them in a way that a person wouldn't ever need to resize them in the
first place.

You may have to limit how small something can be sized, to preserve UI
safety against tactics like "This is safe.txt NOT!.pif"
though for that particular example, a floating tooltip is best and a
status bar expansion less best (as used has to look away to see it)

Not something I really need to particularly worry about in my case. The only
dialogs I have that do any file handling are the OS-Owned file dialogs that
are beyond my control.

My entire application is actually designed to avoid the use of dialog boxes
as much as only possible and only use them where it actually is useful and
necessary.

I find having to use too many dialog boxes to be too disruptive.

I especially get annoyed with applications that nest dialog boxes within
dialog boxes within dialog boxes! One application comes to mind, a
competitor in fact, that has half a dozen ways to get to the same dialog
boxes and it can happen to you that you open Box A via Way A...and 5 other
dialog boxes later, end up with a second copy of Box A via Way B....

--
Stephan
2003 Yamaha R6

å›ã®ã“ã¨æ€ã„出ã™æ—¥ãªã‚“ã¦ãªã„ã®ã¯
å›ã®ã“ã¨å¿˜ã‚ŒãŸã¨ããŒãªã„ã‹ã‚‰

 

[Stephan Rose]

Google( Bart PE ), then (if interested) un-mung and email me, but
you'd need to use a really clear subject line as I won't have you
white-listed out of the spam-bucket yet.

"Bart programming" would prolly catch my eye

Will do, I just googled and briefly looked it over. Looks nice. =) Seeing
how it's 2:30 am right now though, I'll give it a more in-depth look
tomorrow and send ya an e-mail.

There are two things I'd code if I could:
- .INI-driven generic UI front-end for CLI-driven scanners
- .INI-driven log file processor, to extract and combine data

Neither one of those sounds like any particular problem. =)

I already have a wizard that runs a sequence of 8 av scanners from a
non-HD boot, which precludes malware interception. Of those scanners,
about 5 are CLI-driven; one has a GUI front-end, the others I wrap in
a crude batch file front end.

But the real pain is having to wade through and correlate 8 logs, plus
the 3 from the antispyware scanners that follow.

Yup, I can imagine!

Referring to the strong Vista-readiness (i.e. if older Windows, no go)

When the time comes, other vendors will still have to support
compromise OK-on-XP developments, and that's going to hurt.

Well the only thing that Visual Studio links that makes my apps not wanna
play nice with non-Vista systems (unless I statically link) is the crt
stuff which includes all the standard C functions and other functions
needed for the compiler to actually make something useful out of my code.
It's not even anything UI related.

Now MS does have a habit of occasionally upgrading the crt DLLs with new
versions due to bug fixes, optimizations, etc. The pre-Vista version is
verison 7, and the Vista version is version 8.

Now all versions before version 8 could just be re-distributed with the
appliction and things would be fine regardless of what version is installed
on the target system.

But with version 8 that is no longer enough and a seperate installer is
needed to install them. That is where the non-vista compatibility problems
really only come from. Once you install it on XP, it's fine. Statically
link the application to the version 8 crt libraries and it's fine
regardless of installation.

Just can't include the DLLs in the application's directory anymore without
installing them.

It just really made it somewhat more annoying and made me really go "WTF is
wrong now" first time I dealt with it as Visual Studio upgraded to the new
version without even telling me anything about it and i never noticed on my
development system since it automatically got the required DLLs installed
correctly in the upgrade.

I suppose the reason the DLLs can't be included anymore in the same way as
in the past is for security reasons. These things do make kernel calls
(memory allocation, etc.)

I suppose from that perspective it isn't really a bad thing.

--
Stephan
2003 Yamaha R6

å›ã®ã“ã¨æ€ã„出ã™æ—¥ãªã‚“ã¦ãªã„ã®ã¯
å›ã®ã“ã¨å¿˜ã‚ŒãŸã¨ããŒãªã„ã‹ã‚‰

 

[cquirke (MVP Windows shell/user)]

Del WPA.DBA or whatever the file is, and poof! "You must activate in
24 hours (XP Gold) or within 3 days (XP SP1/2)". Pretty easy, given
that the file always has the same name and is easy to locate.

[/QUOTE]

Hmm now there is something I actually didn't know. I realize the activation
info has to be stored on the computer somewhere, but I didn't expect MS to
do it in such a blatantly stupid way.

Same thinking as Sony; 99% of the effort went into ensuring the user
couldn't work around what it does (which is, after all, inherently
user-hostile) and 1% of the effort (if that) into hardening it against
attack. Tho that may be unfair; perhaps that it has so rarely been
attacked, suggests that something behind the scenes is working?

The problem is, once you go user-hostile, the framework of trust
between you and your clients collapses. As clients, we can no longer
count on MS properly documenting what they do, or assist us in getting
it working, etc. Before, failures in these areas may have been
assumed to be a failure of effort; now it's absence of intention.

Good point, though I do wonder how many people actually know about it.

Anyyone who needs to, can and probably does.

