In comp.sys.ibm.pc.hardware.storage Daniel James <wastebasket@nospam.aaisp.org> wrote:
> In article news:<5h158kF3i1djcU1@mid.individual.net>, Arno Wagner wrote:
>> What about Linux on routers, NAS, smartphones, v-servers, etc.? Does
>> it count or not? And how does it count? Per suer? Per installation?
>> Per CPU? You cannot really put Windows onto these things, so is it
>> fair if it counts?
> MS would have you believe that Windows CE was a viable OS platform for
> embedded applications ... so -- as long as CE is included in the figures --
> yes, I think it counts.
Good argument. I agree. Hmm. This would seem to give Linux a
few 10 million or so additional installations...
>> Vista is the first time (unless you count ME), that a new OS from
>> Redmont tries to replace something that is adequate (XP).
> Win2k was adequate. XP brought a few security fixes and built-in driver
> support for USB2, firewire, and SATA (in SP2) -- but nothing that couldn't
> have been put into a service pack. More significantly XP brought the
> inconvenience of activation and a cartload of pointless screen-bling.
I never used w2k, but you are right. OTOH, w2k was not the
mass-market end-user product.
>> Then there is Dell with its continued XP offers (obviously direcly
>> in opposition to MS's wishes) and their Linux offers (which are
>> sort of a revolution and certainly also against MS's wishes).
> Credit and kudos to Dell ... though they don't exactly push preinstalled
> linux outside the US.
They are testing the waters, I think.
>> I have not identified any other reason to move to Vista yet ...
> In fairness, the security of Vista *is* better than XP. The fact that a
> default install does not set all users up as administrators is a definite
> improvement, and the fact that UAC popups (annoying though they are) make
> it possible to operate the machine as an unprivileged user without having
> to log off and log back in as administrator every ten minutes or so is very
> welcome. UAC itself is a pain, though ... I hope they manage to fine-tune
> some usability into it in SP1 or SP2 ...
UAC is in fact a step backwards, since most people will click ''yes''
without reading it after a few of those. The other question is the
security improvement is really necessary. Operating XP with
current patches and a virus scanner, Firefox or Opera as browser
and not Outlook to read mail is pretty secure today.
>> What I actually think could be happening at the moment is the
>> commodization (sp?) of the mainstream OS, ...
> You mean "commoditization". "commodization" would mean turning it into a
> piece of furniture that conceals a chamber-pot ... no, come to think of it,
> if that's what you meant you were right the first time <smile>
Hehe. I was drunk when I tried to spell that. Your spelling
is right.
>> if MS wanted to make an actual server OS (no, they do not have
>> one now by all sane standards), then they would need to ...
> Word is that it will be possible to install Windows NT6 server
> ("Longhorn", if it's still called that) without the GUI ... which
> has to be a step in the right direction.
Definitely. It will help bring the remote-administration capabilities
a bit nearer to Unix, which has full remote admin capability (in fact
as standard mode) for decades now. With Unix (or Linux) you can do
practically everything over ssh (or historically rsh) and it is not
more difficult. Then maybe in a decade or so MS will actually have
something that is fit to put on a server.
"Daniel James" <wastebasket@nospam.aaisp.org> wrote in message
news:VA.0000112b.0a60363c@nospam.aaisp.org...
> In article news:<5h158kF3i1djcU1@mid.individual.net>, Arno Wagner wrote:
>> What about Linux on routers, NAS, smartphones, v-servers, etc.? Does
>> it count or not? And how does it count? Per suer? Per installation?
>> Per CPU? You cannot really put Windows onto these things, so is it
>> fair if it counts?
>
> MS would have you believe that Windows CE was a viable OS platform for
> embedded applications ... so -- as long as CE is included in the figures --
> yes, I think it counts.
>
Huh? WinCE is strictly for embedded apps.
WinCE and WinNT are completely different products.
>> Vista is the first time (unless you count ME), that a new OS from
>> Redmont tries to replace something that is adequate (XP).
>
> Win2k was adequate. XP brought a few security fixes and built-in driver
> support for USB2, firewire, and SATA (in SP2) -- but nothing that couldn't
> have been put into a service pack. More significantly XP brought the
> inconvenience of activation and a cartload of pointless screen-bling.
Firewire was in 2K, SATA is supported by the 2K IDE driver.
>"Daniel James" <wastebasket@nospam.aaisp.org> wrote in message
>news:VA.0000112b.0a60363c@nospam.aaisp.org...
>> In article news:<5h158kF3i1djcU1@mid.individual.net>, Arno Wagner wrote:
>>> What about Linux on routers, NAS, smartphones, v-servers, etc.? Does
>>> it count or not? And how does it count? Per suer? Per installation?
>>> Per CPU? You cannot really put Windows onto these things, so is it
>>> fair if it counts?
>>
>> MS would have you believe that Windows CE was a viable OS platform for
>> embedded applications ... so -- as long as CE is included in the figures --
>> yes, I think it counts.
>>
>Huh? WinCE is strictly for embedded apps.
