Changes between Initial Version and Version 1 of Ticket #20


Ignore:
Timestamp:
08/30/2007 07:03:21 PM (17 years ago)
Author:
Brian Warner
Comment:

Legend:

Unmodified
Added
Removed
Modified

  • Ticket #20 – Description

    initial v1  
    2020I want the contents of the box to be an arbitrary object graph (almost
    2121anything you could pass to {{{callRemote}}}), but things things like live
    22 references may or may not work. The reason that the sealer must be connected
    23 to the Tub is to allow things like references to be handled properly.
     22references may or may not work. Handling live references will certainly
     23require access to a Tub, which is why {{{create()}}} is a method of a tub
     24instance rather than a top-level Foolscap function. {{{seal()}}} and
     25{{{unseal()}}} return Deferreds for the same reason.
    2426
    2527My plan is to implement the serialization by using the normal banana code,
    … …  
    3436cryptographic one.
    3537
     38In E, each sealer/unsealer has a distinct "Brand" object, which also have
     39non-distinct string names (i.e. there could be two Brands with the same name
     40but which are completely unrelated). I'm not sure if I should do that here or
     41just attach a string name to the sealer/unsealer.
     42
    3643Here is a useful thread on cryptographic sealers/unsealers in the objcap world:
    3744http://www.eros-os.org/pipermail/cap-talk/2007-March/007595.html