Tuesday, December 10, 2019

Digest for comp.lang.c++@googlegroups.com - 25 updates in 3 topics

Ned Latham <nedlatham@woden.valhalla.oz>: Dec 09 05:19PM -0600

Lynn McGuire wrote:
> > code that tested the graphics card before every write. For the speedup.
 
> The true speedup was moving the graphics device into ring 0. And the
> cause of many crashes due to buggy video device drivers.
 
Way to go, M$!
Ian Collins <ian-news@hotmail.com>: Dec 10 03:09PM +1300

On 10/12/2019 10:50, Ned Latham wrote:
 
>> WoW used the intel vm86 mode to execute pure win3.1 code.
 
> Amazing. I cannot fathom why anyone would want backwards compatibility
> with such a dud.
 
Games.
 
--
Ian.
Ned Latham <nedlatham@woden.valhalla.oz>: Dec 09 08:30PM -0600

Ian Collins wrote:
 
> > Amazing. I cannot fathom why anyone would want backwards compatibility
> > with such a dud.
 
> Games.
 
That's pathetic. (No wonder they called it WoW.)
David Brown <david.brown@hesbynett.no>: Dec 10 09:06AM +0100

On 10/12/2019 03:30, Ned Latham wrote:
>>> with such a dud.
 
>> Games.
 
> That's pathetic. (No wonder they called it WoW.)
 
That is either blatant snobbery, or you are seriously out of touch with
the technological world.
 
Games are not the only use for old software and emulation, but they are
a very substantial part of it.
Ned Latham <nedlatham@woden.valhalla.oz>: Dec 10 03:58AM -0600

David Brown wrote:
 
> > That's pathetic. (No wonder they called it WoW.)
 
> That is either blatant snobbery, or you are seriously out of touch with
> the technological world.
 
Or maybe I'm just aware of how much better modern games are than the games
of those days were.
 
> Games are not the only use for old software and emulation, but they are
> a very substantial part of it.
 
Your reply implied that in this case they are the whole of it. As a reason
for going to the bother of installing a win3.1 emulator, that's pathetic.
David Brown <david.brown@hesbynett.no>: Dec 10 11:09AM +0100

On 09/12/2019 23:35, Lynn McGuire wrote:
>> code that tested the graphics card before every write. For the speedup.
 
> The true speedup was moving the graphics device into ring 0.  And the
> cause of many crashes due to buggy video device drivers.
 
That was for NT 4.0, IIRC. For NT 3.51, graphics drivers were outside
the kernel privilege ring, which meant ring switches for all access by
the graphics drivers and a significantly slower graphics system than Win
3.x on the same hardware. Since a major use of NT was as a workstation
for CAD and other graphics intensive work, this was a definite problem.
Putting the graphics drivers into the closer ring gave a big speed
boost. The risk of total system crashes due to buggy graphics drivers
was a minor problem - there is relatively little benefit in having the
kernel remain alive while the crashed gui restarts, compared to simply
restarting the whole machine.
David Brown <david.brown@hesbynett.no>: Dec 10 11:18AM +0100

On 10/12/2019 10:58, Ned Latham wrote:
>> the technological world.
 
> Or maybe I'm just aware of how much better modern games are than the games
> of those days were.
 
There are different opinions on that - people vary significantly on the
kind of game they like. You don't get many new players of old games,
but old players can often keep their old games for quite a while - long
after they have changed to newer versions of other software. And WoW
comes from the early NT days - the main consumer OS was 16-bit Windows
(with gradually increasing 32-bit proportions) until Windows XP in 2001.
 
>> a very substantial part of it.
 
> Your reply implied that in this case they are the whole of it. As a reason
> for going to the bother of installing a win3.1 emulator, that's pathetic.
 
That comment implies you have mixed up posters in this thread.
 
But yes, I expect that playing old games is a fairly common reason (but
certainly not the only reason) for installing a Win 3.1 emulator on a
newer system. That does not mean it is something many people will do in
absolute terms.
 
