Red Library - probably a daft question

[PenDragon]

Hi - Bear with me cos' I expect there is a simple answer to this, but at the moment I can't think of it....

In Red Library (V1.9 anyway) the minimum time between checks is 5 minutes.

The maximum links you can verify at a time is 10. ....

Soo ....

There are a maximum of 288 checks you can perform a day, validating 10 links at a time, therefore giving you 2880 links per day.

You are allowed a maximum of 3 machines per username.

Which gives maximum of 8640 links per day validated.

With me so-far...

So how the heck do some users manage to validate about 40,000 links a day... Is someone hacking the client.

To get into the top 20 to need to do more work than is possible ... ?

What is going on ... !

[russkris]

I'm not quite sure on this one.. Have you tried to install it agin and see what happens..

So you have it on 3 machines at the moment? I would be trying to install it on another machine..

Even Meady is getting 5k per day..


PS Congrats Meady for :stomp:ing nano

[Gunslinger]

The previous client (1.18) let you have more than 3 machines at once. At the moment they can still keep on running and getting work, but as soon as the process is stopped and restarted it will insist on an upgrade to 1.19 and then refuse to run more than 3 simultaneous copies.

:gunsling:

[TNT Weazy]

but there are people getting ALOT more than even 4-5 clients could get.

[Gunslinger]

but there are people getting ALOT more than even 4-5 clients could get.Yes - the previous client had no limit on numbers. If I had discovered this project a week or two earlier I could easily have deployed it on over 100 systems, which would have given me around a quarter of a million points per day. Of course that assumes that their server would have been able to cope - the reason they introduced the limit was to relieve the server. Note that from the project point of view they want the checking done from as many different locations as possible to remove systematic errors. Better to have 100 discrete users with 3 clients each than 3 users with 100 clients...

:gunsling:

[TOOCOOL]

Seems to me this creates an uneven playing field time to take this off the vault.
whats to stop someone with the install for the old vers just keep reinstalling it !

[Nanobot]

whats to stop someone with the install for the old vers just keep reinstalling it !When the client connects to the server to return work, and request more work, the server checks the client version number and if it is not the current version it tells the client to stop. So there is no way anyone could keep running with the old version.

If people want to use more than three clients, to boost their teams performance, there is nothing to stop them signing up with multiple accounts :D

[TNT Weazy]

Looks like they are actually going to fix this problem. Of course, the damage has already been done.

[KAMCOBILL]

Hi Y'all,

I've been watching this Project. The daily stats on Project Stats' Page are wrong. If you take the team totals from day to day and check the difference between them, no team has moved like 68K done that day by a User.:rolleyes:

Today seen 1 cap at 9k. Went to user page and it had 3k. :confused: Posted it in the forum to see what response will be. Here It Is (http://www.readyresponse.org/forums/viewtopic.php?t=1899)

It seems like a good project and I think they'll get it straighten out.

Happy Validating :D ,

Bill

[lonewolfz28]

When the client connects to the server to return work, and request more work, the server checks the client version number and if it is not the current version it tells the client to stop. So there is no way anyone could keep running with the old version.

If people want to use more than three clients, to boost their teams performance, there is nothing to stop them signing up with multiple accounts :D
There're already several teams doing this and not even bothering to be discreet about it.:rolleyes:
http://www.readyresponse.org/index.php?item=distributed&opt=12&tid=6
http://www.readyresponse.org/index.php?item=distributed&opt=12&tid=7
http://www.readyresponse.org/index.php?item=distributed&opt=12&tid=14

How long does a project have to be exploited before it's pulled from a vault?:confused: Or, are all the other teams going to be forced to cheat as well in order to make up ground past the #5 position? All of which goes against the projects mission to begin with.

[TOOCOOL]

My vote is kick this outa the vault its been unreliable, its of little use, and open to cheating its just not vault worthy:(

[russkris]

How could having multiple account for the one team on the one account be cheating..

Thats fare in my book.

Would it be the same if someone had several accounts in SETi? What about Borging other peoples computers? Is that not cheating...

[TOOCOOL]

We covered the main reason behind the cap in this thread: http://www.readyresponse.org/forums/viewtopic.php?t=1899

Basically there is no need for someone to be running 20 clients on one internet connection...because they would end up doing unnecessary work which would not be useful to the project.

We will obviously be aiming to detect people who sign up for multiple accounts in order to try and run more clients. We are well-aware that there always a way around the restrictions - and we hope that users respect the project and the reasons for the restrictions being in place.

