WSJT-X/issue-log.txt

wsjtx-1.4.0-rc1
===============
Release candidate abandoned due to packaging issues on Linux.

wsjtx-1.4.0-rc2
===============

Mac installer - g4wjs, Bob n7rjn - Cannot expand archive in RO DMG.

    Move tarball to Desktop or somewhere else before expanding.

Hamlib TS-2000 - k1jt - WSJT-X "Mode" != "None" glitches when Tx QRG changes

       Trying to read "other" VFO mode requires FRn command to "other"
       VFO which is source of glitch when split.

       Current best workaround is to  set "Mode" to "None" in Settings
       which stops WSJT-X trying to manipulate rig mode.

       Won't fix - Closed.

Hamlib FT-847 - Eddy g6uqi - fails to start
Hamlib FT-847 - Mark w7mem - fails to start

       Sent version with CAT trace - waiting log file.

       HamlibTranceiver::init_rig() not coping with  no get VFO and no
       toggle VFO CAT command available, changed to just assume A/MAIN
       is always the Rx VFO on such rigs.

       Sent g6uqi a test version of RC3 with a fix for this.

       g6uqi reposts that this hot fix (r4449) has resolved the issue.

       Fixed wsjtx/wsjtx-1.4 (r4449) - Closed.

Commander TS-2000 - Jim wb6dkh - SPLIT disappearing
Commander TS-2000 - David wd4kpd - possibly the same issue

	  VOX keying might be the factor.
	  Sent logs to Dave aa6yq for comment.

	  Further investigation now points to the auto-atu which swaps
	  VFO and mode while tuning, this may be confusing Commander.

	  Dave  aa6yq  confirms  that WinWarbler  will  interact  with
	  Commander  and cause  this issue  when the  TS-2000 auto-ATU
	  switches to CW during the tuning sequence. Passed info on to
	  wb6dkh & wd4kpd.

	  Not fixed for wd4kpd as he  doesn't use WW. Sent tracing RC2
	  - awaiting log and match Commander ErrorLog.txt.

	  Logs  received from  wd4kpd.  Unexpected drop  out of  split
	  reported by  Commander when setting Tx  frequency. Forwarded
	  to aa6yq for analysis.

	  Dave aa6yq reports that setting Tx frequency clears split on
	  this rig (I guess many others too that use Kenwood style CAT
	  commands). He  suggests setting Tx frequency  before setting
	  split. This  is slightly  problematic because  setting split
	  has side effect  of setting mode on Tx VFO  and on Yaesu and
	  Elecraft changing  from CW to  another mode applies  a small
	  frequency offset :( We may have to accept this as it will be
	  rare and there is no explicit way of setting the mode of the
	  Tx VFO in  the Commander interface. The  other workaround of
	  doing everything  twice (TxFreq, Split, TxFreq,  Split) will
	  glitch the Rx or Tx twice as much.

	  Seems that split needs to be set both before and after every
	  change of Tx frequency. Before to satisfy rigs like the Icom
	  IC-756 and after to satisfy rigs like the Kenwood TS-2000.

	  Sent  out a  test  version  to wb6dkh,  wd4kpd  and aa6yq  -
	  awaiting feedback.

Commander TS-2000 - Jim wb6dkh - Fake split

	  Leaving wrong QRG (Tx) current.
	  Need trace for this issue.

Flex 6000 (TS-2000 emulation) - George kf2t - fails to read QRG
Flex 6700 - Ed k0kc - No CAT no PTT

     Flex doesn't have FB; command.

     Sent kf2t test version that doesn't query TX QRG if not in split.

     Sent RC3 preview with a hotfix (r4434) to kf2t & k0kc.

     Caveats with  SmartSDR & FlexRadio  SDRs:-

     * Data mode cannot be set from WSJT-X, USB should be OK otherwise
       use "None" and set mode manually.

     * Split operation  is not possible,  this should not be  an issue
       given that wide band Tx is available on these radios.

     Initial reports from kf2t indicate a successful fix.

Hamlib FlexRadio (DTTSP UDP & IPC interfaces) - - Need to diagnose and check if these work

       Hamlib has a FlexRadio back end that talks UDP on 19001 or /dev/shm/SDRcommands .

Windows (Finestra WM) - Josh kk1d - Application not staying inside virtual desktop

Hamlib TS-570D - Peter Pauly w0vll - Protocol error while getting TX frequency
Hamlib? TS-570 - ??? n4au - Polling causing errors

       Split not required so r4434 may have resolved this anyway.

       Sent trace version of RC2 to Peter - awaiting log.

