I need to send some data over UDP (cant use TCP), and i want to make sure that the data gets there. I thought that sending som data to the other host, and then that host sends a reply back saying HEY I GOT IT, but then theres a chance that HEY I GOT IT is dropped too..
is there any way i can make sure that the receiver gets it??
>>4
I would do this but, I'm running low on memory as it is.
Name:
Anonymous2008-07-19 20:35
>>5
Maybe it isn't in the spirit of the problem, but why not just incorporate the use fireworks? Only the first message needs to be hidden. The enemy can see the confirmation message because they don't know what the confirmation means. —Preceding unsigned comment added by 80.203.36.250 (talk) 21:52, 13 September 2007 (UTC)
Not the OP, but I have a similar problem related to UDP. It's more of a simple question. I'm using UDP as a crude means of inter-process communication between processes (local UDP socket.) I won't suffer the unreliability problems over localhost, will I?
>>15
Fuck the Sussman. Shared memory is a lot faster too.
Name:
Anonymous2008-07-19 23:42
>>14
I would like to keep things simple, as the idea is that I would like my data to be available to scripting languages with basic socket libraries. Would there be any advantages of using named shared memory over local sockets?
Name:
Anonymous2008-07-20 2:28
My suggestion:
Party A sends party B a message with a random, unique ID with UDP. If party B receives the message, then party B sends an "I got it" message to party A with the same unique ID.
If party A's message fails, then party B will never send a response; after a timeout period, party A will send the message again until it receives a response.
If party B receives the same message from party A multiple times (because party B's responses to party A have repeatedly failed,) then party B will simply keep sending the same response.
Name:
Anonymous2008-07-20 5:14
>>18
In other words, implement rudimentary link-status in UDP. You'd be better off with TCP imo.
Name:
Anonymous2008-07-20 7:46
>>17
In that case shared memory is probably not right for you. But in general, shared memory is a lot simpler as it's just reading an array (or whatever you decide to treat the memory as.)
There's no reason you can't use TCP. You're just being a retard.
Name:
Anonymous2008-07-29 21:36
>>22
They're all faggots that need to learn encryption already - then they can communicate with smoke signals in the clear and it will be too late for the city by the time the defenders even get it decoded to the cipher text.
Name:
Anonymous2008-07-30 9:59
>>1 I need to send some data over UDP (cant use TCP), and i want to make sure that the data gets there. I thought that sending som data to the other host, and then that host sends a reply back saying HEY I GOT IT, but then theres a chance that HEY I GOT IT is dropped too..
>>26 I need to send some data over UDP (cant use TCP), and i want
to make sure that the data gets there. I thought that sending
som data to the other host, and then that host sends a reply
back saying HEY I GOT IT, but then theres a chance that
HEY I GOT IT is dropped too..
>>31
Does it look like /b/ to you here? Do you see a large banner saying ``/b/'' at the top? DO YOU FUCKING UNDERSTAND THAT YOU'RE AT /prog/ AND HERE PEOPLE MAKE ASSES OF THEMSELVES NOT BY SAYING ``FAIL'', BUT BY SOME OTHER MEANS?
I drink all day, I smoke all day, I've done it all but tap the vein.
Name:
Anonymous2009-03-06 12:48
same as who in their right mind would do such tasks in the cellar as a foundation for C When we pass c to car and cdr these call dispatch with a 0 or 50 000 or 50 000 or 50 000 or 50 000 or 50 000 or 50 000 or 50 000 or 50 000 or.