[Logiweb] Process control at logiweb.eu

[Klaus Ebbe Grue]

Hi Frederik,

> Right now I'm working on an experimental pyk file using the input window at 
> http://logiweb.eu:8080/logiweb/tutorial/submit.html. Somewhere there's a bug 
> in my code causing an infinite loop. Therefore, verifying the file brings up 
> the error message:
>
>>>> *** - Program stack overflow. RESET
>
> There's no other response.
>
> Now, when I run the same pyk file at topps.diku.dk, the pyk process enters a 
> LISP read-eval loop which you have to exit manually:
>
>>>> *** - Program stack overflow. RESET
>>>> [1]> (exit)
>>>> [eriksen at localhost ~]$
>
> In other words, the process doesn't terminate by itself.
>
> My question is: Does the process at logiweb.eu terminate by itself? I am a 
> little concerned that I might be flooding logiweb.eu with unterminated 
> processes.

I've checked and there are no 'pyk' processes running at logiweb.eu:8080.

By the way: when you find it convenient, please move from logiweb.eu:8080 
to logieb.eu. I only keep logiweb.eu:8080 alive for the sake of backward 
compatibility. logiweb.eu:8080 and logiweb.eu point to the same machine.

The 'pyk' compiler is supposed to die on all errors, so it should die 
after writing 'Stack overflow ...'.

It is peculiar that the pyk compiler does *not* exit on topps.diku.dk. 
Have you set quit=no somewhere? You can check by issuing
> pyk -- -optionval=quit
Otherwise I will check up on what happens on topps.diku.dk.

I principle, the system could saturate if many pyk compilers loop 
indefinitely. I will do something about it when it becomes a problem or 
when I have time for it, whatever happens first.

Cheers,
Klaus