I'm sure you meant that the other way around: PCIE is faster than AGP, which is faster than PCI. I'm using an Nvidia Quadro FX 540, with a PCIEx16 interface, on two 20" 1600x1200 displays. One display uses the DVI output, and the other the analog, and they work together as a 3200x1200 screen. I've found this to be really useful, as I can have one window of Mcam open on each screen, or Mcam on one side and a PDF print on the other etc. I can even undock a toolbar and set it over on the other screen so that it's accessable but out of the way. No video glitches so far (knocking on wood).

Have you tried doing it as a ramp-contour?

Lou, One thing to try might be to completely uninstall the nethasp drivers, and then maybe reinstall the hasp drivers. Out of curiosity, are you using the Hasp HL driver, or the Hasp 4 driver?

I'm with Cynic all the way. I'm starting to use X once in a while, mostly for file translation, but I go to 9 to get my work done. It's just that much faster, and I already know how to run it. In X I can't even control my view the way I'm used to. Is there any way to make the right-click menu respond to keystrokes like in v9? I'm accustomed to tapping r-click and hitting t, f, s, i, c, r, or d. It takes about a quarter of a second, and I don't have to move the mouse. I know that you can add and remove selections from the r-click menu, but then you still have to click on them. One big advantage of hitting a key (or sequence of keys) is that the mouse can be moving toward the geometry you're about to select while you're hitting the key. I really don't like the [ctrl]+[alt] selections. My hands don't like to play twister on the keyboard, and the [ctrl]+[alt] commands are always active rather than being context-sensitive.

I was able to download the manual from the Haas website, and found that did did indeed contain detailed information on how to do this. It seems that Fanuc compatible controls have a command that will output a text string through the serial port (I'm afraid I don't remember it at the moment). This allows a single NC program in the mill's controller to control both the mill and the indexer. A call to the company that the mill was purchased from confirms that it supports this command. The indexer's parameters can be set to either store the incoming data as a program, or execute it as a series of commands. Each indexer can also be given one of six axis labels, and will only execute moves preceded by it's label. I suppose this would allow multiple indexers to be daisy-chained without getting confused. Now I just need to wire up an RS232 cable correctly for it and find a time when the machine isn't in use.

We have a number of Haas indexers, which we've been programming manually. An M-code sends a pulse through what I think is a two conducter cable to tell the indexer when to go to the next pre-programmed position. Apparently these indexers could be controlled through a serial cable by the mill's controller. What I need to know is what data needs to go through the cable to achieve a specified result, and how to get the machine's controller to send commands through the serial port. The machine I'd like to try this with (our most up-to-date-machine) is a Mori-Seiki NV5000 with an MSX-501 controller. Any help would be appreciated. Thanks, -Matt

In comparing the number of mouse clicks required to perform a function in v9 vs. X, a comparison to keystrokes should also be made. For me, at least, I find keystrokes to take far less time than mouse klicks. I figure that for me, one mouse click takes about the same amount of time as three keystrokes. Keystrokes also have the advantages of not requiring me to move my eyes from the geometry, and allowing me to position the mouse on the geometry as I'm hitting keys. I just got a Demo X disk to install at home, so I can take the time for a few comparisons. Perhaps I'll post them later.