| Re: The Windows 95 chime was created on a Mac [message #356958 is a reply to message #356957] |
Sat, 18 November 2017 03:58   |
mausg
Messages: 2483
Registered: May 2013
Karma: 0
|
Senior Member |
|
|
On 2017-11-18, J Clarke <jclarke.873638@gmail.com> wrote:
> On Thu, 16 Nov 2017 19:45:32 -0800 (PST), Quadibloc
> <jsavard@ecn.ab.ca> wrote:
>
>> On Thursday, November 16, 2017 at 8:38:52 PM UTC-7, J. Clarke wrote:
>>> On Thu, 16 Nov 2017 09:15:58 -0600, Charles Richmond
>>> <numerist@aquaporin4.com> wrote:
>>
>>>> The 4-bit processors are so simple that they take minimal real estate on
>>>> the chip. That leaves room for on-chip ROM, RAM, A-to-D, and even
>>>> programmable logic array elements.. so for example one might implement a
>>>> BCD-to-seven segment decoder or whatever else the application needs.
>>>> And all this on *one* chip to keep down the total chip count.
>>
>>> That might have been a concern a long time ago. A chip the physical
>>> size of the 4004 would today hold about 400 Pentiums.
>>
>> Although it certainly is true that one can put more on a single die today than
>> in the past, die size is still very much a concern. That's why we've only seen
>> eight-core chips enter the mainstream very recently, and chips like the Xeon Phi
>> are still expensive. Die size directly affects yield - and the smaller the
>> features are on a chip, the smaller a defect needs to be to spoil a chip.
>
> Laddie, your information is vastly out of date. The machine I am
> using now has 3500 cores.
What, (in a very general sense) do you need it for?..
--
greymaus.ireland.ie
Just_Another_Grumpy_Old_Man
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356959 is a reply to message #356958] |
Sat, 18 November 2017 07:23 |
|
|
Originally posted by: J. Clarke
On 18 Nov 2017 08:58:20 GMT, greymaus <mausg@mail.com> wrote:
> On 2017-11-18, J Clarke <jclarke.873638@gmail.com> wrote:
>> On Thu, 16 Nov 2017 19:45:32 -0800 (PST), Quadibloc
>> <jsavard@ecn.ab.ca> wrote:
>>
>>> On Thursday, November 16, 2017 at 8:38:52 PM UTC-7, J. Clarke wrote:
>>>> On Thu, 16 Nov 2017 09:15:58 -0600, Charles Richmond
>>>> <numerist@aquaporin4.com> wrote:
>>>
>>>> >The 4-bit processors are so simple that they take minimal real estate on
>>>> >the chip. That leaves room for on-chip ROM, RAM, A-to-D, and even
>>>> >programmable logic array elements.. so for example one might implement a
>>>> >BCD-to-seven segment decoder or whatever else the application needs.
>>>> >And all this on *one* chip to keep down the total chip count.
>>>
>>>> That might have been a concern a long time ago. A chip the physical
>>>> size of the 4004 would today hold about 400 Pentiums.
>>>
>>> Although it certainly is true that one can put more on a single die today than
>>> in the past, die size is still very much a concern. That's why we've only seen
>>> eight-core chips enter the mainstream very recently, and chips like the Xeon Phi
>>> are still expensive. Die size directly affects yield - and the smaller the
>>> features are on a chip, the smaller a defect needs to be to spoil a chip.
>>
>> Laddie, your information is vastly out of date. The machine I am
>> using now has 3500 cores.
>
> What, (in a very general sense) do you need it for?..
Recreation. It's a game machine. The 3500 core processor is designed
as a video processor. However some time in the late 90s or or early
oughts somebody figured out that the cores on a video processor could
also be used for numerical calculations and in 2007 an SDK was
released by one of the chip makers that allows that capability to be
used by any reasonably competent programmer--that SDK, "CUDA", is
specific to nVidia hardware however in 2009 an open source SDK
("OpenCL") that also supports ATI was released.
However I am also slowly working through tutorials on programming
because there are some work processes that could benefit from
massively parallel computation--most of what we do involves the same
calculation performed on a large number of cells with different
inputs, so it should parallelize nicely.