>WinCE and WinNT are completely different products.
>
His point was, if the statistics for Windows "systems"
includes anything deployed running WinCE, then likewise
statistics for Linux "systems" would include embedded apps
like the aforementioned devices.
>>> Vista is the first time (unless you count ME), that a new OS from
>>> Redmont tries to replace something that is adequate (XP).
>>
>> Win2k was adequate. XP brought a few security fixes and built-in driver
>> support for USB2, firewire, and SATA (in SP2) -- but nothing that couldn't
>> have been put into a service pack. More significantly XP brought the
>> inconvenience of activation and a cartload of pointless screen-bling.
>
>Firewire was in 2K, SATA is supported by the 2K IDE driver.
>
As was USB
There were only 3 major revolutionary changes to Windows
once it had become popular enough to be used by the average
person buying a computer, IMO. These were marked by the
introduction of Win95, Win2k, and Vista. Inbetween MS was
just tacking on features and had decided to reserve many as
line-items to promote their then-current OS sales for
98/ME/XP.
In comp.sys.ibm.pc.hardware.storage kony <spam@spam.com> wrote:
> On Sun, 29 Jul 2007 08:55:32 -0700, "Eric Gisin"
> <gisin@uniserve.com> wrote:
>>"Daniel James" <wastebasket@nospam.aaisp.org> wrote in message
>>news:VA.0000112b.0a60363c@nospam.aaisp.org...
>>> In article news:<5h158kF3i1djcU1@mid.individual.net>, Arno Wagner wrote:
>>>> What about Linux on routers, NAS, smartphones, v-servers, etc.? Does
>>>> it count or not? And how does it count? Per suer? Per installation?
>>>> Per CPU? You cannot really put Windows onto these things, so is it
>>>> fair if it counts?
>>>
>>> MS would have you believe that Windows CE was a viable OS platform for
>>> embedded applications ... so -- as long as CE is included in the figures --
>>> yes, I think it counts.
>>>
>>Huh? WinCE is strictly for embedded apps.
>>WinCE and WinNT are completely different products.
>>
> His point was, if the statistics for Windows "systems"
> includes anything deployed running WinCE, then likewise
> statistics for Linux "systems" would include embedded apps
> like the aforementioned devices.
And I was not even talking about embedded Linux. Many of these
small devices run a full Linux kernel.
>>>> Vista is the first time (unless you count ME), that a new OS from
>>>> Redmont tries to replace something that is adequate (XP).
>>>
>>> Win2k was adequate. XP brought a few security fixes and built-in driver
>>> support for USB2, firewire, and SATA (in SP2) -- but nothing that couldn't
>>> have been put into a service pack. More significantly XP brought the
>>> inconvenience of activation and a cartload of pointless screen-bling.
>>
>>Firewire was in 2K, SATA is supported by the 2K IDE driver.
>>
> As was USB
> There were only 3 major revolutionary changes to Windows
> once it had become popular enough to be used by the average
> person buying a computer, IMO. These were marked by the
> introduction of Win95, Win2k, and Vista. Inbetween MS was
> just tacking on features and had decided to reserve many as
> line-items to promote their then-current OS sales for
> 98/ME/XP.
I don't agree that Vista has revolutionary changes.
Arno Wagner wrote:
> And I was not even talking about embedded Linux. Many of these
> small devices run a full Linux kernel.
I'm not trying to jump into your on-going dispute/debate, but I'm
curious as to what you mean by the statement above.
Isn't "embedded Linux" a "full Linux kernel" running on a dedicated
device? I thought that the footprint of such systems were reduced by
paring away libraries, not the most basic part of the OS--the kernel.
See this, possibly non-authoritative, remark from wikipedia:
| Embedded Linux systems combine the Linux kernel with a
| small set of free software utilities.
On Mon, 30 Jul 2007 06:30:35 +0000, Grinder wrote:
> Arno Wagner wrote:
>> And I was not even talking about embedded Linux. Many of these
>> small devices run a full Linux kernel.
>
> I'm not trying to jump into your on-going dispute/debate, but I'm
> curious as to what you mean by the statement above.
>
> Isn't "embedded Linux" a "full Linux kernel" running on a dedicated
> device? I thought that the footprint of such systems were reduced by
> paring away libraries, not the most basic part of the OS--the kernel.
>
> See this, possibly non-authoritative, remark from wikipedia:
>
> | Embedded Linux systems combine the Linux kernel with a
> | small set of free software utilities.
>
> http://en.wikipedia.org/wiki/Embedded_Linux
On an embedded device, you would compile your kernel with only the needed
hardware configured. On a normal distro, the kernel is configured to enable
as much hardware as possible, to serve a variety of users. Usually with
modules that any one computer may never need. So, in my understanding, a
full linux kernel is used for embedded, but only as full as needed to work
on that type of device.
In comp.sys.ibm.pc.hardware.storage Grinder <grinder@no.spam.maam.com> wrote:
> Arno Wagner wrote:
>> And I was not even talking about embedded Linux. Many of these
>> small devices run a full Linux kernel.
