On 09:42 AM 19/12/2002 -0500, Mike Reagan said:
> I certainly hope that Altium do *not* change back to the limited form of
> globals in P99SE.

There were and are on certain occasions that I have found the selection in
99se  PCB  limited.   Those few occasions, I would sacrifice  for the speed
of the current selection process.   I recently had a  friend  (orcad user  )
at my office and I was demonstrating 99SE to him .   His immediate comments
were,  how good the graphics was, how intuitive, and how good the graphical
interface was.   He reflected my long held comments almost to the word.
In addition to the selection  (IS.....),

Longish post follows.

DXP still has the ability to build up selections manually - there have been, and are, discussions about improving this, as one area where DXP dropped the bundle is the loss of the difference between focused and selected. Though the difference in operation, these days, is not as bad as originally. And I think there is more improvement to come. So DXP has manual selections (have been called ad-hoc selections in discussions with Altium bods).

There is still the selection functions like Connected Copper, dragging a selection window, etc.

The queries are used to select classes of objects - not necessarily a class as in a net class etc, but that can be done as well, of course. But I can restrict track selection to just very short tracks, or just components with more than a certain number of pins (or less, or between ...) (CompPinCount). Or what about overlay tracks that are in SMT components only(InLayerClass, InSMTComponent and IsTrack)? Or tracks that are at a 45 degree angle, or just those angled up, or down (IsOblique,IsObliquePositive, IsObliqueNegative)?

I have the ability to query components in a Component Class (InComponentClass). Or nets in a Net Class(InNetClass). Or pads in a Pad Class(InPadClass). Whether the object touches or is within a room (TouchesRoom, WithinRoom). (Rooms can be polygonal.)

The list goes on. So any global you can do with P99SE can be done in DXP (*see below for a caveat here). Plus much more. So there is heaps of power, but that is not your beef, is it.

Can you get to the more limited, but more commonly used sort of things commonly done in P99SE.

For instance, how fast is it to change the width of selected tracks?

Firstly assume the desired tracks are selected, and only these tracks are selected, so this is a test of the operations required just to undertake the global operation itself:
99SE: dbl click on track, click on Global button, drop Selected match by to Same, enter new track width (possibly include arcs), hit enter.
DXP: Hit F11 (to bring up navigator), click in Width field, enter new width, hit enter.

DXP is heaps faster. *But* in DXP the width property will *not* be shown in the Inspector panel if you have other objects selected that do not have widths, such as a fill or even a pad, as could easily happen if you selected a net or connected copper. So a fairer test may be something like....

If the net is not selected both P99SE and DXP have similar functions to solve this, one easy way is Select Net, click in blank space and enter FAT_TRACK. But as discussed this will select fills and pads etc. In P99Se this doesn't matter, in DXP we have a couple of ways about it but probably the easiest is to use the find similar function:
DXP: right click on desired track and choose "Find Similar Objects...", change the match drop boxes as required, in this example modify the Net drop list to Same (rather than Any) and then click OK, then F11 and enter the new track width and hit enter.

This is about the same as P99SE in operations. So for me I am finding that DXP is often faster (less keystrokes), sometimes the same and rarely slower in doing globals...but...read on. Globals in P99SE are essentially constant time and require, basically, the same time for every object type you wish to change. So changing the width of tracks and strings requires twice the time of just doing tracks. DXP can be much faster if you have the selection you wish to change and the selected items all have the property you wish to change. P99SE does not necessarily use selections for globals, DXP needs selections for globals.

Where DXP could use a tweak is that the Find Similar dialog should give the option of matching by selection, I will suggest this. Also, a subtle thing that I am sure you have picked up is that in DXP you have to sometimes modify the selection to get the Inspector to display the property you which to globally change. This is not the case in P99SE. This is the *only* advantage I can see that P99SE has over DXP. For me common globals are faster in DXP than P99SE, but you have to manage selections more carefully, which is a bummer. DXP provides selection memories that can help here. Remembering to use them is the key here...and they require an extra step to store and recall.

I used to get very annoyed by managing selections in DXP, this is getting watered down over time, but I do think that P99SE has something over DXP in that selection is different from focused and selection is a persistent object property. DXP is more standard (think word processors and spreadsheets), but P99SE is better in this regard. I can live with DXP seelction operation but I prefer P99SE's; a number of us are still working on thinking about improvements so who knows what I will think in a years time.

(Actually there are really only four things that P99SE has over DXP in my opinion, 1) By far the biggest is Selected v focused (and the implications for globals as discussed above), 2) faster and lower OS requirements, 3) I know it heaps better, 4) PCB net highlighting in P99SE works better, but there have been discussions on ways ahead for DXP here as well so maybe this one will drop of the list in SP3. There are other things that P99SE has that DXP doesn't - like ddbs - but in general the tradeoff for not having these features is better than the P99SE feature, IMO - like better revision control.)

I pray the global edits are
reinstalled in the program.   It  met 99.9 % of all my criteria.
Do you know that the new way does not also meet your criteria? Globals edits are available - they are just done a different way.

I remember back when I first saw DXP my first question was "AAACK, what happened to globals?"

 But there
is a need for biter selection.         I even got crafty and slick with 99SE
on occasion so there are work arounds for selection.   On the occasions that
it failed me, I doubt if DXP was up to the task of what I wanted to select
anyway.   ie  selecting classes of component, or classes of nets.  Or maybe
99se can do it and I haven't figured out how to select classes
Both can do it easily.

DXP can do stuff like:

Where InSIP, for instance will be true if the entities belong to a component that is SMT and has all pads in a row or column. (Which, it likely not so be all that useful as few SMT SIP connectors, for instance, do not have pads for hold downs that are off the line of electrical pads, but still the point is made).

What about selecting all the local address bus (but assume we do not have net classes set up?
InNet('LA*') will select LA0, LA1 .. even LA_SUXS (oops!).
InNet('LA*") AND (NOT InNet('LA_*')) - there that is better.

Better stop doing Altium's marketing for them - like I said before, the paranoids will have a field day. But seriously I do call things as I see them and for me at least I do not want to go back to the old globals, at least not if it means getting rid of the new stuff. And though I can see lots more being done to make the new stuff easier, I doubt there will be a return to the old way, but I could be wrong.

bye for now,
Ian Wilson

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
* To leave this list visit:
* http://www.techservinc.com/protelusers/leave.html
* Contact the list manager:
* Forum Guidelines Rules:
* http://www.techservinc.com/protelusers/forumrules.html
* Browse or Search previous postings:
* http://www.mail-archive.com/proteledaforum@techservinc.com
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Reply via email to