And no, it is not pathetic. You don't get to judge people's preferences
as "pathetic" just because they have a favourite game that was written
long ago. That is definitely snobbery.
Ned Latham <nedlatham@woden.valhalla.oz>: Dec 10 05:37AM -0600

David Brown wrote:
> after they have changed to newer versions of other software. And WoW
> comes from the early NT days - the main consumer OS was 16-bit Windows
> (with gradually increasing 32-bit proportions) until Windows XP in 2001.
 
I thought about reasons for bothering to create a win3.1 emulator; it
occurs to me that there would have been aq lot of corporate migration
from 3.1 to NT.
 
 
> > Your reply implied that in this case they are the whole of it. As a reason
> > for going to the bother of installing a win3.1 emulator, that's pathetic.
 
> That comment implies you have mixed up posters in this thread.
 
You're right. It was Ian Collins.
 
 
> And no, it is not pathetic. You don't get to judge people's preferences
> as "pathetic" just because they have a favourite game that was written
> long ago. That is definitely snobbery.
 
Don't put words in my mouth. Especially when *my* words are still here, in
this post, giving the lie to your crap.
boltar@nowhere.co.uk: Dec 10 01:12PM

On Mon, 09 Dec 2019 11:39:13 -0600
>of the M$ OS 2 team joined them. Note that DEC had control and use of NT
>until well after delivery. Though it was written *for* M$, it took coding
>far beyond the quality and capabilities that are the usual M$ fare.
 
Are you talking about NT here or some reliable and fast OS that MS never
released? Win NT was a blue screening, slow joke of an OS and frankly I'd have
expected better from people who worked on VMS.
boltar@nowhere.co.uk: Dec 10 01:17PM

On Tue, 10 Dec 2019 11:09:12 +0100
 
>was a minor problem - there is relatively little benefit in having the
>kernel remain alive while the crashed gui restarts, compared to simply
>restarting the whole machine.
 
Clearly you're not familiar with the operational concept of a "Server" - ie you
don't kill the backend just because the ooo-shiny! at the front has gone down.
But then neither is or was Microsoft so you're made for each other.
Bart <bc@freeuk.com>: Dec 10 01:54PM


> Are you talking about NT here or some reliable and fast OS that MS never
> released? Win NT was a blue screening, slow joke of an OS and frankly I'd have
> expected better from people who worked on VMS.
 
IIRC didn't the original NT (3.5?) boot up with a blue screen anyway?
 
But the 'blue-screen' was just how MS chose to report detected system
errors. I'm sure other OSes could also crash, but just didn't do so with
a blue display. (Android might just become unresponsive, or the display
goes crazy. Or just goes completely dead.)
 
MS Windows was also a little unusual as it ran on consumer hardware that
could be assembled from assorted third party components: motherboard,
graphics, controllers, which came with their own third party drivers. It
didn't have as much control as, say, Apple.
 
As to its speed, yes these days Windows is a monstrous bloated mess of
an OS. I use it but just try and ignore most of it as much as I can.
(Eg. right-click on some file icon, and you can count to 10 - or
sometimes even longer - before it shows a menu. WTH is it up to?)
David Brown <david.brown@hesbynett.no>: Dec 10 03:49PM +0100

On 10/12/2019 12:37, Ned Latham wrote:
 
> I thought about reasons for bothering to create a win3.1 emulator; it
> occurs to me that there would have been aq lot of corporate migration
> from 3.1 to NT.
 
Sure, that was critical at the time - lots of software was written for
16-bit Windows, and people needed to run it on 32-bit NT. And it became
steadily less relevant over time. You are unlikely to find many people
wanting to run 16-bit Windows software /now/ in a corporate setting. If
you look at tools aimed at running old software, there is a strong focus
on games - because that is a big use-case.
 
>> long ago. That is definitely snobbery.
 
> Don't put words in my mouth. Especially when *my* words are still here, in
> this post, giving the lie to your crap.
 
