Yes, it is public alpha. But the alpha has a ten day expiration date, and it will refuse to work after that time. And there is also very small number of people using that alpha. We really can't say that it is hurting the network if there is only a few users using that client.



And back to the original issue. If you have read that thread to the end, you would notice that there is a solution for it already, even if the new alpha haven't been released yet. The problem was that the requery interval was set to zero. And because there was no other users complaining about that, the default value was correct with most users (for some reason I had the wrong value). So, actually I was the only user flooding the network with BearShare.

Everybody who change this value is flooding the Gnet, such an option should not be changegable...

Only some users use this alpha? Everybody is able to download it so how do you know how many people are using this alpha?

You should test the alpha in private... like other do too!

Morgwen

P.S.:

The solution was posted after I created this thread here, check the dates!

I'm pretty sure that it will not be changeable in the release version. If it is fixed in next alpha, after the present alpha is expired there are no user flooding the network.

Run any client what is capaple to show what make of host it is connected to. It will be quite a hard work to find a BearShare 3.0.0.aXX user. Or search something, it won't be that easier to find 3.0.0.axx users in that way either.

Releasing the alpha in public gives you a lot of comments from users. Bugs are found much faster than in closed and limited testing enviroments. And because all alphas cam be used only a limited amoun of time, the possible harmful alpha version will not continue harming the network too long.

And it gives a good feeling to the most active users, they can take part to the developing process.

That was my mistake. I'm sorry about that.

someone said: "You should test the alpha in private... like other do too!"

If the alpha were not public then how would this bug have been discovered? There is a difference between INTENTIONALLY requerying at short intervals and a BUG. Luckily, the bug was caught before 3.0 went final, so it does not cause harm to the gnutella network at a large scale. XoloX and friends didn't care about the effects their aggressive re-querying behaviour would have (Don't even try to say that XoloX wouldn't/hasn't harm the the gnutella network. You just said that BearShare would harm the network with agressive requerying). XoloX dont even have public betas.

wrong, look into the xolox forum

To be correct Xolox never released a "final" version AFAIK...

but you can test betas in public (they tested it for some weeks in private now) but not alphas!

Morgwen

Of course now, if Vinnie would label 3.0 aka Defender a beta...

The difference is Xolox tested in private for some weeks, bearshare not!

Normally alphas are versions which are for non-public tests and betas for the public... normally...

Morgwen

Morgwen, do you have possibly something agains the way how BearShare is testing the betas in public? I think there is nothing wrong with the way how BearShare is doing it (ten day expiration date in each beta, new betas out almost daily). It is the best way to get feedback from wide userbase and it helps you to find bugs quickly.

Sure its the best way to find bugs quikly, but while you search for them you might hurt the net...

So the normal process is first to test the version in private (a few users) until you can say the big bugs are found and then release a "public" beta!

Limewire and the others do the it this way...

Morgwen