Fair enough but it is far more trigger happy, and therefore in my mind
unstable, in Vista than XP I'd say.

I'm looking for hard documentation on this, and have to yet to find
any, which is in itself a problem.

Hence why I prefer to keep Vista off my machine and watch the minefield from
a distance. =)

To make a long story short, the question for me was: Am I ready (as a
system builder and freelance tech with an existing Windows client
base) ready to switch from Windows to a Linux, instead of embracing
Vista, for the next OS period of presumably 5 years?

As the answer to this was No, I'm drilling into Vista as thoughroughly
as I can. Staying with XP is IMO a viable solution for the first 6-12
months of the next OS period of 5 or so years; not a realistic option,
more of a token resistance that leaves you with the first 6 months of
your Vista-era client based stranded on an old OS for that period.

However, I may well be faced with the same question at the end of this
period, if not sooner, with even more user-hostility and/or other
reasons why I may not want to stay with Windows. When that time
comes, I want to be better-positioned to switch off Windows than I am
today, and that is why I'm checking out Linux and also why I continue
to favor applications that are available on Windows and Linux.

Meantime, I'll do what I can to favor cooler heads within MS, so that
hopefully the above doomsday scenario doesn't arise. I'd rather stay
with a Windows that doesn't suck than have to switch away from one
that does; I'd don't want to be chained to a platform if that platform
becomes insupportable, be it technically or ethically.

apparently the trip wire is attached to the foot with about 1 mm of slack...

We need details, rather than emotionally-presented general
impressions. Where's Ben E. when we need him? Y'know, the lawyer
dude who scrupulously monitors and documents what commercial malware
vendors get up to... we need folks like him on this stuff.

I don't trust anything DRM. Matter of fact, I despise anything DRM.

It's as clear a crossing of the line as Poland, Pearl Harbour, New
York Trade Center... basically, the folks you pay to serve your
interests are declaring war on you. Even as they take your money,
they are treating you as the enemy.

Frankly, I don't need to watch Hollydross *that* bad.

The irony is, my DVD collection would be four times the size if it
wasn't for the DVD regions nonsense. I travel, and when I do, I shop,
but what I see and want to buy, I can't use because I'm buying it in
the wrong market ghetto. Globalization, eh? Only when it suits them!

Good point

For every one extra icon or intentionally relocated widget, I see
about 10 accidentally-chaotic UIs

This experienced user would like a safety catch that I can flip off
the one minute a year I decide to change the UI, vs. the 200+ times I
do so by accident (and the 2000+ times I get subconsciously stressed
from the expectation of screwing it up accidentally)

So yep, it would be a Good Thing To Add [TM]

Thanks for the suggestion, consider it done. =)

YAY!! Er... you don't want to join the Eudora open source team and
carry the message there, do you?

OK. UI testing can miss rare but (to affected users) crucial "edge
cases", such as the dude running a 320x420 desktop through ZoomText
(one of my clients has to do this, he's almost blind from diabetes) to
the other extreme of someone working on a 2048 x 1600 desktop.

Also, with time, these "edge cases" can become unexpectedly common,
e.g. a move to compact/handheld devices or in the other direction,
some breakthrough LCD technology that lets screens go large.

Generally, I like your approach for dialogs that do not contain
unbounded content, within the caveats above. Looking at existing
apps, though, screw-ups are common - e.g.

http://cquirke.blogspot.com/2006/12/fair-weather-scanners.html

That blog post refers to "antispyware" scanners that one would
generally want to use from Safe Mode or Bart CDR boot, but they ASSume
high resolutions that are unlikely in those environments. So you have
huge dialogs that position "bottom line" UI elements off-screen, where
they can't be clicked, leaving to guess how many Tab kestrokes it
takes to reach what you again have to guess will be there.

The annoying thing (if you needed another reason to be annoyed) is
that this UI space isn't used to any sort of good effect - it's the
stupid "people will think it's easy to use if we use big dummy
buttons" nonsense. Be still, my heaving lunch.

I avoid creating dialog boxes that need to be resized. If there is any
reason to be able to resize it, I do allow it. But in general, I try to
design them in a way that a person wouldn't ever need to resize them in the
first place.

Uhhh... so how do you manage unbounded content? Do you know how
annoying it is having to scroll through a 640 x 480 dialog bot in one
corner of a 1280x1024 desktop? Or are those the cases where you do
reluctantly resort to dialog resiziability?

Not something I really need to particularly worry about in my case. The only
dialogs I have that do any file handling are the OS-Owned file dialogs that
are beyond my control.

Oh, OK. At least those are resizable, yes?

I find having to use too many dialog boxes to be too disruptive.

True, especially losse concurrent ones. The worst are modal dialogs
that are centered (thus may be behind other centered modals from other
apps) that aren't on top - leading to a "pseudo crash"...
- app only appears to be crashed; it's waiting on a modal dialog
- but the modal dialog is hidden, so user thinks it's a real crash
- so user bad-exits the app and/or OS
- and so, real data damage etc. can result

Another "UI sin" is to leave a trail of open dialogs, with only the
last one (usually modal) able to be re-positioned. If possible, allow
"parent" inactive dialogs to be dragged around, in case what the user
needs to enter in the modal "child" is hidden under the "parent".