You have twice said that it is "pathetic" that people would use Win 3.1
support software to run games. How is that to be interpreted other than
as saying it is "pathetic" that people would want to run old games?
 
You can answer that if you like, but I think this thread branch is going
nowhere and is clearly off-topic.
David Brown <david.brown@hesbynett.no>: Dec 10 03:55PM +0100


> Clearly you're not familiar with the operational concept of a "Server" - ie you
> don't kill the backend just because the ooo-shiny! at the front has gone down.
> But then neither is or was Microsoft so you're made for each other.
 
Windows NT was not initially targeted as a server OS, but as a
workstation OS. No one used Windows seriously for servers at the time -
it was Unix or Netware. By the time Windows NT was a common choice for
a server, the integration of the graphics drivers in the kernel layer
was a done deal. And yes, it was a potential problem that a crashed gui
meant a crashed server - but on servers you don't usually install high
performance, high risk graphics drivers.
 
(Don't misunderstand me - I think having the risky third-party driver
code running at kernel level is a bad idea from the reliability
viewpoint, and a terrible idea on a server. And when I set up a server,
I do it using an OS that doesn't have a gui at all.)
Paavo Helde <myfirstname@osa.pri.ee>: Dec 10 05:22PM +0200

On 10.12.2019 15:54, Bart wrote:
> an OS. I use it but just try and ignore most of it as much as I can.
> (Eg. right-click on some file icon, and you can count to 10 - or
> sometimes even longer - before it shows a menu. WTH is it up to?)
 
FYI: these are shell extensions installed by various third-party
software packages and which maybe want to add items in that context
menu. Windows itself is not much to blame here (apart of not
automatically deactivating slow or non-functional extensions).
 
Things become especially interesting (i.e. boring) when some shell
extension wants to access a network drive which is not there any more.
But 10 s seems a bit fast for this.
 
In principle you can clean shell extensions up yourself in the registry.
Probably there are also tools for that.
boltar@nowhere.co.uk: Dec 10 03:26PM

On Tue, 10 Dec 2019 15:55:31 +0100
>workstation OS. No one used Windows seriously for servers at the time -
>it was Unix or Netware. By the time Windows NT was a common choice for
>a server, the integration of the graphics drivers in the kernel layer
 
MS were marketing it as a server from version 3.51 IIRC. Sure, no one took
it seriously, but if that is the best the team from DEC could do then clearly
VMS was a good OS despite them working on it, not because of them. Presumably
the main VMS core was written before they came along. And in that vein lets
hope Linux can survive the mess that Poettering is currently trying to make of
it.
 
>was a done deal. And yes, it was a potential problem that a crashed gui
>meant a crashed server - but on servers you don't usually install high
>performance, high risk graphics drivers.
 
MS command line was such a useless waste of time (and still is really even with
Powershell) that there was little choice but to have some sort of graphics
driver to drive the GUI.
boltar@nowhere.co.uk: Dec 10 03:35PM

On Tue, 10 Dec 2019 17:22:20 +0200
>software packages and which maybe want to add items in that context
>menu. Windows itself is not much to blame here (apart of not
>automatically deactivating slow or non-functional extensions).
 
Allow 3rd party extensions to highjack a fundamental piece of OS shell
functionality is beyond f*cking stupid, but then thats MS for you.
Ned Latham <nedlatham@woden.valhalla.oz>: Dec 10 11:32AM -0600

boltar wrote:
 
> Are you talking about NT here or some reliable and fast OS that MS never
> released? Win NT was a blue screening, slow joke of an OS and frankly I'd have
> expected better from people who worked on VMS.
 
NT, yes. As I said in another post, on taking delivery, M$ set about
"improving" it. It was fine on DEC hardware.
Ned Latham <nedlatham@woden.valhalla.oz>: Dec 10 11:43AM -0600

David Brown wrote:
> Ned Latham wrote:
> > David Brown wrote:
> > > Ned Latham wrote:
 
----snip----
 
 
> You have twice said that it is "pathetic" that people would use Win 3.1
> support software to run games. How is that to be interpreted other than
> as saying it is "pathetic" that people would want to run old games?
 