> I'm not trying to jump into your on-going dispute/debate, but I'm
> curious as to what you mean by the statement above.
> Isn't "embedded Linux" a "full Linux kernel" running on a dedicated
> device? I thought that the footprint of such systems were reduced by
> paring away libraries, not the most basic part of the OS--the kernel.
Well, it is gradual. For example, I can take my Linux-based
router, add some storage and have a full Linux system with
memory protection and everything. On the other hand, an
embedded Linux system has a reduced kernel with features missing,
typically no MMU, sometimes no filesystem write support, very
limited communication channel or none at all, etc..
> See this, possibly non-authoritative, remark from wikipedia:
> | Embedded Linux systems combine the Linux kernel with a
> | small set of free software utilities.
Well, with that definition if I delete most of my installation
on my PC, I have an instant embedded system. Not really helpful.
True, if you go at this from were it is being used, you will
end up with such a definition. But to distinguish it from
ordinary installations, you have to require the significantly
reduced kernel functionality. After all WinCE is not a full
Windows with just less software installed.
In article news:<uj4qa3hdfvrrkmbgs9b61ieuclre2qkp0c@4ax.com>, Kony wrote:
> There were only 3 major revolutionary changes to Windows
> once it had become popular enough to be used by the average
> person buying a computer, IMO. These were marked by the
> introduction of Win95, Win2k, and Vista.
I's have said: NT 3.1, Win95, and XP64. All other changes have been
evolutionary rather than revolutionary. I certainly don't see XP->Vista as
a revolutionary change.
There has also been a trend in that the more recent evolutionary upgrades
have added less and less that is new and good, and more and more that is
pointless (eye-candy) or unwanted (activation, DRM).
In article news:<5h3kcgF3idn73U2@mid.individual.net>, Arno Wagner wrote:
> UAC is in fact a step backwards, since most people will click ''yes''
> without reading it after a few of those.
I agree that UAC is not well implemented in its current incarnation in
Vasti. It does do things that improve security, though, and for most of
those it doesn't matter if the user does just click blindly on "Yes".
UAC is an awkward beast. On the one hand it adds some security by
preventing some questionable activities, while on the other it actually
reduces security by elevating the privileges of some processes to enable
them to run from a limited user account. The way it decides to do that is
questionable ... automatically elevating a process's privilege just because
it's name includes "setup" is an open invitation to malware writers to call
their malwares "setup"!
Given the size of that hole, it seems odd that some of the hoops you have
to jump through to circumvent the 'security' that UAC imposes are very
small and held very high up. For example: UAC normally prevents a process
from installing a JOURNAL_RECORD hook (the sort of hook you'd want to
install to implement a keyboard macro recorder, for example) because such a
hook might be used by a keylogger. Fair enough, keyloggers are bad and
represent a real security liability for people using their PCs for (say)
home banking; stopping keyloggers is a good thing.
Some apps do have legitimate reasons for wanting to use those hooks to
implement macro recorders, though, and to make that possible UAC allows
some apps to continue to use the hooks ... such an app must be code-signed,
must contain a manifest that says it requires elevated privilege, and must
be installed in a standard place (such as a subdirectory of C:\Program
Files). I can see good reasons for all of that ... but given the weakness
of the security of some of the other things that UAC does this is an
incongruously strict set of requirements.
> Operating XP with current patches and a virus scanner, Firefox or Opera
> as browser and not Outlook to read mail is pretty secure today.
Yes ... especially if you don't use an administrator account for
everything. The trouble is that a default XP install is a right PITA if you
don't use an administrator account ... there are too many things that can't
be "Run As" administrator from a limited account, and too many file
permissions that are set up by default to be just a little too restrictive.
UAC does make limited accounts more usable under Vista.
In comp.sys.ibm.pc.hardware.storage Daniel James <wastebasket@nospam.aaisp.org> wrote:
> In article news:<uj4qa3hdfvrrkmbgs9b61ieuclre2qkp0c@4ax.com>, Kony wrote:
>> There were only 3 major revolutionary changes to Windows
>> once it had become popular enough to be used by the average
>> person buying a computer, IMO. These were marked by the
>> introduction of Win95, Win2k, and Vista.
> I's have said: NT 3.1, Win95, and XP64. All other changes have been
> evolutionary rather than revolutionary. I certainly don't see XP->Vista as
> a revolutionary change.
> There has also been a trend in that the more recent evolutionary upgrades
> have added less and less that is new and good, and more and more that is
> pointless (eye-candy) or unwanted (activation, DRM).
I agree. And I frankly doubt that MS is currently capable of
any major innovation. Not that there seems to be much room for
innovation.
In comp.sys.ibm.pc.hardware.storage Daniel James <wastebasket@nospam.aaisp.org> wrote:
> In article news:<13apej953uk0gb7@corp.supernews.com>, Eric Gisin wrote:
>> Huh? WinCE is strictly for embedded apps.
