RE: UMGI testing external testing with Advestigo

From: Neil Saxby <nsaxby_at_mediadefender.com>
Date: Tue, 17 Jul 2007 10:11:01 -0700

Will do - Paul is out until some point next week, but once I've spoken to them I'll let you know.

Cheers,
Neil

-----Original Message-----
From: Randy Saaf
Sent: Tuesday, July 17, 2007 6:09 PM
To: Neil Saxby; Tabish Hasan; Ben Grodsky; Octavio Herrera
Cc: Jay Mairs
Subject: Re: UMGI testing external testing with Advestigo

Ok. Let us know how discussions continue. We can hire baytsp or audible magic as a last resort.


----- Original Message -----
From: Neil Saxby
To: Randy Saaf; Tabish Hasan; Ben Grodsky; Octavio Herrera
Cc: Jay Mairs
Sent: Tue Jul 17 09:55:11 2007
Subject: RE: UMGI testing external testing with Advestigo

All of their reports state that they used eMule v0.47c

I think Mark and Paul will accept that this evidence indicates that eMule' tests may have been unfair. That said, they're still concerned about the large number of fails on the other networks. Advestigo are claiming that they didn't use any IP filtering, PG, etc...

I'm happy to explain to Mark and Paul that this issue with the .met file makes it harder for us to believe Advestigo's results for the other networks too. I've already hinted politely that they may have ulterior motives for finding bad results but am willing to go into more detail with them if needs be. Please let me know.

The main issue for UMGI is convincing their labels that we are effective outside of the U.S. If we could run some tests with an independent company based in Europe it might help bring them back on board, or at least strengthen our argument.

Cheers,
Neil

-----Original Message-----
From: Randy Saaf
Sent: Tuesday, July 17, 2007 5:35 PM
To: Neil Saxby; Tabish Hasan; Ben Grodsky; Octavio Herrera
Cc: Jay Mairs
Subject: Re: UMGI testing external testing with Advestigo

Just read this.

Could it be that they are not using emule but another edonkey client that never updates its server met file.

If they r usng emule we caught them red handed. If they ran emule once in the past year razorback would have been booted from the list.

I don't see how this could be a debatable issue with marc.


----- Original Message -----
From: Neil Saxby
To: Tabish Hasan; Ben Grodsky; Octavio Herrera
Cc: Jay Mairs; Randy Saaf
Sent: Tue Jul 17 05:50:43 2007
Subject: RE: UMGI testing external testing with Advestigo

I explained about the .met file to Mark and mentioned that it’s becoming a difficult situation to resolve. I’m beginning to suspect more and more that Advestigo might have some slightly underhanded reasons to undermine the value in our services. Looking at the services they provide this seems even more likely: HYPERLINK "http://www.advestigo.com/protection.php?men=2&rub=2&LANG=ANG"http://www.advestigo.com/protection.php?men=2&rub=2&LANG=ANG. It also made me consider how they analyzed the files… using their fingerprint technology?

 

If this situation continues to cause UMGI to lose faith in our abilities I might see if there’s any way to get time stamped copies of the files that Advestigo downloaded.

 

Octavio also mentioned that it might be a good idea to run some tests with Advestigo ourselves. I’m trying to get some contact details without disclosing whether we will run tests with them or not. If we are going to do this we’ll need some way of ensuring that they aren’t using other techniques to circumvent our techniques. Please let me know if this is something you’d see value in.

 

Cheers,

Neil

 

From: Neil Saxby
Sent: Monday, July 16, 2007 9:45 PM
To: Tabish Hasan; Ben Grodsky
Cc: Jay Mairs; Randy Saaf
Subject: RE: UMGI testing external testing with Advestigo

 

Thanks Tabish,

 

I’ll politely mention this to Paul and Mark and let you know how it goes.

 

Cheers,

Neil

 

From: Tabish Hasan
Sent: Monday, July 16, 2007 6:37 PM
To: Neil Saxby; Ben Grodsky
Cc: Jay Mairs; Randy Saaf
Subject: RE: UMGI testing external testing with Advestigo

 

Neil,

 

Thanks for providing the update and the server.met.