I especially get annoyed with applications that nest dialog boxes within
dialog boxes within dialog boxes! One application comes to mind, a
competitor in fact, that has half a dozen ways to get to the same dialog
boxes and it can happen to you that you open Box A via Way A...and 5 other
dialog boxes later, end up with a second copy of Box A via Way B....

That's almost a problem of hypertext theory, and yes; it can be
"powerful", but confusing for support walk-throughs, etc.

What is worse is the FUD that arises when different dialogs alter the
?same settings data, or different processes do the ?same thing.
Classic example; Windows Update vs. Automatic Update.

When writing hypertext, the challenge is to keep "target" content
context-neutral, so that it makes sense no matter what route you took
to reach it. It's quite easy to mess this up...
- you write an article, and realise a keyword needs a backgrounder
- so you spawn a new page and write that detail
- later, you write an unrelated article that uses the same keyword
- you link that to the backgrounder you wrote earlier
- but from that context, the backgrounder doesn't make sense

This is almost "code reusability" for humans ;-)

-------------------- ----- ---- --- -- - - - -

Tip Of The Day:
To disable the 'Tip of the Day' feature...

 

[cquirke (MVP Windows shell/user)]

cquirke (MVP Windows shell/user) wrote:

Will do, I just googled and briefly looked it over. Looks nice. =) Seeing
how it's 2:30 am right now though, I'll give it a more in-depth look
tomorrow and send ya an e-mail.

Fair enough. I've done some time in Bart, so I can quickly bring you
up to speed with at least some of how it works if we get serious.

Neither one of those sounds like any particular problem. =)

Heh - first time I've had that response!

Yup, I can imagine!

Hence ".INI-driven log parser", with the .INI cueing a generic engine
on how to extract info from a particular log. The idea is that one
just creates a new .INI when adding a new scanner to the mix... though
it can be more powerfully general, capturing any other states that
generate logs or parseable log-like material.

Ultimate goal; self-generated service logs...

Well the only thing that Visual Studio links that makes my apps not wanna
play nice with non-Vista systems (unless I statically link) is the crt
stuff which includes all the standard C functions and other functions
needed for the compiler to actually make something useful out of my code.
It's not even anything UI related.

Now MS does have a habit of occasionally upgrading the crt DLLs with new
versions due to bug fixes, optimizations, etc. The pre-Vista version is
verison 7, and the Vista version is version 8.

Is that MSVCRT.DLL?

If so, we're old 'combat buddies' a la...

http://cquirke.mvps.org/9x/dllhell.htm

I suppose the reason the DLLs can't be included anymore in the same way as
in the past is for security reasons. These things do make kernel calls
(memory allocation, etc.)
Ew...

I suppose from that perspective it isn't really a bad thing.

....if it prevents the GDIPlus.dll debacle, I'd be happy. Ever tried
to rid a "mature" consumer PC installation of exploitable GDIPlus.dll?

Chuck Berry knew about this...

--------------- ----- ---- --- -- - - -

My DLL, my DLL, I want to play with *my* DLL

 

[Stephan Rose]

Same thinking as Sony; 99% of the effort went into ensuring the user
couldn't work around what it does (which is, after all, inherently
user-hostile) and 1% of the effort (if that) into hardening it against
attack. Tho that may be unfair; perhaps that it has so rarely been
attacked, suggests that something behind the scenes is working?

The problem is, once you go user-hostile, the framework of trust
between you and your clients collapses. As clients, we can no longer
count on MS properly documenting what they do, or assist us in getting
it working, etc. Before, failures in these areas may have been
assumed to be a failure of effort; now it's absence of intention.

Agreed, hence one of the reason's I am reducing my dependance on anything MS
as much as possible.

I'm looking for hard documentation on this, and have to yet to find
any, which is in itself a problem.

Some hard documentation would be nice, Agreed.

To make a long story short, the question for me was: Am I ready (as a
system builder and freelance tech with an existing Windows client
base) ready to switch from Windows to a Linux, instead of embracing
Vista, for the next OS period of presumably 5 years?

Well you *might* wanna take a peek at the next release of Ubuntu come April
19th. I've heard extremely good things about it with quite a few people
saying that it is good enough to actually really be competition to Vista.

I haven't played with it yet, but I will once it comes out.

I think they simplified a lot of the things and added new UI stuff for quite
a bit of stuff.

One thing I heard about, which I find quite nice is that installing the
nVidia driver is down to basically a single mouse click. Stuff like that..

As the answer to this was No, I'm drilling into Vista as thoughroughly
as I can. Staying with XP is IMO a viable solution for the first 6-12
months of the next OS period of 5 or so years; not a realistic option,
more of a token resistance that leaves you with the first 6 months of
your Vista-era client based stranded on an old OS for that period.

