Trust me, I had that idea long before Rob mentioned it

@rob-fitzpatrick I am strongly considering it.

I like that you said “if”

Go to your usual website and grab a download!

Release highlights that I remembered to add to the notes:

1. Added a column to the "s" (Remote Server IO) screen to show the -S port that the server is associated with.

2. Also changed the "type" to reflect ABL, SQL or BOTH.

3. Fixed some issues with the occasional unexpected display of negative numbers. These sometimes occur when certain integer fields overflow or when an instance of a metric "changes hands". That is annoying but not worth getting too excitied about -- the fix is just to actually show "-" rather than letting the error (74) occur.

4. Made "zippy" less cowardly by adopting the standard mkFlag() routines.

5. Added a DLC age field to report the number of days old your Progress release is.

6. Added code from Mike at Consultingwerk to add a Windows registry entry recording the install directory. This simplifies integration of ProTop with 3rd party tools such as the Smart Component Library Desktop. (Only works with oe11+.)

7. Further simplification to enabling session temp-table information aka lib/ttinfo.p, by removing the need to modify lib/ttdebug.i and by renaming the environment variable from "PTDEBUG" to "TTDEBUG". Now you only need to set the -tt* parameters in etc/protop.pf and TTDEBUG=yes in bin/localenv.

8. Added LatchReq (latch requests) and Lat2LogRd (latches to logRd ratio) to the dashboard.

9. Added con_BlockRec and con_BlkOther to distinguish between connections blocked on record locks vs connections blocked by internal resource contention. Record locking issues are generally coding problems while resource contention is more of a DB and OS tuning situation.

10. Modified "programmer mode" (control-p) to toggle back and forth and to display "programmerMode" among the indicators found in the upper left part of the screen when it is enabled.

11. Modified the -*rangesize checking routine to write suggested values to $TMPDIR/$FRNAME.range.pf to make it easier for people to copy them to their startup parameters.

12. Added SKIPLGSCAN=yes to bin/localenv.x if you'd like to skip the scanning of the .lg file at startup. Post oe11.5 it should be redundant, prior to that the impact might be that some startup parameter values may not be available any other way.

13. Replaced the conversion of an internalProDataSet to JSON/XML when passing data to the UI (lib/dynscreen.p) and the subsequent re-conversion back to a ProDataSet. This eliminates the need for the ptdefs/*.xsd files, improves UI performance somewhat and reduces memory usage.

14. Enhanced the startup debugging messages for protop.p (the interactive client) to help to identify the source of startup issues (if there are any). You can change this at any time before or after starting a session (even while running). For instance:

    echo 3 > tmp/protop.s2k.dbg   
    

will set the debugging level to 3 (which is more than enough for most purposes). You must delete the tmp/protop.*.dbg file to return to the normal default debug level.

You probably should not set this unless you have a craving for annoying debug messages.

15. ESC-D will allow the debug level to be changed from within a session. (I have no idea if that works with Windows -- DELETE-WORD is the keyfunction used.) This will NOT be persisted but it may be useful if change the startup debug level as described above and do not want to modify the .dbg file for some reason.

16. Removed the "top blocker" sub-screen from blocked users because
    a) it was wrongly labelled -- it was actually the blocked session's stack and
    b) the blocker info was added to the notes some time ago so the need just isn't there and
    c) it takes up a lot of space for no good reason.

17. Improved interoperability of bin/syncio.sh and bin/maintenance.*

18. Released as pt3.14 in honor of pi day. Plus I was running out of letters to mark minor releases with

Something like that will probably occur. I need to start mucking about with oe12 now that it is actually available.

@kynan great!

@rohan That's helpful, thanks!

The tailoring script is supposed to copy the contents of ustub to bin. So something must have gone wrong. pt3inst.err or pt3inst.log might reveal what.

I have found that I sometimes just need to run the install twice. I think I might have done something silly like replace the running pt3inst.sh while it is running.

@rohan what did you need to change? We have found some unexpected OS and shell dependent issues with the "time" command.

Older 32 bit releases sometimes have attitude problems about allocating shared memory that don't become apparent until a batch client or character client attaches to them. Try starting your databases with -shmsegsize 128.

(Ok, actually the problem is probably that the old oe10 _progres.exe was not improved to handle larger segments at the same time that the db server executables were.)

Zing!