> My comment was made in reply to Arno Wagner's question:
>> What about Linux on routers, NAS, smartphones, v-servers, etc.?
> Routers and NAS devices are very definitely embedded apps. Smartphones are
> also arguably embedded apps -- and they're certainly things that use WinCE.
Hmm. Byt my Linux NAS happens to be my NAT, firewall and communication
server too. And sometimes compute server. You would call that embedded?
I don't. I call it ''server'' and count it as a full Linux
installation.
The thing is that Linux looses far less functionality in small
installations that anything from MS does.
On 30 Jul 2007 05:53:17 GMT, Arno Wagner <me@privacy.net>
wrote:
>> There were only 3 major revolutionary changes to Windows
>> once it had become popular enough to be used by the average
>> person buying a computer, IMO. These were marked by the
>> introduction of Win95, Win2k, and Vista. Inbetween MS was
>> just tacking on features and had decided to reserve many as
>> line-items to promote their then-current OS sales for
>> 98/ME/XP.
>
>I don't agree that Vista has revolutionary changes.
>
>Arno
I'm not suggesting "revolutionary" is necessarily a good
thing, but there are enough differences (including things
that were being developed but simply not ready by the time
it was released, but expected to be released eventually)
that it is distinctly different enough to effect users/uses.
On 30 Jul 2007 05:53:17 GMT, Arno Wagner <me@privacy.net>
wrote:
>> There were only 3 major revolutionary changes to Windows
>> once it had become popular enough to be used by the average
>> person buying a computer, IMO. These were marked by the
>> introduction of Win95, Win2k, and Vista. Inbetween MS was
>> just tacking on features and had decided to reserve many as
>> line-items to promote their then-current OS sales for
>> 98/ME/XP.
>
>I don't agree that Vista has revolutionary changes.
>
>Arno
I'm not suggesting "revolutionary" is necessarily a good
thing, but there are enough differences (including things
that were being developed but simply not ready by the time
it was released, but expected to be released eventually)
that it is distinctly different enough to effect users/uses.
In comp.sys.ibm.pc.hardware.storage kony <spam@spam.com> wrote:
> On 30 Jul 2007 05:53:17 GMT, Arno Wagner <me@privacy.net>
> wrote:
>>> There were only 3 major revolutionary changes to Windows
>>> once it had become popular enough to be used by the average
>>> person buying a computer, IMO. These were marked by the
>>> introduction of Win95, Win2k, and Vista. Inbetween MS was
>>> just tacking on features and had decided to reserve many as
>>> line-items to promote their then-current OS sales for
>>> 98/ME/XP.
>>
>>I don't agree that Vista has revolutionary changes.
>>
>>Arno
> I'm not suggesting "revolutionary" is necessarily a good
> thing, but there are enough differences (including things
> that were being developed but simply not ready by the time
> it was released, but expected to be released eventually)
> that it is distinctly different enough to effect users/uses.
Well, yes. There is the UI, that is finally almost as fancy as
some window-managers for X-windows were years ago. The security
model is less archaic that that of XP. But I would say these are
gradual changes, not revolutionary ones. Especially as you can
very well go on using the old product without massive
disadvantages.
Of course MS wants you to think this stuff is revolutionary.
It is not. But they are growing desparate, since their OSes
begin to approach reasonable usability, reliability and
performance. MS used to be a master to build products just
good enough to customers would have to upgrade, but bad
enough that there was ample room for further imrovement.
Today, it is well understood what the capabilities of a
modern OS should be. And although MS still has to go far,
it is mostly in things not visible to the end user.
In article news:<5h66g6F3ivdh6U1@mid.individual.net>, Arno Wagner wrote:
> I agree. And I frankly doubt that MS is currently capable of
> any major innovation. Not that there seems to be much room for
> innovation.
I don't see this as an issue of capability ... I don't see how Microsoft
could possibly be /motivated/ to innovate much on the OS front. They have a
huge user base running all kinds of apps -- some from MS themselves and
some third-party -- and those users would HOWL with rage if MS were to
change the OS in any way that stopped those apps being usable.
I see MS patching security bit by bit, and I see them adding more DRM-like
things -- I see them heading towards a software rental licensing model
rather than outright licence sale over the next several years.
As you said: MS used to have a continuous income stream as they brought out
new versions and charged for the upgrade, but now that Windows changes less
and less between versions fewer and fewer people are bothering to upgrade.
Software rental must look very attractive to them right now.
In article news:<5h66kmF3ivdh6U2@mid.individual.net>, Arno Wagner wrote:
> Hmm. Byt my Linux NAS happens to be my NAT, firewall and communication
> server too. And sometimes compute server. You would call that embedded?
> I don't. I call it "server" and count it as a full Linux
> installation.
Yes, certainly, that's a full linux installation and so it does 'count'
toward the total of linux boxes ... but that wasn't the question. The
question was whether the 'commodity' router, firewalls, NAS servers, etc,
running linux (very possibly from ROM) 'counted' ... and those certainly
*are* embedded.
