Does anyone know (including QNX) if there will ever be native SATA support
for QNX 4? Many newer compters are abandoning IDE emulation.
Thanks
Not a solution; just some food for thought:
Guess you’ve probably got the suggestion to try vertualization? I mean
look at this logically: if the computer abandons IDE emulation, it is
probably of a generation powerfull enough to implement a VM which is
quite suiteable to run QNX4 on. (QNX4 being a very efficient OS capable of
running perfectly on computers of yesteryears.)
That said, I do understand that in some cases you can’t use VM’s. (Embeded
realtime systems, for instance.)
regards,
rick
Jeffrey Adler jeffadler_at_bigfoot_dot_com@127.0.0.1 wrote:
Does anyone know (including QNX) if there will ever be native SATA support
for QNX 4? Many newer compters are abandoning IDE emulation.Thanks
What is vertualization? Where can I get such software?
“Rick Lake” <rick@private-domain.invalid> wrote in message
news:eqvunr$6ak$1@inn.qnx.com…
Not a solution; just some food for thought:
Guess you’ve probably got the suggestion to try vertualization? I mean
look at this logically: if the computer abandons IDE emulation, it is
probably of a generation powerfull enough to implement a VM which is
quite suiteable to run QNX4 on. (QNX4 being a very efficient OS capable of
running perfectly on computers of yesteryears.)That said, I do understand that in some cases you can’t use VM’s. (Embeded
realtime systems, for instance.)regards,
rickJeffrey Adler jeffadler_at_bigfoot_dot_com@127.0.0.1 wrote:
Does anyone know (including QNX) if there will ever be native SATA
support
for QNX 4? Many newer compters are abandoning IDE emulation.Thanks
“Jeffrey Adler” <jeff.nospam.adler@bigfoot.com> wrote in message
news:er0dev$erd$1@inn.qnx.com…
What is vertualization? Where can I get such software?
I think rick is talking about software such as VMWare?
Problem with this is you can’t use any PCI devices aside the one
supported/emulated by VMWare. If you have an IO card for example you are
out of luck.
“Rick Lake” <> rick@private-domain.invalid> > wrote in message
news:eqvunr$6ak$> 1@inn.qnx.com> …
Not a solution; just some food for thought:Guess you’ve probably got the suggestion to try vertualization? I mean
look at this logically: if the computer abandons IDE emulation, it is
probably of a generation powerfull enough to implement a VM which is
quite suiteable to run QNX4 on. (QNX4 being a very efficient OS capable
of
running perfectly on computers of yesteryears.)That said, I do understand that in some cases you can’t use VM’s.
(Embeded
realtime systems, for instance.)regards,
rickJeffrey Adler jeffadler_at_bigfoot_dot_com@127.0.0.1 wrote:
Does anyone know (including QNX) if there will ever be native SATA
support
for QNX 4? Many newer compters are abandoning IDE emulation.Thanks
\
It is always possible for there to be Native SATA support for QNX 4,
however it is very very unlikely to come from QSSL. There are at
least a couple old timers out there with the expertise to do this,
but I think one of them would need an incentive. This type of
thing did occur with QNX 2, with respect to (for instance) the EIDE
LBA driver. 
“maschoen” <maschoen@pobox-dot-com.no-spam.invalid> wrote in message
news:er1b00$4eo$1@inn.qnx.com…
It is always possible for there to be Native SATA support for QNX 4,
however it is very very unlikely to come from QSSL. There are at
least a couple old timers out there with the expertise to do this,
but I think one of them would need an incentive. This type of
thing did occur with QNX 2, with respect to (for instance) the EIDE
LBA driver. >
I am faced with choices.
With support for QNX4 going the way of QNX2, yes, it is possible to buy a
couple of more years of QNX4, but ultimately I think that the computer
hardware from the big names (Dell, for example) will keep moving away from
the tried and true.
The big issues for me this last “new computer go-around” was 1) SATA drives
with no ide emulation, and 2) USB keyboard with its own emulation problems.
SO…
Without a reasonable commitment from QNX to continue some rudimantary
support for QNX4 on the newer hardware (been there with QNX2 already), I
must now (sooner, not later) look at porting my application to something.
The option of providing some incentive ($) to someone for a native SATA
driver might buy a year or two of product life, but it doesn’t appear to be
the solution; only the band-aid.
My reasonable choices are QNX6, Linux, or Windows.
