Re: [Xlog-discussion] xlog-0.7 release candidate 1 !!

[Joop Stakenborg]

On Wed, 9 Oct 2002 22:45:21 +0200
Stephane Fillod <address@hidden> wrote:

> 
> Hi Joop,
> 
> Some ramdom reports before 0.7 releases:
> 

Aaah, something to chew on.... :-)

> * segfault:
>   When I do Ctrl-O to open a new log, from the "select a log" dialog
>   box, I go up in the directories (../..), then down to my logfile
>   directory (still using double click), select a file,
>   and then, whe I click the combo box to change type to "adif",
>   xlog seg faults. IOW, it's because of the file loading.
>   Must be related to readdir somehow, because the last strace line
>   is the following:
> 
>   open("/home/fillods/RA/logfile/wb5nhl.adi", 
> O_RDONLY|O_NONBLOCK|O_LARGEFILE|O_DIRECTORY) = -1 ENOTDIR (Not a directory)
>   --- SIGSEGV (Segmentation fault) ---
>   +++ killed by SIGSEGV +++
> 
>   The seg fault seems to come out of libqt (like passing a bad pointer).
>   gdb makes my X server crash..
> 
>   If you can't reproduce it, I can spend more time on it.
> 

Yes, it reproduces. Will try to fix this one.

> * hamlib capabilities
>   My rig does not support LEVEL_RFPOWER retrieval. However, when I click
>   on the "Power" button in the QSO window of Hamlib, I get a big
>   "UNKNOWN". Would it be possible to check for 
> rig_has_get_level(RIG_LEVEL_RFPOWER)
>   and put a default default (from prefs) value in the field (thus Ctrl-K is 
> still handy).
>   The same applies for other fields.
> 

Okay. Would you mind if I put this in 0.8beta1? At this point
I don't want to introduce any changes (and possible bugs) to the GUI.
But will put it on the top of the TODO list.

> * RST?
>   The TX(RST) field in the QSO window sometimes reports "510", and no,
>   I'm not operating CW at that time. This simply happens when
>   rig_get_strength reports over S9.. Would it be possible to chech the
>   ceiling?
> 

Haven't seen this one. I will try to reproduce/fix it.

> * QSO list
>   when moving around in the QSO list (I happen to do that), using the 
>   page down first and the cursor down make the highlighted QSO to be
>   different thant the selected one. This does not break anything,
>   but it is not expected behaviour (I guess page up/down is at fault).
>   Pressing space key makes it okay again.
> 

I see what you mean. Should be fixed.

> * dup checking
>   I don't see exactly how Dupe check works. It looks like it stops on
>   first callsign dupe found.

This means you only have one dupe in the log (same station worked on the
same band/mode)?

>   That'd be neat for future releases to have the already-seen callsign flag
>   as you typin, you were talking about with the kpsk guys.
> 

Agreed, live dupe checking is a must. It's one of my first wishes for version 
0.8.

> * hamlib.m4
>   before releasing xlog, I'd like to send you a patch to support non
>   standard installation of hamlib (ie. arbitrary install dir), 
>   including support for forthcoming pkg-config.
> 
> 

Okay, will wait for that.

> Let me know if you need help with the french translation.

Jean-Luc took the job.

> So far xlog brought everything I needed. I'm a satiated man :)
> 

Me too. Liblogfile performs nicely!
I will probably kick out a second release candidate after I have
fixed the bugs you found... Thanks for your time.

> Cheers,
> Stephane
> 
> 

Regards,
Joop PA4TU