p2psp team mailing list archive

[Cristóbal Medina López <cristobalmedinalopez@xxxxxxxxx>]

Hi there,

Ilshat is right. In IRC, the p2psp user is a bot which collects logs from
conversation. However, the logs are deleted each week.

BTW, I prefer email too :)

Regards!

El mié., 8 de abril de 2015 a las 17:46, Ilshat Shakirov (<
im.shakirov@xxxxxxxxx>) escribió:

> Yes, we can. But, a question, (I have used an IRC server one or two times
>> before): Do the "conversations" remain in the IRC server between sessions?
>> If the answer is "no", I would prefer the e-mail, if you don't mind.
>
> I think no, but we can attach a bot to the #p2psp channel, which will be
> collect logs from conversation. To be honest, I thought that user p2psp is
> a such bot =)
>
>> OK. Do you think that this could affect to the results?
>>
> I think no, because the number of poisoned chunks doesnt depend from time
> of execution of PoisonedChunksObserver.
> To be honest, I have no idea why the simulation shows such results at 50
> peers. Im trying to understand smth from log files, but it quite slow
> process =)
>
>> Please, note that this email has not been sent to the P2PSP e-mail list.
>>
> My mistake, I have restored CC.
>
> Thanks!
>
> 2015-04-08 20:32 GMT+05:00 Vicente Gonzalez <
> vicente.gonzalez.ruiz@xxxxxxxxx>:
>
>>
>>
>> On Tue, Apr 7, 2015 at 1:26 PM Ilshat Shakirov <im.shakirov@xxxxxxxxx>
>> wrote:
>>
>>> ps Also I think we can move to IRC?
>>>
>>
>> Yes, we can. But, a question, (I have used an IRC server one or two times
>> before): Do the "conversations" remain in the IRC server between sessions?
>> If the answer is "no", I would prefer the e-mail, if you don't mind.
>>
>
>>
>>>
>>> 2015-04-07 16:14 GMT+05:00 Ilshat Shakirov <im.shakirov@xxxxxxxxx>:
>>>
>>>> So, as far as I have understood, in your simulator, two (or more)
>>>>> attackers cannot be rejected in less than n ticks of simulation? In other
>>>>> words, how many attacker are excluded in n ticks?
>>>>
>>>>
>>
> I was wrong. Module for computing num of poisoned chunks is executed every *n
>>>> *ticks. Also, the process of streaming launches at *m* ticks. It can
>>>> be that we exclude malicious peer before *n* ticks, but we can't
>>>> exclude peers before *m* ticks. Attacker is excluded from team when he
>>>> sends poisoned chunk to trusted peer. It can be that we excluding 2 or more
>>>> attackers in gap between 2 executions of stats module.
>>>> Source of stats module is here :
>>>> https://github.com/ishakirov/sim/blob/master/src/PoisonedChunksObserver.java
>>>> Also, you can see the excluding attacker from team here:
>>>> https://github.com/ishakirov/sim/blob/master/src/Peer.java#L70 and
>>>> https://github.com/ishakirov/sim/blob/master/src/Source.java#L118
>>>>
>>>
>> OK. Do you think that this could affect to the results?
>>
>>
>>>
>>>>  does an attacker poison all the chunks?
>>>>>
>>>> yes, in simulation, attacker poison all the chunks he receives
>>>>
>>>>
>> OK.
>>
>> Please, note that this email has not been sent to the P2PSP e-mail list.
>>
>> Best,
>> Vi.
>>
>
>