Op problem !

Technical discussion about the NMDC and <a href="http://dcpp.net/ADC.html">ADC</A> protocol. The NMDC protocol is documented in the <a href="http://dcpp.net/wiki/">Wiki</a>, so feel free to refer to it.

Moderator: Moderators

Locked
Kandjar
Posts: 4
Joined: 2003-03-12 06:16

Op problem !

Post by Kandjar » 2003-03-14 19:35

I have a problem (on my Hub), when an OP is arriving to the server, the other users see him like an standart user ! The incoming user get the op flags after a "refresh user"... why ?

mindpirate
Posts: 3
Joined: 2003-03-15 17:39

Post by mindpirate » 2003-03-15 17:43

I've had that problem as well since 0.24 was released, but I thought it was due to some changes I had just made to my hub's source code (Shasta hub). Which hub are you using?

GargoyleMT
DC++ Contributor
Posts: 3212
Joined: 2003-01-07 21:46
Location: .pa.us

Post by GargoyleMT » 2003-03-15 22:09

changelog.txt wrote:* Updated H: tag. It now looks like H:x/y/z where x, y, and z are normal, registered and op hubs respectively. It is updated roughly once a minute, but if you're disconnected, the count won't go down until the window is closed (i e hubs you're trying to reconnect to count as well, to avoid too many updates being sent to the hub). Also, note that there's a problem with the op count. All op's will first be counted as registered users, and then, when they're fully logged in the count will be updated. This is a protocol limitation and there's not much to do about it (although, most of the time you shouldn't notice, if the hub sends the op list fast enough...). This release also fixes a bug where the count was not correctly updated across all hubs.
It sounds like at least some of this was expected. If the behavior is limited to a specific hub, then perhaps we ought to look at that. If there's no common denomenator other than DC++, then...

mindpirate
Posts: 3
Joined: 2003-03-15 17:39

Post by mindpirate » 2003-03-15 23:50

GargoyleMT wrote:
changelog.txt wrote:* Updated H: tag. It now looks like H:x/y/z where x, y, and z are normal, registered and op hubs respectively. It is updated roughly once a minute, but if you're disconnected, the count won't go down until the window is closed (i e hubs you're trying to reconnect to count as well, to avoid too many updates being sent to the hub). Also, note that there's a problem with the op count. All op's will first be counted as registered users, and then, when they're fully logged in the count will be updated. This is a protocol limitation and there's not much to do about it (although, most of the time you shouldn't notice, if the hub sends the op list fast enough...). This release also fixes a bug where the count was not correctly updated across all hubs.
It sounds like at least some of this was expected. If the behavior is limited to a specific hub, then perhaps we ought to look at that. If there's no common denomenator other than DC++, then...
Thanks, I completely skipped over that little blurb in the changelog. Sounds like all I need to do is change the order of a few functions in Shasta's login methods, which is what I've already been jockeying around with lately anyhow. I'll be back with an update.

mindpirate
Posts: 3
Joined: 2003-03-15 17:39

Post by mindpirate » 2003-03-16 19:50

I seem to have solved the problem in Shasta. $OpList is now sent immediately after $Hello is sent, and the sending of the MOTD has been postponed until after those messages. My ops now appear where they should upon login, viewed with DC++ 0.24 as well as earlier clients.

Locked