Talk:PeerGuardian

Known Issues

Updating with information regarding BF2 and HL:Source games not working with PeerGuardian 2. This is a KNOWN ISSUE. See proof in bug list here: http://sourceforge.net/tracker/?group_id=131687&atid=721926

This is neither a bug nor an issue. The servers of games companies are blanket blocked to avoid issues related to DRM and numerous anti-piracy measures from these companies. While I'm not impartial here, from npov this is an invalid criticism since this is what PG is DESIGNED to do. I comment on this issue in the section about the Lists in criticism.

I say: "Occasionally IP addresses of "friendly" organisations and groups may be added to the blocklist. This is sometimes a case of over-zealous or incorrect moderation at blocklist.org, but can also be for other reasons. For example, a lot of ranges owned by Valve Software are blocked, meaning that some users must allow these IP addresses in order to connect to game servers. The reasons behind Valve's addition to the blocklist follows a number of occasions where the organisation posted fake BitTorrent downloads."

to back this up see the BBC article at http://news.bbc.co.uk/1/hi/technology/4041289.stm

It is not valid to make criticisms without full understanding. As I also say in the article, any IP address can be allowed by right clicking on it and choosing to allow the IP. This means when you connect to counter-strike, etc you can just switch to pg, right click on the Valve IPs and choose to allow them. It'll never do it again.

This application is not simplistic, it's built from several elements and has often been hyped by non-savvy media orgs - there are misconceptions about the application, this being one of them. In the persuit of fact I have reverted your entry, and I think have justified myself from the evidence above. If you wish to hear exactly the same argument again you can post on the forums about it.

I'm trying incredibly hard to keep this article accurate and comprehensive, and I am actually setting out the largest section of the article to the analysis of criticism of it. That is neutral (I think), while this is, while an issue people discover, not an actual problem with the application. In short: it's meant to do that.

--JFM 14:00, 8 September 2005 (UTC)[reply]

You're involvement in the project itself is skewing your positioning of this article. The fact that this program breaks Battlefield 2 and Steam VAC connectivity is a huge concern considering its audience, doubly so because it gives no indication to its gaming audience that it will purposefully break such applications. The desire to play legitimate games far exceeds that of software piracy. The developers themselves have been reluctant to comment on the STEAM and BF2 issues and the fact that this software breaks them.

Unfortunately it is impossible to distinguish between the different parts of an orgs webservers. Have you read the manual of the program? It makes these issues very clear, and I have already stated that a simple choice to "allow" an IP address permenantly will make the application work correctly. This is similar to the operation of most firewalls (where the user is prompted to allow an application to connect to the internet). I have already stated my involvement with this project, so that there can be no doubt or suspicion about this - I am being open about that and thus far I have refrained from writing a longer article about this application because I have feared that it would be considered biased. It does not break the applications; it only prevents them from connecting to IP addresses in the blocklist, the same thing will happen to any other application where the IP is in the blocklist. The reason for an IP being in the blocklist is usually clearly explained (unfortunatley after the dispute over control our blocklist information page is down temporarily). To call this a bug or an issue shows a misunderstanding about the purpose and operation about the program. --JFM 12:36, 26 October 2005 (UTC)[reply]

Well, thanks for all the info, JFM. It would be great if every software developer could be as clear as you are. —The preceding unsigned comment was added by 201.19.201.199 (talk) 16:56, 30 April 2007 (UTC).[reply]

History

When did this become open sourced, I had thought that it was closed.

As well we can add more information about PeerGuardian 2's new features and enchancements if anyone knows (I do not currently use this software I am here because I noticed it was on the 10 ten downloads on sourceforge and had thought it was closed so i am investigating). --ShaunMacPherson 07:26, 23 July 2005 (UTC)[reply]

- The software was open sourced right before the closure of PeerGuardian.net about 2 years ago. The software had always been free and closed, we reckoned open source was the best way forward. It was licensed GPL originally, but the new version 2 has been zLib from day one.

I don't know the exact date, and I don't want to fill in too much here because anything I write would certainly not be a neutral point of view.

