We encountered an interesting situation and could use some clarity and/or confirmation.

Using a Linux server (RHEL 7.9) with PxPlus v15.1 via WindX we open a test file and EXTRACT a record. In another session, open the same file and READ the record. No Error 0.

Upon closer examination, we noticed the Linux file permissions were "-rw-r--r--" and the file was "owned" by a different username. When we changed the permissions to "-rw-rw-rw-", the EXTRACT/READ above worked as expected.

So, it seems that an EXTRACT on a file where a WRITE cannot actually happen is treated as a READ, even though everything, including FIN(ch,"Extract"), suggests the EXTRACT really happened. Except it didn't.

Is this proper? Or is there something missing along the way?

Any guidance would be greatly appreciated.

Thanks!

Until I started here a couple of years ago, the NOMADS panels that were being used were being SPAWNed as separate processes off our menus. Since that would allow for multiple 'sessions' using the same FID(0), it also led to occasional file conflict problems. We've been working to remove the SPAWNing part of the sequence and just have the menu PROCESS the panel directly after hiding itself ['show'(-1)]. With a little work to undo some of the added 'hooks' for the SPAWNing, this works pretty well.
The problem we're running into is timing. When the panel was SPAWNed, it loaded and displayed VERY quickly. However, directly PROCESSing the panel generally takes quite a bit longer, sometimes 6-8 seconds or more compared to the 0-2 seconds previously. And it doesn't necessarily correlate with the number of controls on the panel. Perhaps it can sometimes be chalked up to server utilization, but it seems to be a pretty consistent differential.
We're running RHEL 7.9 (64-bit) with PxPlus v15.1 and WindX via VPN. Most of the client PCs have been upgraded from Win 10 to Win 11.
After the latest wave of updates, we're starting to get some clear complaints about this. Any suggestions would be greatly appreciated.
Thanks, all!

We have a Report View listbox with 4 text lines per entry. When we include accented characters as part of a line, it breaks the line and we now have 5 text lines (see attached). If I translate the accented characters into un-accented characters, it's back to being a 4-line entry.

Is there an option I'm missing somewhere? Is this something with Windows or PxPlus and listboxes?

Suggestions welcome.

We're starting to get new people using our application, and instead of using the F-keys we display, they'll use then ESCape key. Normally, we wouldn't expect any problems, but it seems that PxPlus maps then ESCape key to ctl=4 (F4 key) by default. Is this correct? Or was there a change a while back? Or am I missing a parameter setting?

I've changed 'my' keyboard configuration to have ESC return $1B$ and it works as expected.

BTW, the example program for 'Format 3' on the SETESC page does not work exactly as described in the table.

Is it possible to get both a single and double click in a Report View listbox? I've been able to get the single by setting Signal All Changes, but when I do that, I can't double-click; it always responds as if it's a single click. When I don't set it, I can get the double but not the single.
Any ideas?

We've got a panel with a TitleBar, an embedded panel, and several other controls. The TitleBar has 1 dependency, the embedded panel has 2, and there is one for one of the buttons on the panel. We had some timing issues with this, and I had previously had some duplicated entries, but those have been removed, so we're down to what I've listed.

I set up a trace file, and it looks like all of the dependencies for the TitleBar and embedded panel are processed twice for each cycle, while the panel button's dependency is only processed once.

Also, if the panel is set up as Resizable/Custom (the original plan), all of those dependencies seem to be re-doubled. It's as if the setup procedure is called multiple times and/or the counter is not reset to 0.

Is there something I should/not be doing?

(PxPlus v15 on Linux)

We've got a file with keys defined as:

External key size ............: 6
Alt. key 1 ...................: [7:1:10]+[1:1:6]
Alt. key 2 ...................: [12:1:8]+[1:1:6]
Alt. key 3 ...................: ["ADDRESS1":3:1:30]

When creating an Auto-Complete definition for Alt. key 3, there seems to be a problem. When I set the "Key Name/Number" field to 3, all is well. But when I set it to "ADDRESS1" (with or without quotes), the auto-complete doesn't work; it operates as a standard multi-line.

Is there something I'm missing?

(PxPlus v15.1 on Linux)

I've been wracking my brain trying to figure out what's going on, with only minor insights. It doesn't help that I don't really have any REST interfacing experience.

We're trying to connect with a service for address validations. There are two steps involved: (1) Get an oauth2 token from them, then (2) submit the actual request using the token. I haven't been able to get to step 2 because of failures in step 1.

Working with someone else in the company, we were able to get a curl command to work:

curl --location 'https://example.com/oauth2/vn/token' --header 'Content-Type: application/json' --data '{ "client_id": "xxxxxxxxxxxx", "grant_type": "client_credentials", "scopes": "urn:tr:onesource:auth:api:IndirectTaxDetermination", "client_secret": "xxxxxxxxxxx" }' > /tmp/result.txt

But running the related information through the *plus/web/request, I was ALWAYS receiving a 400 error in response. (Actually, I had to make some adjustments, since it always added ";secure" to the OPEN, and that generated a PxPlus error.) Apparently, we need to be able to submit the request down at the "token" page level, but the *plus/web/request breaks that up, and PxPlus won't accept a filename beyond the primary site (.com).

I would like to keep all of the programming within PxPlus, but if I need to curl things to get them to work, it's what I'll have to do. Any suggestions would be greatly appreciated.

Currently running v15 (2018) with WindX on Linux.

Hi All!
Has anyone had any experience interfacing PxPlus with Thomson Reuters OneSource for Address Validation? I've been stumbling a little and was wondering if anyone has any pointers.