With QNX6, the porting development costs are reasonably low but the
deployment costs to upgrade my existing customers are high. Also, there are
few guarantees of long-term support.
With Linux or Windows, there is a much better chance of hardware support
continuing (MUCH higher with Windows).
I am looking at trying to implement some kind of similar S/R/R model under
Windows. I think that is where my path will lead me.
Thanks, all.
On Thu, 15 Feb 2007 21:33:38 +0300, Jeffrey Adler
<jeff.nospam.adler@bigfoot.com> wrote:
I am looking at trying to implement some kind of similar S/R/R model
under Windows.
Why not on NetBSD?
Tony.
In a nutshell: virtualization SW runs on a host. With it you create
virtual machines on that host, which emulate real machines. Some of them
are pretty convincing: they boot up with a familiar BIOS and then you
install your favorite OS on them. I know for a fact that QNX4 runs on it.
And I wouldn’t be surprised if even QNX2 can run on it. [Can anyone verify
this?] Just google a bit and I’m sure you’ll hit all the info you need.
I didn’t mention any virtualization vendors at first because I wanted to
stay objective. But yes, as Mario said in another follow-up, VMWare is one
of them. They have a free version (I believe for non-commercial use), and
also the more professional versions for serious work for which you have to
pay license costs. Their software runs both on Windows and Linux hosts,
and they also have a native stand-alone version. (Which is essentially
based on a modified Linux.) They’re probably the biggest player on the X86
platform and they had the monopoly for a long time.
If you choose Linux as the host, you also have another option: Xen, which
is an open source virtualization sw solution. Haven’t really checked it
out, but it’s a fast growing competitor. (If you ask me, IMHO that’s what
prompted VMWare to release a free version of their sw.)
Of course, as Mario pointed out, because of the nature of VM’s, you don’t
have full control over special hardware. Your VM depends on what the
virtualization software offers it. But for applications which depend on
standard resources such as cpu, memory, storage, network, mouse, keyboard
and soundblaster, VM’s are ideal. Plus you get the bonus of manageability:
you can duplicate VM’s just by copying the VM file. (Talk about easy
backups 
Thus creating a duplicate test machine to play with. Or you
can just take a snapshot of your running machine, try out something very
dangerous, and if it fails, just revert back to your snapshot and your
machine is back in the state it was at the time you took the snapshot.
(Pretty cool . A second bonus is consolidation: you only need one heavy
duty host (or two to avoid single point of failure).
I can go on and on, but I think it’s better to check out the vendor’s
sites and let them speak for themselves.
Good luck,
regards,
rick
Jeffrey Adler <jeff.nospam.adler@bigfoot.com> wrote:
What is vertualization? Where can I get such software?
“Rick Lake” <> rick@private-domain.invalid> > wrote in message
news:eqvunr$6ak$> 1@inn.qnx.com> …
Not a solution; just some food for thought:Guess you’ve probably got the suggestion to try vertualization?
You have an interesting perspective on your porting possibilities,
which I don’t share exactly.
I agree that the cost to move to QNX 6 is probably the highest. If
you require real time in your application, neither Windows or Linux
will work.
As far as future support, the difference in the way QNX and Windows
work is illusary. For example, if you want USB support under Win98,
you are sol. Like QNX, Microsoft periodically does a major upgrade
that can be incompatible to legacy software. To their credit,
their interface has been fairly backward compatible. It is unlikely
that QNX will undergo another major move like QNX 4 to QNX 6. While
Dan might be dreaming something like this up, it is doubtful that the
new owners would go for it. He’ll just have to start a new company.
Your biggest concern about QNX should be the viability of the company.
In spite of the fact that they are now part of a major public
company, I feel that this has diminished. If the market were to
turn against QNX it is possible that Harmony would cut and run.
This was probably less so when QNX was private. Windows on the other
hand has what I see as a more serious problem. The periodic major
upgrades, WinME->XP, XP->Vista seem to always come with another
requirement, an order of magnitude increase in processor speed. I
was in a Best Buy last week looking at Dual Core laptops running
Vista. I was shocked at how slow they were. Neither QNX nor
Linux provide this ah, feature. So while you can count on future
hardware support, you might need to replace your
processor/motherboard with something closer to the state of the art,
even though your application has no such requirement.
So I would say that Linux is a good bet if you don’t need real time.
Otherwise you will need to look at QNX and its real time competitors,
eg. Wind River.