However, I may well be faced with the same question at the end of this
period, if not sooner, with even more user-hostility and/or other
reasons why I may not want to stay with Windows. When that time
comes, I want to be better-positioned to switch off Windows than I am
today, and that is why I'm checking out Linux and also why I continue
to favor applications that are available on Windows and Linux.

Well I can tell ya one thing, there is one type of posts I am seeing quite
frequently on the Ubuntu newsgroup.

"Hi, I am a newbie..."

Meantime, I'll do what I can to favor cooler heads within MS, so that
hopefully the above doomsday scenario doesn't arise. I'd rather stay
with a Windows that doesn't suck than have to switch away from one
that does; I'd don't want to be chained to a platform if that platform
becomes insupportable, be it technically or ethically.



We need details, rather than emotionally-presented general
impressions. Where's Ben E. when we need him? Y'know, the lawyer
dude who scrupulously monitors and documents what commercial malware
vendors get up to... we need folks like him on this stuff.

Don't know who that guy is but...someone definitely needs to be on this
stuff, agreed!

It's as clear a crossing of the line as Poland, Pearl Harbour, New
York Trade Center... basically, the folks you pay to serve your
interests are declaring war on you. Even as they take your money,
they are treating you as the enemy.

Frankly, I don't need to watch Hollydross *that* bad.

The irony is, my DVD collection would be four times the size if it
wasn't for the DVD regions nonsense. I travel, and when I do, I shop,
but what I see and want to buy, I can't use because I'm buying it in
the wrong market ghetto. Globalization, eh? Only when it suits them!

Do what I do, use your computer as your DVD player. Easy to do if you have a
TV that can accept DVI Input or a video card with TV Out.

In case of linux, you don't need to worry about regions...period. It'll just
play *anything*....just need to install libdvdcss (easily installed via
automatix) to be able to decode CSS protected DVDs, or you can also
actually buy a decoder commercially if the legal aspect bugs libdvdcss bugs
you.

In case of windows, there is a utility out there called DVD Region + CSS
Free, essentially does the same thing as libdvdcss, that allows you to
watch any region DVD under windows. Doesn't cost too terribly much and well
worth the investment.

I myself too have DVDs from many different regions. US, Europe and Japan.

YAY!! Er... you don't want to join the Eudora open source team and
carry the message there, do you?

Hahahaha! Ummm...not really lol =)

OK. UI testing can miss rare but (to affected users) crucial "edge
cases", such as the dude running a 320x420 desktop through ZoomText
(one of my clients has to do this, he's almost blind from diabetes) to
the other extreme of someone working on a 2048 x 1600 desktop.

Well a 2048x1600 desktop my application would just embrace, a 320x420
desktop on the other hand I don't think is going to be very suitable for
*any* CAD application. =)

Also, with time, these "edge cases" can become unexpectedly common,
e.g. a move to compact/handheld devices or in the other direction,
some breakthrough LCD technology that lets screens go large.

Generally, I like your approach for dialogs that do not contain
unbounded content, within the caveats above. Looking at existing
apps, though, screw-ups are common - e.g.

http://cquirke.blogspot.com/2006/12/fair-weather-scanners.html

That blog post refers to "antispyware" scanners that one would
generally want to use from Safe Mode or Bart CDR boot, but they ASSume
high resolutions that are unlikely in those environments. So you have
huge dialogs that position "bottom line" UI elements off-screen, where
they can't be clicked, leaving to guess how many Tab kestrokes it
takes to reach what you again have to guess will be there.

The annoying thing (if you needed another reason to be annoyed) is
that this UI space isn't used to any sort of good effect - it's the
stupid "people will think it's easy to use if we use big dummy
buttons" nonsense. Be still, my heaving lunch.

Oh yea I hate UI's with huge buttons. My boss at a company I worked at used
to do that all the time. He'd make these huge dialog boxes with extreme
over-sized buttons on them....say, got a spare bucket for me? My dinner
wants out...

Uhhh... so how do you manage unbounded content? Do you know how
annoying it is having to scroll through a 640 x 480 dialog bot in one
corner of a 1280x1024 desktop? Or are those the cases where you do
reluctantly resort to dialog resiziability?

If I have a dialog box large enough to need 640x480 pixels I am doing
*something* wrong as far as I am concerned. Like I said, I generally try to
avoid dialog boxes totally where possible and if I do make them, I try to
keep their size to a minimum.

Oh, OK. At least those are resizable, yes?

Depends on the OS, nothing I have control over. But so far, I've yet to see
an OS that doesn't allow resizing those.

True, especially losse concurrent ones. The worst are modal dialogs
that are centered (thus may be behind other centered modals from other
apps) that aren't on top - leading to a "pseudo crash"...
- app only appears to be crashed; it's waiting on a modal dialog
- but the modal dialog is hidden, so user thinks it's a real crash
- so user bad-exits the app and/or OS
- and so, real data damage etc. can result

Oh yea, I've dealt with that problem many times. Drives me nuts.

