"...it just occurs to me that a basic patch"...

This should have read "virtual patch". Sorry.

-gt

On Mon, Jun 20, 2011 at 2:49 AM, George Toledo <[email protected]> wrote:

>
>
> On Mon, Jun 20, 2011 at 2:19 AM, vade <[email protected]> wrote:
>
>>
>> Now, considering you are not making an app, you are almost totally out of
>> luck, unless you write some custom plugins to load an encrypted composition
>> within an un-encrypted one. So you'd be loading two comps, one unencrypted
>> plain one which is simply a single instance of your "de-crypter" plugin.
>> That plugin points to a URL of an *encrypted* composition. Now, the issue
>> is, I dont think Keynote will happily run 3rd party plugins since its 'safe
>> mode'. If it does, thats your solution.
>>
>>
> A feature request for some variation of "virtual patch" or something like a
> composition importer, as opposed to loader, to do that would probably be a
> reasonable way of handling that, it seems like. A patch could embed the
> contents of the qtz as encrypted data in the qtz, and then decrypt it, and
> that should be currently possible, but wouldn't work because of safe mode,
> or would work but would be non-standard API.
>
> This is basically the same thing you're saying, it just occurs to me that a
> basic patch is sort of a black box in the qc editor, not so much in plist
> editor, and there's the ability to flatten it. If it was encrypted data with
> no ability to flatten, like a "locked" virtual patch, it would probably be
> useful to a lot of people, for reasons other than protection even.
>
>
>
>>
>> On Jun 19, 2011, at 9:19 PM, Alastair Leith wrote:
>>
>> Thanks Christopher
>>
>> Fundamentals I still don't *get* (apologies to the wise):
>>
>> If I was to employ someone using this QCComposition  +compositionWithData
>> method what do I end up with? Are we building a plugin that I incorporate
>> into a standard .qtz file in the QC Editor app, or can Xcode actually build
>> a .qtz file that references encrypted data so when this .qtz file is opened
>> in the QCEditor most of the contents remains hidden in an encrypted file. I
>> always thought a .qtz file was just a .plist describing patches and noodles
>> and a bit of user data.
>>
>> The issue is I'm not building my own app, I *need* a .qtz (or .mov) file
>> to provide the Theme user something to import onto the Keynote slides. It is
>> desirable that this .qtz file only reveal user adjustable inputs and little
>> else when opened in QCEditor. It doesn't want to be loading from another
>> .qtz file that is a plain text .qtz file obviously, that's all I'm capable
>> of today.
>>
>>
>> On 20/06/2011, at 1:14 AM, Christopher Wright wrote:
>>
>> It's seems important to me that the decrypted composition resides in
>> memory not written, even temporarily, to HD in plain-text. A force quit of
>> hosting app like Keynote or QC Editor then search ".qtz" & sort by date will
>> turn up the secret composition in plain text form otherwise (as pointed out
>> by Christopher on the Kineme thread).
>>
>>
>> In memory isn't much better than on-disk to a determined adversary.  And
>> to be honest this doesn't require a particularly determined adversary at all
>> (more than a casual finder-only attacker, but if that's all you're trying to
>> thwart then vade's "use an encrypted NSData" is more than sufficient - in
>> fact, that's what the NSData API was added for.  If you're wanting to
>> prevent even less casual attackers, you'll need to do some leg work to find
>> out how they do what they do, and then how to prevent that where possible
>> and how to make it inconvenient where not possible.  There's a cost/benefit
>> curve between complexity and breakability, and you'll need to find where
>> you're comfortable on that curve based on your constraints (how much
>> complexity are you willing to write/purchase from a developer, how much
>> illicit activity do you intent to prevent, etc.)
>>
>>
>> Okay so feature request to Apple stands, as good as Apple can make it, all
>> constraints considered. :-) (*Ignorance is bliss)*
>>
>> Casual sharing is the lesser issue, although, if a disincentive to finder
>> copying is possible, great. Effortless competitor pilfering of techniques is
>> the main concern. If a competitor can hire a ninja aquatinted in the
>> black-arts to can-open whatever we can produce in half a day, then I guess
>> I'm no better off bothering.
>>
>>
>>
>> Are you talking about writing a QC composition in Xcode that can load and
>> decrypted the NSData blob from another file?  Or are we talking about a
>> writing a QC plugin that loads the NSData and decodes it and somehow creates
>> a QCComposition object that can bind to the plugin patch's ports in QC.
>>
>>
>> QCComposition has documented API initializers, including
>> +compositionWithFile:   and +compositionWithData:.  you can create a mutable
>> data from a file on disk, decrypt it (using any one of a variety of
>> algorithms), and then feed it to +compositionWithData:.
>>
>>
>> Your requirements aren't particularly well-defined:
>>
>>  The encryption key remains secret from a inquisitive competitor! Where
>> would it be stored or would it auto-generated at runtime and stay secret
>> somehow? Or would it be stored in the Keychain at installation time thereby
>> becoming safe.
>>
>>
>> how inquisitive?  Also, you're totally missing out on numerous attack
>> vectors that happen _after_ decryption in memory, regardless of how locked
>> down the key is.  Stuff in the keychain is user-visible (via the "show
>> password" check box + enter their keychain password).
>>
>>
>>  *The machine does not isolate man from the great problems of nature but
>> plunges him more deeply into them.**
>> **Antoine de 
>> Saint-Exupery*<http://www.brainyquote.com/quotes/quotes/a/antoinedes131176.html>
>> * *
>>
>>
>>
>>  _______________________________________________
>> Do not post admin requests to the list. They will be ignored.
>> Quartzcomposer-dev mailing list      ([email protected])
>> Help/Unsubscribe/Update your Subscription:
>>
>> http://lists.apple.com/mailman/options/quartzcomposer-dev/gtoledo3%40gmail.com
>>
>> This email sent to [email protected]
>>
>>
>
>
> --
> George Toledo
> [email protected]
> www.georgetoledo.com
>
>
>


-- 
George Toledo
[email protected]
www.georgetoledo.com
 _______________________________________________
Do not post admin requests to the list. They will be ignored.
Quartzcomposer-dev mailing list      ([email protected])
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/quartzcomposer-dev/archive%40mail-archive.com

This email sent to [email protected]

Reply via email to