Soon the client limits will be removed...but until then we will enforce it in every way we can Smile

Thanks.

Wang

[russkris]

Thanks TOOCOOL.

We look forward to 2.0....

[Toutouf]

Seen @ http://s15.invisionfree.com/Crunchers_Inc/index.php?showtopic=122&st=125 :
If you make a copy of the existing client and just rename it by putting a 1, 2, 3 etc... the user name remains the same but there is a different client running. There is also only 1 Links.xml file in the folder that holds ALL the links info for ALL the clients running within that folder... Looks like it tags the number of Link.xml folders with the same username... so far so good... upped to 6 clients on 1 machine & 4 clients on 2 machines under cswchan & 1 machine under cswchan1 is running 8 clients...
Hmmmm. you consider 20 "just a few more"?...

As you know this proj limited to 3 clients for same user because what is interesting is having clients on different lan.

Some (I hope they're not all agreeing that kind of doing) Crunchers_Inc members : Stupid or babies ? :(

Conclusion : I agree on what Toocool wrote :
My vote is kick this outa the vault its been unreliable, its of little use, and open to cheating its just not vault worthy

[russkris]

This is something you should take up with Red Libary...

It's the same with Seti Classic, there was cheating going on there to, and now that it past, now the 'cheaters' have had the credits taken from them.

[TOOCOOL]

If the DC-Vault knowingly keeps stats for a project that is being abused by cheaters even if it is by a minority, it devalues the point of the vault and tarnishes the reputation of the Nija's. It should be policy to remove any project that is being abused from the vault until the project can get their shit together.

[Nanobot]

According to this post (http://www.readyresponse.org/forums/viewtopic.php?t=1899) production is capped at the server at 9000 so anyone generating more is just wasting CPU cycles.

[Ragnarog]

If the DC-Vault knowingly keeps stats for a project that is being abused by cheaters even if it is by a minority, it devalues the point of the vault and tarnishes the reputation of the Nija's. It should be policy to remove any project that is being abused from the vault until the project can get their shit together.

Um... we never forced you to use the DC-Vault ;)

We are not responsible for unreliable clients :)

It requires work and time to add/remove projects even if you may think it all miraculously fixes itself over night :D

So we remove... 90% of the vault projects coz there are probably known exploits for all of them? :cunning:

If the Vault did anything in this case, it was to motivate people to find ways around the limitation. That is now public knowledge and can be fixed by the project admins asap. Its in their hands to fix it.

I agree that we could set a timetable, say give them 1 week to straighten it out and see what happens. If its fixed and we notice the game field is leveled, then it stays in the vault. Otherwise its removed on a temporary basis until it is fixed.

How does that sound?

Rag

[NeoGen]

D2OL and TSC have been straightening their act together lately too, which has been causing alot of rising and dropping of teams on the Vault.
According to what they say, now on phase II they say that corrected several problems and are much more cheater-proof.

I didn't know there was cheating on those two before their admins went public and admitted it, but I'm sure plenty of people knew and no one complained here about it that I remember...

There are no cheat-proof projects, and I bet there's people at all projects finding ways around to get more points. The only thing we can do is spot them and alert the admins so they can correct the situation.

About keeping them or not on the Vault... well, that's a decision for the Vault Masters to take. Whatever their decision is, I just ask that it be a general rule applied to all projects.

[russkris]

As i said in post 16..

This needs to be taken up with the project itself..

I know where your comming from. Having 4 account to the 1 team might be cheating but then again is it? Maybe it just plan old good sence... We'll never know.

The Vault was created to record Team Stats.

Red Libary has stated that running any more then 3 client from the same location is of no benifit to them, so why havent they fixed that problem?

[TOOCOOL]

Um... we never forced you to use the DC-Vault ;)

We are not responsible for unreliable clients :)

It requires work and time to add/remove projects even if you may think it all miraculously fixes itself over night :D

So we remove... 90% of the vault projects coz there are probably known exploits for all of them? :cunning:

If the Vault did anything in this case, it was to motivate people to find ways around the limitation. That is now public knowledge and can be fixed by the project admins asap. Its in their hands to fix it.

I agree that we could set a timetable, say give them 1 week to straighten it out and see what happens. If its fixed and we notice the game field is leveled, then it stays in the vault. Otherwise its removed on a temporary basis until it is fixed.