Hi Jeffrey,
We are finding ourselves in the same situation. Plus it seems that QNX are
really only interested in the REALLY big contracts (we are a pretty big
contract I think! but sometimes one wonders). This is a flawed buisness
concept in my eyes.
Anyway I have had QNX here and they have explained that they are going to
continue supporting QNX as long as they are making money from it.
Look at this - http://www.qnx.com/developers/qnx4/CustomerQNX4letter.pdf
I dont know if you have already read that letter, but it seems to hint at
continuing hardware support.
We probably sit in an even worse situation as we are using QNX Windows and
finding cards this works on is becoming alot harder.
I think porting to Linux is more attractive than Windows or QNX6 at this
point. Due to price (one can always create ones own ‘distribution’ to keep
costs down).
Maybe we should make our own Linux distro suited to porting qnx4
applications (there are implementations of Send() Received() IPC for linux
out there and other libraries that would easy qnx4 porting). One could even
get a Photon look and/or Photon emulation. and write some nice skin
libraries. One could even use the watcom compiler instead of gcc for that
authentic feel. Q-Linux anyone?
QNX better start coming up with some backported hardware drivers from qnx6
(this is the way they usually do it soon) or some big qnx4 customers are
going to get nervous.
comments?
/Darel
“Jeffrey Adler” <jeff.nospam.adler@bigfoot.com> wrote in message
news:er292g$nde$1@inn.qnx.com…
“maschoen” <> maschoen@pobox-dot-com.no-spam.invalid> > wrote in message
news:er1b00$4eo$> 1@inn.qnx.com> …
It is always possible for there to be Native SATA support for QNX 4,
however it is very very unlikely to come from QSSL. There are at
least a couple old timers out there with the expertise to do this,
but I think one of them would need an incentive. This type of
thing did occur with QNX 2, with respect to (for instance) the EIDE
LBA driver. >
I am faced with choices.With support for QNX4 going the way of QNX2, yes, it is possible to buy a
couple of more years of QNX4, but ultimately I think that the computer
hardware from the big names (Dell, for example) will keep moving away from
the tried and true.The big issues for me this last “new computer go-around” was 1) SATA
drives with no ide emulation, and 2) USB keyboard with its own emulation
problems.SO…
Without a reasonable commitment from QNX to continue some rudimantary
support for QNX4 on the newer hardware (been there with QNX2 already), I
must now (sooner, not later) look at porting my application to something.The option of providing some incentive ($) to someone for a native SATA
driver might buy a year or two of product life, but it doesn’t appear to
be the solution; only the band-aid.My reasonable choices are QNX6, Linux, or Windows.
With QNX6, the porting development costs are reasonably low but the
deployment costs to upgrade my existing customers are high. Also, there
are few guarantees of long-term support.With Linux or Windows, there is a much better chance of hardware support
continuing (MUCH higher with Windows).I am looking at trying to implement some kind of similar S/R/R model under
Windows. I think that is where my path will lead me.Thanks, all.
Darel Cullenwrote:
Hi Jeffrey,
Look at this -
http://www.qnx.com/developers/qnx4/CustomerQNX4letter.pdfI dont know if you have already read that letter, but it seems to
hint at
continuing hardware support.
It clearly doesn’t just hint at continuing hardware support, but you
have to read between the lines. Gordon Bell told me years ago that
the added cost to provide a QNX 4 driver when there was a QNX 6
driver available was close to zero, do to dual source methods they
were using. This clearly refers to at least Photon video drivers,
and CAMS disk drivers. So the letter, while promising a lot, is not
cost QNX much. But QNX abandoned support of QNX Windows years ago,
so it is doubtful you will see video drivers unless they are just the
same as the Photon drivers. QNX’s support for disk controllers also
has been a little spotty. As far as I know, there are dozens of
SCSI RAID controllers out there, but none are supported to my
knowledge. This is probably becauseQNX’s current target market,
embedded systems, has not seen a need for them. The same problem
probably exists for SATA. So, no SATA for QNX 6 = no SATA for QNX
4.
It clearly doesn’t just hint at continuing hardware support, but you
have to read between the lines. Gordon Bell told me years ago that
the added cost to provide a QNX 4 driver when there was a QNX 6
driver available was close to zero, do to dual source methods they
were using. This clearly refers to at least Photon video drivers,
and CAMS disk drivers. So the letter, while promising a lot, is not
cost QNX much. But QNX abandoned support of QNX Windows years ago,
so it is doubtful you will see video drivers unless they are just the
same as the Photon drivers.