Another "UI sin" is to leave a trail of open dialogs, with only the
last one (usually modal) able to be re-positioned. If possible, allow
"parent" inactive dialogs to be dragged around, in case what the user
needs to enter in the modal "child" is hidden under the "parent".

My competitor software product comes to mind on that one...

--
Stephan
2003 Yamaha R6

å›ã®ã“ã¨æ€ã„出ã™æ—¥ãªã‚“ã¦ãªã„ã®ã¯
å›ã®ã“ã¨å¿˜ã‚ŒãŸã¨ããŒãªã„ã‹ã‚‰

 

[cquirke (MVP Windows shell/user)]

cquirke (MVP Windows shell/user) wrote:

Well you *might* wanna take a peek at the next release of Ubuntu come April
19th. I've heard extremely good things about it with quite a few people
saying that it is good enough to actually really be competition to Vista.

In another thread, I responded to a poster who claimed MS forced you
to churn your Windows versions in order to stay "supported", by
pointing out that Linux doesn't mean longer supportable lifetimes
either. I pointed to Ubuntu as an example, mentioning thier 2008 and
2009 "supported until..." dates for current downloads.

I was flamed for not knowing about an extended-support version of
Ubuntu, so I went back and looked - that's the one that's supported to
2009, as opposed to 2008. Better, but still not better enough to
point fingers at MS's "second-latest-SP" support horizon.

One thing I heard about, which I find quite nice is that installing the
nVidia driver is down to basically a single mouse click. Stuff like that..

Controversial, that. Wikipedia's article on Ubuntu is very
interesting, referring to two controversies; one, that ubunto may
"fork" far enough from Debian that the same installation packages may
no longer work for both (I'd heard about this before), and two, that
including proprietary drivers (such as nVidia, I guess) undermined the
"software should be free and open" ethos.

So while some of us (myself included) may applaud the pragmatism in
accepting and including such drivers even if they come with strings
attached, others may see Ubuntu as failing open source ethics.

Don't know who that guy is

Let's see if I can find him... Search( Ben spyware )... bingo!

http://www.benedelman.org/

Ben Edelman is his name, and I love what he does.

Do what I do, use your computer as your DVD player. Easy to do if you have a
TV that can accept DVI Input or a video card with TV Out.

I don't have a TV, and I can't be bothered to figure out how to get
"All Regions" functionality. It's OK; if the media pimps don't want
me to buy and watch their products, I'll comply with that

In case of linux, you don't need to worry about regions...period. It'll just
play *anything*....just need to install libdvdcss (easily installed via
automatix) to be able to decode CSS protected DVDs, or you can also
actually buy a decoder commercially if the legal aspect bugs libdvdcss bugs
you.

The Wikipedia article on Ubuntu mentioned Automatix, in the context of
possible compatibility issues such as failed updates.

In case of windows, there is a utility out there called DVD Region + CSS
Free, essentially does the same thing as libdvdcss, that allows you to
watch any region DVD under windows. Doesn't cost too terribly much
and well worth the investment.

If I spend money on content, I'd like that money to go to whoever
created the content, not a gang of "conflicting" pimps.

I know that buying a CD or DVD pays the label more than the artist,
but still, it seems ethically preferable to paying someone else who if
anything is set to reduce the artist's revenue.

My stance could change, if the labels get to be a problem to the
extent that destroying them becomes the best pro-artist thing to do.

Well a 2048x1600 desktop my application would just embrace, a 320x420
desktop on the other hand I don't think is going to be very suitable for
*any* CAD application. =)

This is true

Oh yea I hate UI's with huge buttons. My boss at a company I worked at used
to do that all the time. He'd make these huge dialog boxes with extreme
over-sized buttons on them....

There's a thin line between "user-friendly" and "user-condescending",
and the big boys just miss this. Seems like the more graphics and
marketing professionals they call in, the worse it gets.

The best computer book I have EVER read, was the original manual that
came with the ZX Spectrum. This book took you from plugging in the
wires and pressing your first key, to defining how a line of BASIC was
held in RAM in case you wanted to hack the bytes.

But the miracle is that the manual spanned that range without ever
changing gear; there was never a sense of "this is baby stuff, this is
too tough unless you're really keen". It all unfolded, one step after
another, at the same friendly pace. Supurb tech writing!

The old PICK R83 manual was pretty good too; it also went all the way
to hacking file system structures directly, from a fairly low
assumption base, and could be used as your sole documentation. I
wrote a BASIC disassembler (i.e. reading the raw compiled pcode and
displaying this as the stack-based Forth-like language it was) based
on the manual alone. It was resolutely typewriter font ASCII
throughout, though; even the underlining was done with hyphens!

If I have a dialog box large enough to need 640x480 pixels I am doing
*something* wrong as far as I am concerned.

But if you want to list (say) 10 000 files in a folder, surely the
content dictates as much screen area as you can provide, to reduce the
amount of scrolling? Same for a list of anything, where the
horizontal length of each item and the number of items (vertical
height) are unbounded. Often there are so many items that you'd want
a multi-colum "List" view, rather than a "Details".