I said, and I quote, "[the] reply indicated that in this case [games] are
the whole of it. AS A REASON FOR GOING TO THE BOTHER OF INSTALLING A
WIN3.1 EMULATOR, that's pathetic".
Cholo Lennon <chololennon@hotmail.com>: Dec 10 03:17PM -0300

On 10/12/19 12:22, Paavo Helde wrote:
> But 10 s seems a bit fast for this.
 
> In principle you can clean shell extensions up yourself in the registry.
> Probably there are also tools for that.
 
Autoruns from MS Sysinternals helps with that awful problem.
 
--
Cholo Lennon
Bs.As.
ARG
Keith Thompson <kst-u@mib.org>: Dec 10 10:19AM -0800

> David Brown wrote:
[...]
 
> I said, and I quote, "[the] reply indicated that in this case [games] are
> the whole of it. AS A REASON FOR GOING TO THE BOTHER OF INSTALLING A
> WIN3.1 EMULATOR, that's pathetic".
 
And this has what exactly to do with C++, the topic of this newsgroup?
 
--
Keith Thompson (The_Other_Keith) kst-u@mib.org <http://www.ghoti.net/~kst>
Working, but not speaking, for Philips Healthcare
void Void(void) { Void(); } /* The recursive call of the void */
Ned Latham <nedlatham@woden.valhalla.oz>: Dec 10 12:50PM -0600

Keith Thompson wrote:
> > the whole of it. AS A REASON FOR GOING TO THE BOTHER OF INSTALLING A
> > WIN3.1 EMULATOR, that's pathetic".
 
> And this has what exactly to do with C++, the topic of this newsgroup?
 
Shit happens.
Bonita Montero <Bonita.Montero@gmail.com>: Dec 10 11:28AM +0100

#include <iostream>
#include <algorithm>
 
using namespace std;
 
struct Inner
{
Inner() = default;
Inner( Inner &&other )
{
cout << "Inner( Inner &&other )" << endl;
}
Inner &operator =( Inner &&other )
{
cout << "Inner &operator =( Inner &&other )" << endl;
return *this;
}
};
 
struct Outer
{
Inner i;
};
 
int main()
{
Outer o1, o2;
swap( o1, o2 );
}
 
The Inner-object of Outer is moved although there is no move-constructor
or move-assignment-operator of Outer. Can anyone tell me where the C++
-standard does specifify this?
Paavo Helde <myfirstname@osa.pri.ee>: Dec 10 12:39PM +0200

On 10.12.2019 12:28, Bonita Montero wrote:
 
> The Inner-object of Outer is moved although there is no move-constructor
> or move-assignment-operator of Outer. Can anyone tell me where the C++
> -standard does specifify this?
 
15.8.1/8 [class.copy.ctor]
 
"If the definition of a class X does not explicitly declare a move
constructor, a non-explicit one will be implicitly declared as defaulted
if and only if
(8.1) — X does not have a user-declared copy constructor,
(8.2) — X does not have a user-declared copy assignment operator,
(8.3) — X does not have a user-declared move assignment operator, and
(8.4) — X does not have a user-declared destructor.
"
 
Your struct Outer matches all these points.
Bonita Montero <Bonita.Montero@gmail.com>: Dec 10 01:44PM +0100

> (8.4) — X does not have a user-declared destructor.
> "
> Your struct Outer matches all these points.
 
That's what I thought about an hour after posting.
T <T@invalid.invalid>: Dec 10 04:07AM -0800

On 2019-12-04 01:19, Geoff wrote:
> }
> return (FALSE);
> }
 
Thank you!
You received this digest because you're subscribed to updates for this group. You can change your settings on the group membership page.
To unsubscribe from this group and stop receiving emails from it send an email to comp.lang.c+++unsubscribe@googlegroups.com.
Posted by at

No comments:

Post a Comment

Subscribe to: Post Comments (Atom)