concept, Forget the D and C stuff. It's just more confusing. The file is resident in memory, be it system or virtual, as soon as it is opened. Read the article in the link above. It explains that even though your system may still have physical (system) memory, things are still being placed in virtual memory. As far a why it helped Tony, I couldn't tell you exactly. It must have been a whopper of a file and maxed out all of his memory resources, making things take longer to refresh. He (and probably the IT guy) is the only one who's seen the before and after. I can honestly say that I have no idea how Mastercam takes advantage of system resources. I see alot of threads on the forum about what's stable, not what's fast. From my personal experience a Quardo FX 500, GeForce 7800GTX, and Radeon 9700 Mobility all perform at the same level. But I don't work on 20,000+ surface files, either. I was just trying to help those, who were eagerly waiting for his reply, get some information incase it sounded like something they might want to try for themselves. If his IT department is anything like mine, the forum folk could have been waiting a while.

concept/ocean, They are and they are not. The easiest thing to overlook is throughput. Try this: When copying files between partitions the head is constantly moving reading (and being interrupted by the system), then writing ( and being interrupted by the system). Constantly back and forth. When copying files between drives (separate hard drives) one drive reads, while the other writes. There is supposed to be much less latency when copying between disks than copying between partitions of the same physical disk. Make sense? Maybe someone else can translate that dribble into something coherant...

concepttool, Optimally, I believe you would want to set-up a completely separate disk, not a partition. I say this because read/writes by applications, back ground apps and the page file will cause bottle necks. Reading/Writing to disk (MB/s) is much slower than RAM (GB/s). So having a disk dedicated to the Page File should be more beneficial than a partition on the primary drive. Perhaps having a partition for the page file on a secondary storage drive would work just as well. Sounds like a project for the weekend. Page file info can be found here.

Tony, Your "page file" is also known as Virtual Memory. As a basis for comparison, how much system RAM does your computer have? To view your settings; Right click on My Computer, Properties, Advanced Tab, Performance Settings, Advanced Tab, Change Custom size: Initial size (MB): XXXX (usually 1.5 x system RAM) Maximum size (MB): XXXX (usually 3 x system RAM) Having adequate room on the hard drive to allow the maximum size of the page file is a must. I realize this shouldn't be a problem with today's HDD capacities. It might also help if your disks are tidied and defragged before trying this. I'm pretty sure that having a seperate drive just for the page file is the best way to set it up for those of you who are really interested in performance gains. HTH

On our prints the drill call outs are Diameter "X" Depth. Flat bottom holes are called out when needed, as well as MAX DEPTHs or DO NOT BREAK THRUs. We always go by full diameter to depth unless otherwise specified. In the case that a depth (for either drill or tap) is not specified, we have a company standards chart to refer to. How to interperate a drill depth? You'll probably get the correct answer from the person who designed the part. But that might be easier said than done... Good luck David.

MR1 was released in January wasn't it? One down, two to go.

Rakesh, FYI, you should open the Machine (MMD, LMD, etc.) and Control Defs with Mastercam, not a text editor [hence square blocks]. PST files can be opened in the text editor of your choice. More info on the Machine/Control Defs can be found here. Like jmparis suggested, make back-ups before modding. HTH

pbld, n$, "G90 G53", "X20.", "Y20.", e$ should do it.

quote: preselecting has the same problemIf you mean my click select/click un-select, yes. I don't consider that a "problem". It beats unselecting all and then selecting again or switching to another funtion to un-select. Would using the "delete duplicates" function first take care of some of that headache? quote: Otherwise I have to delete, hit the green and then start all over againNo "green" required when pre-selecting then deleting. If you pre-select your entities, then delete, it's a one time deal. The delete funtion does not remain the active function. So you can just keep on picking/deleting without "end selection".

Now that that's over.. back to business.

Paul, After going back to look at the problem I this is what I know: in the pheader$ post block of file Dynapath System_10_20_Conversational 4X Mill.pst, when defining eob$ = 36, no "$" characters are output during post.

