Big Muscle
Member-
Content count
702 -
Joined
-
Last visited
Posts posted by Big Muscle
-
-
no they doesn't :-P they normally send their filelist to every client which asks for it... but our client doesn't ask for it :)
-
the only result will be that our client will be banned because of upload blocking :-P
ask your hub owner to block these non-TTH clients in the hub
-
But actually first you removed (changed) something from (in) the client, and for this reason Strong DC++ 2.0 is incompatible with oDC, right? Hubowners did not do anything. :fear:no, I haven't removed anything.
a) only simple condition has been added to disallow downloads without TTH tree
change hasn't been done by me, but by DC++ developers.
What ??that you are crying on a bad grave :)
-
I agree with you that they should go away from DC, but it should be ensured by hubowners and OPs.
And no, they are not leechers, because they don't block anything :)
-
BUT WHY THE HELL ARE THEY STILL DOWNLOADING FROM ME ??BECAUSE WE ARE NOT THE LEECHERS!!!!!!!!!!!!! GO AWAY FROM DIRECT CONNECT IF YOU DON'T WANT TO UPLOAD!!!!!!!!!!
-
THEN WHY CAN THEY STILL DOWNLOAD ?because we are not leechers to block uploads
I'm asking to COMPLETELY remove compatibility between these types.ask your hubowner to clock these clients. There's nothint to be removed in the client.
-
DC is about uploading and not downloading, so we won't block uploads to anyone, because we are not leechers.
-
I would like to see anything that stops...Just look at the website of another provider and you will see it :)
-
does this memory leak disappears from detector when you use your fix? because if you check the code, you will see that "clearUserList()" is called in onClose routine in HubFrame.cpp
-
Wouldn't it be easier to simply ask for "maximum connections from same IP"?yeah! it would be nice option but I don't have much time to write it at this time :)
-
if I remember well, than apexDC doesn't use overlapping because its code is overriden by super-seeding feature
-
There is a 'chunk overlapping' function already in SDC++ witch is suppose to leave fastest sources when there is "no free block" - it disconnects slower and faster user continues his segment. I suppose it's not working perfectly since no one mention it earlier?It works perfectly... erm.. but nothing works perfectly it's just because:
a) it works only for sources with known speed
it requires that the current chunk would be downloaded below 15 seconds with new overlapping source
c) if the source was very fast before, it can be very slow now or it doesn't have to have a slot already
but I admit that this feature still needs some improvements especially in forcing the connection to fast sources when there's a slow one remaining etc..
-
I prefer if you leave me to decide if I want to risk corrupted files or not.But it's not your right here. DC++ is about sharing and not only about downloading and there's big probability that you will share downloaded files.
-
Er, this is what I see in Transfers... The client gets TTH, then downloads files.xml.bz2 and then starts to download from the same users...it has nothing to do with multisource. It's due to feature Autosearch automatch queue (or how is it exactly called)
Also multisource has nothing to do with hubs, so it can't affect them in any way.
-
no, it's not possible and there's no reason why it should be possible
-
I try 20- 30 segment and works a little bit faster and better.it's really nice. You take 2x - 3x more slots in network, so many other users aren't able to download anything and it works only A LITTLE BIT faster ^_^
-
So I won't use your clinet. Live with itIt's your choice. You will get a lot of fake files, corrupted files and you will not be able to download anything from 99% of users. Live with it
I also think hashing is stupid.I do.
What's stupid with that? That you can't share fake files? Or that you can't upload corrupted files? Yes, it's really stupid :)
-
I think, it behaves like fulDC in DC++ emulation mode ;-)
-
IP address in "External WAN/IP" field is used only when the check "Don't allow override IP" is checked, else it uses IP address which has been used for connecting to hub.
-
I have something similar in my TODO on my HDD, but it's not to easy to write it.
-
A good point for adding different emulations to Apex.emulation is to emulate DC++. If you add anything else, then it won't be emulation, so it don't have to be there, because you will be kicked too.
-
possible implementaion is to send $RevConnectToMe to all active users and $ConnectToMe to all passive, then collect ips from answers (not actually connecting in first case and closing port in second)which causes hub overload
-
it always checked the share before opening else you wouldn't be able to connect to any hub
-
English:The same problems are observed at attempt to receive a file a leaf from linux the client (if I am not mistaken that at the user costs valknut. Its tag - <DCGUI V:0.3.7, M:A, H:1, S:3>)
p.s. Forgive for my English =)
it was already written - remote client doesn't have full TTH support so we won't download from it. It's a feature and not a problem!
[Support] Does Multisource adversely affect hubs?
in Support
Posted
yes, it's true
this is false. Pause is maximally 10 ms and you can't lose the slot.