I know you use the OS dialogs for file ops, and these are hopefully
resizable (as you say, may depend on the OS, with or without caller
control), but other items can be a problem too.

------------ ----- ---- --- -- - - - -

Sig Not Found: Bored with the old one,
can't think of a new one

 

[Charlie Wilkes]

In another thread, I responded to a poster who claimed MS forced you to
churn your Windows versions in order to stay "supported", by pointing
out that Linux doesn't mean longer supportable lifetimes either. I
pointed to Ubuntu as an example, mentioning thier 2008 and 2009
"supported until..." dates for current downloads.

I was flamed for not knowing about an extended-support version of
Ubuntu, so I went back and looked - that's the one that's supported to
2009, as opposed to 2008. Better, but still not better enough to point
fingers at MS's "second-latest-SP" support horizon.

The big difference is that it costs nothing to upgrade Ubuntu.

Charlie

 

[Guest]

Except you can no longer do anything you want. Unlike MS's license Open
Source tries to control people's activities. Write a program in a unix clone
and you may not be able to sell it.

 

[Charlie Wilkes]

Except you can no longer do anything you want. Unlike MS's license Open
Source tries to control people's activities. Write a program in a unix
clone and you may not be able to sell it.

Sure, freedom is no better than tyranny for a starving man. I suppose
from a mercantile standpoint Windows is more desirable. But from a user
standpoint, Windows is on its last legs. Even if Vista is a hit,
Microsoft will face the same problem again in just a few years, i.e., the
desktop market is saturated, and Microsoft's core business is the desktop
OS. So what will they do next? Write an OS that occupies 50 gb of drive
space and takes 8 gb of RAM to run properly, at a cost of $100 billion?

Open source is more in sync with the present-day marketplace. Linux
developers don't waste the work that has already been done. They aim for
incremental refinements, and collectively they offer many specialized
flavors at a time when IT is migrating to devices like music players,
cell phones, automotive gadgets, etc. No one company, however big and
powerful, can dominate all those segments at once. Experience shows that
the most successful businesses are those that narrow their focus to a few
core competencies, and Microsoft is no different. Already it is clear
that Microsoft just another player, and not necessarily the most
successful one, in game platforms, Internet services, etc.

Charlie

 

[cquirke (MVP Windows shell/user)]

The big difference is that it costs nothing to upgrade Ubuntu.

Yup - like an SP. The other big difference is that if source code is
available, others can carry on patching even if the folks who wrote
Ububtu and delevered the support, lose interest.

But there's a 20th-century element of trust in that. How do you know
what the real purpose of patches from arbitrary sources may be?

What bothers me more, is the impact of potentially having to re-do the
OS every six months. How well-behaved is Ubuntu in this respect? Can
you install a newer OS over old and lose no settings or apps, etc.?

 

[cquirke (MVP Windows shell/user)]

I'm not sure if that's true. If you leverage the open source work of
others, then you may inherit restrictions on how you can "sell" it,
but if you write something that doesn't use other people's work, you
could prolly sell it however you like. You may not get included with
the distros, but that wouldn't be what you want anyway.

I suppose from a mercantile standpoint Windows is more desirable.

Maybe, maybe not. Right now it probably is, if you're writing
something to be used on general-purpose PCs, because the PC most users
already have will be on Windows. But if you write a black-box app to
which a PC has to be entirely dedicated, then you could so it either
way... and if you want to provide the hardware, doing it on Linux
means not having to sell Windows as a component.

Windows is on its last legs. Even if Vista is a hit, Microsoft will face
the same problem again in just a few years, i.e., the desktop market
is saturated, and Microsoft's core business is the desktop OS.

As long as PCs are replaced in 3-year cycles, then MS won't have that
problem. The question is, will the need (or "need") to replace PCs
every few years, continue - and if so, why? Will there be compelling
new things to do with PCs that today's boxen can't do?

Open source is more in sync with the present-day marketplace. Linux
developers don't waste the work that has already been done.

Uhhh... I've been told by another poster that MS needs to start from
scratch, writing Windows from hte ground up, in order to "kill all the
bugs". Now I'm told that a Linux strength is that it leverages an
existing code base with a minimum of "scorched earth".

If anything, I might agree with you rather than the other guy, in that
code that works is a very valuable commodity.

But I have to say, I think Linux is a bit like the Missisippi...
Miisissipi... er, Nile, in that it's a slowly-evolving flow that bears
all the kinks and dings of its evolution. Most things are done via
CLI, and most CLI syntax has developed on an ad-hoc "least number of
keys to press" basis. That makes it difficult to use, but also can
lead to slip-ups where the user doesn't get the syntax quite right.

-------------------- ----- ---- --- -- - - - -

Tip Of The Day:
To disable the 'Tip of the Day' feature...

 

[Charlie Wilkes]

As long as PCs are replaced in 3-year cycles, then MS won't have that
problem. The question is, will the need (or "need") to replace PCs
every few years, continue - and if so, why? Will there be compelling
new things to do with PCs that today's boxen can't do?