> The thing is that Linux looses far less functionality in small
> installations that anything from MS does.
That's true ... but a little unfair. Wince is actually not a bad little
embedded OS -- it wouldn't be my first choice for an embedded system but I
could work with it -- but MS and their licensees have always chosen to cut
down the functionality of the user interface and apps of every device I've
seen that used it. For example: the pocket office applications in Windows
Mobile didn't *have* to be such an emasculated parody of the desktop office
apps that they insult the intelligence of the user, that's just how
Microsoft chose to specify them -- blame Microsoft, but don't blame the
Wince platform for that, it's capable of better.
(I could make the same criticism of the Hansol Office apps on the
linux-based Sharp Zaurus, for that matter ... why do software vendors
appear to believe that a handheld computer is just a toy, and should only
be allowed to run toy software? Of all vendors, only Psion have ever
displayed an inkling of understanding of their own marketplace)
In article news:<5h66g6F3ivdh6U1@mid.individual.net>, Arno Wagner wrote:
> I agree. And I frankly doubt that MS is currently capable of
> any major innovation. Not that there seems to be much room for
> innovation.
I don't see this as an issue of capability ... I don't see how Microsoft
could possibly be /motivated/ to innovate much on the OS front. They have a
huge user base running all kinds of apps -- some from MS themselves and
some third-party -- and those users would HOWL with rage if MS were to
change the OS in any way that stopped those apps being usable.
I see MS patching security bit by bit, and I see them adding more DRM-like
things -- I see them heading towards a software rental licensing model
rather than outright licence sale over the next several years.
As you said: MS used to have a continuous income stream as they brought out
new versions and charged for the upgrade, but now that Windows changes less
and less between versions fewer and fewer people are bothering to upgrade.
Software rental must look very attractive to them right now.
In comp.sys.ibm.pc.hardware.storage Roger Blake <rogblake10@iname10.com> wrote:
> In article <MPG.21143020883df94698a33b@news.individual.net> , Conor wrote:
>> You've been using Linux since 1994? Riight.
> Sounds reasonable to me. That's around the time I started playing around
> with Linux, using Slackware. (For that matter I've worked with Unix and
> Unix-like systems since the 1970s.)
My first Distro was the "April' 94" SuSE", which I bought together
with some friends the day it hit the stores. Therefore I am pretty sure.
Note: Think first, then check facts, then shoot off mounth.
In comp.sys.ibm.pc.hardware.storage Daniel James <wastebasket@nospam.aaisp.org> wrote:
> In article news:<5h66g6F3ivdh6U1@mid.individual.net>, Arno Wagner wrote:
>> I agree. And I frankly doubt that MS is currently capable of
>> any major innovation. Not that there seems to be much room for
>> innovation.
> I don't see this as an issue of capability ... I don't see how Microsoft
> could possibly be /motivated/ to innovate much on the OS front. They have a
> huge user base running all kinds of apps -- some from MS themselves and
> some third-party -- and those users would HOWL with rage if MS were to
> change the OS in any way that stopped those apps being usable.
So what? MS already has their money. But without significant innovation
MS cannot sell new OS copies...
> I see MS patching security bit by bit, and I see them adding more DRM-like
> things -- I see them heading towards a software rental licensing model
> rather than outright licence sale over the next several years.
This has been their vision for at least a decade now.
Understandably, slavery is better for the slave-holder than
the alternatives.
> As you said: MS used to have a continuous income stream as they
> brought out new versions and charged for the upgrade, but now that
> Windows changes less and less between versions fewer and fewer
> people are bothering to upgrade. Software rental must look very
> attractive to them right now.
Indeed. But I do not think they can actually make it. People
will just use old copies of XP and install it on new hardware.
Drivers are written by manufacturers anyways. Memory growing
beyond 4GB may become an issue, at some point.
But seriously, for a standard office PC, the only reason to
use MS products is ignorance of the alternatives. With higher
pressure that is bound to change.
In comp.sys.ibm.pc.hardware.storage Daniel James <wastebasket@nospam.aaisp.org> wrote:
> In article news:<5h66kmF3ivdh6U2@mid.individual.net>, Arno Wagner wrote:
>> Hmm. Byt my Linux NAS happens to be my NAT, firewall and communication
>> server too. And sometimes compute server. You would call that embedded?
>> I don't. I call it "server" and count it as a full Linux
>> installation.
> Yes, certainly, that's a full linux installation and so it does 'count'
> toward the total of linux boxes ... but that wasn't the question. The
> question was whether the 'commodity' router, firewalls, NAS servers, etc,
> running linux (very possibly from ROM) 'counted' ... and those certainly
> *are* embedded.
Not that clear-cut. Sorry. And there are 'commodity' routers that do
filesharing, web-servering, DNS-caching and other server-stuff.
>> The thing is that Linux looses far less functionality in small
>> installations that anything from MS does.
> That's true ... but a little unfair.
I don't believe so. I think it is entirely fair to point
out that Linux is not the bloatware MS produces.
