B
bitrex
Guest
On 6/20/19 8:46 PM, Clifford Heath wrote:
global variables?
what YEAR is this???
global variables?
what YEAR is this???
Navigation
Install the app
How to install the app on iOS
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
More options
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
B
bitrex
Guest
On 6/20/19 8:46 PM, Clifford Heath wrote:
Functional programming is all the rage among the kids these days. No
mutable state. Or at least as close to none (interacting with the real
world usually requires some state)
On 21/6/19 10:23 am, bitrex wrote:
On 6/20/19 6:36 PM, John Larkin wrote:
On Thu, 20 Jun 2019 16:14:34 -0400, bitrex <user@example.net> wrote:
On 6/20/19 3:44 PM, John Larkin wrote:
https://www.theregister.co.uk/2019/06/20/ge_lightblulb_reset/
In other news, a few months ago my microwave oven keypad broke; the
first two rows of numbers quit working, 012 and 345. That made things
tricky. Had to cook things for 6:66 minutes and seconds. Setting time
was hard. I assumed that some trace had broken in the membrane thing.
So last night we had a brief power failure, and now it all works! I
don't understand how that's possible.
Software written by EEs
How can one bit of code break, and be fixed by a reset?
I guess I just don't have the chops to be a programmer.
Also ask Toyota they prolly know
Reports say their code was littered with thousands of static/global
variables, creating almost an infinitely complex (but certainly
un-analysable) state space.
The first thing I do before starting to look for bugs in a client's
embedded device is make a comprehensive review of all global data. It's
amazing how many potential bugs you can find just by looking at the
linker map of global data.
Clifford Heath.
Functional programming is all the rage among the kids these days. No
mutable state. Or at least as close to none (interacting with the real
world usually requires some state)
C
Clifford Heath
Guest
On 21/6/19 11:00 am, bitrex wrote:
I wish that was true. There are some adopting it, but I wouldn't say it
is "all the rage". The Haskell community has no desire to reach a mass
audience, because they're so special and mass adoption would make them
less special, so they wouldn't be able to sit back and sneer and snipe
any more, depriving them of their favourite hobby.
A few enlightened folk are trying to blend ideas from other systems,
e.g. Elixir is a Ruby-like FP language that runs on the Erlang VM. That
has more chance of success.
Clifford Heath.
On 6/20/19 8:46 PM, Clifford Heath wrote:
On 21/6/19 10:23 am, bitrex wrote:
On 6/20/19 6:36 PM, John Larkin wrote:
On Thu, 20 Jun 2019 16:14:34 -0400, bitrex <user@example.net> wrote:
On 6/20/19 3:44 PM, John Larkin wrote:
https://www.theregister.co.uk/2019/06/20/ge_lightblulb_reset/
In other news, a few months ago my microwave oven keypad broke; the
first two rows of numbers quit working, 012 and 345. That made things
tricky. Had to cook things for 6:66 minutes and seconds. Setting time
was hard. I assumed that some trace had broken in the membrane thing.
So last night we had a brief power failure, and now it all works! I
don't understand how that's possible.
Software written by EEs
How can one bit of code break, and be fixed by a reset?
I guess I just don't have the chops to be a programmer.
Also ask Toyota they prolly know
Reports say their code was littered with thousands of static/global
variables, creating almost an infinitely complex (but certainly
un-analysable) state space.
The first thing I do before starting to look for bugs in a client's
embedded device is make a comprehensive review of all global data.
It's amazing how many potential bugs you can find just by looking at
the linker map of global data.
Clifford Heath.
Functional programming is all the rage among the kids these days.
I wish that was true. There are some adopting it, but I wouldn't say it
is "all the rage". The Haskell community has no desire to reach a mass
audience, because they're so special and mass adoption would make them
less special, so they wouldn't be able to sit back and sneer and snipe
any more, depriving them of their favourite hobby.
A few enlightened folk are trying to blend ideas from other systems,
e.g. Elixir is a Ruby-like FP language that runs on the Erlang VM. That
has more chance of success.
Clifford Heath.
