MantisBT - Zandronum
View Issue Details
0003099Zandronum[All Projects] Suggestionpublic2017-04-27 20:042017-04-27 21:19
Korshun 
 
normalmajoralways
newopen 
MicrosoftWindowsXP/Vista/7
3.0-beta 
 
0003099: ReliablePuke
Add new latent functions ReliablePuke and ReliablePukeNamed to easily puke scripts reliably and without double puking.
No tags attached.
Issue History
2017-04-27 20:04KorshunNew Issue
2017-04-27 20:06KorshunNote Added: 0017376
2017-04-27 20:07KorshunNote Edited: 0017376bug_revision_view_page.php?bugnote_id=17376#r10437
2017-04-27 20:16KorshunNote Edited: 0017376bug_revision_view_page.php?bugnote_id=17376#r10438
2017-04-27 21:16jdagenetNote Added: 0017378
2017-04-27 21:19jdagenetNote Edited: 0017378bug_revision_view_page.php?bugnote_id=17378#r10440

Notes
(0017376)
Korshun   
2017-04-27 20:06   
(edited on: 2017-04-27 20:16)
Oh no, severity should be "feature", not "major". Also, it is fine if this request is postponed indefinitely due to lack of a reliable client -> server communication system in Zandronum. It can be worked around if you write ACS correctly, but most modders don't know how to do that.

(0017378)
jdagenet   
2017-04-27 21:16   
(edited on: 2017-04-27 21:19)
I'd imagine this would require a complete overhaul of the netcode and script puking system -- something that's not exactly a concern nor a demand at the moment.

From what I've found, puked scripts that fail are the result of unstable connections from the client. That is, frequently dropped connections and packets not ever reaching the server.

As you mentioned, stacked executions of puke commands will lower the chance of the packet not ever reaching the server. If the server STILL doesn't receive those packets then there's nothing you can really do on your end but code in fail-safes and sanity checks to account for situations like that.