I have a very different sight of the_GDF and their role.

Basically it's one valuable source of knowledge beyond others. I take the their proposals as proposals, which means I try to understand and improve what could be done better fom my programming knowledge and point of view. If ppl don't like it, fine, but discussing should be still allowed here. Some months back (before Mike Green cleaned up the chaos), the GDF was very chaotic and sometimes discussing things dead until they didn't move anymore. We gained no improvement in network topology or for the endusers for months, other P2P systems where much more innovative and gained more users every week. Currently the GDF is more attractive, I wonder positive, great for Gnutella! But I still see LW and BS implementing things, make it a "defacto standard" and come with a "eat or die" mentalaty. I still see the GDF hiding with a lack of documentation for new coders who would like to contribute or to program a client that follows the best available Gnutella technology. With this high society behaviour, it is easy to say... oh you are bad for "the network" b/c you don't follow our rules. About votes from important developers -which developers- we have only 4-5 clients in active development. So I don't see a large active balanced developers community, compared to other open source places. The invisble pushing force is LW and BS = commercial interests (Note: For exmaple I see the name "ultrapeer" as marketing, but well some say it's because FT's supernode sounds so close... do we name hashs ultrahashs or smarthashs soon?). Finding an end, the GDF was thought for a forum of coordination between developers: but I think promotion of new concepts and new coders is behind implementation question inside the GDF. Implementation isn't everything, the GDF will not take you serious until you coded an(other) client?

I feel uncomfortable with your description of wish-they-were developers forum here. I and others spend a lot free time in writing articles and helping new Gnutella programmers + bringing new ideas and concepts to Gnutella. I don't want to turn Gnutella into FastTrack, eDonkey... but I get inspired from all, not only from the GDF.

Please dont take it personally. I really knew that I shouldnt have put the comment in there but it was really more of a swing at myself than anyone. I dont know any programing and can only contribute to possible theories on how to improve gnutella.

I was not aware that anyone here actaully had a client (besides cultiv8r, which I have yet to see).

If you are a programer/developer great. What languages do you know and what projects are you working on? Gnucleus, Phex, LW...?

I spoke in defense for this forum, most important is we all find and work together.

About coders lurking here at the moment, I know there are Cultiv8or, Tama, GodX to name someone. I have a C++ background, there's no Gnutella client I contribute with code.

Greets, Moak

Let me just add this, the current mails on the GDF are perfectly explaining the "high society" and "commercial" attitude I mentioned:

Raphael wrote: "Let's vote on this issue. I'll manually conclude the vote when I see all the active servent developpers have voted. Please, only Gnutella developpers should vote, and only ONE vote per servent. I.e. LW gets ONE vote."

Vinnie wrote: "Besides, regardless of the poll results, BearShare will follow whatever scheme LimeWire uses in the interest of not creating divergent implementations."

Vinnie wrote: "Contrary to popular belief, the GDF is not a democracy, and servent developers are under no obligation to "adhere" to GDF proceedings. This having been said, cooperation is *recommended* but not required."

Sorry i was confused to.. i mean HASHes...

I was only wondered why using SHA1 because from my point of view i find it "too much" for the Gnutella Protocol and for the purpose that is used. I read some articles for reducing network traffic for Gnutella like (Pong/Query Hashes or Ultrapeers). These "extensions" are used to save bandwidth and speed up downloads. But implementing SHA1 is like going to the opposite direction.

I am not a Gnutella developer, although i know how to program. So i am not trying to "push" my ideas and make other developers to adopt them. I just like to discussing about an "on development" protocol like Gnutella.

gnutellafan you said that this forum is for gnutella developers wannabies well IMHO (always ) this forum is more oganized than the GDF, and i have found more information for Gnutella protocol by following the links posted in here than downloading some files from the GDF. I also dont like registering to Yahoo.

Both of you didnt answered if not using HASHes for text files is a good idea. I know that they are not too many out there downloading text files but its a possibility that it could be best to exclude.

When someone sends a query then all the other clients respond with query hits thath have HASHes inside them or you have to send a query with HASH inside to filnd alternatives?

Moak and veniamin, since you both know programing languages why dont you help out on an existing open source project. Any help would be very valuable to the development of the network.


As for the GDF they of course have no power to enforce anything. However if developers dont work together the network will fall apart and become fragmented.

Speaking about myself: I do only things that make fun, especially in my free time. Haven't found a free gnutella client yet I feel comfortable with. (Perhaps when Mutella becomes better documented, Max *g*, or Xolox goes OpenSource)

PS: Veniamin, currently I prefer sending a hash always when the other side requests it (0.7 proposal appendix B). Small files are not that common in Gnutella these days, so it's not a big traffic issue IMHO, also hashs help finding alternative locations when a host is "firewalled" or drops connection.

Gnutellafan!

Why you think that these developers here must be wannebies?

I tell something... I know "one" of these wannebies - his name is godXblue! He developed a client within three months... all people who tested this client so far (they used bearshare, Limewire and Xolox before!) said this prog is great!

So lets compare this with a GDF developer - for example Vinnie... he is working on bearshare for about 15 months...

So our wannebe have developed in 20% of the time a much better client! Our wannebe don´t use open source...

Hmm...

now I wonder who is the wannebe...?

Perhaps many of the GDF developers are thinking about more important things - how to make money!

This is no offense Gnutellafan, only a little story about "wannebies"!

Morgwen

Hmm, I believe I said that I was the wannabie!! I was not aware of any developers here besides cultiv8r (which I still have yet to see, same goes for client by godXblue). I apoligised for any offense, removed it, let it drop!

Moak, if you dont like the direction of a client take it in a new direction.

I would really like to see someone with the know-how add data encryption and encrypted file caches (as well as partial file sharing).

OK!

Morgwen