ChangeLog 2000


Jan 2000

Jan 1 to 10:

   - afoc_init - added shared memory percentage and /tmp/afoc.txt output
     during initialization process.
   - fixed broken year on DATE_OBS FITS header. format is now yyyy-mm-dd.
   - John asked for a list of all outstanding spex work to be done prior to
     telescope installation. Here is my input:
      000106.tony_email.txt.
   - Added into the calmir code support for controlling the baytech RPC3
     power switch.

   - Many fixes on the Spex Facility web site.

Jan 10-14: Spex CoolDown in irlab.

   - I did not install the new version as to not contribute to any problems.
     (Will debug motor code and install new version after cold test).

   - I sampled temperature data from the Jan 10 cooldown. There is the
       spectrograph's data (with graph).
     The guider tc330 program had an error so it only sampled a
     few points (data).

   - I took new ArrayFocus Calibrations for the array focus mechanism
     during the cool down.
     Applied these values to the spex software.

Jan 18-21:
   - Applied Sun's cluster_patch dated 1/1/00 to bigdog & guidedog on 1/18.
   - showing move decimal values in DV display for smaller data values.
   - Fixes to the XUI's STOP function.
   - Worked on temper cntl bugs: Sim display message & guidedog wasn't working.
      (However tempcnlt code needs work - separate rcp.tc330 for each controller )
   - modifed calmir so out/in position are: 252,000 & 10,000 (from 256,000 & 800).
     hopefully these position are farther away from the limits.
   - Fixed SmartMotor_Not_Powering_Down problem.
   - Recompiled and Install as new spex version 1/20.
   - Tested Array Focus mechanism.

  Here is my email on the installation of version 2000.01:  000121.tony_email.txt.

Jan 24-28:
   - Add TCS communications in the IC
   - Added code to make the 'Set Position Angle to Parallactic Angle' possible.
     Button & TCS communction is there. Needs postion_2_Angle calculations.

   - Found a problem with the use of message queue. Fixed in spex & DV.

Feb 2000

Jan 32-Feb 4:
   - Worked code to support the ParameterSave & ParameterRestore command.
     Working with some spex parameters, but I need specific command to be restore from JohnR.

   - configure the IC so that it starts up the ldisplay program.

Feb 7 - Feb 11:

   Monday - worked in irlab. Install the video splitter and monitor switch.
            tested 2 DV on one system (using different port numbers).
            Made new user account (spex) on jeans, bigdog, guidedog.
            Recompiled and installed new version ( 0002 ) as current.

   Tues - off (RUTA for Jan Sunday Drill)

   Wed - return to office.

   started learning IDL again..  As a learning exercise, will try to write an IDL
   version of the xgfits.

Feb 14 - 18:

   Finished my IDL version of xgfits.
   archive source as spex.0002.tar.Z
   This is the version used for the Feb 17 Spex Cold test using Actual Science Arrays.

   Starting version 2000.03:
   Will start on the 206 mux modifications.

   tested pattern preserve mask in ADS21060 simulator. We can use AND and XOR to implement
   the preserve mask. Some cost to array readout speed on slow mode. Fast mode is unchanged.
   Does require pattern file format change.

   Modifed cpat & dpat tools need to create and view patterns to support the new spex pattern format.

   Started re-writting the pattern definition '.p' files to implement the new clocking scheme.

Feb 21 - 25:

   Short week due to holiday(21) and helping Phi pull fibers at new building on thurs(24).
   Continued with DSP Clocking / Buffering
      - finished making new clocking patters.
      - Hi Lib mod/new stuff to support multiple-overlapping subarrays:
          Load Pattern routines modified for new patterns.
          New HiCvTable functions to help manage clocking & sorting
          HiCkLoadTable rewritten to create new sequence tables.
          The libHi.a compiled.

March 2000

Feb 28 - Mar3:

   Continued with DSP Clocking / Buffering
      - change execute_pattern.asm code & adjusted timing parameters.
      - change all bcard arc_*.asm function parameter from (npixels, reps, addr) to (nconverts, reps, addr).
      - change all bcard cds_*.asm function parameter from (qwid, qhgt, addr) to (nrow_pair, cvt_rowpair, addr).
        The old code counted pixel in individual subarray. Since subarray can overlap, we need to count
        by just using number of convert in a array readout.
      - change the interface structure between Host and DSP to pass new variables.
      - changed all DSP .c files - compiled all DSP code & libHi.a

    Worked at IRLab Manoa (2 & 3rd): Continue the new DSP code changes:
      - Able to clock. Unable to sample anything yet.

Mar 6 - Mar 11:

   Make a new version 0002a. Basiclly its version 0002 with updated menu
   name for the OSF, GFilter, and Grating Turret. Also Grating position updates.

   Continue with version 0003 (new DSP code).
     - Added fixes for calmir/lamp (new QTh&IR menu item, XUI icon value bug fix).
     - Able to image using MUX206 code in diag using D16 basic mode.

