On Wednesday 16 Mar 2016 18:24:25 Nico Bouthoorn wrote:
> I've the same issue. Version owfs 2.8p15-1 is working fine, running Debian
> GNU/Linux 7 on a beaglebone.
It's sort of reassuring that someone else has problems :(
> This schematic: https://fstab.nl/en/lcd_owfs
That appears to be a 0x29
On Wednesday 16 Mar 2016 18:24:25 Nico Bouthoorn wrote:
> I've the same issue.
Guess it would have been useful to ask what 1-wire host adapter you are using
in my previous reply :/
Andy
--
Transform Data into Opportuni
I've the same issue. Version owfs 2.8p15-1 is working fine, running Debian
GNU/Linux 7 on a beaglebone.
This schematic: https://fstab.nl/en/lcd_owfs
When i switch to version owfs-3.1p1 it doesn't work anymore.
Commands like this:
owwrite /29.09C21100/LCD_H/clear 1
owwrite /29.09C21
So.. Just to sum things up before digging further.
We have a 0x29 (DS2408-based) LCD, which works in 2.8p15, but not on 3.1p1
We have a 0xFF (L. Swart-based) LCD, which works in 2.9p3, but not in 2.9p5
Both use DS2482-800-based i2c controller.
Both exhibit the same issue: the first character wri
On 16/03/16 21:10, Johan Ström wrote:
> For L.Swart-LCD, do you have the issue when writing to both screenXX and
> lineXX? Please include full commands used when writing.
>
Actually, if you can run this it would probably help even more:
Use the latest version of owfs (either git or latest releas
On Friday 18 Mar 2016 21:49:49 Johan Ström wrote:
> I've patched OWNet.pm in master now, to not send a null-terminated
> payload. The main issue was that that the payload field held the
> null-terminated length of full payload, but the size field held the
> non-null-terminated length (of the dat
On 19/03/16 10:27, Andy Carter wrote:
> On Friday 18 Mar 2016 21:49:49 Johan Ström wrote:
>> I've patched OWNet.pm in master now, to not send a null-terminated
>> payload. The main issue was that that the payload field held the
>> null-terminated length of full payload, but the size field held the
Hello,
> You've come a far bit by pinpointing which version it broke in (2.9p4 or
> p5), let's see if we can fix this :)
As 2.9p5 is only an bugfix-realease for the damaged version 2.9p4
(Regarding information on the changelog), the fault should to be
found between 2.9p3 and 2.9p5. Version 2.9p
On Wednesday 16 Mar 2016 21:10:24 Johan Ström wrote:
> So.. Just to sum things up before digging further.
>
> We have a 0x29 (DS2408-based) LCD, which works in 2.8p15, but not on 3.1p1
> We have a 0xFF (L. Swart-based) LCD, which works in 2.9p3, but not in 2.9p5
> Both use DS2482-800-based i2c co
On Saturday 19 Mar 2016 10:51:53 Johan Ström wrote:
> > I've just tried reinstating the raspbian jessie owfs etc. to check that
> > also works with the new OWNet.pm but I seem to have a broken system -
> > back later :/
> >
> > Many thanks for this
>
> Great! Let's hope this solves it for Nico to
On Friday 18 Mar 2016 05:49:28 Andy Carter wrote:
> > "git bisect" is your friend.
>
> Thanks, looks very useful. I'll try that over the weekend, won't have time
> today.
Well, I tried, got a useful looking set of files but had no idea what to do
with them.
My very limited experience of git ha
Hello,you can look a the following link by buildroot, where the packed owfs is now also includefor an build with it. There is the patch already active:https://git.busybox.net/buildroot/tree/package/owfs/0002-compat.h-only-include-compat_netdb.h-if-needed.patchBest regards,Roland Gregg Levine hat
On Wednesday 16 Mar 2016 21:26:43 Johan Ström wrote:
> On 16/03/16 21:10, Johan Ström wrote:
> > For L.Swart-LCD, do you have the issue when writing to both screenXX and
> > lineXX? Please include full commands used when writing.
Should have mentioned that LS LCD doesn't have those exact commands
On Wednesday 16 Mar 2016 22:25:30 Johan Ström wrote:
> When testing these old versions, you may have to build with ./configure
> --enable-debug --enable-traffic (or something similar, cannot recall
> exactly), as these was not default until just a few releases ago.
>
> With that, can you please
ds2482-800 i2c
see https://fstab.nl/en/beaglebone
Andy Carter wrote:
> On Wednesday 16 Mar 2016 18:24:25 Nico Bouthoorn wrote:
>> I've the same issue.