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356960 is a reply to message #356959] |
Sat, 18 November 2017 08:11 |
mausg
Messages: 2483
Registered: May 2013
Karma: 0
|
Senior Member |
|
|
On 2017-11-18, J Clarke <jclarke.873638@gmail.com> wrote:
> On 18 Nov 2017 08:58:20 GMT, greymaus <mausg@mail.com> wrote:
>
>> On 2017-11-18, J Clarke <jclarke.873638@gmail.com> wrote:
>>> On Thu, 16 Nov 2017 19:45:32 -0800 (PST), Quadibloc
>>> <jsavard@ecn.ab.ca> wrote:
>>>
> Recreation. It's a game machine. The 3500 core processor is designed
> as a video processor. However some time in the late 90s or or early
> oughts somebody figured out that the cores on a video processor could
> also be used for numerical calculations and in 2007 an SDK was
> released by one of the chip makers that allows that capability to be
> used by any reasonably competent programmer--that SDK, "CUDA", is
> specific to nVidia hardware however in 2009 an open source SDK
> ("OpenCL") that also supports ATI was released.
>
> However I am also slowly working through tutorials on programming
> because there are some work processes that could benefit from
> massively parallel computation--most of what we do involves the same
> calculation performed on a large number of cells with different
> inputs, so it should parallelize nicely.
-
I updated motherboerds almost every year, until I now see no point.
(I do not play computer games, but would think that good communications
would be the first priority)-
greymaus.ireland.ie
Just_Another_Grumpy_Old_Man
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356961 is a reply to message #356959] |
Sat, 18 November 2017 09:22 |
Quadibloc
Messages: 4399
Registered: June 2012
Karma: 0
|
Senior Member |
|
|
On Saturday, November 18, 2017 at 5:23:50 AM UTC-7, J. Clarke wrote:
> Recreation. It's a game machine. The 3500 core processor is designed
> as a video processor.
I am aware that GPUs have high "core" counts. And that they can be used to perform
some operations on vectors, thus helping to accelerate computation.
John Savard
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356962 is a reply to message #356961] |
Sat, 18 November 2017 09:26 |
Quadibloc
Messages: 4399
Registered: June 2012
Karma: 0
|
Senior Member |
|
|
On Saturday, November 18, 2017 at 7:22:58 AM UTC-7, Quadibloc wrote:
> On Saturday, November 18, 2017 at 5:23:50 AM UTC-7, J. Clarke wrote:
>
>> Recreation. It's a game machine. The 3500 core processor is designed
>> as a video processor.
>
> I am aware that GPUs have high "core" counts. And that they can be used to perform
> some operations on vectors, thus helping to accelerate computation.
....and you definitely have me beat at the moment, as my gaming laptop has merely
192 cores in its GPU, being an out-of-date unit.
John Savard
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356963 is a reply to message #356961] |
Sat, 18 November 2017 10:26 |
Charles Richmond
Messages: 2754
Registered: December 2011
Karma: 0
|
Senior Member |
|
|
On 11/18/2017 8:22 AM, Quadibloc wrote:
> On Saturday, November 18, 2017 at 5:23:50 AM UTC-7, J. Clarke wrote:
>
>> Recreation. It's a game machine. The 3500 core processor is designed
>> as a video processor.
>
> I am aware that GPUs have high "core" counts. And that they can be used to perform
> some operations on vectors, thus helping to accelerate computation.
>
Are these actual cores or just parallel integer execution units??? For
many years (from the 1980's), processors could have two, three, or even
four integer units to do integer operations in parallel... if operands
did *not* depend on the result of an immediately previous operation.
