carette
2019-5-16 12:53:39

So I have managed to get the VirtualBox up, and ‘make run-gmm–25’ is running.


carette
2019-5-16 12:54:56

Things I was also really expecting to see: 1. some indication (in the instructions) of what the expected run should be 2. some indication of how long each ‘make XXX’ is expected to take 3. some indication of the stages of the compiler being run


carette
2019-5-16 12:56:11

For example, I don’t have any idea if Maple is being used for anything at all here. Or even, for that matter, Haskell. The only thing I see “for sure” being used is Racket.


carette
2019-5-16 13:00:31

In fact, I have never even seen the instructions given to the artifact evaluators — where is it? Why isn’t it with the paper (as well as stashed away wherever else)?


carette
2019-5-16 13:02:08

Lastly, @rjnw gave a password, but I never had to use it.


carette
2019-5-16 13:02:45

[I’m going to be not-on-slack for a couple of hours]


ccshan
2019-5-16 15:48:28

It would be nice if people at the talk in August can try out the example using the Maple code in our artifact this week.


carette
2019-5-16 16:52:54

And I had to use it when that window went to sleep. Ok, that would have been nice to know.


carette
2019-5-16 16:53:09

Still in the dark about the rest of the questions.


rjnw
2019-5-16 17:09:06

Maple is currently not being used we have precompiled code checked in and the benchmarks use that. I will try to calculate expected time for each target. There is also make allbench that does everything, should we give that? I also need to give the location of output pdf files in instructions. I can give some indications of what is running but if augur produces a lot of noise on stdout and it will just hide everything, so I don’t know how to tackle that.


carette
2019-5-16 19:09:11

I’m sorry, I can’t parse your sentence, as it seems to involve time travel. Can you try again please?


carette
2019-5-16 19:10:06

Are there global instructions somewhere? I’m still looking for those.


ccshan
2019-5-16 19:27:30

It would be nice if the artifact we submit this week supports the example that we present in August, so that people at the talk can try stuff out if they’re interested. If you agree then this Maple debugging is actually more urgent than reading related work…


rjnw
2019-5-16 19:45:28

I will write a detail set of instructions tonight after work. But for now it’s just the readme of hakaru-benchmarks repository


carette
2019-5-16 20:49:07

Clearer, thanks. Ok, changed priorities.


carette
2019-5-16 23:20:29

Hmm, so I left the first thing running on the VM on my home machine this morning - and when I came back, it had rebooted itself (not the VM, the whole computer). Not sure if they are related, but it’s not a good sign.


ccshan
2019-5-17 00:47:17

Is it expected for make run-nb to look like this unmoving for a long time?


rjnw
2019-5-17 01:17:19

that’s jags running it takes around 6hrs



rjnw
2019-5-17 02:05:08

I reran all the benchmarks once more to make sure they are working. I don’t know what the issue with @carette machine is maybe check the crash log of the operating system. I also delete the log files after generating pdf’s otherwise it fills up the whole virtual hard drive, I hope that is ok.