Aaaagh! Promises are hurting my brain.

I'm trying to figure out how to provide a useful subset of E's reference mechanics in newpb/foolscap. Specifically, one of the clever things that E does is to provide Promise Pipelining, a limited form of remote code execution, in which I can ask you for an object and tell you to deliver a message to that object in a single round trip (rather than the usual two). So I want to be able to do something like:

# target, record, and results are all Promise objects
target = tub.getReferenceAsPromise(sturdyref)
record = send(target).getRecord(args)
results = send(record).getField(otherargs)
def printResults(r):
  print r
when(results).addCallback(printResults) # when() returns a Deferred

You can also include Promises as arguments:

record = send(target).getRecord(args)
send(laserprinter).printRecord(record)

So I'd like to provide this feature in python/foolscap, both because using Promises as a programming technique holds a lot of promise (as it were) for being a cleaner asynchronous style, and because it opens up the possibility of doing pipelining (which is an actual performance win).

The challenge is that E has very different reference mechanics than python. In E, any reference could be a Promise. (specifically, each reference is in any one of 5 states: LocalPromise, RemotePromise, Near, Far, and Broken). Whereas in python, references are always Near, and we have to fake everything else with wrapper objects.

My current approach is to have the Promise class be the wrapper and have it handle everything except Near references. The basic Promise is created with a matching resolver:

promise, resolver = foolscap.makePromise()
resolver(result) # resolves the promise

But the most common way to get one is to do an eventual send to something:

from foolscap import send
class Adder:
  def add(arg):
    return arg+1
a = Adder()
promise = send(a).add(4)

There are only two things you can do with a promise: send it more messages, and wait for it to resolve. The former is done with send (which accepts either a promise or a regular object, and always does an eventual-send), the latter is done with when:

from foolscap import when
d = when(promise)
d.addCallback(printResults)

The when always returns a Deferred that will fire with the resolution of the Promise. So send moves us from the synchronous world to the asynchronous+promise world, while when and addCallback move us back to the synchronous one. (when by itself moves us from the asynchronous+promise world to the asynchronous+Deferred world).

So far so good. But here are some problems:

  • can Promises be resolved multiple times? I don't think so. The state transitions between LocalPromise and RemotePromise don't count.
  • how should this interact with eventual-send? certainly when you do send(a).foo(), and a is a normal reference (not a Promise), that a.foo() call should not happen in the same call stack (i.e. Vat turn, i.e. Reactor tick). But should all the events be sent in a big batch as soon as the promise is resolved? Or should they be sent out one at a time somehow? I suppose if Promises can only ever be resolved once, this is not as complicated as I'd first thought.
  • Should we try to support immediate calls to resolved Promises? In E, if you have a Near reference, you can do both immediate and eventual sends. In python, it would look like:
p = send(obj).foo(args)
# later.. p is probably resolved
send(p).bar() # eventual send
p.baz() # immediate send

Hm, maybe that isn't such a great idea.