--
numerist at aquaporin4 dot com
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356964 is a reply to message #356960] |
Sat, 18 November 2017 11:13 |
Ahem A Rivet's Shot
Messages: 4843
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
On 18 Nov 2017 13:11:30 GMT
greymaus <mausg@mail.com> wrote:
> On 2017-11-18, J Clarke <jclarke.873638@gmail.com> wrote:
>> On 18 Nov 2017 08:58:20 GMT, greymaus <mausg@mail.com> wrote:
>>
>>> On 2017-11-18, J Clarke <jclarke.873638@gmail.com> wrote:
>>>> On Thu, 16 Nov 2017 19:45:32 -0800 (PST), Quadibloc
>>>> <jsavard@ecn.ab.ca> wrote:
>>>>
>> Recreation. It's a game machine. The 3500 core processor is designed
>> as a video processor. However some time in the late 90s or or early
>> oughts somebody figured out that the cores on a video processor could
>> also be used for numerical calculations and in 2007 an SDK was
>> released by one of the chip makers that allows that capability to be
>> used by any reasonably competent programmer--that SDK, "CUDA", is
>> specific to nVidia hardware however in 2009 an open source SDK
>> ("OpenCL") that also supports ATI was released.
> I updated motherboerds almost every year, until I now see no point.
> (I do not play computer games, but would think that good communications
> would be the first priority)-
Depends on the game, for online real time games then yes good low
latency communications are important (not quite so important as for high
frequency trading - but some gamers may disagree). Many modern games make
very heavy use of 3D rendering graphics hardware to render complex scenes
with high realism at high frame rates. As a result the most powerful
computing engines easily available are the high end gamers graphics cards
(mostly nVidia and ATI cards).
--
Steve O'Hara-Smith | Directable Mirror Arrays
C:\>WIN | A better way to focus the sun
The computer obeys and wins. | licences available see
You lose and Bill collects. | http://www.sohara.org/
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356965 is a reply to message #356963] |
Sat, 18 November 2017 11:43 |
mausg
Messages: 2483
Registered: May 2013
Karma: 0
|
Senior Member |
|
|
On 2017-11-18, Charles Richmond <numerist@aquaporin4.com> wrote:
> On 11/18/2017 8:22 AM, Quadibloc wrote:
>> On Saturday, November 18, 2017 at 5:23:50 AM UTC-7, J. Clarke wrote:
>>
>>> Recreation. It's a game machine. The 3500 core processor is designed
>>> as a video processor.
>>
>> I am aware that GPUs have high "core" counts. And that they can be used to perform
>> some operations on vectors, thus helping to accelerate computation.
>>
>
> Are these actual cores or just parallel integer execution units??? For
> many years (from the 1980's), processors could have two, three, or even
> four integer units to do integer operations in parallel... if operands
> did *not* depend on the result of an immediately previous operation.
>
>
>
Which would mean compiling for every seperate job?
--
greymaus.ireland.ie
Just_Another_Grumpy_Old_Man
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356966 is a reply to message #356963] |
Sat, 18 November 2017 13:08 |
|
|
Originally posted by: J. Clarke
On Sat, 18 Nov 2017 09:26:20 -0600, Charles Richmond
<numerist@aquaporin4.com> wrote:
> On 11/18/2017 8:22 AM, Quadibloc wrote:
>> On Saturday, November 18, 2017 at 5:23:50 AM UTC-7, J. Clarke wrote:
>>
>>> Recreation. It's a game machine. The 3500 core processor is designed
>>> as a video processor.
>>
>> I am aware that GPUs have high "core" counts. And that they can be used to perform
>> some operations on vectors, thus helping to accelerate computation.
>>
>
> Are these actual cores or just parallel integer execution units??? For
> many years (from the 1980's), processors could have two, three, or even
> four integer units to do integer operations in parallel... if operands
> did *not* depend on the result of an immediately previous operation.
The organization is a bit complex. There are 28 streaming
multiprocessors, each of which can if I understand correctly has four
control units each with 32 parallel execution units. Parallel
execution is possible on integer, single-precision floating point, and
double-precsion floating point operands. The instruction set on the
parallel execution units is, as I understand it, quite limited but the
control units have a wide range of functionality.
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356974 is a reply to message #356965] |
Sat, 18 November 2017 17:41 |
Charles Richmond
Messages: 2754
Registered: December 2011
Karma: 0
|
Senior Member |
|
|
On 11/18/2017 10:43 AM, greymaus wrote:
> On 2017-11-18, Charles Richmond <numerist@aquaporin4.com> wrote:
>> On 11/18/2017 8:22 AM, Quadibloc wrote:
>>> On Saturday, November 18, 2017 at 5:23:50 AM UTC-7, J. Clarke wrote:
>>>
>>>> Recreation. It's a game machine. The 3500 core processor is designed
>>>> as a video processor.
