Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!csd4.csd.uwm.edu!jgreco From: jgreco@csd4.csd.uwm.edu (Joe Greco) Newsgroups: comp.sys.cbm Subject: Re: VIP term Message-ID: <460@uwm.edu> Date: 12 Oct 89 21:53:07 GMT References: <20859@gryphon.COM> Sender: news@uwm.edu Reply-To: jgreco@csd4.csd.uwm.edu (Joe Greco) Organization: Interstellar Telephone, Telegraph, and Telepath, Inc. Lines: 55 In comp.sys.cbm article <20859@gryphon.COM>, rhebi@pnet02.gryphon.com (Roy Hebi) wrote: ]I just read the message saying that 2400 on the C64 with VIP is impossible. ]Well, as I recall, I have been using the VIP for over two years now (or maybe ]I have an old version). It does not support 2400 though it claims to. I ]recently bought an IBM-PC, and decided to transfer about 3.5Megs of textfiles ]that I have to the IBM, and I use VIP to do that. The VIP can send 2400 error ]free, as far as the IBM had noticed... Let's not get picky. VIP term cannot do a heck of a lot at 2400. File transmission is not only a one-way transmission, the 64 also does not need to do major screen updating. And file transfers hardly comprise the mainstream of terminal program use. Heck, I can write a program to transmit at 9600 baud. ] Regarding taking up 120% of the processor time, well, I wrote a 80-columns ]driver that will print CBM-ASCII or regular ASCII, which prints (if scrolling ]is not needed) over 1200 chars per second. This means each char takes .00083 ]of a second to print. 2400 bps is 240 chars per second, which are approx. 4166 ]clock cycles per char. My 80-columns needs 2083 clock cycles. Now, if you add ]these up and lkook at the execution times of the RS-232 routines, you'll see ]that it is impossible. BUT!!! I also have a source code in 6502 ML for ]enhanced RS232 routines which give good 2400. And their execution times are ]such that I believe that 2400 and even full-ANSI emulation is possib le. Everything is possible. As far as processing time goes, there are of course operations that do not take as much time as others. I seriously doubt that somebody using a terminal program would run into a situation where the screen rarely needed to be scrolled. For ANSI emulation, screen manipulation is a MAJOR function! This means that you had better take scrolling time into consideration. I actually prefer to use benchmarks that are scrolling-intensive, beyond what one might consider reasonable. This helps to ensure that there are not problems. VIP does not have the specialized drivers. Therefore, that is a null point. Even with the specialized drivers, the processor drag is severe. I've tried it. You have also failed to consider other operations. RS232 transmission takes time. Interrupt processing. ANSI emulation takes some time, too! You can't just print the incoming data stream. 2400 baud RS232 is certainly possible on the 64. I'm hesitant to claim that it is reliable, but it may even be so. However, I do not believe that a decent ANSI terminal emulation can be done in hi-res 80 columns at 2400 baud and be reliable. Either it will lose incoming data, be very sluggish, drop keystrokes, or all of the above. Now, if you are using an 80 column card, it's a slightly better forecast... and if you are using a 6551 ACIA, it is a pretty picture indeed. -- jgreco@csd4.csd.uwm.edu Joe Greco at FidoNet 1:154/200 USnail: 9905 W Montana Ave PunterNet Node 30 or 31 West Allis, WI 53227-3329 "These aren't anybody's opinions." Voice: 414/321-6184 Data: 414/321-9287 (Happy Hacker's BBS)