An ambitious project has gone underway to improve the DCC (Direct Client to Client) protocol on IRC. DCC, as many will know, is used for transferring files and also to establish 'DCC chats' - a type of query message that does not go through the server. The 'new' protocol is to be called
DCC2.
Several authors of popular IRC applications/clients have taken interest in this project and signed themselves up as members. The full member list can be found
here. The people who started the project are the authors of IRC client
dIRC. Other clients which are members include popular
IRCle (Macintosh) and
Klient (Windows), aswell as developers of WinBot, BersIRC, ChatZilla and Visual IRC.
I emailed several of the developers of these clients and asked them 3 questions about the project. The following is the 3 questions I asked, followed by the answers given:
Questions 1. What made you join the DCC2 project?
2. Are you at all concerned about the file sharing advancement in the DCC2 project?
3. Do you think this will be a successful project and is it the future for the DCC protocol?
Answers Tom McAlee, author of Windows IRC Client 'Klient':
1. "I received an e-mail from Dan Smith a few months ago. It seemed interesting, as the current protocol can certainly be improved. It was also nice to see a coordinated effort. Many efforts to improve protocols are done in open forums but there are usually only one or two people involved, and they usually aren't people that can have much of an impact on the eventual implementation. The DCC2 project still needs some of the major players to be onboard, but it is the best effort I've seen thus far."
2. "Yes. Klient's community is comprised of people who primarily use IRC for chatting. Thats who I write it for and its who I want to write for. Occasionally there is a need to send a file, and I'm all for improvements to that feature. For example, eliminating the acknowledgement that is not necessary in a TCP connection is a good idea. Or, the ability to send multiple files sequentially might be nice. But, I view these as welcome improvements for sending files in an application that is primarily concerned with chatting on IRC; I don't advocate using such features to turn clients/scripts into better file sharing programs."
3. "I think its too early to tell. Like it or not, the fact is that it can't be successful unless it is adpoted by mIRC and bitchx. I'm sure Khaled was approached to join the group, but he has not done so. That could either mean that he just isn't interested in it, or that he would rather be pragmatic. So, I guess we should hope that he's just being pragmatic and if it is successful elsewhere he will consider it more seriously."
Nicholas Copeland, owner/developer of IRC client 'BersIRC':
1. "The DCC2 Project founders sent me an email shortly after I purchased Bersirc from Jamie Frater. Later, Theo Julienne and I pledged Bersirc support for DCC2 but have yet to implement it as we are waiting for a final draft."
2. "I am not. Is anyone concerned with file sharing over AIM? I've read a lot of the comments on Slashdot's DCC2 post. It seems to me there is a lot of speculation about IRC becoming the next major "file pool" like Kazaa or Napster or Gnutella that users will abuse with piracy. IRC users already abuse file sharing like that. Fixing a broken protocol isn't going to make the abuse go away or continue en masse.
DCC2 is just going to standardize solutions to technical issues caused by modern networking."
3. "Yes, I do. It can only be successful if a majority of IRC clients support DCC2. So far with eleven IRC projects' leaders committed to DCC2, I think the project will be successful."
Craig Edwards, developer of IRC bot, 'WinBot':
1. "I am the developer of a popular Windows IRC bot, and very much into IRC development, therefore the planning of any new protocol is interesting to me. Also, a few years back, I showed great interest in making my client (WinBot) support Trillian's 'secure dcc' system, however the system was severely lacking in documentation and at the time the developers were too busy to provide me with enough details to do this. As you may be aware, when you start a communication with a bot, this communication should be secured. I am hoping that by participating in the DCC2 project I'll be able to give some input which will allow for secure dcc chats which may be used in my (and others) software to produce an open standard for secured dcc chats. The fact that it provides secure (and possibly more reliable) dcc file sends is a separate bonus to me."
2. "I am concerned about the fact that filenames are not sent out onto IRC, making filtering difficult. However, there are privacy issues here and some would say it's none of the IRC admins business what you send over IRC, as the file transfer itself doesnt go over the IRC server. I'm not very much into the "warez and movies" scene on IRC, but I still use DCC a lot, sometimes to send very sensitive files, which I would love to be able to secure. Also, anyone who has ever been bombed with 10 DCC sends from a friend who wishes to send you 10 files at once will appreciate the ability to send 10 files in one request, but the main point to this is... IF it's done properly. By participating, I hope to be able to provide feedback that will ensure it is done in a way that benefits us all."
3. "At this point its only in planning and i really can't say. If the specification is finalized, i most certainly will add support to WinBot as a plugin, however its uptake depends solely on wether others add support too. We could find, a year down the line, that nobody decides to support it, but if that happens, at least we tried, and the next attempt will have better ground to start on due to the efforts of people before."
As you can see, a range of answers, but with the same ideals in mind.
A notable exclusion from the members list is the participation of the most popular IRC client - mIRC. Khaled Mardam-Bey, author of mIRC, has not said what his stance is on the protocol. This is understandable due to the huge influence mIRC's participation would have, and of course, the still unclear future of this project. The current draft(s) of the DCC2 protocol can be found
here.
Thanks to those people who answered our questions, it was much appreciated. You may wish to keep checking this article as some replies are still being waited upon.
Comments