 

Apparently the server.met they provided is super old and not reflective of what they’re really using. For example, they still have Razorback 2.0, which hasn’t been active in more than a year. And the most recent ping dates in their list are with a ping date of February 06, which indicates they haven’t used this server.met since February 2006. And the chance of their system time being incorrect, thus causing the old ping dates, is not possible because of the fact they have Razorback 2.0 in their list.

 

I’ve attached the parsed server.met for you.

 

-TH

 

   _____

From: Neil Saxby
Sent: Thursday, July 12, 2007 4:28 PM
To: Neil Saxby; Ben Grodsky; Tabish Hasan
Cc: Jay Mairs
Subject: RE: UMGI testing external testing with Advestigo

 

I got a copy of the .met Advestigo used for the UMGI testing back today and a few answers to some other questions:

 

1. They connect to the server that has the more users and files

2. They use global search

 

Cheers,

Neil

 

From: Neil Saxby
Sent: Thursday, June 21, 2007 6:52 PM
To: Ben Grodsky; Tabish Hasan
Cc: Jay Mairs
Subject: RE: UMGI testing external testing with Advestigo

 

They’ve only been contracted to test us once so far, so I reckon that the .met they used is the same as the one available at the link below. I’ll see if I can get an exact copy of the .met and find out:

 

1. How they connect to a sever: auto vs. handpicked

2. How they search: server vs. global (Please let me know if it’s ok to check this with Mark)

 

Apparently they’ve abided by all the other rules outlined in the testing methodology, but if you have any other queries let me know. It might take a little while for the answers, as I want to ask Mark casually, but I’ll let you know as soon as I get some more info.

 

Cheers,

Neil

 

From: Ben Grodsky
Sent: Thursday, June 21, 2007 6:22 PM
To: Neil Saxby; Tabish Hasan
Cc: Jay Mairs
Subject: RE: UMGI testing external testing with Advestigo

 

Can we get a copy of Advestigo's current server.met? Can we find out how they connect to ed2k servers? Do they auto-connect? Do they hand-pick a particular server?

   _____

From: Neil Saxby
Sent: Thu 21-Jun-07 10:19
To: Tabish Hasan
Cc: Ben Grodsky; Jay Mairs
Subject: RE: UMGI testing external testing with Advestigo

Hi all,

Mark Denby got back to me today. Advestigo are claiming that they don't use any IP filtering technologies. They also said that they obtained their server.met from here: HYPERLINK "http://www.emugle.com/server.met"http://www.emugle.com/server.met

Any feedback would be much appreciated. Their major concern is the usual: that we're not effective outside of America. I know that this isn't the case and it is highly likely that Advestigo got lucky (as I have done in the past), but I need something more substantial to keep the marketing people (with a little tech knowledge) happy/interested.

Many thanks,
Neil

-----Original Message-----
From: Neil Saxby
Sent: Friday, June 15, 2007 7:37 PM
To: Tabish Hasan
Cc: Ben Grodsky; Jay Mairs
Subject: RE: UMGI testing external testing with Advestigo

Cheers mate! I'll get back to you when UMGI come back with some info.

-----Original Message-----
From: Tabish Hasan
Sent: Friday, June 15, 2007 7:14 PM
To: Neil Saxby
Cc: Ben Grodsky; Jay Mairs
Subject: RE: UMGI testing external testing with Advestigo

Hi Neil,

Thanks for the update. Regarding your questions in red on the attached sheet:
1) Projects with "the" in the beginning usually hit either way (ie with "the" or w/out "the"). That particular project, "The Fratellis" did not get setup properly for some reason. Other projects with "the" should be ok

2) Piolet has been flakey lately. Our developers are aware of the issues and we should get back to normal in the coming weeks.