You have put your finger squarely on the main problem. 1n 1997, every PC
user benefited immensely by upgrading their 1993 (386/486) box to a
Pentium box running Windows 95. But what is the rationale for upgrading
a 2003 rig, with an Athlon XP 2000+ running XP, to a 2007 dual-core setup
running Vista? Many consumers enjoy the vanity and fun of using the
latest technology. But vanity and fun isn't enough to win over corporate
managers and others who take their IT investments seriously.

In truth, the basic problems of desktop productivity have been solved.
Users have all manner of customized shells from which to choose, and any
computer made in the present decade has more than enough power to support
what most people do. Vista, unfortunately, is evidence that Microsoft
does not feel it can afford to accept this reality.

Eventually, new technologies will emerge that require more RAM and CPU
power than today's systems can deliver. One example is 3d printing/rapid
prototyping, which is available now for commercial applications. But,
when such technologies arrive in the mass market, will they require the
services of a massive operating system like Vista? Or will they call for
a nimble OS that supports specialized software, perhaps embedded in a
stand-alone device?

Uhhh... I've been told by another poster that MS needs to start from
scratch, writing Windows from hte ground up, in order to "kill all the
bugs". Now I'm told that a Linux strength is that it leverages an
existing code base with a minimum of "scorched earth".

Well, I think that individual's opinion is poorly thought out. Vista
represents a the work of thousands of programmers and designers working
under great pressure over a compressed time frame. Obviously that is not
a formula to "kill all the bugs."

If anything, I might agree with you rather than the other guy, in that
code that works is a very valuable commodity.

But I have to say, I think Linux is a bit like the Missisippi...
Miisissipi... er, Nile, in that it's a slowly-evolving flow that bears
all the kinks and dings of its evolution. Most things are done via CLI,
and most CLI syntax has developed on an ad-hoc "least number of keys to
press" basis. That makes it difficult to use, but also can lead to
slip-ups where the user doesn't get the syntax quite right.

Well, people have been designing shells to fit over the CLI since the DOS
era. Microsoft has taken it too far, trying to make the PC as simple as
an ATM or a TV set. But computers are inherently complex, and problems
do arise, as millions of Vista early adopters are discovering to their
chagrin. The CLI doesn't over-promise, but it is a powerful tool once a
user takes some time to learn how it works.

Charlie

 

[Stephan Rose]

Except you can no longer do anything you want. Unlike MS's license Open
Source tries to control people's activities. Write a program in a unix
clone and you may not be able to sell it.

That is not correct. I can target Linux as a platform all day long and sell
software for it. *Nothing* in place preventing *anyone* from doing that.

If I directly include some GPL'ed code in my own code, said code is subject
to the GPL license...but I don't need to do that to write code for Linux.

For those exact concerns however the LGPL license was born by the way which
allows integration of libraries into commercial products *without* the
requirement of making any source code available.

That for example allows me to use wxWidgets in my commercial project as it
is LGPL licensed. So by doing that, I can support 3 platforms (Win, Linux,
Mac) and keep my project commercial and my code proprietary.

So you are wrong. You *can* do anything you want.

--
Stephan
2003 Yamaha R6

å›ã®ã“ã¨æ€ã„出ã™æ—¥ãªã‚“ã¦ãªã„ã®ã¯
å›ã®ã“ã¨å¿˜ã‚ŒãŸã¨ããŒãªã„ã‹ã‚‰

 

[Stephan Rose]

I'm not sure if that's true. If you leverage the open source work of
others, then you may inherit restrictions on how you can "sell" it,
but if you write something that doesn't use other people's work, you
could prolly sell it however you like. You may not get included with
the distros, but that wouldn't be what you want anyway.
Correct.


Maybe, maybe not. Right now it probably is, if you're writing
something to be used on general-purpose PCs, because the PC most users
already have will be on Windows. But if you write a black-box app to
which a PC has to be entirely dedicated, then you could so it either
way... and if you want to provide the hardware, doing it on Linux
means not having to sell Windows as a component.

Also in case of embedded projects I am switching to Linux away from
WindowsCE. Saves the licensing fee on every unit sold for WinCE and I can
adapt it to different hardware easier.

As long as PCs are replaced in 3-year cycles, then MS won't have that
problem. The question is, will the need (or "need") to replace PCs
every few years, continue - and if so, why? Will there be compelling
new things to do with PCs that today's boxen can't do?

I honestly doubt it. I mean seriously...PCs are already overpowered for 90%
of the *average user's* tasks! You do not need 2 cores to check e-mail and
write a word document!

The only things I can really see that can benefit from any further
development:

- Games, 3D Graphics still have room to grow as do AI and Physics.
- Commercial / Industrial needs
- CAD, CAD, FEA and the like
- Science needs

The average home user though? They are already smacking flies with a
sledgehammer with *today's* hardware.

