Install and ProxiedDevicesControlField

[PERALDI Olivier]

Hello,

1/ I have installed the last version of OLA on a new VMware machine, and I got an error while I was doing the "make check" step. The function was looking for files. I have been obliged to copy the ola and the olad directory from /usr/share/ to /usr/local/share/

2/ When I am running the rdm test, I have a problem with "ProxiedDevicesControlField" function. The http windows freezes, and the terminal is definitively looking for something (see attached). Since I don't know what this function is looking for, I have some problem to troubleshoot it. Can you give me some clue.

Thank you. Best regards.

Olivier

[Peter Newman]

Hi Oliver,

We'd need to see some details to assist with 1.

When you say latest, do you mean 0.10.3, or from Git or something else?

Can you try running the command line version with debug, see http://docs.openlighting.org/ola/man/man1/rdm_responder_test.py.1.html but likely something like this based on your log:
rdm_responder_test.py -s -u 0 -t ProxiedDevicesControlField -d 4d44:010c0000

Then we can try and work out what's happening.

[PERALDI Olivier]

Hello Peter,

Thank you for your answer.

1/ I have used "git clone https://github.com/OpenLightingProject/ola ola" command. From all the logs, it corresponds to the 0.10.3 version.

2/ I have run the command as requested. The answer is attached. 

This does not help me so far. 

- GetSupportedParameters does not include any Proxy function (We don't use any proxy in fact).

- MuteDevice answer with a ControlField equal to 0, meaning it is not a proxy and does not manage any proxy. 

Normally, I would expect to see this command ending with an unMute, but the soft seems to be trying to do something with a Proxy function, but does know what to do!!!! Am I right?

Hope this help. Have a nice day.

Olivier

[Peter Newman]

Okay, we'd actually call that master. It's version 0.10.3 plus a lot of unreleased changes. It should work, but it won't have been as widely tested as a proper release like 0.10.3. Think of it as the bleeding edge of OLA development.

In this case, I think the changes had a bug, which I've already found and fixed. Specifically you want the change to ResponderTest.py here:
https://github.com/peternewman/ola/commit/2b84c31c372366feff387baeeb7962aaf20742b8

Then it should work as intended.

[PERALDI Olivier]

1/ I understand the fact that last is a beta, but I thought that git pull was really downloading the last stable version. Why last is released as 0.10.3 if it is a version different from 0.10.3? It is a bit confusing... :-)

2/ Modifications solved the problem. Thank you.

Best regards.

Olivier

[Peter Newman]

Hi Olivier,

If you wanted to download the official 0.10.3 release via git there is a tag for it here:

https://github.com/OpenLightingProject/ola/releases/tag/0.10.3

There is also our 0.10 branch, which is the 0.10.3 code (or the latest 0.10.* release) with any additional bugfixes only on top (but no new features):

https://github.com/OpenLightingProject/ola/tree/0.10

I agree the versioning isn't perfect, but without adding an extra bit to the version to flag if it's a release candidate or stable or whatever, there's not much we could do. It's just simply that no-one has released the new version yet/updated the number. We prefer to do it just as we release, so the commit where we bump the version number ( https://github.com/OpenLightingProject/ola/commit/07f94dae3864656277563b110a08ca6dde54ebf5 ) is the tag for that version. The other alternative would be to increment it as soon as we release 0.10.3, but then how could you tell whether the code now, or the code next week, is the full 0.10.4. At least this way, you get some extra features on 0.10.3, rather than not enough.

Glad they fixed you're issue. They'll be in 0.11.0 when it's released at some point in the future.

[PERALDI Olivier]

Hello Peter,

My comments was based on the following issues:

1/ With vMware, I was planning to rebuild a stable OLA/RDM_responder virtual machine.

2/ From the linux tutorials (which are not up to date, by the way... "git clone https://code.google.com/p/open-lighting/ ola")l, I ended with a 0.10.3 version improved, but not stable. I would have been the same problem by using the "git pull" command you recommend to upgrade the program.

For those who are running the rdm tests from time to time (and thus upgrade the OLA from time to time), to test our new firmware, it is a problem to upgrade to a version which is not stable.... First, we may have problem to recompile it (and when you are not a linux specialist, it is a tremendous problem!!!!). Then we don't know if the rdm issue is coming from our code or from he rdm/ola test program. When I have time, it is not an issue (your help is really appreciated, by the way), but when the timing is critical, it is quite a problem.

I have now 4 different versions of OLA on Ubuntu (14, 16), Mint, Raspberry PI just to be sure to be able to run my tests even if one version fails...

You are doing a great job with that suite of programs. Thank you to all of you.

Olivier

[Peter Newman]

I've you're running the RDM responder tests, I'd personally be inclined to say that master is probably the best bet. You're presumably trying to test some bit of kit, so you want it to be as thoroughly tested as possible. Hence you probably want to run the latest tests available. We do test new tests before we commit them, or changes, it was just unfortunate this one only triggered on some types of device, so got missed before. To put it another way, if this test hadn't failed, would we be having this conversation?

In your case, I'd be inclined to run the old/current version of tests you have against the new firmware, then you could upgrade and see what, if anything, breaks at a time that suits you.

Remember that with both the VM and the Pi you can easily take a snapshot/backup of the entire machine, so you've always got something to roll back to if the upgrade fails or there is a bug with the tests.

Were you referring to this page? I've made some updates to hopefully improve it:
https://www.openlighting.org/ola/tutorials/ola-on-windows-via-vmware/

You really shouldn't need two different versions of Ubuntu for running the tests, I'd just stick with the later one if I were you.