B
bitrex
Guest
On 6/20/19 3:44 PM, John Larkin wrote:
Dig the stack overflow exploit that was discovered in this WiFi-enabled
smart plug:
<http://www.devttys0.com/2014/05/hacking-the-d-link-dsp-w215-smart-plug/>
It runs Linux and unpatched you can get root access without auth and log
into a shell and just switch whatever is connected on and off over WiFi
https://www.theregister.co.uk/2019/06/20/ge_lightblulb_reset/
In other news, a few months ago my microwave oven keypad broke; the
first two rows of numbers quit working, 012 and 345. That made things
tricky. Had to cook things for 6:66 minutes and seconds. Setting time
was hard. I assumed that some trace had broken in the membrane thing.
So last night we had a brief power failure, and now it all works! I
don't understand how that's possible.
Dig the stack overflow exploit that was discovered in this WiFi-enabled
smart plug:
<http://www.devttys0.com/2014/05/hacking-the-d-link-dsp-w215-smart-plug/>
It runs Linux and unpatched you can get root access without auth and log
into a shell and just switch whatever is connected on and off over WiFi
B
bitrex
Guest
On 6/20/19 10:25 PM, bitrex wrote:
int content_length, i;
char *content_length_str;
/* Certain input data from the incoming TCP/IP packet isn't sanitized
before being just dumped into this buffer on the stack so send a lot! */
char post_data_buf[500000];
content_length = 0;
content_length_str = getenv("CONTENT_LENGTH");
if(content_length_str)
{
content_length = strtol(content_length_str, 10);
}
memset(post_data_buf, 0, 500000);
for(i=0; i<content_length; i++)
{
post_data_buf = fgetc();
}
On 6/20/19 3:44 PM, John Larkin wrote:
https://www.theregister.co.uk/2019/06/20/ge_lightblulb_reset/
In other news, a few months ago my microwave oven keypad broke; the
first two rows of numbers quit working, 012 and 345. That made things
tricky. Had to cook things for 6:66 minutes and seconds. Setting time
was hard. I assumed that some trace had broken in the membrane thing.
So last night we had a brief power failure, and now it all works! I
don't understand how that's possible.
Dig the stack overflow exploit that was discovered in this WiFi-enabled
smart plug:
http://www.devttys0.com/2014/05/hacking-the-d-link-dsp-w215-smart-plug/
It runs Linux and unpatched you can get root access without auth and log
into a shell and just switch whatever is connected on and off over WiF
int content_length, i;
char *content_length_str;
/* Certain input data from the incoming TCP/IP packet isn't sanitized
before being just dumped into this buffer on the stack so send a lot! */
char post_data_buf[500000];
content_length = 0;
content_length_str = getenv("CONTENT_LENGTH");
if(content_length_str)
{
content_length = strtol(content_length_str, 10);
}
memset(post_data_buf, 0, 500000);
for(i=0; i<content_length; i++)
{
post_data_buf = fgetc();
}
C
Chris Jones
Guest
On 21/06/2019 08:36, John Larkin wrote:
If the register that controls the pin direction (input / output) of the
IO pins that scan the keypad were to get corrupted, that would do it.
Most people don't reinitialise the IO registers except at startup. There
is an argument for forcing a reset from time to time, e.g. each time the
appliance is used, but that is hard if you also keep the real time clock
in software.
> I guess I just don't have the chops to be a programmer.
If it had been written by real programmers, then it would need a CPU
clock frequency higher than the oscillation frequency of the magnetron,
and it would still be laggy. Similarly, all-vacuum-tube televisions warm
up faster than a modern television can boot up, and the battery life of
a modern digital radio is worse than the ones that had separate filament
accumulators and HT batteries.
On Thu, 20 Jun 2019 16:14:34 -0400, bitrex <user@example.net> wrote:
On 6/20/19 3:44 PM, John Larkin wrote:
https://www.theregister.co.uk/2019/06/20/ge_lightblulb_reset/
In other news, a few months ago my microwave oven keypad broke; the
first two rows of numbers quit working, 012 and 345. That made things
tricky. Had to cook things for 6:66 minutes and seconds. Setting time
was hard. I assumed that some trace had broken in the membrane thing.
So last night we had a brief power failure, and now it all works! I
don't understand how that's possible.
Software written by EEs
How can one bit of code break, and be fixed by a reset?
If the register that controls the pin direction (input / output) of the
IO pins that scan the keypad were to get corrupted, that would do it.