Hamlib FT-450 - Rudy ve3bdr - Protocol error while getting mode of split VFO

       May be  fixed by  r4434. I  have asked Rudy  to try  again with
       "Mode" set to "None".

       Not fixed. Sent trace version of RC2, awaiting log.

LXDE desktop Ubuntu - Craig wb3gck - No application menu bar

     Suggested installing or un-installing appmenu-qt5 package.

     Un-installing appmeny-qt5 resolved this issue.

     Closed.

Hamlib IC-7100 - Erik on4pb - CAT PTT method not available on IC-7100

       This  has been  identified as  a HamlibTranceiver  registration
       defect. PTT type categories incorrectly registered.

       Sent hot fix version of RC3 to on4pb for testing.

       Further problem reported with poll  of PTT status failing. This
       appears   to  be   a  hamlib   defect  with   ptt  type   being
       RIG_PTT_SERIAL_CAT   instead   of    RIG_PTT_RIG.   Asked   for
       clarification on hamlib-devel ML.

Hamlib IC-746 - Doc kb3mow - Wrong Tx QRG used after dbl-click on decode

       Asked for more details of exact Rx/Tx dial QRGs.

       Explanation appears  to show Tx QRG  out by 500Hz. I  can't see
       any obvious  way this can  happen. kb3mow has reported  that he
       can no longer reproduce this issue.

       On Hold.

Hamlib ANAN-100D (PowerSDR mRX PS v3.2.19) - Ton pa0tbr - split mode not working

       Not urgent as split mode not required due to wide band Tx available.

       Sent hot  fix version  of v1.4 with  mods for  SmartSDR TS-2000
       emulation.

OmniRig TS-590S - Randy k5rhd - OmniRig server going offline

	Issues with  clearing the error.  Probably the "Test  CAT" not
	always able to start OmniRig client cleanly issue.

	Sent a tracing version of RC2 - awaiting log.

Hamlib R8500 - Chris (UK SWL) airguzzler@btinternet.com - Hamlib IO error

       Probably port mis-configuration. Awaiting update.


CW synthesiser - Jim n4st - Distorted CW Id

   Sounds horrible, something is broken in the envelope shaper.

Commander KX3 - Neil kb3tvu - occasionally stuck in TX after normal TX period

	  Advised to  build a  CAT trace executable  and send  log for
	  analysis - awaiting log.

Log files - Wayne m0way - Would like links on desktop to log files

    This is probably a good idea, I'm not sure that the desktop is the
    best place for them but there  must be somewhere sensible and more
    visible  to  put  such  links.  Need  to  investigate  CPack  NSIS
    generator to find the best way to do this.

    A similar feature on Linux/Mac should be possible.

    Proposed for v1.5.

Waterfall - Jean f8rz - Requesting different labelling

	  Would like frequency ticks to be at absolute multiples of
	  500Hz even when start frequency is not such a multiple. E.g.
	  for start 200Hz ticks (and labels) at 200, 500, 1000, 1500,
	  ...

	  Proposed for v1.5.

Decoding - Jean f8rz - Abandoned Tx should not disable decoding for a period

	 This is an outstanding issue in at least v1.3 also.

	 Needs changes to Detector class  to allow audio buffers to be
	 not cleared by transmitting.  Complications arise in that any
	 data from prior receive periods  must be cleared. Might be as
	 simple as clearing the audio buffer after decoding but may be
	 complicated  when  decoding   overlaps  next  period.  Double
	 buffering may  be an  option. Question:  is the  audio buffer
	 needed after the last symspec call?

Decoded text - Albert kc8kjf - Report or grid last character truncated

	This occurs with larger fonts sizes on long decoded text lines
	e.g.  two   6  character   callsigns  and   a  R+   report  or
	locator. There is a defect  somewhere in the DisplayText class
	with font metric calculations. I'm not sure why we even try to
	truncate  the message  text given  that  it is  rendered in  a
	<pre/> element and therefore doesn't wrap.

	This  can  be  resolved  by  choosing a  small  font  size  or
	increasing the main window width a little.

	Proposed for v1.5.

HRD FTDX3000 - Jean-Michel f6eao/p - Not working

    This as  yet untested  rig with  WSJT-X +  HRD clearly  needs some
    adjustment to the initialization in the HRDTransceiver class.

    Requested the "HRD Interface Information.txt" file for analysis.

Multi-instance support - Bill g4wjs - Lock mechanism needed to avoid duplicates

	       There is no check that multiple instances are being run
	       with the same rig-name.  Some mechanism to control this
	       is desirable.  A simple  mechanism that isn't  prone to
	       false positives  (i.e. not relying on  lock files being
	       deleted on exit) needs to be developed.

	       Implemented in v1.5. - Closed.