I don't doubt it Paul. All I can say is it worked before the update, and that's the explanation I got as to why it no longer functioned. This problem started the same time that all this rpd_typ_v7$ business started too. I'm fairly certain you're aware of that one. All's well that ends well..

I think i found it... code: pxycalc # Calculate Output variables X, Y, and Z @x$, @y$, @z$, @xs$, @ys$ if not(vtoolx$ = 0 & vtooly$ = 0 & (vtoolz$ = 0 | vtoolz$ = 1)) & (opcode$ <> 16 & (opcode$ <> 3 | (opcode$ = 3 & startflg = 1))) & rotaxis$ =0, pmx0$ xabs = x$ yabs = y$ zabs = z$ xinc = x$ - prv_x$ yinc = y$ - prv_y$ zinc = z$ - prv_z$ if rotaxis$ <> zero & add_rad <> zero, pgetz if rotaxis$ = one, aabs = xs$ if rotaxis$ = two, babs = ys$ xtotal = xtotal + xinc ytotal = ytotal + yinc ztotal = ztotal + zinc !x$, !y$, !z$, !xs$, !ys$,e$ <--I think it's seperating the line here

Tex, The way it was explained to me was ever since Update 3 for SP1, changes made to MP.dll made it difficult to properly output the ascii character 36 ($, which used to be defined in the PST) as an EOB character. As you stated above, the $ character is also used to define a post processor system variable.

OH, sorry! I just re-read the problem. Your getting the "L" just on the line above. Debugging's a PITA. Looks like another post block is calling an e$ before outputting the position and completing the pdrill_2 post block. Look to see which of two other post blocks call pxycalc. If you want to, email it to me, and I'll take a look at it.

Mike, That "L" should do the trick. I would temporarily change the "FMAX" to "WTF" to verify that is line is being output. If it doesn't change in the code at least you can move on to a different section. Which post is that?

Mike, Look in the pdrill_2 post block for *strl. If it's not there try adding it after the n,. You may just be missing the *. code: pdrill_2$ #Canned Drill Cycle, additional points pdrlcommonb n$, *strl, pfxout, pfyout, strr0, smaxrate, strm99, e$ pcom_movea HTH

Jack, I had the same problem going to SP1 update 3. There are a couple of solutions to the problem and which one you choose depends on how much editing you like to do. In the Contol Definition you can change you EOB character to "36" which will output a "$" on every line, including the first (12345678)$ and last END$. The alternative is adding the "$" to every output line in the post. This is easier than it sounds and requires no extra edits after posting. To do the latter: 1. Back up your current PST file. 2. Open the PST file in the editor of your choice 3. Edit, Replace (ALL) e$ with "$" , e$ 4. Save the PST file 5. Post code and verify all of your lines have an EOB character HTH

Regarding Gene's link above: There is a PDF that explains how the Machine Definition, Control Definition and POST all interact. It's the most indepth literature I've seen so far, though I may have missed it in other documentation supplied with MR1/SP2. link to Working_with_MD_CD.pdf

Pre-selecting the items before deleting them would be more productive.

Are you trying to merge toolpaths from different Machine Groups?

Not again!?!

You may have to drop into single channel mode in the BIOS, in which case you will probably take a performance hit. I don't think you can run dual channel mode with three sticks of RAM. You would be better off buying another pair of 512MB sticks and slapping them in the open slots. That's assuming the 925 is Dual Channel capable, which it probably is.

From what I can tell, when only a Z value is entered in the Fast Point dialog the line will end up at the proper Z coordinate and double the D coordinate. A multi-line starting at D0. Z0. going to D1. will create a vertical line to D1. Z0., then going to Z-1 will end up at D2.0 Z-1. in a Lathe Diameter C-Plane. Is this what you are seeing as well?

Maybe having the "Fast Point" dialog be "pin-able" would be a reasonable option. I normally prefer the Fast Point dialog over the XYZ/DZY fields on the Auto Cursor toolbar for everything any how.