--
Stephan
2003 Yamaha R6

å›ã®ã“ã¨æ€ã„出ã™æ—¥ãªã‚“ã¦ãªã„ã®ã¯
å›ã®ã“ã¨å¿˜ã‚ŒãŸã¨ããŒãªã„ã‹ã‚‰

 

[Stephan Rose]

What bothers me more, is the impact of potentially having to re-do the
OS every six months. How well-behaved is Ubuntu in this respect? Can
you install a newer OS over old and lose no settings or apps, etc.?

I will tell you on April 19th when I upgrade to the new release, if not even
earlier if I get tempted and impatient and upgrade to the current beta
release. =)

--
Stephan
2003 Yamaha R6

å›ã®ã“ã¨æ€ã„出ã™æ—¥ãªã‚“ã¦ãªã„ã®ã¯
å›ã®ã“ã¨å¿˜ã‚ŒãŸã¨ããŒãªã„ã‹ã‚‰

 

[Stephan Rose]

In another thread, I responded to a poster who claimed MS forced you
to churn your Windows versions in order to stay "supported", by
pointing out that Linux doesn't mean longer supportable lifetimes
either. I pointed to Ubuntu as an example, mentioning thier 2008 and
2009 "supported until..." dates for current downloads.

Well even if some posters have a problem with that, in neither windows nor
ubuntu do I have a problem with that. I mean new versions are released,
support for old versions eventually is and ultimately has to be dropped.

My primary concerns rather are...what does upgrading mean? Is it a painless
easy process or would I prefer calling my dentist and getting a root kanal?

Controversial, that. Wikipedia's article on Ubuntu is very
interesting, referring to two controversies; one, that ubunto may
"fork" far enough from Debian that the same installation packages may
no longer work for both (I'd heard about this before), and two, that
including proprietary drivers (such as nVidia, I guess) undermined the
"software should be free and open" ethos.

Both are true I suppose but I *think* they are budging on the driver issue.
I almost wanna say, but I am not 100% sure yet, that nVidia's proprietary
driver may actually be included in the new upcoming Release in april.

As far as the packages go, most places that do have debian packages for
ubuntu when dowloading software already list them as Ubuntu packages. Even
seperated by version # (6.06 / 6.10) if necessary.

So while some of us (myself included) may applaud the pragmatism in
accepting and including such drivers even if they come with strings
attached, others may see Ubuntu as failing open source ethics.



Let's see if I can find him... Search( Ben spyware )... bingo!

http://www.benedelman.org/

Ben Edelman is his name, and I love what he does.

I'll have to check that out. =)

The Wikipedia article on Ubuntu mentioned Automatix, in the context of
possible compatibility issues such as failed updates.

Quite possible, Automatix isn't really Ubuntu specific, it supports all
types of distributions. Personally the only thing I use it for is to get my
codecs installed and that is it.

Automatix automatically adds the proper repositories that ubuntu doesn't
have included by default and installs the packages I need for the missing
codecs.

I could actually also do this myself by adding the repositories by hand and
installing the packages myself via Synaptic. Same result...

But if you want to list (say) 10 000 files in a folder, surely the
content dictates as much screen area as you can provide, to reduce the
amount of scrolling? Same for a list of anything, where the
horizontal length of each item and the number of items (vertical
height) are unbounded. Often there are so many items that you'd want
a multi-colum "List" view, rather than a "Details".

It's rare I have such scenarios in my applications. If that happens though,
yes I do use a resizable dialog with a list view. It is rare I have need
for that though.

Usually the only time that will happen in my app is in File open / save
dialogs and those are OS owned anyway.

--
Stephan
2003 Yamaha R6

å›ã®ã“ã¨æ€ã„出ã™æ—¥ãªã‚“ã¦ãªã„ã®ã¯
å›ã®ã“ã¨å¿˜ã‚ŒãŸã¨ããŒãªã„ã‹ã‚‰

 

[Charlie Wilkes]

Yup - like an SP. The other big difference is that if source code is
available, others can carry on patching even if the folks who wrote
Ububtu and delevered the support, lose interest.

But there's a 20th-century element of trust in that. How do you know
what the real purpose of patches from arbitrary sources may be?

I go by intuition. Example... last year at Xmas I helped a kid get her
creative zen to work with XP after the driver install failed. Creative
didn't have anything for me, so I went to a 3d party site with a busy
forum, and from there found my way to a site where someone had put
together a script. I had no way of knowing what might happen... but here
is this guy, volunteering his time to help people out for free, getting
strokes for it, and putting up a site with helpful information. Instinct
told me not to worry, and the script worked like a charm.

Communism has proven to be wholly dysfunctional in the real world, but
something similar works pretty well on the Internet, eh? Bricklayer
today, poet tomorrow, programmer/IT consultant the day after that.
People seem powerfully motivated to collaborate in open source
development and help others solve their IT problems.

What bothers me more, is the impact of potentially having to re-do the
OS every six months. How well-behaved is Ubuntu in this respect? Can
you install a newer OS over old and lose no settings or apps, etc.?

Good question. So far I have installed a large number of upgrades with
no problems. I will upgrade to Feisty on my test rig and then decide if
it makes sense for my "serious" computer.

Charlie