> Wince is actually not a bad little
> embedded OS -- it wouldn't be my first choice for an embedded system but I
> could work with it -- but MS and their licensees have always chosen to cut
> down the functionality of the user interface and apps of every device I've
> seen that used it. For example: the pocket office applications in Windows
> Mobile didn't *have* to be such an emasculated parody of the desktop office
> apps that they insult the intelligence of the user, that's just how
> Microsoft chose to specify them -- blame Microsoft, but don't blame the
> Wince platform for that, it's capable of better.
> (I could make the same criticism of the Hansol Office apps on the
> linux-based Sharp Zaurus, for that matter ... why do software vendors
> appear to believe that a handheld computer is just a toy, and should only
> be allowed to run toy software? Of all vendors, only Psion have ever
> displayed an inkling of understanding of their own marketplace)
On 01 Aug 2007 02:45:28 GMT, Roger Blake
<rogblake10@iname10.com> wrote:
>In article <MPG.21143020883df94698a33b@news.individual.net> , Conor wrote:
>> You've been using Linux since 1994? Riight.
>
>Sounds reasonable to me. That's around the time I started playing around
>with Linux, using Slackware. (For that matter I've worked with Unix and
>Unix-like systems since the 1970s.)
Well Linux, err Unix, was what anyone was using if they took
even most 101 level college computer classes 12+ years ago.
On 1 Aug 2007 07:37:35 GMT, Arno Wagner <me@privacy.net>
wrote:
>In comp.sys.ibm.pc.hardware.storage Daniel James <wastebasket@nospam.aaisp.org> wrote:
>> In article news:<5h66g6F3ivdh6U1@mid.individual.net>, Arno Wagner wrote:
>>> I agree. And I frankly doubt that MS is currently capable of
>>> any major innovation. Not that there seems to be much room for
>>> innovation.
>
>> I don't see this as an issue of capability ... I don't see how Microsoft
>> could possibly be /motivated/ to innovate much on the OS front. They have a
>> huge user base running all kinds of apps -- some from MS themselves and
>> some third-party -- and those users would HOWL with rage if MS were to
>> change the OS in any way that stopped those apps being usable.
>
>So what? MS already has their money. But without significant innovation
>MS cannot sell new OS copies...
You're ignoring something pretty substantial - most people
do not buy or otherwise independantly seek an OS, rather it
comes preinstalled on their PC and thus the MS monopoly does
guarantee that even without any minor innovations they will
continue to sell new OS copies so long as the industry
remains in it's present state.
Maybe you weren't considering that as a "sale" but it is how
MS makes most of their money on Windows.
>But seriously, for a standard office PC, the only reason to
>use MS products is ignorance of the alternatives. With higher
>pressure that is bound to change.
To the head of IT in an office, the situation is a bit
different, that the system has to run an OS the employees
are productive using without excessive training and minimal
IT intervention to constantly fix problems. If it's the
beancounters making the decision, it could come down to a
per-seat license for the server since Windows itself is a
trivial tech cost (as a %) on a client system coming from an
OEM.
If you're talking about a personal system or one limited to
only use by one technically inclined, the situation
changes... but is that user only using the system as a
"standard office PC"?
On Aug 1, 5:20 pm, kony <s...@spam.com> wrote:
> On 1 Aug 2007 07:37:35 GMT, Arno Wagner <m...@privacy.net>
> wrote:
>
> Well Linux, err Unix, was what anyone was using if they took
> even most 101 level college computer classes 12+ years ago.
>
> If you're talking about a personal system or one limited to
> only use by one technically inclined, the situation
> changes... but is that user only using the system as a
> "standard office PC"?
2 trivial points. Most of my college classes, '90-94 were done on a
VAX or a PC with Win 3.x. My last semester one lab installed
Slackware so they could access Netscape (and Doom;)) and that was my
first experience with any seriously Unix-derived Operating System.
Also, as well as being technically inclined, I'm an artist. As such,
I've been sensitized to how people actually use things, because that's
how I was trained and how so many people I know make their livings:
dyne:bolic is a nifty little package but I neither use it much (having
discovered most of the packages in it through other distros) nor
recommend it because most artists I know would find the packages more
annoying than useful for what they do --'cept gimp which is equally
both. Even as far as a "standard office PC" goes, there's no such
thing. Most people perceive and use their computers so superficially
they do not even know what's on their computers and even I've been
surprised when somebody pokes around.
On Aug 1, 5:20 pm, kony <s...@spam.com> wrote:
> On 1 Aug 2007 07:37:35 GMT, Arno Wagner <m...@privacy.net>
> wrote:
>
> Well Linux, err Unix, was what anyone was using if they took
> even most 101 level college computer classes 12+ years ago.
>
> If you're talking about a personal system or one limited to
> only use by one technically inclined, the situation
> changes... but is that user only using the system as a
> "standard office PC"?
