MantisBT - Zandronum
View Issue Details
IDProjectCategoryView StatusDate SubmittedLast Update
0000993Zandronum[All Projects] Suggestionpublic2012-08-26 21:032014-06-11 21:50
Reporterjummama 
Assigned To 
PrioritylowSeverityfeatureReproducibilityhave not tried
StatusfeedbackResolutionopen 
PlatformAllOS-OS Version-
Product Version1.0 
Target VersionFixed in Version 
Summary0000993: "sv_reporthostname" or similar setting
DescriptionI think it would be a "nice-to-have" now that the code is open, to have something like "sv_reporthostname" where one could put a string in a form of "<hostname-or-ip>:[portnumber]", which the master server could then verify, and then choose to publish on condition of success, which helps by making sure every server the master gives clients such as Doomseeker are actually capable as advertised of speaking to game clients. This would probably break compatibility with the currrent skulltag protocol, but I believe it would also represent an easier way to find servers that have upgraded from Skulltag vs Servers on Zandronum, and help users verify port forwards with log messages telling people when their servers don't have working portforwards, and other odditys.
Steps To Reproduce1. Grab the source myself
2. RFC in a bug tracket
3. ????
4. Etc.

Wish me luck. Also, I'm open to related suggestions, since I plan to dig around in that area.
Additional Information
TagsNo tags attached.
Relationships
related to 0000992feedback Torr Samaho Master server code makes incorrect assumptions about the source port of updates. 
Attached Files
Issue History
Date ModifiedUsernameFieldChange
2012-08-26 21:03jummamaNew Issue
2012-08-27 20:19Torr SamahoNote Added: 0004488
2012-08-27 20:20Torr SamahoRelationship addedrelated to 0000992
2012-08-27 20:21Torr SamahoAdditional Information Updatedbug_revision_view_page.php?rev_id=2455#r2455
2012-08-27 20:22Torr SamahoNote Edited: 0004488bug_revision_view_page.php?bugnote_id=4488#r2457
2012-08-27 20:30jummamaNote Added: 0004489
2012-08-27 20:32jummamaNote Added: 0004490
2014-06-11 21:50WatermelonNote Added: 0009124
2014-06-11 21:50WatermelonStatusnew => feedback

Notes
(0004488)
Torr Samaho   
2012-08-27 20:19   
(edited on: 2012-08-27 20:22)
Ok, I see why you would want to publish a hostname instead of an IP, but why would you want to override the IP?

BTW: The master already verifies that a server can be reached, so you likely won't need to extend the protocol for this.
(0004489)
jummama   
2012-08-27 20:30   
I think that having a hostname reported along with a port would be nice for people hosting with dynamic dns, so that the server list on the client end can be populated with things like doomhost.dyndns.org instead of ugly-isp-assigned-host.with-an-ip-1-2-3-4.dhcp.us.wa.everett.isp.net

I did notice that the master server connects to verify already now that I've looked at it. A hostname setting like this should do a lookup on the provided hostname so it can make sure it really points at the same ip it's receiving from, and initiate a connection on the supplied port instead of the source port.
(0004490)
jummama   
2012-08-27 20:32   
As far as IP overrides, I can't see much reason for it, but if a hostname is used there, an ip would be technically valid. I think I'll set it up so that you can use the form of *:(port) to override the port and not report a host name.
(0009124)
Watermelon   
2014-06-11 21:50   
Still wanted?