>>>
>>> I am aware that GPUs have high "core" counts. And that they can be used to perform
>>> some operations on vectors, thus helping to accelerate computation.
>>>
>>
>> Are these actual cores or just parallel integer execution units??? For
>> many years (from the 1980's), processors could have two, three, or even
>> four integer units to do integer operations in parallel... if operands
>> did *not* depend on the result of an immediately previous operation.
>>
>>
>>
>
> Which would mean compiling for every separate job?
>
>
ISTM that the CPU has an input queue that provides a look-ahead. The CPU
can decide if more than one integer unit can be used. At least that's
how I heard it...
--
numerist at aquaporin4 dot com
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356975 is a reply to message #356964] |
Sun, 19 November 2017 15:07 |
Charlie Gibbs
Messages: 5313
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
On 2017-11-18, Ahem A Rivet's Shot <steveo@eircom.net> wrote:
> Depends on the game, for online real time games then yes good low
> latency communications are important (not quite so important as for high
> frequency trading - but some gamers may disagree).
In the minds of some, high-frequency trading _is_ a game.
--
/~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs)
\ / I'm really at ac.dekanfrus if you read it the right way.
X Top-posted messages will probably be ignored. See RFC1855.
/ \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign!
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356976 is a reply to message #356975] |
Sun, 19 November 2017 15:34 |
Anne & Lynn Wheel
Messages: 3156
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
Charlie Gibbs <cgibbs@kltpzyxm.invalid> writes:
> In the minds of some, high-frequency trading _is_ a game.
somewhat from being involved in doing "electronic commerce",
we were brought in to x9a10 financial standard working group
(which had been given the requirement to preserve the integrity
of the financial infrastructure for *ALL* retail payments). One
of the participants was from NSCC who then brought us in to
look at improving integrity of trading floor transactions.
I work on it for awhile and then get a call saying that it
was being suspended because a side-effect would have greatly
improved transparency and visibility (anathema to wallstreet
culture). This was before NSCC merges with DTC, becoming
DTCC
https://en.wikipedia.org/wiki/Depository_Trust_%26_Clearing_ Corporation
Rhetoric on flr of congress was that Sarbanes-Oxley would prevent future
ENRONs and guarantee executives and auditors did jail time, but it
required SEC to do something. Possibly because even GAO didn't believe
SEC was doing anything, it started doing reports of public company
fraudulent financial filings, even showing increase after SOX goes
into effect (and nobody doing jailtime) ... past posts
http://www.garlic.com/~lynn/submisc.html#financial.reporting .fraud
Sarbanes-Oxley posts
http://www.garlic.com/~lynn/submic.html#sarbanes-oxley
enron posts
http://www.garlic.com/~lynn/submisc.html#enron
this is interview about illegal activity goes on all the time and nobody
has to worry about SEC (before HFT really kicks in)
http://nypost.com/2007/03/20/cramer-reveals-a-bit-too-much/
In the congressional Madoff hearings they had person that had tried
unsuccessfully for a decade to get SEC to do something about Madoff.
Congress asked him if new regulations were needed. He replied that
while new regulations might be needed, much more important would
be transparency and visibility (possibly since SEC wasn't doing
anything anyway) ... some past posts
http://www.garlic.com/~lynn/submisc.html#madoff
HFT then really starts to kick in ... and frequenctly used to
further obfuscate illegal activity. a number of more recent
HFT posts:
http://www.garlic.com/~lynn/2014.html#89 Royal Pardon For Turing
http://www.garlic.com/~lynn/2014b.html#65 Royal Pardon For Turing
http://www.garlic.com/~lynn/2014d.html#93 New York seeks curbs on high-frequency trading
http://www.garlic.com/~lynn/2014e.html#18 FBI Investigates High-Speed Trading
http://www.garlic.com/~lynn/2014e.html#60 FBI Investigates High-Speed Trading
http://www.garlic.com/~lynn/2014e.html#71 FBI Investigates High-Speed Trading
http://www.garlic.com/~lynn/2014e.html#72 Three Expensive Milliseconds
http://www.garlic.com/~lynn/2014f.html#3 Three Expensive Milliseconds
http://www.garlic.com/~lynn/2014f.html#20 HFT, computer trading
http://www.garlic.com/~lynn/2014f.html#41 System Response
http://www.garlic.com/~lynn/2014g.html#1 HFT is harmful, say US market participants
http://www.garlic.com/~lynn/2014g.html#64 HFT is harmful, say US market participants
http://www.garlic.com/~lynn/2014g.html#107 The SEC's Mary Jo White Punts on High Frequency Trading and Abandons Securities Act of 1934
http://www.garlic.com/~lynn/2014g.html#109 SEC Caught Dark Pool and High Speed Traders Doing Bad Stuff
http://www.garlic.com/~lynn/2014h.html#64 Dark Pool Greed Drove Barclays to Lie to Clients, N.Y. Says
http://www.garlic.com/~lynn/2015.html#58 IBM Data Processing Center and Pi
http://www.garlic.com/~lynn/2015b.html#26 What were the complaints of binary code programmers that not accept Assembly?
