Archived discussion about features (predating the use of Bugzilla as a bug and feature tracker)
Moderator: Moderators
-
acc
- Posts: 5
- Joined: 2003-01-14 23:46
Post
by acc » 2003-02-04 21:07
How about if it would work like this: If the user doesnt come back
in the hub he is downloading from in 10mins, one will be disconnected.
I think it would be better when downloading from crappy unstable hubs

-
BSOD2600
- Forum Moderator
- Posts: 503
- Joined: 2003-01-27 23:47
- Location: USA
Post
by BSOD2600 » 2003-02-04 23:18
Are you talknig about this?
-- 0.20 --
* Added automatic disconnection of uploads if a user leaves the hub. This is a somewhat controversial feature that disconnects any uploads to a user if the user leaves the hub (or is kicked...). If an upload is cancelled in this way, a free slot is granted to the user being kicked, in case that the user was unwillingly disconnected from the hub and returns within 10 minutes. This effectively prevents the "jump into hub, start a bunch of downloads, and exit the hub"-cheat, without the cheater being able to do anything about it. It is off by default, turn on in the settings.
-
fregu334
- Posts: 5
- Joined: 2003-02-15 19:35
Post
by fregu334 » 2003-02-15 21:01
The time (10min at the moment) the user *have* to get back should be lower than the tempban time of the hub, so that users that get kicked for breaking rules will loose their slots before they can get back in. As it is now, kicked users is not harmed in any way, which they should be! This feature does much good, but it could do more!
CyberAL @ TropiCo.se
-
sarf
- Posts: 382
- Joined: 2003-01-24 10:43
- Location: Sweden
Post
by sarf » 2003-02-16 21:01
And how, pray tell, do you tell the client what duration there is on tempbans? This varies from hub to hub, correct?
Or do you mean that the time should be user-definable?
Sarf
---
I came, I saw, she conquered.
-
GargoyleMT
- DC++ Contributor
- Posts: 3212
- Joined: 2003-01-08 02:46
- Location: .pa.us
Post
by GargoyleMT » 2003-02-17 00:48
sarf wrote:And how, pray tell, do you tell the client what duration there is on tempbans? This varies from hub to hub, correct?
Or do you mean that the time should be user-definable?
Oh c'mon, it's obvious that he wants a dynamic, intelligent system. All you have to do is keep track if kicks bans on each hub, notice when the user came back online, and compare that to the existing temp ban estimate and take the lower figure. Once your estimate is stable, make a new class of reserved slots just for those disconnected users, and set the time (per hub) to be about 75% of the estimated temp ban size. "It should be a just a simple change."
wink, wink, nudge, nudge, say no more
And on my primary hub, 15 minutes are the length of the temp bans, so the 10 minute expiration granted slots works fine for me.
-
ivulfusbar
- Posts: 506
- Joined: 2003-01-03 12:33
Post
by ivulfusbar » 2003-02-17 07:16
but what do you do if you have random-temp-ban time? with a distribution with very high volatility? ,))
Everyone is supposed to download from the hubs, - I don´t know why, but I never do anymore.
-
GargoyleMT
- DC++ Contributor
- Posts: 3212
- Joined: 2003-01-08 02:46
- Location: .pa.us
Post
by GargoyleMT » 2003-02-17 13:13
ivulfusbar wrote:but what do you do if you have random-temp-ban time? with a distribution with very high volatility? ,))
You're twisted. I like it.
Users browsing this forum: Google [Bot] and 0 guests