2 trivial points. Most of my college classes, '90-94 were done on a
VAX or a PC with Win 3.x. My last semester one lab installed
Slackware so they could access Netscape (and Doom;)) and that was my
first experience with any seriously Unix-derived Operating System.
Also, as well as being technically inclined, I'm an artist. As such,
I've been sensitized to how people actually use things, because that's
how I was trained and how so many people I know make their livings:
dyne:bolic is a nifty little package but I neither use it much (having
discovered most of the packages in it through other distros) nor
recommend it because most artists I know would find the packages more
annoying than useful for what they do --'cept gimp which is equally
both. Even as far as a "standard office PC" goes, there's no such
thing. Most people perceive and use their computers so superficially
they do not even know what's on their computers and even I've been
surprised when somebody pokes around.
In comp.sys.ibm.pc.hardware.storage kony <spam@spam.com> wrote:
> On 1 Aug 2007 07:37:35 GMT, Arno Wagner <me@privacy.net>
> wrote:
>>In comp.sys.ibm.pc.hardware.storage Daniel James <wastebasket@nospam.aaisp.org> wrote:
>>> In article news:<5h66g6F3ivdh6U1@mid.individual.net>, Arno Wagner wrote:
>>>> I agree. And I frankly doubt that MS is currently capable of
>>>> any major innovation. Not that there seems to be much room for
>>>> innovation.
>>
>>> I don't see this as an issue of capability ... I don't see how Microsoft
>>> could possibly be /motivated/ to innovate much on the OS front. They have a
>>> huge user base running all kinds of apps -- some from MS themselves and
>>> some third-party -- and those users would HOWL with rage if MS were to
>>> change the OS in any way that stopped those apps being usable.
>>
>>So what? MS already has their money. But without significant innovation
>>MS cannot sell new OS copies...
> You're ignoring something pretty substantial - most people
> do not buy or otherwise independantly seek an OS, rather it
> comes preinstalled on their PC and thus the MS monopoly does
> guarantee that even without any minor innovations they will
> continue to sell new OS copies so long as the industry
> remains in it's present state.
> Maybe you weren't considering that as a "sale" but it is how
> MS makes most of their money on Windows.
For the OS, perhaps. For Office I higly doubt that.
>>But seriously, for a standard office PC, the only reason to
>>use MS products is ignorance of the alternatives. With higher
>>pressure that is bound to change.
> To the head of IT in an office, the situation is a bit
> different, that the system has to run an OS the employees
> are productive using without excessive training and minimal
> IT intervention to constantly fix problems. If it's the
> beancounters making the decision, it could come down to a
> per-seat license for the server since Windows itself is a
> trivial tech cost (as a %) on a client system coming from an
> OEM.
Actually I was thinking about all the ''improvements''
in Office, that will require retraining. The retraining to
OOffice could be significantly ceahper and its development
theam does not have any reason to fundamentally change the
interface. In fact FOSS tends to stay static or change very
slowly once it is good enough. It also tends to stay available
for a very long time.
> If you're talking about a personal system or one limited to
> only use by one technically inclined, the situation
> changes... but is that user only using the system as a
> "standard office PC"?
I don't think so. However while MS has used the home PC to get people
to want the same thing at their workplace, the removal of the MS
monopoly could well come from the other direction.
AFAIK MS is earning its money primarily with Office. The OS is very
cheap for OEMs. Office is not. The gaming division (not surprisingly
with this one desaster afer another strategy) is not profitable, and
my guess is it never will be. What else do they have to offer?
In article news:<5harrgF3hu34fU3@mid.individual.net>, Arno Wagner wrote:
> > Yes, certainly, that's a full linux installation and so it does 'count'
> > toward the total of linux boxes ... but that wasn't the question. The
> > question was whether the 'commodity' router, firewalls, NAS servers, etc,
> > running linux (very possibly from ROM) 'counted' ... and those certainly
> > *are* embedded.
>
> Not that clear-cut. Sorry. And there are 'commodity' routers that do
> filesharing, web-servering, DNS-caching and other server-stuff.
I think we both know what I'm trying to say here ... there is a class of
device, that we may call an 'appliance', that uses an embedded linux system to
perform just one function -- be that router, DNS server, DHCP server,
firewall, NAS, or (more usually) some combination of these.
Yes, there are other devices that don't fit that pattern, but they're not what
we were talking about.
> >> The thing is that Linux looses far less functionality in small
> >> installations that anything from MS does.
>
> > That's true ... but a little unfair.
>
> I don't believe so. I think it is entirely fair to point
> out that Linux is not the bloatware MS produces.
I said it was unfair because you're not comparing like with like. When you
compare the linux used to run an 'appliance' you can't compare that with NT --
you have to comapare it with WinCE. CE is quite a decent little RTOS. It's not
fair to tar that with the same 'bloatware' brush that one might apply to XP.
Not that I'm talking about the core WinCE OS, not the "Windows Mobile" crap
that tends to be piled on top of it ... after all, when we talk about embedded
linux we're not talking about Qtopia.