(Joseph Farthing, Methlabs.org)

--JFM 21:45, 31 July 2005 (UTC)[reply]

Very interesting and I am glad that the people of Methlabs choose to make it free / open source. For the longest time I had thought it was closed for some reason so of course I avoided it (since I hardly ever trust closed source applications for security or privacy applications). Thanks, --ShaunMacPherson 06:30, 15 August 2005 (UTC)[reply]

Well a lot of people thought that a program that was so hyped in the media (a lot of what they said was wrong - people STILL believe that PG is meant to hide your IP address somehow, as shown by some of the edits made to this page) could only be a scam. When people find out that we're not charging money or claiming "magic solutions" I think the tone changes a bit. It would certainly be impossible (or very morally questionable) to make this sort of app shareware! I've kind of gone back on what I said before, I've made quite a few additions to the article, trying to keep this as neutral as possible. The problem is that there are a lot of misconceptions about the program, so I'm trying to edit to keep it truthful, etc. IE; I'm considering criticisms and problems with the application, but I am not going to talk about ones which are invalid, such as the "PG does not hide your IP address" argument, since the application is not designed to do this. Hope everyone is OK about me working on this article, there are a lot of other things I'd like to write about, but I'm determined to stay as neutral as possible. Thanks for the comments anyway.

--JFM 16:37, 31 August 2005 (UTC)[reply]

Binary search speed

I removed the suggestions that a binary search was a fast way to check each IP address the block list, since it's not close to being accurate. A few faster options include:

hash-based quick-lookup table, say 32k entries with a single bit per entry to indicate if an IP address a possible candidate based on the matching "bad" ranges in the list. Since few will be blocked, a single fixed-time lookup eliminates most scans. 32k is probably too large since it takes 4k to store and it's desirable to cache in L1 or L2.
bin-based short-circuiting of the first few binary search entries. Say 256 bins with the highest byte or the IP addres if the data is in that order.
hash-based recent IP list with quick yes/no result checking, similar to the quick lookup table but with the actual result for IPs which have been checked. This exploits the comparatively small number of IPs any end user system is likely to encounter.

Jamesday 07:17, 13 October 2005 (UTC)[reply]

That's an interesting idea. I read that Bloom filters were used for early spell checkers (in a C book whose name I have forgotten). IP addresses are easy to hash too - since the ranges are generally linear - i.e. 111.111.012.* to 111.111.013.*. If you only try and get class C plus maybe 1 bit - that would give you 48 million address buckets - say 1% are bad - you'd only need a 600 k bloom filter to get a 1% false positive rate. And an 2400 k filter would give you a 0.001% false positive rate. Megapixie 01:03, 14 April 2006 (UTC)[reply]

Restructuring needed

Seems like the bulk of the "Criticisms" section talks a lot about the basic functionality of the software -- rather than addressing this as a criticism, the information should simply be added to the main body of the article to explain what it does (using a flat, neutral tone). If there is room to criticize what the application does, then that should go in the Criticism section. Additionally, if someone can add details about what information can/can't be retrieved by a peer, with and without using PeerGuardian (so a comparison chart of sorts), that would be very helpful in explaining the function of this program and what benefit it provides to privacy/security. VanishingUser 04:36, 23 January 2007 (UTC)[reply]

Beta 6c is where?

Three edits ago, Leeisl changed the latest version number to 2.0 Beta 6c, instead of 6b. I wonder if this is really an official release, because I can't find it on either Phoenix Labs or their project page on SourceForge... --Unsound 08:35, 9 February 2007 (UTC)[reply]

Beta 6c hasn't officially been released, Phrosty of the Phoenix Labs forums has it on his site: [1] Strider01 14:47, 13 February 2007 (UTC)[reply]

Whoops, looks like Phrosty removed it from his site. Wait for an official release. Strider01 15:53, 30 March 2007 (UTC)[reply]

As it seems it's not an official version, and noone knows where to obtain it, I will change the text in the article back to 6b... --Unsound 12:10, 8 April 2007 (UTC)[reply]

Privacy

I added a section on privacy since the article does not mention the issue. Users of PeerGuardian should be aware the software does not offer any guarantee to users that they will operate anonymously.Goodpaster 15:20, 14 August 2007 (UTC)[reply]

The thing is, it isn't intended for anonymity as far as I can see, it is designed to try to keep instances like the RIAA out of the door. Of course it can not guarantee anything, nothing is to stop above and other organisations from using different IPs, but on the other hand the blocklists keep getting updated. This section should be expanded. I'll have a closer look at the software and see if I can do something with this. --Wouter de Groot 11:41, 15 August 2007 (UTC)[reply]

Here's my question: can anything that is intended to be included in this section be cited? Do the creators of PeerGuardian clearly make the claims that it's being suggested they're not living up to? Just because people have made false expectations of a piece of software does not mean the software is at fault, unless the creators of the software clearly claim such expectations as reasonable. Smeggysmeg 05:38, 22 August 2007 (UTC)[reply]

Image

The image description claims that it shows PeerGuardian 2 running in Vista but the PeerGuardian website states that there is currently no Vista support. Which is correct? Is it a beta? What? 98.223.170.167 (talk) 06:09, 10 April 2008 (UTC)[reply]