Most people don't reinitialise the IO registers except at startup. There
is an argument for forcing a reset from time to time, e.g. each time the
appliance is used, but that is hard if you also keep the real time clock
in software.
> I guess I just don't have the chops to be a programmer.
If it had been written by real programmers, then it would need a CPU
clock frequency higher than the oscillation frequency of the magnetron,
and it would still be laggy. Similarly, all-vacuum-tube televisions warm
up faster than a modern television can boot up, and the battery life of
a modern digital radio is worse than the ones that had separate filament
accumulators and HT batteries.
W
whit3rd
Guest
On Thursday, June 20, 2019 at 1:13:45 PM UTC-7, Rick C wrote:
Sounds like the MCU should have been given a finite-state-machine program, but
got a rat's-nest instead. Real-time (interrupt driven) software is an
art that doesn't have a lot of literature and cirriculum support.
On Thursday, June 20, 2019 at 3:47:21 PM UTC-4, John Larkin wrote:
https://www.theregister.co.uk/2019/06/20/ge_lightblulb_reset/
In other news, a few months ago my microwave oven keypad broke; the
first two rows of numbers quit working, 012 and 345....
So last night we had a brief power failure, and now it all works! I
don't understand how that's possible.
Are you joking? ... The tiny MCU needed a good reset. The MCU probably is never turned off ...
It may have been a short power glitch that knocked it silly in the first place
Sounds like the MCU should have been given a finite-state-machine program, but
got a rat's-nest instead. Real-time (interrupt driven) software is an
art that doesn't have a lot of literature and cirriculum support.
J
John Larkin
Guest
On Fri, 21 Jun 2019 15:03:52 +1000, Chris Jones
<lugnut808@spam.yahoo.com> wrote:
OK, that would do it, a flipped port control bit taking out an entire
row of a matrix scan of the keypad. Somehow. ESD maybe.
Well, it's easier to poach eggs now. (In a water bath, to slow things
down.)
This microwave, an official Amana RadarRange (tm) is quirky as hell, a
smart appliance programmed by dumb people. Like the light bulb.
The RadarRange was Raytheon's name for the first microwave oven,
introduced in 1947.
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
<lugnut808@spam.yahoo.com> wrote:
On 21/06/2019 08:36, John Larkin wrote:
On Thu, 20 Jun 2019 16:14:34 -0400, bitrex <user@example.net> wrote:
On 6/20/19 3:44 PM, John Larkin wrote:
https://www.theregister.co.uk/2019/06/20/ge_lightblulb_reset/
In other news, a few months ago my microwave oven keypad broke; the
first two rows of numbers quit working, 012 and 345. That made things
tricky. Had to cook things for 6:66 minutes and seconds. Setting time
was hard. I assumed that some trace had broken in the membrane thing.
So last night we had a brief power failure, and now it all works! I
don't understand how that's possible.
Software written by EEs
How can one bit of code break, and be fixed by a reset?
If the register that controls the pin direction (input / output) of the
IO pins that scan the keypad were to get corrupted, that would do it.
Most people don't reinitialise the IO registers except at startup.
OK, that would do it, a flipped port control bit taking out an entire
row of a matrix scan of the keypad. Somehow. ESD maybe.
Well, it's easier to poach eggs now. (In a water bath, to slow things
down.)
This microwave, an official Amana RadarRange (tm) is quirky as hell, a
smart appliance programmed by dumb people. Like the light bulb.
The RadarRange was Raytheon's name for the first microwave oven,
introduced in 1947.
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
R
Rick C
Guest
On Friday, June 21, 2019 at 2:31:27 AM UTC-4, whit3rd wrote:
Actually I was taught state machine design including async state machines. I only ever once saw such a state machine, but I could understand it and could have reworked it if needed.
I'm not going to second guess a design I know nothing about. It did it's job with no issues for a couple of decades before needing to be unplugged for a fix. I think that is a pretty good track record.
--
Rick C.
-- Get 1,000 miles of free Supercharging
-- Tesla referral code - https://ts.la/richard11209
On Thursday, June 20, 2019 at 1:13:45 PM UTC-7, Rick C wrote:
On Thursday, June 20, 2019 at 3:47:21 PM UTC-4, John Larkin wrote:
https://www.theregister.co.uk/2019/06/20/ge_lightblulb_reset/
In other news, a few months ago my microwave oven keypad broke; the
first two rows of numbers quit working, 012 and 345....