http://www.garlic.com/~lynn/2015c.html#17 Robots have been running the US stock market, and the government is finally taking control
http://www.garlic.com/~lynn/2015e.html#58 Time to Fire Mary Jo White: SEC Covers Up for Bank Capital Accounting Scam Promoted by Her Former Firm, Debevoise
http://www.garlic.com/~lynn/2015g.html#48 seveneves
http://www.garlic.com/~lynn/2015g.html#66 Michael Hudson's New Book: Wall Street Parasites Have Devoured Their Hosts -- Your Retirement Plan and the U.S. Economy
http://www.garlic.com/~lynn/2015h.html#23 It A "Liquidity Mirage": New York Fed Finally Grasps How Broken The Market Is Due To HFTs
http://www.garlic.com/~lynn/2016b.html#68 Eric Hunsader Explains To CNBC That "Markets Are Always Rigged"
http://www.garlic.com/~lynn/2016g.html#18 Bundesbank Confirms HFTs Reduce Liquidity, Contribute To Flash Crashes, Withdraw At Times Of "Market Stress"
http://www.garlic.com/~lynn/2017c.html#22 How do BIG WEBSITES work?
http://www.garlic.com/~lynn/2017c.html#23 How do BIG WEBSITES work?
--
virtualization experience starting Jan1968, online at home since Mar1970
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #356977 is a reply to message #356975] |
Mon, 20 November 2017 02:13 |
Mike Spencer
Messages: 997
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
Charlie Gibbs <cgibbs@kltpzyxm.invalid> writes:
> On 2017-11-18, Ahem A Rivet's Shot <steveo@eircom.net> wrote:
>
>> Depends on the game, for online real time games then yes good low
>> latency communications are important (not quite so important as for high
>> frequency trading - but some gamers may disagree).
>
> In the minds of some, high-frequency trading _is_ a game.
The Toronto Globe & Mail business section is instructive. The front
pages have reports on mergers, IPOs, newly installed CEOs, ventures
undertaken in China and so on. The language is sprinkled with
terminology from sports, gambling and games.
A few pages further along are the columns offering advice for poor
people, viz. those who can't pay a team of lawyers and traders to
manage their humble investments and have to do it themselves. There,
the language is all about wise management of risk, cautious strategies
and a balanced portfolio. No metaphors from the casino or the
gridiron.
High-speed trading? Read "Traveller's Rest" by David I Masson
_From The Caltraps of Time_. It's on Google Books. The style is a
little dated but it's all about the mentality of high-speed trading,
embodied in a bit of unexplained science-fiction physics.
--
Mike Spencer Nova Scotia, Canada
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #359673 is a reply to message #356721] |
Tue, 26 December 2017 21:26 |
|
|
Originally posted by: J. Clarke
On Wed, 15 Nov 2017 05:14:08 GMT, usenet@only.tnx (Questor) wrote:
> On Mon, 13 Nov 2017 17:18:32 -0500, Andreas Kohlbach <ank@spamfence.net> wrote:
>> On Mon, 13 Nov 2017 19:28:08 GMT, Scott Lurndal wrote:
>>> Jorgen Grahn <grahn+nntp@snipabacken.se> writes:
>>>> On Mon, 2017-10-23, Peter Flass wrote:
>>>
>>>> > 64-bit Linux runs 32-bit programs just fine, [...]
