Talk:DC++

NPOV concerns

The article's wording is too heavily in praise of the software.--24.114.252.183 14:00, 11 Feb 2005 (UTC)

Agreed. This sounds more like a plug for the software rather than a neutral description or comparison. 69.227.23.179 15:53, 17 Feb 2005 (UTC)

Have you ever considered the notion that maybe DC++ is all-around better than Direct Connect? 24.177.11.130 01:25, 27 Feb 2005 (UTC)

I have rewritten this article somewhat, do you think the NPOV dispute tag can be removed now? Solver 16:51, 5 Mar 2005 (UTC)

I don't think so, no. Your changes were definitely an improvement, but there is still a ways to go. As I see it, there are two major problems vis-à-vis NPOV: things given as facts that may or may not be factual, but that aren't verifiable either way (e.g., the assertion that most people connect to DC networks using DC++), and things given as facts that are out-and-out opinions (e.g., the characterization of various features of DC++ as advantages over NMDC). Ruakh 03:36, 11 Mar 2005 (UTC)

The assertion that most connect using DC++ is verifiable. When I connect to a DC hub, I can see what client is used by every other user. A quick look through the user list reveals that differnet versions of DC++ by far make up the majority. I couldn't find any statistics online to link to, though, but I wanted to include such a link in the article. Solver 22:15, 11 Mar 2005 (UTC)

You can verify that most users on a given hub use DC++, but I don't see how you could verify it Direct-Connect-wide. After all, this is a prime example of selection bias: since people find out about a given hub from other people (likely their friends), and since people find out about a given client from other people (likely their friends), I'd assume that people on the same hub are more likely to use the same client than people on different hubs. Ruakh 04:19, 12 Mar 2005 (UTC)

Your selection bias does not hold true in this case. As an hublist (dc.selwerd.nl/hublist[.xml].bz2) admin I have visited countless hubs while assessing the hublist pinger (a bot that scans the hubs for user count and share size) performance, and can confirm that the _vast_ majority uses DC++ or a derived version. As a hub owner, I have spoken to several other owners and seen the state of their hub. And as a Direct Connect veteran I can confirm that historically Neo Modus Direct Connect usage dropped dramatically around the introduction of DC++ version 0.181. Yes, it is possible to modify the pinger bot to get fairly accurate client usage statistics, but this is not at all necessary. Fetch a hublist, and go join a random couple of hubs. You _will_ find an overwhelming percentage of DC++ and derived clients. (The "friends" argument holds only partially true as well, the hub list system allows users that become a bit familiar with the software to join any random hub on the list, spreading the population drastically.) Sedulus 12:55, 2005 Mar 28 (UTC)

Regarding out-and-out opinions: "Many users feel that DC++'s interface is easier to use. Differences from NMDC include: (..three points..)" This can be scrapped yes. After a long period of silence from Jon Hess (NMDC) he did update his client. I have seen it once (I don't like to install software that has ads/adware), and his version 2 is able to connect to multiple hubs and I assume able to resume files from a different user. I can't even confirm that it still has ads/adware (the old version 1 most certainly did though). So, I'll remove that and will wait for others to do a comparison against current NMDC. Will this un-NPOV-dispute it? Sedulus 13:04, 2005 Mar 28 (UTC)

I've removed the NPOV tag, read through the article and I think it's neutral now. If anyone disputes this feel free to put the tag back, but state the reasons why, because I've addressed all the NPOV issues described here. Splintax 11:48, 2005 Apr 21 (UTC)

It looks much more neutral now, thanks. Now all it needs is a good copy-edit. :-) Ruakh 17:23, 21 Apr 2005 (UTC)

How is this link: Commercial Managed DC Hub Hosting Provider related to DC++? Removing it. Gustavb 19:49, 2005 Jun 26 (UTC)

From the linked faq at the end of the article:

"The Neo-Modus Direct Connect website was sold and is no longer is operational. While the NMDC client still works, you won't find anyone using it anymore. The DC++ client (and its mods) are now the de facto standard for the Direct Connect protocol."

http://www.dslreports.com/faq/6511

Be Bold In Edits (talk) 08:16, 24 May 2009 (UTC)

SababaDC

I removed the addition of SababaDC for two reasons. Firstly, the list of mods is becoming a bit long. It already takes up a hefty chunk of the page. Secondly, there were just too many things listed for it. It looked a bit ridiculous. In that vein, I'm tempted to chop out some of the bullets from StrongDC as well ... --TheParanoidOne 23:00, 1 February 2006 (UTC)

The guy who wrote SababaDC constantly spams his programs in forums, so I don't doubt for a minute that he has added all those "features" to his list. Also from the looks of how SabbaDC is worded I believe this is a form of self advertising. I removed most of the so called features... 24.174.96.139 (talk) 05:55, 3 February 2008 (UTC)

This is the guy advertising http://en.wikipedia.org/wiki/Special:Contributions/84.108.189.199

Hubs

Why are hubs advantageous? Why can't the entire userbase of dc++ be a single hub, so that one can locate just about any file? It would be so much stronger if everybody on the network where connected, no? —The preceding unsigned comment was added by 68.80.214.158 (talk) 14:59, August 30, 2006.

The design of the protocol does not lend itself to that. It could be argued that one of the Direct Connect's network advantages is being decentralized, allowing users to form their own communities based on whatever criteria they're comfortable with. Your point about the lack of a global search is one I've seen before, and it has merit from an end user point of view. --GargoyleMT 18:41, 12 September 2006 (UTC)

Try to search with at least 10'000 users, and you will be overwhelmed with results, if your search-string is popular enough. You'll see the needed file with very first results, but the line would be immediately scrolled out of the screen (by sorting) as more and more results arrive. You just won't be able to select the file and start downloading until you finish to receive all those answers, which could take very long. Just image what if web search engines listed every site in a couple of million findings on a single page — that would be disastrous. :) 217.172.21.161 (talk) 08:03, 23 May 2008 (UTC)

Supported hub address URLs

The NMDC part of the article is incorrect. DC++ support even more specific information than that. IIRC, file names and users are supported as well. Anyway, it's an easy check in the source. My point is, does the NMDC protocol support everything that DC++ does, or does NMDC just support dchub://hub:port (as this info would be nice to know in the NMDC article)? I don't have a copy of the NMDC client so I can't check with that (and most results on Google seem to be malware). --Ullner 13:47, 22 August 2007 (UTC)

Vandalism

I don't know if its just me not knowing terms but it seems there was vandalism added at the same time the information itself was added in the greylink section, there are multiple mentions of hubs having balls and drugs. --ryan_the_leach (talk) 15:45, 27 May 2009 (UTC)

It was translated with google translate from russian, since the features are all in russian, the client is 100% in english —Preceding unsigned comment added by 86.121.239.93 (talk) 22:22, 14 June 2009 (UTC)