>
> Guess it would have been useful to ask what 1-wire host adapter you are using
> in my previous reply :/
>
> Andy
>
>
On 03/16/16 00:00, Arnout Vandecappelle wrote:
>
>
> On 03/15/16 04:30, Matthias Urlichs wrote:
>> On 10.03.2016 09:38, Arnout Vandecappelle wrote:
>>> Unless of course we decide that a system-supplied getaddrinfo (and
>>> inet_ntop)
>>> is required, and remove the compat stuff completely. I
I spotted another difference:
owwrite messages are NOT null terminated, but perl OWNet ones yes.
I remember a short discussion on this list in which I asked if owserver
messages should be null terminated. I think that Paul changed the code to
reflect my observation, see the thread:
https://sou
On 18/03/16 09:00, Stefano Miccoli wrote:
I spotted another difference:
owwrite messages are NOT null terminated, but perl OWNet ones yes.
I remember a short discussion on this list in which I asked if
owserver messages should be null terminated. I think that Paul changed
the code to reflect
Just a simple point:
with owwrite you are writing ‘/FF.xxx’ but with perl 'uncached/FF.xx’, so your
test are not 100% equivalent.
Please check if this is the reason for the different behaviour (which could
give a clue on the bug).
S.
> On 17 Mar 2016, at 12:17, Andy Carter wrote:
>
> On We
Hello!
Can you compress the patch? Either use Bzip2 or even ordinary Zip? Or
even our old friend Gzip. There's a peculiar 40K limit on attaching
files to this list.
Oh and how big was it originally?
-
Gregg C Levine gregg.drw...@gmail.com
"This signature fought the Time Wars, time and again."
On 16.03.2016 14:44, Johan Ström wrote:
> git log v2.9p3..v2.9p5 -p
>
> shows every row changed between 2.9p3 and 2.9p5.
> I don't have time this second to go through this myself, may find some
> timer later if you haven't found anything until then.
"git bisect" is your friend.
--
-- Matthias
On Wednesday 16 Mar 2016 14:52:04 Jan Kandziora wrote:
> Am 16.03.2016 um 13:29 schrieb Andy Carter:
> There is not much of a change between 2.9p4 and 2.9p5, so I think it
> would be "not OK" either.
I guessed that to be the case.
> > I have no idea where to look for a possible cause but willing
If you are familiar with programming in general you could always skim
over the changes and see if you see anything standing out:
git log v2.9p3..v2.9p5 -p
shows every row changed between 2.9p3 and 2.9p5.
I don't have time this second to go through this myself, may find some
timer later if you hav
On 16/03/16 22:11, Andy Carter wrote:
> On Wednesday 16 Mar 2016 21:10:24 Johan Ström wrote:
>> So.. Just to sum things up before digging further.
>>
>> We have a 0x29 (DS2408-based) LCD, which works in 2.8p15, but not on 3.1p1
>> We have a 0xFF (L. Swart-based) LCD, which works in 2.9p3, but no
Am 16.03.2016 um 13:29 schrieb Andy Carter:
> I've built a number of versions this morning. Results are...
>
> 2.8p15 OK
> 2.9p0 OK
> 2.9p2 OK
> 2.9p3 OK
> 2.9p5 not OK
>
> 2.9p4 sources don't seem to be available
>
There is not much of a change between 2.9p4 and 2.9p5, so I think i
On Friday 18 Mar 2016 04:45:05 Matthias Urlichs wrote:
> On 16.03.2016 14:44, Johan Ström wrote:
> > git log v2.9p3..v2.9p5 -p
> >
> > shows every row changed between 2.9p3 and 2.9p5.
> > I don't have time this second to go through this myself, may find some
> > timer later if you haven't found an
On 18/03/16 09:03, Johan Ström wrote:
>
> On 17/03/16 12:17, Andy Carter wrote:
>> On Wednesday 16 Mar 2016 22:25:30 Johan Ström wrote:
>>> When testing these old versions, you may have to build with ./configure
>>> --enable-debug --enable-traffic (or something similar, cannot recall
>>> exactl
On Wednesday 16 Mar 2016 14:44:56 Johan Ström wrote:
> If you are familiar with programming in general you could always skim
> over the changes and see if you see anything standing out:
>
> git log v2.9p3..v2.9p5 -p
>
> shows every row changed between 2.9p3 and 2.9p5.
A little coding experience
28 matches
Mail list logo
