[WriteLog] "its broken"

[Ford Peterson]

> If we want get to a solution, it might make some progress to make a posting
> that says something like this: "I pressed this button and setup this menu
> entry, and did <this and that>. And then I expected the machine to <do this
> or that other thing>. But, instead <this is what happened>".
> 
> 
> 
> Wayne, W5XD

Fair enough.  How's this...

All of these relate to a clean boot of XP-Home with all updates available current and loaded.  The ONLY software running was WL 10.59DFULL downloaded the week before Field Day.

When one of the ops hits the "save configuration" selection, it appears to save something.  But when the system goes down, the configuration is not available.  The F keys require re-entry, and the path to wave files is gone.

When the error message "WL encountered an error and must shut down" a shut down proceeds to bring you back to the desktop with no WL running.  Then, on execution of WL, the network configuration requires relinking, the local file path is no longer valid for local backup copy saves, the system repeatedly disconnects.

For no apparent reason, the cntrl-W would simply cease functioning.  Not on one machine, but any one of several--seemingly randomly.  The only fix is to shut down and experience all the nonsense indicated above.

When a new op sits down at the station to change his/her name in the data file, this act has no apparent purpose other than to place a marker in the flat file log with that new name in the field.  Any attempt to program paths to wave files is completely useless as the system will no longer recognize where the local backup copy is now to be saved.  

Something is clearly broken in the coding that involves any one of a number of "save" activities.  The "save configuration" can and does effect where the "Save As" will place the log.  Using the "browse" button to go locate the Wave files will also effect the "save configuration" and "Save File".  It isn't any one key stroke, as you describe.  It is any one of a number of keystrokes that appear to break what should be a seamless restart.

Changing Op names can and does effect some aspects of the GUI.  For example, I heard lots of complaints about having the "on enter sends" suddenly appear active after changing an op name.

For whatever reason, using TCP/IP network, the operating positions would lose track of it's peers.  Network Gab would disconnnect.  Networked frequencies would disappear.  You could tell who was logged into the network by looking at the frequencies of the various positions.  Suddenly one or another op position would disappear.  If the Op had saved his/her local backup in the right place and under the right name, these Qs would be added to the log once the system reconnected.  More than once, operator error?  Coding problems?  who knows.  The system should know who is connected where and the ops should not have to browse all over the place to locate shared files each and every time one of the PCs goes into brain death mode.  Currently, reboot WL on power up takes almost 10 minutes of reconfiguration of F keys, wave files, paths to the log.  File names for local backup.  

We also experienced one op name getting attached to Qs logged by another op at another position.  This does not give one warm fuzzy feelings about the log dealing with the data in a responsible manner.

There are 20-30 minute gaps in our log after FD due to networks that simply do not continue talking to one another for reasons that are inexplicable.  The behavior is completely unacceptable.

RF into the system can be the root cause of some system problems.  But our problems last weekend were really due to the way in which WL requires manual intervention continuously.  For example, why the heck does an op have to save anything after 20 Qs?  Don't ask!  Just do it!  Do it after every Q!  And why on reboot does the error message "Cannot save log" appear on the screen when they hit that little floppy disk icon after being pestered about saving the file after 20 Qs?  Why does an op have to decipher which of the many files on the "most recently used" log window after a reboot?  If you save the local log, then that is what appears on top.  If you browse to the main log on another machine, then that's the one.  If some body typed in a file name wrong, then THAT's the one that appears first on that screen.  It is clear to this op that after the contest starts, there is no switching to a different log file.  Why the hell do you ask the op to decide which of the many test files, real files, mistyped name files, are to be used?  There is one ONE log.  Use that one and stop pestering the ops about anything else!  If somebody presses the wrong button, then, well, you have a mess.  JUST DO IT DAMMIT!

When a new op sits down at the log, taking over somebody elses shift, they should change their name so that the ops name gets attached to the Qs.  Then, if the previous op used "on enter send" and the current op will only use F keys manually, why doesn't WL know that the op wants this or that configuration and just set up the interface the way they want?

Why, on reboot, must the op spend several minutes repositioning the screens to make them all fit?  This used to be saved under the configuration.  But no, some windows, like super check partial, end up off the map.  Net work gab is close to where it is supposed to be, but not quite--it covers up the scores.  Geez this is simple stuff guys.  

I'm reminded of a friend who took a Mercedes for a test drive.  A beautifully crafted automobile.  Details that would make one passionate about driving it.  After the thrill of pushing a 2 G turn and enjoying some real performance, it started to rain.  The dang windshield wipers didn't work for a tinker's damn.  Windshield wipers were perfected back in the 60's.  Why the heck would a superior performance automobile have wipers that suck?   Why?  Because the author didn't take the time to make them work correctly.

Nuff said...

Ford-N0FP
ford at cmgate.com