How does that sound?

Rag
We were'nt forced to use the vault but you did invite us

the timetable part sounds good :)

[russkris]

OK this is going nowhere fast..

This is a problems that isn't going away.

Whatever the project, there will be cheating or attempts to get around limitation, we cant help that nor can we fix that.

I feel it is up to the projects administration. If a problem or problems are found within a project then that problem(s) should really be taken up with the project admin not the DC-Vault, we reflect the stats output by the Project.

This problem is not going to go away and yes I agree having account like, cruncher1 and cruncher2 is probably not cheating in it's highest form, so until the problems is resolved at the highest level, i.e: The Project Admin, then the Vault will continue to record the stats output.
I also think that the project or projects that have 'cheaters' should not be removed from the DC-Vault, it would't be fare to all the rest of the Teams that are doing the right thing by only running the 3 clients..

<This subject is closed for conversation, until the DC-Vault admin and owner can discuss what is to happen and will be re-opened after.>

[Ragnarog]

Seems like its reopened again so i'll just add my *agree* to rusty's comments :)

Rag

[N.V.M.]

if i may add some insight here. i started a thread, HERE (http://www.readyresponse.org/forums/viewtopic.php?t=1900) (over at the project forums) a while ago and the admin guy over there,Wang, said this to me:

Soon the client limits will be removed...but until then we will enforce it in every way we can

so, i guess this whole thread will be moot soon enough anyways.

[vaio]

Smap

DLV cheater (retired) checking in........those who have the bandwidth should just run Majestic 12 instead.:p

Anyway, DLV repeatedly say they are gonna remove the client limit.......won't that devalue the thing even more???

Plus, if added clients is of no benefit, why remove the limitation?

Bleugh

[russkris]

Thank for the Link NVM

[JustCuz]

Seems like its reopened again so i'll just add my *agree* to rusty's comments :)

Rag


I agree too, but I'd also like to suggest that you consider a "no points cap" criteria when adding projects to the Vault in the future. I know the points cap was a recent addition at DLV and they say it will be removed soon, so just a suggestion when adding future projects.

If DLV doesn't remove the points cap, I think the project should be removed from the Vault.

[Gunslinger]

I don't see anything in the DC Vault rules that says a project should be excluded if it limits the total number of clients per user or if it has a cap on what a user can score in a day.

The 'science bit' of this particular project is best served by having as many discrete client locations as possible and not duplicating effort from any given end point, and so I can fully understand why the researchers want to discourage what is (at best) redundant background noise.
I can also understand the three client limit (eliminates some spurious bad link errors without weighting the entire project due to local connectivity issues).

Since the stats points are effectively a reward for participation, surely they are entitled to say they won't reward something that is of no value to them?

If the DLV administration comes back and says that the project had a points cap based on IP address (or even IP address block) rather than user name, would you still argue that the project should be excluded?

:cowboy:

[JustCuz]

I don't see anything in the DC Vault rules that says a project should be excluded if it limits the total number of clients per user or if it has a cap on what a user can score in a day.

me neither



Since the stats points are effectively a reward for participation, surely they are entitled to say they won't reward something that is of no value to them?

Yes, as far as I can tell, they own the project so they can make rules that best suit their needs.


If the DLV administration comes back and says that the project had a points cap based on IP address (or even IP address block) rather than user name, would you still argue that the project should be excluded?

:cowboy:

Yes I would. The points posted on their stats page should be reward enough for your participation. I don't think a capped project should be included in the Vault.

[Angus]

Isn't this project long dead? Why is it still included?

[russkris]

I have posted on the RR forum, in hope I will get an answer to what the heck is going on...


PS: :welcome: to the Forum

[Angus]

http://www.readyresponse.org/forums/viewtopic.php?t=1847

and all mention of Distributed Computing is gone from their site. The link at the very top of site is blank.

[umccullough]

there is this news post:

http://www.readyresponse.org/index.php?item=article&id=84

But otherwise, it definitely appears that the 1.0 project is done. That definitely suggests that it be eligible for removal I think...

[russkris]

I think it should and will remove it soon.. I am still a bit mad that they just canned the project and never sent an email or anything....

[russkris]

So, If anyone is following the DLV thread I started over at RR.

It looks like 2.0 might not be that far away..

I am just waiting for a post back from WANG or anyone for that matter to when 2.0 will be released..

Then the Vault admin will have a PM war and make a choice to keep it there or remove it