   Attended Cisco course on Mar 8,9,10.

Mar 13 - Mar 17:

    Continued with DSP work: Added diag MUX206 code to IC, worked on simulation
       mode support, and scram/unscram data functions.
    Worked at IRLab Manoa (15,16,17):
       tested&debuged 206MUX in D16 basic mode & Movie Mode on BigDog - Everything seems to
       be working.
       added RemoveExtra206MuxData, SyncFocusToGrating on the GUI.
       compiled the guidedog executables.
       Small DV changes: Image Parameters were improved (BoxScale).

Mar 20 - Mar 24:
    More fixes: NAXIS1 sort on D16 movie, SyncArray2Grating OK, Qth&IR lamp Ok,
     Save&Restore includes lamp, DV truecolor ps output OK, New Icons for GUI/DV.
    Change afoc from menu to dialog box on XUI.
    Tested MovieMode with Jim Eliot - Test went well...both systems worked.
    Tested 206 code in GuideDog/basic mode..seems OK.
    Ready to install v2000.03!
    Movie Mode now displays the last 'image' in each file on the local display.

Mar 20 - Mar 24:
    Update Mechanism.doc.
    Added 206Explained in UserGuide Part II.
    Added parallatic angle calculation in rot_usePA.

Mar 27 - Mar 31:
    Wrote support code for 208 temperature controller (simulated only as
     guidedogm was off-line all week). Also changed the way the temperature
     data is handled for tc330 & tc208. All value sampled in littledogm using
     new tcmonitor program.  Code was tested on venom.
    Update GTK on rayleigh, bigdog and guidedog to verion 1.2.7.
    Recompiled all code.

April 2000

Apr 03 - Apr 07:

   Updated IC dictionary of commands.
   Began preparing littledog replacement disk with 2 boot partitions: Copy of
      current littledog disk, and update linux (sw70).
   Finished rpc, IC, XUI, and tool code for 208 temperature control.
   Installed new spex version 2000.03 as current on Apr 6.

Apr 10 - 15:

   Code Fixes to 2000.03:
      XUI icon status for lamp/calmir not correct.
      Added tc208 infor to fits header.
      Added HIST01 to HIST12  keywords to fits header for Mike C's IDL code.
      Update grating and array focus position per J.Rayner.
      -- compiled -- 11Apr2000.
   Starting v2000.05 (engineering run release):
      Fixed DV 180 degree rotateImage bug.

Apr 17 - 21:
   More fixes:
      - Added Parallatic Angle Summary (this requires constant TCS polling).
      - Redesigned Rotator Dialog box.
      - Position Angle is now used to position Rotator.
      - SyncFocus2Grating implemented in a stronger enforcement (to force sync policy).
      - Mechanism update were (incorrectly) refreshing widget on XUI - fixed.
      - Added platescale & position angle information to FIT files, and modified
        DV to use FITS position & Angle in its offset function.
      - Misc stuff: tc330 reversed on XUI; Grating menu names.
      - OBS mode needs a pair AB like nsfcam.
      - Recompile v2000.05 as engineering version.
   21st is a holiday (good friday).


Apr 24 - 28:
      - 24th Manoa Travel: Backup computer and prep for shippment.
      - Spex is backup & shipped to the summit.
      - Software change during shipment:
         + Putting in engineering restrictions.
         + Started setting up accounts for Bigdog,GuideDog on summit.
         + Compile/Install GTK development libraries + theme support on summit: stefan, max, wien,
              kelvin, helmholtz. (Not boltzmann, it is lacking some development tools)
         + Install DV, Spex Tools' on summit machines.

May 2000

May 1 - 5 & 6/7 (Sat/Sun):
   - Software change during shipment (continues):
      + Added call-back for focus-out event for entry widgets on DV.
      + Changed the calmir to accept hotdog hostname from littledog shared memory.

    - Spex arrived at the summit:
       + rayleigh's harddisk failed. Brought it backup down to hilo and
         replaced the IDE 4GB disk with a lab spare. Re-installed OS.
       + Bigdog SCSI disk experience some SCSI warning. Ordered new disk.
       + Other computers (guidedog, littledog, 2-rpcs) are OK.
       + Reconfigure systems for summit use.

May 8 - 12:
   + Rayleigh fixed and installed as planck on summit for spex testing.
   + Spex & MIM electronic moved from prep room and installed on MIM (9th).
   + Moved bigdog/Guide from prep room to TCS room. Cable up spex for
     its final configuration (ethernet, fiber, power, keyboard, video). (10th).
   + Finished cabling/Moved Planck. Got spex software integrated & running in observers area. (11th).
   + Spex Party in Hilo Office on the 12th.
   Software changes:
      + Added call-back for focus-out event for entry widgets on XUI.
      + Add TCS offset commands to guider.

