These are chat archives for CORE-POS/IS4C

2nd
Aug 2016
Finn
@thefinn93
Aug 02 2016 00:01
oh
i think we can turn that on
testing
yeah so that worked
Aug 01 17:04:48 jebian-dessie mono[30306]: RECV FROM SCALE: 08A079340687219
Aug 01 17:04:48 jebian-dessie mono[30306]: PASS TO POS: 079340687219
didn't actually get passed to the POS though
Andy Theuninck
@gohanman
Aug 02 2016 00:06
Hmmm. Is it dropping files into ss-output? Easiest to check with the browser closed
Finn
@thefinn93
Aug 02 2016 00:09
yeah it is
 ✘ ⚡ root@jebian-dessie  /var/www/IS4C/pos/is4c-nf/scale-drivers/drivers/NewMagellan/ss-output   master ●  find 
.
./tmp
./tmp/blah.txt
./00000244dac3b-a4e2-4727-91a4-f56b781775b4
./0000262137f39-2eda-4be8-9c20-fde617e37a2d
./0000335991221-ae15-4314-b83a-19dbc3a74a89
./0000120661845-8d43-481f-b4ad-66d4511cfe97
hrm
re-opening it the POS appears to have picked them up?
yeah, scanning something doesn't show up in the list of items, but refreshing the page a few times makes it show up
oh wait wait i think it magically started working
Andy Theuninck
@gohanman
Aug 02 2016 00:14
Most likely culprit would be an error in PHP after adding the record but before sending a response (AJAX request gets an HTTP 500) or something extra in the input parsing's ouput makes it invalid JSON.
Finn
@thefinn93
Aug 02 2016 00:25
hrm
i didn't see anything in the PHP error log
oh hrm
Aug 1 17:20:18 jebian-dessie lane[2204]: (debug) stream_socket_client(): unable to connect to tcp://127.0.0.1:15674 (Connection refused) Line 117, /var/www/IS4C/pos/is4c-nf/lib/MiscLib.php
it's doing that a fair bit
Andy Theuninck
@gohanman
Aug 02 2016 00:27
That's normal - it's checking whether a RabbitMQ server is present and the connection failure triggers a warning
Finn
@thefinn93
Aug 02 2016 00:28
ok
Asa
@asapdx
Aug 02 2016 00:36
We ran into another bug, when items are rung up the total isn't updating. screenshot
Andy Theuninck
@gohanman
Aug 02 2016 00:37
does the total update on page refresh or is it just always zero?
Dan
@danry25
Aug 02 2016 00:38
@asapdx real professional, eh?
Asa
@asapdx
Aug 02 2016 00:38
Refreshing seems to update the total
Dan
@danry25
Aug 02 2016 00:38
Need to have Lukevers build some decent logos for this
Asa
@asapdx
Aug 02 2016 00:39
looks like some JS stuff
Andy Theuninck
@gohanman
Aug 02 2016 00:39
Yeah
Asa
@asapdx
Aug 02 2016 00:40
oh, cut off the right but it's from ajax-parser.js:21:9
Andy Theuninck
@gohanman
Aug 02 2016 00:41
Right. It's re-drawing the items then trying to redraw the items on the 2nd monitor / customer window. redrawing the footer would happen next if that 2nd window update didn't error
Finn
@thefinn93
Aug 02 2016 00:42
oh, maybe it's related to that bug we observed but ignored in the customer display where it doesn't open the page
just a blank page
Andy Theuninck
@gohanman
Aug 02 2016 00:43
That'd make sense. customerWindow.$ would be undefined if that window never loads jQuery
Finn
@thefinn93
Aug 02 2016 00:47
oh i think it was some weird state we managed to get the settings into
going to the Extra page and adjusting the height (i was curious to see if it'd use more of the display) then hitting save magically made the customer display work
¯_(ツ)_/¯
Andy Theuninck
@gohanman
Aug 02 2016 00:53
RE: height: you'd have to create a store.css and override the height on div.baseHeight with !important to make room for additional lines (yes, CSS for the lane is an antiquated mess)
Finn
@thefinn93
Aug 02 2016 00:53
k, don't care that much :P
i'll do it eventually
but my point is it magically fixed the customer display
Finn
@thefinn93
Aug 02 2016 04:08
Using xdotool and R-Kiosk, I've gotten the customer display and pos display looking pretty good
Not 100% yet, need to have some way to determine from the title of the window which is the customer display
As best I can tell they both have the same title, but the customer display wasn't loading when I was testing it
I also had to set browser.links.open_newwindow to 2 to make Firefox launch new windows in new windows
Lol
Finn
@thefinn93
Aug 02 2016 05:58
filed #788 with some fixes to the customer screen
Dan
@danry25
Aug 02 2016 08:03
so CorePOS seems to not be picking up scanned items
I see the ncr driver writing it to a file for corepos to pickup
also the 7878 stops scanning randomly
gonna try to debug that
Asa
@asapdx
Aug 02 2016 08:04
Also, when scanning an item that uses age verification the item triggers the password prompt to verify their age, but the item doesn't actually show up afterwards.
