We've got a printer selection panel that we can flag to run in "Silent Mode" - a default selection is made, the open processing is executed, and returns to the calling program without displaying the actual panel.

Some of our calling programs will call this routine multiple times to create disk files instead of using an actual printer. While debugging some program mods, we noticed that one of our system files that gets opened for this panel doesn't seem to always be closed on exit, and sometimes dozens of channels are in opened for this single file.

The Pre-Display is to Perform ";Init_Prog"
The Post-Display (which in this case should not be used) is to Perform ";Start_Up"
The On Exit is to Perform ";Wrap_Up"

There are 2 system files opened in Init_Prog. If running in Silent Mode, the end of the Init_Prog routine GOSUBs to the open-printer routine, sets Cmd_Str$="END", then RETURNs.
The Wrap_Up simple closes the 2 system files, resets a global variable, and RETURNs.

The curious part of it all is that the 2 system are both opened and closed in adjacent lines, but only 1 seems to get closed -- most of the time. The channel variable names are PRT_FILE and FORM_FILE, with the latter being the one that seems to remain open.

For the debugging today, it seems to have been called about 1300 times, and left behind about 100-125 open files that should have been closed. Most of them were the FORM_FILE, with a handful of the PRT_FILE.

Has anyone run into anything like this before? Is there something that's not getting set correctly when bypassing the Display Panel process?

Thanks in advance for (1) reading through this, and (2) any suggestions you may have.

While the forum seems to have made it to 2019, the main PxPlus.com website doesn't respond as being alive. This includes the online manual.

May 2019 be a Wondrous and Prosperous Year for Everyone!

In upgrading our legacy system to Nomads, we'll be having some inquiry programs with folders having quite a few tabs. Has anyone developed a method of customizing the sequence of tabs based on criteria like used id?
For example, the Item Inquiry has tabs like General Info, Pricing, Inventory, Costing, Open Orders, etc. Different users have different priorities for the information they seek, and would want different tabs earlier in the list (without having to scroll, especially). Our warehouse people may want to see Open Orders, then Inventory, then General Info, whereas our buyers may want to see Inventory, then Costing, the Open Orders, and so on.
Has anyone developed a method of doing this within the standard Nomads processing, or would I need to fill the Tab and Folder objects on a more manual basis?
Thanks, all!

And magical holidays to you, too!

Can NOMADS have a maximized window on a secondary monitor be persistent?

Every time I try to have a maximized window show up on monitor #2, it goes to #1. I'm trying to determine whether this is standard NOMADS processing, or a result of my fiddling with *winpnl.

Thanks!

We seem to have a lot of zombie processes on our Linux server, e.g.:
     user_id  26008 21801  0 13:37 ?        00:00:00 [date] <defunct>
Most of them seem to be a result of:
Temp$=""; open (hfn)"<date"; read (lfo,err=*proceed)Temp$; close (lfo)
When the user logs out, then yes, the zombie disappears, but until then, it just hangs around, getting its time updated whenever that date call is processed. According to some online comments, these hang around "because their parent has not destroyed them properly." Also, "On Unix and Unix-like computer operating systems, a zombie process or defunct process is a process that has completed execution but still has an entry in the process table. This entry is still needed to allow the parent process to read its child's exit status."
Is there anything we need to do to ensure these will disappear promptly, or do we simply have to wait for the user to log out?
Thanks, all!

I've noticed on my dual monitor PC that panels on the primary window appear (essentially) instantaneously, whereas panels on the secondary monitor take just a little more time to appear (extra 1/10 second?). It's hardly noticable onsite (the menu program fades out completely before the panel appears), but for remote PCs on the nationwide network, it may show up a bit more for the few dual-monitor systems out there.

Any insights as to why this would happen? I've tried different panels and they all seem to do this.

Running PxPlus 13.10 on RHE Linux 5 with WindX via *nthost/*ntslave

One of our users just had a problem I was working to resolve. I was at command level, and when I tried to string search:
*[" "]
the entire Windows process just disappeared! Totally gone, dropping the WindX connection to Linux with it.
She logged in again, got to the same error location as earlier, and this time, I entered:
*["/"]
Same results: totally gone. When I give the same command on my PC, I get the typical reponse: a list of statements with the desired text, or:
Error #71: String not found
Has anybody run into this before? Any ideas?
Thanks!

Here's an interesting one: I've been creating a working command window on my 1920x1080 screen using:

print 'dialogue'(66,16,82,25,"Work Window",opt="-cX"),'show'(1),'sr','cs','black','_white','cs','df','black','_white','cs','df',

It works fine, but no menu bar. When I change the opt= to "-cMX", I get an Error message box window stating only "Create failure". However, it does seem to be partly created, as the cursor is now at the top of the creating window, but nothing is erased, and nothing echoes when typed, though the cursor moves. Popping the window returns me to where I started.