Besides that, I don't have any other questions for you to ask UMI in the meanwhile. Thanks

-TH

-----Original Message-----
From: Neil Saxby
Sent: Friday, June 15, 2007 6:44 AM
To: Tabish Hasan
Cc: Ben Grodsky; Jay Mairs
Subject: RE: UMGI testing external testing with Advestigo

Hi all,

Just had a quick call with Mark Denby at UMGI about these results. I explained that our internal testing is showing a max of 2 failures and no way represents what Advestigo have reported. We had a little chat about IP block lists and I asked him if he could find out if they were using one and if we could get a copy of it. I pointed out that we are generally affective even when these lists are used, but it would be good to see what Advestigo are using in case it hadn't yet been picked up on our radar. I also asked if we could get a copy of the .met list they were using with eMule.

He should hopefully get back to me within the next few days. In the mean time, please let me known If you've got any other questions/other reasons relating to why the Advestigo results were so dire. I've attached a doc summarising all discussion/suggestions so far.

Hope that's all good,
Cheers,
Neil

-----Original Message-----
From: Tabish Hasan
Sent: Saturday, June 09, 2007 1:29 AM
To: Neil Saxby
Cc: Ben Grodsky; Jay Mairs
Subject: RE: UMGI testing external testing with Advestigo

Hi Neil,

Thanks for the info. Some servers are harder to protect on than others, and it seems like they picked some of our more challenging servers. I took a look at your server.met. Seems like you have a pretty fresh install of eMule, correct? We'd be more effective on your particular machine after you've ran it for a while, and jumped onto a few different servers.

Our main issue is their Ares and Gnutella results. I don't have many questions to ask them, except possibly to find out if they're using any kind of IP block lists? Maybe you can casually ask them?

-TH

(Jay/Grodsky CCed)

-----Original Message-----
From: Neil Saxby
Sent: Friday, June 08, 2007 11:46 AM
To: Tabish Hasan
Subject: RE: UMGI testing external testing with Advestigo

Hey Tab,

I don't know how helpful this is but I did a little test myself in the interest of testing overseas. There are several variables that could have made this test inaccurate or unfair but nonetheless I thought it would be worth sharing the results with you:



 

Network: ed2k and KAD
Client: eMule v0.48a
Testing time GMT: 1.30pm - 3.30pm
Search string: Take That Patience
Search type: Global
Results: 18 pirated, 2 corrupt, 1 wrong audio, 24 failed to DL

I've also attached a copy of the server.met I'm using which came with the client download.

Connection details are as follows:

eD2K Network
Status: Connected
IP:Port: 86.144.79.251:4661
ID: 4216295510
        High ID

eD2K Server
Name: DonkeyServer No5
Description: www.First-Load.de / dual opteron 16gb / germany
IP:Port: 62.241.53.4:4242
Version: 17.14
Users: 239,626
Files: 36,208,850
Connection: Normal

Kad Network
Status: Open
IP:Port: 86.144.79.251:4671
ID: 1452298235
Buddy: None

Cheers,
Neil

-----Original Message-----
From: Tabish Hasan
Sent: Friday, June 08, 2007 2:40 AM
To: Randy Saaf; qa
Subject: RE: UMGI testing external testing with Advestigo

From what I read this morning...sounds like they're a European based Audible Magic type company. They have digital audio fingerprinting .

-----Original Message-----
From: Randy Saaf
Sent: Thursday, June 07, 2007 6:20 PM
To: Tabish Hasan; qa
Subject: Re: UMGI testing external testing with Advestigo

Who is advestigo?

 

----- Original Message -----
From: Tabish Hasan
To: qa
Sent: Thu Jun 07 17:59:06 2007
Subject: RE: UMGI testing external testing with Advestigo

Even our internal tests today did not match what they reported. We re-did 2 tests each of James Morrison and Take That and pretty much all of them passed.

The biggest discrepancy is the Ares results between us and them. I don’t really see how we failed as badly as they reported. We tested using the same version and everything.

2 possible explanations:

Either the version of Ares they reported using is inaccurate. They’re actually using 2.09+.

Or they’re using something like Peer guardian.

Neil, I’ll try to come up with some questions for you to ask them tomorrow.