>>>>
>>>> Although, running with such a mismatch indicates you're either running
>>>> badly broken software (which, for some reason, after decades of 64-bit
>>>> Unix still isn't portable to a modern system) or you're running
>>>> non-free software produced by a really backwards company and you're a
>>>> sucker.
>>>
>>> Or you care about icache/dcache footprint. If an application doesn't
>>> require a large memory footprint, what's wrong with compiling it
>>> for 32-bit on an intel 64-bit host?
>>
>> I think he was talking about running 32Bit programs on a 64Bit Linux OS.
>>
>> I have a 64Bit CPU but run 32Bit Linux just fine.
>>
>> The problem with 32Bit (any OS) will show in about 21 years (Y2K38
>> bug). If course by then it's not too likely anyone still runs 32Bit
>> software. Like today almost no one runs 16Bit or even 8Bit on productive
>> systems.
>
> Are you excluding embedded systems? I have read that 8-bit processors like the
> Z80 still vastly outnumber everything else.
In any case, the correct statement is that any code that uses a 32-bit
signed integer to represent seconds elapsed since 1 Jan 1970 has this
issue. I'm pretty sure that lets Z/OS out as it existed in some form
prior to 1970 (although that does not mean that it won't simply have
an issue earlier, although I haven't heard any suggestion that it
would).
If the code doesn't use a date (why would, say, a control system for a
machine tool care about the date?) then it's a non-issue.
We have a great deal of code that deals with events beyond 2038 that
works just fine. Some of it dates back to the '60s.
>> Of course we love nostalgia (wouldn't otherwise read and write here) we
>> might have old 16Bit or 8Bit hardware, or emulate that.
>
|
|
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #359709 is a reply to message #359673] |
Wed, 27 December 2017 16:27 |
Andreas Kohlbach
Messages: 1456
Registered: December 2011
Karma: 0
|
Senior Member |
|
|
On Tue, 26 Dec 2017 21:26:18 -0500, J. Clarke wrote:
>
> On Wed, 15 Nov 2017 05:14:08 GMT, usenet@only.tnx (Questor) wrote:
>
>> Are you excluding embedded systems? I have read that 8-bit processors like the
>> Z80 still vastly outnumber everything else.
>
> In any case, the correct statement is that any code that uses a 32-bit
> signed integer to represent seconds elapsed since 1 Jan 1970 has this
> issue. I'm pretty sure that lets Z/OS out as it existed in some form
> prior to 1970 (although that does not mean that it won't simply have
> an issue earlier, although I haven't heard any suggestion that it
> would).
>
> If the code doesn't use a date (why would, say, a control system for a
> machine tool care about the date?) then it's a non-issue.
>
> We have a great deal of code that deals with events beyond 2038 that
> works just fine. Some of it dates back to the '60s.
That should depend on the underlying counting system. UNIX counts seconds
elapsed since 1/1/1970, thus 32Bit versions are vulnerable. Not sure
about MS-DOS, but it seems it counts in 10th/ms since 1/1/1980 and covers
until 12/31/2099, according to <https://en.wikipedia.org/wiki/DATE_(command),
look for "DOS".
--
Andreas
You know you are a redneck if
you've ever used lard in bed.
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #359710 is a reply to message #359709] |
Wed, 27 December 2017 16:57 |
|
|
Originally posted by: J. Clarke
On Wed, 27 Dec 2017 16:27:29 -0500, Andreas Kohlbach
<ank@spamfence.net> wrote:
> On Tue, 26 Dec 2017 21:26:18 -0500, J. Clarke wrote:
>>
>> On Wed, 15 Nov 2017 05:14:08 GMT, usenet@only.tnx (Questor) wrote:
>>
>>> Are you excluding embedded systems? I have read that 8-bit processors like the
>>> Z80 still vastly outnumber everything else.
>>
>> In any case, the correct statement is that any code that uses a 32-bit
>> signed integer to represent seconds elapsed since 1 Jan 1970 has this
>> issue. I'm pretty sure that lets Z/OS out as it existed in some form
>> prior to 1970 (although that does not mean that it won't simply have
>> an issue earlier, although I haven't heard any suggestion that it
>> would).