Any ideas?

PxPlus 13.10, Linux 5 server, WindX on Win7Pro

There are times I'd like to have an ENUMeration-type directive available. I know the LET gets things done, but not having to repeat variable names would help.

E.g., instead of:

Opt.1=1,Opt.2=Opt.1+1,Opt.3=Opt.2+1,Opt.4=Opt.3+1

we would have something like:

ENUM Opt.1,Opt.2,Opt.3,Opt.4

Possibility?

Since a NOMADS Data Element marked as External-Only is not supposed to take up space in the record, does it matter where in the list of data elements such a field is identified? Beginning, end, middle, mixed?

Thanks!

I wanted to help someone understand what was available with *next, *same, etc. Unfortunately, when I type *next or "*next" into the Search For box, it responds:

QuoteWords in your query may not start with the wild-card characters "?" or "*".

Can something be done to allow this type of entry for a search?
Thanks!

A couple of questions on this:

• Is it possible to Disable or Enable an entire Radio Button group with one statement, other than setting the group up in Group Assignment and using *wingrp?
• Is it possible to "uncheck" all of the Radio Buttons in a group once one of them has been selected?

Thanks!

I'm setting up a panel with a set of Radio Buttons to control which Tab on a Tabless Folder to use. I've included a Perform routine for the RB's "When Receiving Focus" logic, but it never seems to execute. (ESCAPE as first statement doesn't happen.)
Is there something I'm missing?

I'm using a generated File Maint program to work on a file that has numeric fields in it. When I want to refer to these values in Logic procedures in a multi_line, I seem to get inconsistent results. Which of these should I actually use?

• Field_Name
• Field_Name.Ctl'Value
• Field_Name$
• Field_Name.Ctl'Value$

Thanks!

After a bunch of trial and error, I've been able to get (most of) our Nomads panels to scale everything fairly well; this tracks pretty well along the lines of the text-based windows from our legacy software. (Most of the users run maximized text windows, or close to it.) The thing that doesn't do so well is a check_box with right-justified text to the left of the box.
I asked this last year, and got an answer from Mike King, which looked good, but involved a bit more work than I was in a position to do at the time. Also, our scaling wasn't happening, so just setting the positions and sizes for 100% worked well. It was put on the back burner.
Now that we've got a few things going "right", I've revisited the situation. Following his instructions, I've got it doing pretty well, but the images I'm using get shrunk REALLY small - see attached. The large checkbox is the standard one used by PxPlus. The small one is from the library: {!32X32/Controls/Check_Box_Blue}
When I scale up the window, it starts to approach reasonable size and position. But at 100% (this sample), the image is tiny, and the spacing to the text is rather large (compared to the Fonted Text above in the sample). If I use a base library image (not 32X32), the 100% image is OK, but stays that size when scaled, so it looks tiny against the enlarged fonts.
The Attribute settings for this are: Tab Stop, Button to the Right of Text, Flat Button, Flat - No Border, and Transparent. Is there anything I'm missing to get this looking decent throughout the range of sizes?
Thanks, all!

Does anyone know if there is a way to adjust the line spacing in a list_box? (Maybe a grid, too?)

We're looking for something like a leading setting that would increase readability. I know we can set the list'LinesPerRow to something like 1.1 or 1.25, but that also enables text wrap, which we don't really want.

I've seen this behavior in the Nomads designer, where it's normally a bit spaced, but if the designer errors out and I restart it, the lines are cramped together.

Suggestions?

We use the Status Bar to provide certain info to the users. We also allow them to re-scale the window size in specified increments. When the window is resized, we also reset the 'option'("MessageBar".##) value, along with the segment sizes, all scaled to the new window size. The "problem" is that the new MessageBar information doesn't take effect until either another program is run, or the window is manually minimized/restored.
Is there something I'm missing, or is that "just the way it is", or can something be done about it?
Thanks!

There are several methods within PxPlus to generate an ESCape character: CHR(27), $1B$, esc variable, and 'ES' mnemonic. In most cases, the 'ES' mnemonic is (and should be) set to a value of $1B$.
However, certain printers may use a different "attention" character, e.g., Printronix typically uses $01$. Would this be a better use of the 'ES' mnemonic, since a $1B$ wouldn't necessarily do what one would expect?
Thoughts?

Is there any expected timeline for any of the maillist content to be available on the forum? While the maillist is still functioning in a Q&A mode only (no interactions or searching available), there is a deep fount of knowledge that would ultimately be lost if it doesn't get transferred - even in abridged form.

In a quick review on my browser bookmarks, I noticed some items that I had marked that were in the forum, but alas are no longer accessible. (I guess I should have reviewed earlier while it was still up, so I could download the details. Too late...for now.)