-TH

(UMGI results from today are attached).

________________________________

From: Tabish Hasan
Sent: Thursday, June 07, 2007 11:13 AM
To: qa
Subject: RE: UMGI testing external testing with Advestigo

I’m attaching our internal results for these same tracks. They don’t look anywhere close to what they reported. We also did some quick searches using their version of Ares, and looks like we still hit.

I will also address each of the projects one by one below:

James Morrison - you give me something:

* Their reported search string on Limewire is not accurate. The full search string “james morrison you give me something” can not fit. They must have used another string.

* On 5/22, we found a problem with the track keywords and got it fixed. Their tests were on 5/17 & 5/23

The Fratellis - Chelsea Dagger:

* Didn't get marked as a single until 5/22. Their tests were on 5/17 & 5/23.
* No longer active so we can’t retest.
* Besides being a non-single, we most likely failed because they used the search string “fratellis” w/out the “the” and our PM was setup with the search string “the fratellis”

Snow Patrol - Chasing Cars:

· Got terminated on 6/5, so we can't test any longer

Take That – Patience

· We passed most of our internal tests (except for Piolet). We

We’re going to run full DL tests on the projects that are still active, using the versions and search strings they used…and send another report out.

-TH

________________________________

From: Ben Grodsky
Sent: Thursday, June 07, 2007 9:30 AM
To: qa
Subject: FW: UMGI testing external testing with Advestigo

QA,

Neil has gotten us UMGI's testing results. He's trying to get more info from them now to nail down their testing methodology. Please run tests on the same titles UMGI tested and see if there's anything that may be wrong with how the projects are setup. The results from them are atrocious.

Thanks,

Ben

________________________________

From: Neil Saxby
Sent: Thu 07-Jun-07 08:42
To: Octavio Herrera; Randy Saaf; Ben Grodsky
Subject: UMGI testing external testing with Advestigo

 

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.472 / Virus Database: 269.8.11/837 - Release Date: 6/6/2007 2:03 PM

 

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.472 / Virus Database: 269.8.11/838 - Release Date: 6/7/2007 2:21 PM
  

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.472 / Virus Database: 269.8.11/838 - Release Date: 6/7/2007 2:21 PM
  
  

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.472 / Virus Database: 269.8.11/838 - Release Date: 6/7/2007 2:21 PM
  

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.472 / Virus Database: 269.8.16/849 - Release Date: 6/14/2007 12:44 PM
  
  

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.472 / Virus Database: 269.8.16/849 - Release Date: 6/14/2007 12:44 PM
  

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.472 / Virus Database: 269.8.16/849 - Release Date: 6/14/2007 12:44 PM
  

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.472 / Virus Database: 269.9.1/857 - Release Date: 6/20/2007 2:18 PM
  

 

No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.5.472 / Virus Database: 269.9.1/857 - Release Date: 6/20/2007 2:18 PM

 

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.472 / Virus Database: 269.9.1/857 - Release Date: 6/20/2007 2:18 PM

 

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.476 / Virus Database: 269.10.4/897 - Release Date: 7/11/2007 9:57 PM

 

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.476 / Virus Database: 269.10.6/902 - Release Date: 7/15/2007 2:21 PM

 

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.476 / Virus Database: 269.10.6/902 - Release Date: 7/15/2007 2:21 PM


No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.476 / Virus Database: 269.10.8/904 - Release Date: 7/16/2007 5:42 PM
 

No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.5.476 / Virus Database: 269.10.8/904 - Release Date: 7/16/2007 5:42 PM
 

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.476 / Virus Database: 269.10.8/904 - Release Date: 7/16/2007 5:42 PM
 

No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.5.476 / Virus Database: 269.10.8/904 - Release Date: 7/16/2007 5:42 PM
 

No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.476 / Virus Database: 269.10.8/904 - Release Date: 7/16/2007 5:42 PM
 
Received on Fri Sep 14 2007 - 10:55:54 BST

This archive was generated by hypermail 2.2.0 : Sun Sep 16 2007 - 22:19:46 BST