Hi guys..
Heard of winmx.. file sharing util.. mp3..ne stuff ? .
I am intersted in writing such an app for linux. Only after march 2003, coz my final exams are coming up ? wanna try it out ????

have you tried running winmx with wine?

sounds cool,
have you done much research in the way of P2P or would this be your first stab at it? I would be interested helping out, but It would take some time for me to get up to speed as I am not yet familiar w. the different P2P protocals. However I am interested in learing and contributing because I think it is a cool idea as you never can have enough open source filesharing software out there.

cludwin

no, i'm not saying I want to do a P2P, I'm suggesting that you should try running winmx with a program called wine(the windows emulator) under linux, before trying to do anything, it may already work under linux?


I don't think winmx protocol is documented, unless you only want to connect to the non-winmx stuff which means you want to do a napster clone.

Hi there
I am in fact trying to write a clone fo napster. i havent done great research in P2P.. but i believe we can implement our own custom protocol for it . The design part is just starting.. the actual coding will start only in march (mid) i can send u the specs i have done till now .. just the scratch specs. I want to learn new things thats the whole purpose of it

I would be interested. ....just started network programming though...

I just want to learn new things so I'm going to go and rob a bank.

LOL

Cheers,
Tink :)

you would be better doing a clone of Kazaalite.

And trying to use both the gnutella networks nad also the type of protocols that Kazaa uses to widen the search engins.

Also with the GTK- and Qtella, and winmx, you tend to find that hte search engines arent as powerful as thoughts of the Kazaalite.

so you might want to look into that as well.

I cant actually program, but I was talking to a freind about this last night, because it is something that I want to learn.

I am learning how to you Qt Desinger so i could do your GUI for you, but that would be about it!

Interesting.
What language are you planning to use?

I dont actually program as such.

But he Qt Designer, is a Gu based c complier.

So I am at the very least going to create the UI for the program, and then if anyone else would like to get involved to get it working then they are more than welcome.

jsut drop me a mail - kenster@blueyonder.co.uk

I'm totally for it! That would be very cool. Although there will probably be a scarce amount of linux files on it ...

Qt Designer is actually just a UI designer for the Qt toolkit. You'll still have to write a bunch of code outside of Designer, but it is quite easy to connect the signals/slots and implement the slot code with Designer. Or even without. Doesn't matter.

You do have to know some C++ to use Qt Designer, but you don't need to be a guru.

Qt has a bunch of networking classes, so you may want to read the Qt docs before you go mad writing networking code if you decide to use Qt.

doc.trolltech.com

This is just my opinion, but it might be a lot better to go with an established protocol than to create your own. With a new protocol, you are essentially starting with a user base of 0, and since so many people are already using software built on established protocols, you might have a hard time leading them away to your own system. If you don't like Gnutella (who does?), consider OpenNap.

Of course, if you are just looking for a learning experience and don't really care whether a lot of people ever use your program, then it might be more instructive to design your own protocol. It all depends on what you really want.

Going along with Dave, you will have problems with a user base, if trying to write your own protocol. Check out giFT and OpenFT - you'll have the protocol, so you could write a giFT replacement to that (why would you?) or you could write a giFTcurs replacement.

Also, if you rea up on the giFT/openFT project - Kazaa (more exactly, the FastTrack protocol) keeps changing its code - exactly so projects like giFT can't use it (that's why they stopped supporting it).