CityScope / CS_Cooper-Hewitt

meta repo/sandbox repo for keeping everything related to the Cooper Hewitt exhibition
5 stars 2 forks source link

status check #101

Closed cpgmedialab closed 5 years ago

cpgmedialab commented 5 years ago
time who status
11/24 Sat 2:10pm Yasushi all green 👍 (??)
11/25 Sun 2:10pm Yasushi *1 scanner version miss match
11/25 Sun 5:00pm Yasushi & Arnaud scanner reboot, flipping corners, all green
yasushisakai commented 5 years ago

1 :

cpgmedialab commented 5 years ago

@yasushisakai 's remark came from https://github.com/RELNO/CityScope_Scanner_Python/issues/6

RELNO commented 5 years ago

@yasushisakai my best guess is that @popabczhang @Carsonsmuts edit to add a separated UDP channel for Unity was never PRed/pushed. Therefore, if the mac pro machine is now running the Scanner's master or any other commit, unity won't display interaction.

popabczhang commented 5 years ago

@yasushisakai my best guess is that @popabczhang @Carsonsmuts edit to add a separated UDP channel for Unity was never PRed/pushed. Therefore, if the mac pro machine is now running the Scanner's master or any other commit, unity won't display interaction.

That sounds right. Let's keep it running for stress test until tomorrow morning. Tomorrow let's have a try to see if Processing and Unity can listen to the same port (highly likely because they are on the different machines now). I will update you @RELNO if I succeed or not.

cpgmedialab commented 5 years ago

ok, at least we have a guess. Status is red now (not sure from when)

popabczhang commented 5 years ago

time who status 11/24 Sat 2:10pm Yasushi all green 👍 11/25 Sun 2:10pm Yasushi *1

@yasushisakai, does "11/24 Sat 2:10pm Yasushi all green" mean the building types changed correctly on Sat?

yasushisakai commented 5 years ago

thing is, not sure. I remember I saw it has responded. Don't remember if the buildings were synced (for both).

Friday night (time unknown) it was working.

RELNO commented 5 years ago

@yasushisakai if the mac pro is running the same scanner version as on your last status check, the problem is NOT what I've described. If otherwise you've pulled a new commit, then it might explain. Can you verify?

Carsonsmuts commented 5 years ago

I was in this morning Sunday 25th 11:30am. Everything was working fine. I moved the slider and both Processing and Unity followed the slider line, and as far as I could tell everything looked good for a stress test.

Carsonsmuts commented 5 years ago

@yasushisakai my best guess is that @popabczhang @Carsonsmuts edit to add a separated UDP channel for Unity was never PRed/pushed. Therefore, if the mac pro machine is now running the Scanner's master or any other commit, unity won't display interaction.

Everything was reacting normally to interaction this morning. Testing in person at the lab 11:30am Sunday 25th

cpgmedialab commented 5 years ago

@Carsonsmuts slider was and is working fine. Did you observe the building mismatch?

Y.S.

Carsonsmuts commented 5 years ago

@Carsonsmuts slider was and is working fine. Did you observe the building mismatch?

Y.S.

I see.... I was preoccupied with the communication pipeline and the stability of the software. I only checked to see that the two computers were still talking over the wireless router and that both processing and unity were updating. Whether the data for the buildings was correct, I did not observe any mismatch, but keep in mind I was looking elsewhere for issues.

yasushisakai commented 5 years ago

@Carsonsmuts Yup, the network was stable. I think we just want the commit id or a tag for each component, so we make sure we run the right one.

Table state has recovered, by flipping the corners of keystone.py of the scanner

agrignard commented 5 years ago

duplicate with #83 (needs to keep only one issue) even if a status ceck is not really an issue it's more a report.

RELNO commented 5 years ago

Putting everyone on the same page: https://github.com/RELNO/CityScope_Scanner_Python/commit/70255a2a34f9f566dc61c3edb37ea04df6c1630c merged two [sort of] major changes which needed to be stress-tested:

agrignard commented 5 years ago

hard to maintain as an issue (not reproductible, nothing to "fix")

I think now the status is the table is working by default if not a dedicated issue is created