| TIP: Click on subject to list as thread! | ANSI |
| echo: | |
|---|---|
| to: | |
| from: | |
| date: | |
| subject: | win95 mailslots |
TC> DG> I've got two computers networked together in win95. I decided to make TC> DG> a program to chat between the computers. So I read the win32 sdk and TC> DG> thought, gee, pipes look good. So I did that for two hours wondering TC>Two things come to mind here: TC>1. The 486 may be running on half-duplex while the 686 on Full Duplex. TC> Check your network settings. TC>2. Remember that a 486 is not a "true 32-bit" machine. I was told that TC> only a Pentium (ie. 586 and 686) have true 32-bit data transfers. TC>For #2, you have to realize that the data being send from the 486 is TC>16 bits long anyway because 486s are all 16-bit motherboards. Your 686 TC>recieves/sends 32 bits. So the answer here is to do a bit of TC>bit-chunking (I call it bit-banging, really :) 386 and up are all 32-bit. DOS is only 16-bit. WinNT is 32. Win95 is mostly 32. My point: the 486 is 32-bit. Except, you said 'true 32-bit'. What does this mean? And told by whom? (actually I prefer who, but I'm told that whom is correct). Another word for bit-chunking/banging is bit-thunking as used by the win32 dox. Where's ANSI when you need them? :) TC>On the 686 end: TC>1. For incoming transmissions to the 686, have the program get only TC> the first 16 bits from the mailslot. Disregard the rest. TC>2. For outgoing transmissions from the 686, have it send only 16 bits TC> at a time. TC>One the 486 end: TC>1. For incoming, have it wait for TWO transmissions, then have the TC> program combine and use the two 16-bit chunks for the total data. TC>2. For outgoing, you really can't do anything. That's where #1 above TC> for the 686 will have to be written to handle the 16-bit incoming TC> transmissions. This won't work anyway. I wasn't too accurate in my original message. I meant that if I send a message, I get the message twice, not just each character. eg: I send 'A' I get 'AA' I send 'TIKA' I get 'TIKATIKA' I was thinking that maybe it's for error detection. But then, how would you know a) if 1 of the messages was left out and b) which was correct? b) would be solved easily but just resending, but a). If you skipped every second message, you'd have the chance to skip the message completely. So that's where I was stuck. TC>Tika Interesting name. How is it pronouced? Tie-ka, tee-ka, other TC> Phantom's Gate - Home of the Hacker's Haven TC> http://www.frontiernet.net/~phantom7 * phantom7@frontiernet.net What's here? I tried to get there but the connection was screwed up at college today. (teachers _teach_, they don't _know_) Is this NY New York? Let's check... Yep, sysop Ken Serikstad. Wow, this message reached NY quickly. What's the weather like there? So what do you anyway? Uni/work? Also, if it was cause of 16/32 bit hassles, then wouldn't other network stuff screw up too? Like the winpopup thing. Either it's a setting in the mailslot init stuff, or it does it all the time. Another guy I talked to says all the time. I think it's time to goto the library and do a bit of research. Later, Danial. ___ X SLMR 2.1a X --- Maximus/2 3.00 ---------------TC> * Origin: Knight Moves - Rochester,NY 716-865-2106 (1:2613/313) * Origin: Hunter Connection OS/2 BBS 24hrs (049) 57-1801 (3:622/403) |
|
| SOURCE: echomail via exec-pc | |
Email questions or comments to sysop@ipingthereforeiam.com
All parts of this website painstakingly hand-crafted in the U.S.A.!
IPTIA BBS/MUD/Terminal/Game Server List, © 2025 IPTIA Consulting™.