Asa
@asapdx
Aug 02 2016 23:00
Scale driver seems to have just stopped parsing data?
-Dan
I can see the item get scanned in cat /dev/ttyS0, but the driver appears to not be passing the live weight or barcodes read to the point of sale
Andy Theuninck
@gohanman
Aug 02 2016 23:08
I'm not sure what to add besides look at logs and whether files appearing in ss-output. I can't debug in depth with a device that I don't have and have never worked with
Finn
@thefinn93
Aug 02 2016 23:10
i went and SSH'd into his box, looks like the driver isn't receiving it, no "PASS TO POS" line
Andy Theuninck
@gohanman
Aug 02 2016 23:11
Well, it'd always list RECV FROM SCALE if the device is sending data with the expected message delimiter. PASS TO POS would follow if the driver understands the message
Finn
@thefinn93
Aug 02 2016 23:12
hrm, lemme check without the grep
yeah, no RECV FROM SCALE
Andy Theuninck
@gohanman
Aug 02 2016 23:13
Then the ETX byte changed or the driver somehow lost connection to the serial port (there are no known bugs or circumstances wherein that could happen though some I'm not sure where you'd even start investigating)
Finn
@thefinn93
Aug 02 2016 23:14
hrm... okay
that was my thought as well
oh, just VNC'd in, he's running cat /dev/ttyS0
would that mess up the connection for the driver?
Andy Theuninck
@gohanman
Aug 02 2016 23:17
No clue. I don't know if bytes get delivered to both processes. On Windows you simply can't do that. If one process has the port open a 2nd attempt to access it will fail
Asa
@asapdx
Aug 02 2016 23:17
I've stopped doing that and restarted the service -Dan
Finn
@thefinn93
Aug 02 2016 23:21
okay, got interceptty up and running, it appears this is the scanner being wierd
sometimes prefixing it with 8A, other times 08A
other times just not the full scan
Andy Theuninck
@gohanman
Aug 02 2016 23:36
does it output hex or some other non-ascii to see where the STX/ETX are and if a checksum byte is present?
Finn
@thefinn93
Aug 02 2016 23:36
it's ending in 0x03 byte... i didn't see a checksum byte but lemme check
Andy Theuninck
@gohanman
Aug 02 2016 23:37
K. Code isn't expecting a checksum
Asa
@asapdx
Aug 02 2016 23:37
>     0x32 (2)
>     0x30 (0)
>     0x37 (7)
>     0x34 (4)
>     0x30 (0)
>     0x38 (8)
>     0x03
Finn
@thefinn93
Aug 02 2016 23:37
that's one of the weird short barcodes it received
i believe that was when a barcode ending in 207408 was scanned
Andy Theuninck
@gohanman
Aug 02 2016 23:40
Looks cut off to me, too. 0x2 should be the STX byte regardless of whether 08 vs 8A vs 08A follows
Maybe different bytes are still going to different processes?
Finn
@thefinn93
Aug 02 2016 23:40
hrm, maybe
i gotta run, dan will likely be playing around with this for a while
Andy Theuninck
@gohanman
Aug 02 2016 23:45
You could put Console.WriteLine(b); after https://github.com/CORE-POS/IS4C/blob/master/pos/is4c-nf/scale-drivers/drivers/NewMagellan/SPH_NCR_Scale.cs#L129. The driver's reading one byte at a time anyway so logging each one received is simple enough. Then it's only one process trying to read from the port