We have them custom engineered at QNX (they use photon drivers as the base
and apparently its pretty much the same)
We have for example a Qnxwindows driver for the intel extreme 2 chipset.
QNX’s support for disk controllers also
has been a little spotty. As far as I know, there are dozens of
SCSI RAID controllers out there, but none are supported to my
knowledge. This is probably becauseQNX’s current target market,
embedded systems, has not seen a need for them. The same problem
probably exists for SATA. So, no SATA for QNX 6 = no SATA for QNX
4.
I am sure SATA will come soon… (although they could let us know about it to
keep us from worrying).
Knowing the “next big move” is really quiet difficult. I have to say that
Linux is very attractive, and there are real time variants of it (montevista
for example).
Also… Windows has realtime addons, But I don’t like windows, its too much
uneeded software and embedded semi-embedded is not really their forte
(though MS did present windows .micro to me)
">
I am sure SATA will come soon… (although they could let us know about it
to keep us from worrying).
Yes it will, but as it comes out they will already be a generation behind.
SCSI is going away to make room for SAS and there aren’t any SAS driver in
the pipe that I know of.
Knowing the “next big move” is really quiet difficult. I have to say that
Linux is very attractive, and there are real time variants of it
(montevista for example).Also… Windows has realtime addons, But I don’t like windows, its too much
uneeded software and embedded semi-embedded is not really their forte
(though MS did present windows .micro to me)
The QNX 4.25 RTOS Fsys.atapi Block Driver update available from the myQNX
download site does contain support for native SATA controllers. Please refer
to the release notes, linked below, for a list of supported SATA
controllers.
http://www.qnx.com/developers/articles/rel_1351_1.html
This list will be updated later this year as we port the additional SATA
support from QNX6.3.0 SP3 back to QNX4. While I do not have a firm date at
this time, it is intended that this port will be available later this year.
As well QSS will continue to back port QNX6 drivers to QNX4, specifically
targeting newer graphics (Photon drivers), Ethernet and hard disk
controllers.
Thanks,
Dave Nickerson
QNX Software Systems
“Jeffrey Adler” jeffadler_at_bigfoot_dot_com@127.0.0.1 wrote in message
news:eqvhvd$s92$1@inn.qnx.com…
Does anyone know (including QNX) if there will ever be native SATA support
for QNX 4? Many newer compters are abandoning IDE emulation.Thanks
There are also industrial motherboard solutions that use older hardware (ie:
chipsets in INTEL’s extended support embedded roadmap). They are more money
but offer things like ISA and lots of PCI slots that you don’t find anymore
in retail boards. We have had to move to PortWell motherboards with P4’s
and 6 PCI as we need lots of slots.
http://www.portwell.com/products/embedded.asp#ATX
–
Allan Williams
“Jeffrey Adler” <jeff.nospam.adler@bigfoot.com> wrote in message
news:er292g$nde$1@inn.qnx.com…
“maschoen” <> maschoen@pobox-dot-com.no-spam.invalid> > wrote in message
news:er1b00$4eo$> 1@inn.qnx.com> …
It is always possible for there to be Native SATA support for QNX 4,
however it is very very unlikely to come from QSSL. There are at
least a couple old timers out there with the expertise to do this,
but I think one of them would need an incentive. This type of
thing did occur with QNX 2, with respect to (for instance) the EIDE
LBA driver. >
I am faced with choices.With support for QNX4 going the way of QNX2, yes, it is possible to buy a
couple of more years of QNX4, but ultimately I think that the computer
hardware from the big names (Dell, for example) will keep moving away from
the tried and true.The big issues for me this last “new computer go-around” was 1) SATA
drives with no ide emulation, and 2) USB keyboard with its own emulation
problems.SO…
Without a reasonable commitment from QNX to continue some rudimantary
support for QNX4 on the newer hardware (been there with QNX2 already), I
must now (sooner, not later) look at porting my application to something.The option of providing some incentive ($) to someone for a native SATA
driver might buy a year or two of product life, but it doesn’t appear to
be the solution; only the band-aid.My reasonable choices are QNX6, Linux, or Windows.
With QNX6, the porting development costs are reasonably low but the
deployment costs to upgrade my existing customers are high. Also, there
are few guarantees of long-term support.With Linux or Windows, there is a much better chance of hardware support
continuing (MUCH higher with Windows).I am looking at trying to implement some kind of similar S/R/R model under
Windows. I think that is where my path will lead me.Thanks, all.