Made changes to config files to build repyV1 runtime

[choksi81]

Updated config files to build repvV1 runtime for nodemanager build.

[choksi81]

#112

[aaaaalbert]

Test results?

[choksi81]

For repyV1 runtime, results were:

Chintans-MacBook-Pro:runnable chintan$ grep -c 'PASS' result.txt 
155
Chintans-MacBook-Pro:runnable chintan$ grep -c 'FAIL' result.txt 
69

Many of the Fails were thrown by virtual_namespace.py. They were of type:

Full debugging traceback:
  "repy.py", line 448, in <module>
  "repy.py", line 179, in main
  "/Users/chintan/repy_v1/RUNNABLE/virtual_namespace.py", line 78, in __init__

User traceback:

Exception (with type 'exceptions.ValueError'): Code failed safety check! Error: ("<type 'exceptions.SyntaxError'> invalid syntax (line 12)",)

[aaaaalbert]

Have a go and dig into that problem! (Maybe there is a V1/V2 incompatibility with one of the runtime files?)

[choksi81]

The problem is with the emulcomm file. All the unit-tests where communication is done (i.e. emulcomm is used) are failing. Others are getting passed.
If I remove emulcomm's method from the unit-tests, they pass.

[aaaaalbert]

Let's make sure we are talking about the same issues here.

First, the init config needs to include seattlelib_v1, see above.

Second, I can't see any problems in the nodemanager unit tests that would relate to RepyV1 being included.

However, checking out SeattleTestbed/repy_v1, building, and running the unit tests causes lots of SyntaxErrors. This is the problem that you mention. Take a look at any of the problematic source code lines --- you'll agree that we need to run testportfiller over the unit tests.

[aaaaalbert]

Instead of this PR, we chose an alternative implementation path: The RepyV1 directory is now included in SeattleTestbed/installer-packaging. It makes much more sense to include it there, because SeattleTestbed/nodemanager's build script can only build a nodemanager for testing anyway.