>>
>> If the code doesn't use a date (why would, say, a control system for a
>> machine tool care about the date?) then it's a non-issue.
>>
>> We have a great deal of code that deals with events beyond 2038 that
>> works just fine. Some of it dates back to the '60s.
>
> That should depend on the underlying counting system. UNIX counts seconds
> elapsed since 1/1/1970, thus 32Bit versions are vulnerable. Not sure
> about MS-DOS, but it seems it counts in 10th/ms since 1/1/1980 and covers
> until 12/31/2099, according to <https://en.wikipedia.org/wiki/DATE_(command),
> look for "DOS".
Our code doesn't work on dates for anything except the time stamp on
saved files used for record-keeping. It works on "durations" instead.
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #359712 is a reply to message #359709] |
Wed, 27 December 2017 18:09 |
|
|
Originally posted by: Bob Eager
On Wed, 27 Dec 2017 16:27:29 -0500, Andreas Kohlbach wrote:
> On Tue, 26 Dec 2017 21:26:18 -0500, J. Clarke wrote:
>>
>> On Wed, 15 Nov 2017 05:14:08 GMT, usenet@only.tnx (Questor) wrote:
>>
>>> Are you excluding embedded systems? I have read that 8-bit processors
>>> like the Z80 still vastly outnumber everything else.
>>
>> In any case, the correct statement is that any code that uses a 32-bit
>> signed integer to represent seconds elapsed since 1 Jan 1970 has this
>> issue. I'm pretty sure that lets Z/OS out as it existed in some form
>> prior to 1970 (although that does not mean that it won't simply have an
>> issue earlier, although I haven't heard any suggestion that it would).
>>
>> If the code doesn't use a date (why would, say, a control system for a
>> machine tool care about the date?) then it's a non-issue.
>>
>> We have a great deal of code that deals with events beyond 2038 that
>> works just fine. Some of it dates back to the '60s.
>
> That should depend on the underlying counting system. UNIX counts
> seconds elapsed since 1/1/1970, thus 32Bit versions are vulnerable. Not
> sure about MS-DOS, but it seems it counts in 10th/ms since 1/1/1980 and
> covers until 12/31/2099, according to
> <https://en.wikipedia.org/wiki/DATE_(command),
> look for "DOS".
The usual problem is with file dates. MS-DOS only does file times to a
resolution of 2 seconds. However, it has a seven bit field for the year
part of a file date, taking it up to 2099 (as it fails at the end of the
century).
--
Using UNIX since v6 (1975)...
Use the BIG mirror service in the UK:
http://www.mirrorservice.org
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #359714 is a reply to message #359712] |
Wed, 27 December 2017 19:26 |
Charlie Gibbs
Messages: 5313
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
On 2017-12-27, Bob Eager <news0006@eager.cx> wrote:
> The usual problem is with file dates. MS-DOS only does file times to a
> resolution of 2 seconds.
That's why I've formatted the flash drive I use for transporting
my development stuff as NTFS. File times remain accurate to the
second, so version comparisons and makefiles don't get weird.
NTFS on a flash drive really sucks, though.
--
/~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs)
\ / I'm really at ac.dekanfrus if you read it the right way.
X Top-posted messages will probably be ignored. See RFC1855.
/ \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign!
|
|
|
|
| Re: The Windows 95 chime was created on a Mac [message #359743 is a reply to message #359712] |
Thu, 28 December 2017 15:28 |
Andreas Kohlbach
Messages: 1456
Registered: December 2011
Karma: 0
|
Senior Member |
|
|
On 27 Dec 2017 23:09:26 GMT, Bob Eager wrote:
>
> The usual problem is with file dates. MS-DOS only does file times to a
> resolution of 2 seconds. However, it has a seven bit field for the year
> part of a file date, taking it up to 2099 (as it fails at the end of the
> century).
Where none of us should care about. Unless the cryo-system our bodies
might be in runs on DOS. ;-) Just had to think about Futurama...
--
Andreas
You know you are a redneck if
your four-year-old is a member of the nra.
|
|
|
|
Members
Search
Help
Register
Login
Home