In comp.sys.ibm.pc.hardware.storage Daniel James <wastebasket@nospam.aaisp.org> wrote:
> In article news:<5harrgF3hu34fU3@mid.individual.net>, Arno Wagner wrote:
>> > Yes, certainly, that's a full linux installation and so it does 'count'
>> > toward the total of linux boxes ... but that wasn't the question. The
>> > question was whether the 'commodity' router, firewalls, NAS servers, etc,
>> > running linux (very possibly from ROM) 'counted' ... and those certainly
>> > *are* embedded.
>>
>> Not that clear-cut. Sorry. And there are 'commodity' routers that do
>> filesharing, web-servering, DNS-caching and other server-stuff.
> I think we both know what I'm trying to say here ... there is a
> class of device, that we may call an 'appliance', that uses an
> embedded linux system to perform just one function -- be that
> router, DNS server, DHCP server, firewall, NAS, or (more usually)
> some combination of these. Yes, there are other devices that don't
> fit that pattern, but they're not what we were talking about.
I know indeed what you are trying to say. My claim is that
with Linux installations there is a large grey area that does
usually not exist with commercial OSes.
>> >> The thing is that Linux looses far less functionality in small
>> >> installations that anything from MS does.
>>
>> > That's true ... but a little unfair.
>>
>> I don't believe so. I think it is entirely fair to point
>> out that Linux is not the bloatware MS produces.
> I said it was unfair because you're not comparing like with
> like. When you compare the linux used to run an 'appliance' you
> can't compare that with NT -- you have to comapare it with WinCE. CE
> is quite a decent little RTOS. It's not fair to tar that with the
> same 'bloatware' brush that one might apply to XP.
Why, it is. The thing is that you do not have to resort to
a "little RTOS" to get something Linux-like onto a smaller device.
You can just use the same kernel, that also is put on larger and
very large systems.
> Not that I'm talking about the core WinCE OS, not the "Windows
> Mobile" crap that tends to be piled on top of it ... after all, when
> we talk about embedded linux we're not talking about Qtopia.
Aha. I think there is the misunderstanding. You see, the WinCE kernel
is not the Windows kernel. But the Linux kernel is the Linux
kernel. There is no cut down and adapted version, because the original
thing can be used in an embedded context. Consequentially you have the
option to run a lot of standard Linux software on embedded Linux
installations. This is what makes defining what a ''Linux
installation'' is much more difficult.
In article news:<5he1qkF3kgs69U1@mid.individual.net>, Arno Wagner wrote:
> I know indeed what you are trying to say.
I thought I had succeeded in saying it ... but as long as you understand,
that's the main thing.
> My claim is that with Linux installations there is a large grey area
> that does usually not exist with commercial OSes.
"Grey area"? I'd say rather that linux has the potential to support a wide
range of diverse applications that Windows is not able to address because
of its closed nature, proprietary control, and costly licensing ... but you
can call it a "grey area" if you like <smile>.
> Why, it is. The thing is that you do not have to resort to
> a "little RTOS" to get something Linux-like onto a smaller device.
> You can just use the same kernel, that also is put on larger and
> very large systems.
The whole notion of "same kernel" seems a little strange to me ... I have
three boxes here that spend some or all of their time running linux and I
don't run the same kernel on any two of them: they're all 2.6.19 (at
present) but the II and C3 boxes don't have multiprocessor support compiled
in, the C3 and P4 boxes don't have SCSI drivers compiled in, the PIII and
P4 boxes don't have v4l support compiled in ... and my Zaurus (on 2.4.6) is
different again ...
However, I do know what you mean ... NT and CE aren't the same OS, but the
linux on my Zaurus is substantially the same as the linux on all three
Gentoo boxes. I happen to configure the kernel differently for each box,
but I could run one kernel configuration that would work on all three boxes
(and there would be advantages in doing so, at least for the two that run
IDE drives and could then boot off each others' disks).
> Aha. I think there is the misunderstanding. You see, the WinCE kernel
> is not the Windows kernel. But the Linux kernel is the Linux kernel.
No, I understand that ... I just wasn't sure whether you were going to
insist on making the distinction. This all started with the observation
that if CE counts as Windows then embedded linux devices must count as
linux, for the purposes of counting the size of the installed base. You're
now taking that a step further and saying that CE devices are not what is
meant, generically, by "Windows" but linux appliances *are* linux ... I'd
agree with that but many (Windows Mobile users, especially) would not.
I've a feeling that there are many more linux applicances than there are
Wince devices, so I say we count 'em all!
( ... Tivo boxes are linux appliances ... that's another.)
> This is what makes defining what a ''Linux installation'' is much more
> difficult.
On the contrary ... it makes defining a "linux installation" easier, it
makes defining a "Windows installation" more difficult.
In comp.sys.ibm.pc.hardware.storage Daniel James <wastebasket@nospam.aaisp.org> wrote:
> In article news:<5he1qkF3kgs69U1@mid.individual.net>, Arno Wagner wrote:
>> I know indeed what you are trying to say.
Re: Windows XP optimization tricks 