Thanks in advance!

It seems that when we try to post a new topic on the Forum, it takes quite a while before an error is returned that isn't really an error. The message actually gets posted but the error message says not. As a result, several people (including myself early on) have posted the topic again, only to find it creates a second entry.

Is there any way to get this cleared up?

Is including an ERR= clause in the 'DROP' mnemonic a valid construct? e.g.:
print 'drop'(Win_Num,err=*next),
It doesn't generate an error, but it also doesn't seem to trap error #57 if the specified (non-zero) window# doesn't exist.
Any insights?

It seems that the Google Maps object is only usable within Nomads. We're looking into using the GeoCode functionality to verify addresses being entered in various application programs. Is there a way to do that with the current object? Is there something else we can use? Do we need to develop something ourselves?

Thanks!

We've got a F/M panel with a menu bar, TitleBar, and 7-tab folder. I'm trying to have a popup menu with variable file-driven data appear in most situations when the user tabs out of a specific field. I've got most of the positioning down, but while the control's 'Col and 'Line positions seem to reflect the accurate position relative to the window, when we apply those coordinates to the popup_menu directive, they seem to try to position the popup relative to the folder's sub-panel. As a result, the popup appear several lines below the control it's supposed to be associated with. I've been able to compensate a bit by subtracting the sub-panel's 'Col and 'Line positions, but that doesn't seem to take into account the TitleBar or menu bar.

Is this a problem with the window building process, or do I need to do these compensations, and how do I determine the heights of the bars?

Thanks in advance!

I'm updating a panel that has single-line multi_lines and multiple-line multi_lines. The fonts are different sizes: the multiple-line entries' fonts are 1/6 larger than the single-line entries (see attached snip). All of the font info and attributes are the same.

As a test, I shrank a multiple-line entry to 1 line and the font became the smaller size.

Anyone have any ideas?

Thanks, all!

While testing an update to a panel, I opened the old version in one session and the new version in a second session on the same workstation. Whichever session I first access for the autocomplete seems to lock out the other session; it's still a multi-line, but doesn't create the auto-complete drop-down.
Is there something I'm missing/not setting up? Any suggestions welcome.

We're running a character-based menu system that will run an existing CUI program or PROCESS a NOMADS panel, based on file parameters. I've created a new NOMADS F/M and when I exit by clicking the appropriate button, all is well. However, when I use the Alt-X hotkey I have set up for the Exit button, I get unexpected characters in the menu program's typeahead buffer.

We have other panels that are PROCESSed off the menu without incident, but this one, for some reason, works differently. (We're not using the NOMADS Security processing.) It's a panel with a few controls, a TitleBar, an embedded panel, and a folder on it. We have 2 options: Full Access and Inquiry Only, based on a file-based permissions list. The menu starts the Full Access version with 7 tabs on the folder, but if the user is identified as Inquiry Only, part of our Pre-Display routine PROCESSes a 3-tab version and exits with Cmd_Str$="END". While it doesn't seem to make any sense that there would be a difference, the 7-tab Full Access version drops "BN" into the input buffer, while the 3-tab version drops "BL" in. If I include a PRINT 'CI', in the Exit button's logic, the characters don't show up.

Any insights would be appreciated!

We're running PxPlus v15.10 on Linux RedHat with WindX on Windows 10 Pro workstations.

We have a multi-keyed file with 2 keys; the first has 8 components, the second has 9. When we tried to manually remove an entry, the REMOVE wouldn't work unless we EXTRACTed the record. It turns out that the primary key had a null key field element. I wouldn't think that should cause it to require an EXTRACT to be REMOVEd. (BTW, the file has almost 14M records.)
Any insights?

Is there a way to synchronize the horizontal scrolling of 2 list_boxes? The panel isn't wide enough to display the full width of the main list, which has the detail lines. We're setting up a second list of 1 line that has the totals for certain columns in the main list. By synchronizing them, the totals will always appear below the columns they refer to.

Similarly, if the user changes a column's width, is there are way to detect that and replicate it in the other list?

Thanks in advance, all!

When we set up many of our panels, we set them as Resizable/Custom with the various controls anchored how we want them. Part of this is to ensure we don't have any (including a logo picture) trying to be set up beyond the edges. I've noticed that this essentially sets the minimum size of the panel based on the designed size. This only allows the user to make the panel larger, but not smaller.

Has anyone got an idea about how to set a small minimum size, but initialize the panel at a somewhat larger "default" size, which would allow the user to resize the panel larger or smaller?

Thanks, all!

While testing/preparing to use split files, I ran into a situation with an embedded io routine. In multifile.pvc, the Do_Pre_Write routine does its thing until it gets to:
if IoType=1 then write lock record (Chnl[_Found_chnl],key=_key$)_rec$ else write lock record (Chnl[_Found_chnl])_rec$
!
if and(_opt,64)=64 then gosub Extract_Update else extract release (lfa); gosub Extract_Free

Our embedded io routine does some post-write processing. This is called by the write lock record directive. It opens a reference file, reads it, then closes it. (There may also be some additional writing, based on results of this read.) Since _opt is 0, the next code line tries an extract release for lfa, but lfa was not the Chnl[_Found_chnl] in the write line. Changing lfa to Chnl[_Found_chnl] in the extract release line works fine in this situation.

However, there are other references to extracts and lfa elsewhere that I haven't gone through yet. These may be affected by post-read and post-write routines in embedded io routines.

Thought you'd like to know...