So last night we had a brief power failure, and now it all works! I
don't understand how that's possible.
Are you joking? ... The tiny MCU needed a good reset. The MCU probably is never turned off ...
It may have been a short power glitch that knocked it silly in the first place
Sounds like the MCU should have been given a finite-state-machine program, but
got a rat's-nest instead. Real-time (interrupt driven) software is an
art that doesn't have a lot of literature and cirriculum support.
Actually I was taught state machine design including async state machines. I only ever once saw such a state machine, but I could understand it and could have reworked it if needed.
I'm not going to second guess a design I know nothing about. It did it's job with no issues for a couple of decades before needing to be unplugged for a fix. I think that is a pretty good track record.
--
Rick C.
-- Get 1,000 miles of free Supercharging
-- Tesla referral code - https://ts.la/richard11209
J
Jan Panteltje
Guest
On a sunny day (Fri, 21 Jun 2019 02:50:40 -0700) it happened John Larkin
<jjlarkin@highlandtechnology.com> wrote in
<5f9pge5lt0dfg0uf9elk55qli93fucf876@4ax.com>:
Using at at work?
Could it be some of your pulse experiments upset it?
OTPH what is secure?
I wrote some software for the Dlink DCS-900 IP camera many years ago:
http://panteltje.com/panteltje/mcamip/
and found that I could just read from it without username and password...
I contacted them, and got a nice reply asking how I could write that soft..
Replied I used 'snort' (open source network intrusion detection system)
looking at what browser send...
They did seem very curious, wanted all my test files, I asked why,
never heard from them again.
I was a good camera, still using it,
but light sensitivity of CMOS sensors has since improved dramatically.
Nothing is really secure if you are motivated,
I was motivated because I did not want to run MS windows,
and they had no Linux software.
<jjlarkin@highlandtechnology.com> wrote in
<5f9pge5lt0dfg0uf9elk55qli93fucf876@4ax.com>:
On Fri, 21 Jun 2019 15:03:52 +1000, Chris Jones
lugnut808@spam.yahoo.com> wrote:
On 21/06/2019 08:36, John Larkin wrote:
On Thu, 20 Jun 2019 16:14:34 -0400, bitrex <user@example.net> wrote:
On 6/20/19 3:44 PM, John Larkin wrote:
https://www.theregister.co.uk/2019/06/20/ge_lightblulb_reset/
In other news, a few months ago my microwave oven keypad broke; the
first two rows of numbers quit working, 012 and 345. That made things
tricky. Had to cook things for 6:66 minutes and seconds. Setting time
was hard. I assumed that some trace had broken in the membrane thing.
So last night we had a brief power failure, and now it all works! I
don't understand how that's possible.
Software written by EEs
How can one bit of code break, and be fixed by a reset?
If the register that controls the pin direction (input / output) of the
IO pins that scan the keypad were to get corrupted, that would do it.
Most people don't reinitialise the IO registers except at startup.
OK, that would do it, a flipped port control bit taking out an entire
row of a matrix scan of the keypad. Somehow. ESD maybe.
Well, it's easier to poach eggs now. (In a water bath, to slow things
down.)
This microwave, an official Amana RadarRange (tm) is quirky as hell, a
smart appliance programmed by dumb people. Like the light bulb.
The RadarRange was Raytheon's name for the first microwave oven,
introduced in 1947.
Using at at work?
Could it be some of your pulse experiments upset it?
OTPH what is secure?
I wrote some software for the Dlink DCS-900 IP camera many years ago:
http://panteltje.com/panteltje/mcamip/
and found that I could just read from it without username and password...
I contacted them, and got a nice reply asking how I could write that soft..
Replied I used 'snort' (open source network intrusion detection system)
looking at what browser send...
They did seem very curious, wanted all my test files, I asked why,
never heard from them again.
I was a good camera, still using it,
but light sensitivity of CMOS sensors has since improved dramatically.
Nothing is really secure if you are motivated,
I was motivated because I did not want to run MS windows,
and they had no Linux software.