[Bf-committers] How can I send performance data reliably?

Shaul Kedem shaul.kedem at gmail.com
Tue Mar 11 08:52:37 CET 2008


Ok,
 How about only blender.org registered users will be able to post
their perf. data ?

Shaul

On Tue, Mar 11, 2008 at 9:35 AM, Eric Cosky <eric at cosky.com> wrote:
> When all the code is on a machine outside your control, your options are
>  limited and none are going to guarantee 100% that nobody has messed with the
>  data. You do have options though. Obfuscate as much as you can is a good
>  start, but be sure to add marker values to detect when the 'obvious' values
>  are messed with.
>
>  Unfortunately if it is important to people, they will probably reverse
>  engineer it. If your obfuscation & marker techniques don't change very often
>  or very much then it's probably not worth the bother.
>
>  The only real solution to this problem that I can think of is to require a
>  connection to a server, have the server generate some unique 'seed' data (a
>  .blend file with a generated image, for example) that must be processed by
>  the client and automatically submitted, and the server can then timestamp
>  the submission & verify the results are what they should be given the seed
>  data.
>
>  A lot of bother imho..
>
>
>
>  -----Original Message-----
>  From: bf-committers-bounces at blender.org
>  [mailto:bf-committers-bounces at blender.org] On Behalf Of Shaul Kedem
>  Sent: Monday, March 10, 2008 11:05 PM
>  To: bf-blender developers
>  Subject: Re: [Bf-committers] How can I send performance data reliably?
>
>  Thank you for the idea - but what happens if using an editor, or a
>  debugger (or other means) I change the information right after the
>  measurement ?
>
>
>  On Tue, Mar 11, 2008 at 12:15 AM, Jasper Mine <jaspermine at earthlink.net>
>  wrote:
>  > Just a quick idea. Obfuscated key in code to encrypt data, some kind
>  >  of hash code to coincide with database and perhaps a bit overkill,
>  >  public and private keys for client and dB.
>  >
>  >  Jasper
>  >
>  >
>  >
>  >  On Mar 9, 2008, at 8:35 AM, "Shaul Kedem" <shaul.kedem at gmail.com> wrote:
>  >
>  >  > Hi,
>  >  > I've started thinking about using the work done with the system
>  >  > information script and python scripts enabling automatic processes to
>  >  > be executed by blender in order to gather performance data. The next
>  >  > step would be sending all of this information to a central DB, but I
>  >  > would like to make sure that no one is tampering with anything in the
>  >  > process - memory, python scripts, blend files, performance data and so
>  >  > on - is there some way of doing it either in python or in C?
>  >  >
>  >  > The reason I think this is important is that performance testing
>  >  > tends to become a competitive thing with some people (and companies)
>  >  > and I want to make absolutely sure no one is tempering with the data,
>  >  > so the figures will be accurate,
>  >  >
>  >  > Thanks,
>  >  > Shaul
>  >  > _______________________________________________
>  >  > Bf-committers mailing list
>  >  > Bf-committers at blender.org
>  >  > http://lists.blender.org/mailman/listinfo/bf-committers
>  >  _______________________________________________
>  >  Bf-committers mailing list
>  >  Bf-committers at blender.org
>  >  http://lists.blender.org/mailman/listinfo/bf-committers
>  >
>  _______________________________________________
>  Bf-committers mailing list
>  Bf-committers at blender.org
>  http://lists.blender.org/mailman/listinfo/bf-committers
>
>  _______________________________________________
>  Bf-committers mailing list
>  Bf-committers at blender.org
>  http://lists.blender.org/mailman/listinfo/bf-committers
>


More information about the Bf-committers mailing list