May 15-19:
   Spex engineering night are 16,17,18 (tues-thur).
   + Install new software version (2000.05) for engineering run.
   + I say the run went well for 1st time. See
     John's lukewarm email
     and
     Bill's more upbeat email
     I like Bill's slightly better point-of-view.
   + On the last night, John reported an unrecoverable crash of the bigdog/guidedog
     computer. Went to summit on 19th to diagnose problem. Eventually found the
     vme bus resets were electronically connected via the reset button on the
     VME crate/power supply. Peter advised Imai to disconnect.

May 22-26:
   Spex engineering night are 24,25,26 (wed-fri).

   + On the 24th, the hard disk in littledog failed. Was able to
     replace the disk/OS using spare disk and recent backup.
   + Installed new spex software version (2000.05.24) for engineering run.
   + New position angle code for rotator position, DV offsets,
     and slow guiding is working! New step-to-positionAngle develop
     on 1st engineering night.
   + New dialog box implemented for lamps/shutter/calmir. And littledog can
     set/monitor rpc for HotDog (lamps & shutter).
   + Changed vf/Dv/xgfits so they all used the same data file: /tmp/xgfit.dat.
     New dv, vf, & xgfit compiled & installed on summit suns.
   + command 'rot 180' & 'shutter open' commmands issued error: fixed.
   + recompiled all software on the 25th.

May 22-26:
   + Monday Hoilday & I took Thur/Fri off (RUTA) as comp-time for summit support.
   + Added support for B-Beam Guiding with GuideBoxB and TCS query for beam position.
   + Fixed bug in Parallactic Angle code.
   + Changed ShortXD (menu 0) step position from 0 to 719600.
   + Added FITS zero points information for Guider filter.

June 2000

5th-9th:
   + changed the method for controlling hotdog outlets to allow multiple
     lamp & shutter commands by removing rpc3_set, and letting the rpc3monitor
     program compare requested outlet vs real outlet, then setting them.
     It will set calmir busy until all outlet match.
   + compiled & installed new version (v2000.06) for 6/5 observing. This
     version worked very wlll. Hotdog communication fixed & AB guiding works!

   + installed summit photo on 2000 2nd quarter Spex news.
   + Starting 'to-do' list for next run.
   + dv:
      - Made XOR line thicker for objbox on images.
      - Beamswitch & offset button too close - re-arranged widgets.
   + slowguide:
      - removed rotation angle. Use rotator's position angle during guiding.
      - Guide AB changes: A bean to ignore offse when in B beam.
      - local display is always TRUE during guiding.
   + added beamswitch widget in setup. Default: guidedog=2, bigdog=5.
   + get save_parameter--restore_parameter working - done.
   + installed gtk (plus all support libs) & dv on minerva.

12th-16th:
   + fixed up ldog_show to work on 80 column xterm for lars.
   + Summit guest account: modify guest account for observers.
   + Worked on 'to-do-item' from john rayner's email
     Besides some fo the above fixes, John suggested the following:
      - Make dtime & slowcnt an obs parameter on the XUI OBS page.
      - Added set subarray buttons to DV's TCS panel
      - Fixed FITS header - had 2 END statements.
      - Itime also indicates minutes remaining.
      - Issue beep at completion of integration.
      - made autosave message yellow - talk to john about his comment.
    + setup guest accounts.
    + updated IC command dictionary.
    + new autosave taps to easily notice state of autosave.

19th-23th:

   + moved macro on bigdog/guidedog in ~/Oldstuff directory.
   + Recompiled and Installed Version 000623 on BigDog/Guidedog/LittleDog
     computer. Theres are now the current version as of Jun 20. 
   + changed the account 'spex' to 'littledog' on the littledog PC.
   + Added support for bad pixel mask for guiding.
   + updated the DV command dictionary.
   + Recompiled spex for 6/23/00 observing.

26th-30th:

   + recompile xgfit on bigdog/guidedog.
   + various tweek request by B.Vacca:
      XUI: Moved guidebox A/B outside of subarray tabs
      DV:  remove comma on TCS offset To/From.
      DV:  changed 'DivByCoadds' to 'DivByDivisor'.
      xgfit: recompiled/installed  (came up with wrong filename)
   + Added coadds to changeable slowguide options. Need testing.
   + message on XUI ask to check IC status for IPC errors..
   + changed defaults bgr.flag, ms, min_ms, cnt to ON, 1000, 1, 800 (20000ns).
   + XUI's 'Go.Init' just prints a message telling users to run array_init macro.
   + Recompiled spex for Fri night's observing.

July 2000

3rd-14th:
   + 4th is holiday, Military Lv on 6,7,11,14th.
   + Added 'AutoGuideBoxSetup' command and Button. This will set the
     GuideBoxA/B and issue the ABOffset command to the TCS based on the
     slit and rotator. Plus it outline these item on the DV image.
   + Need to added DrawLine and DrawBox command to DV to support
     the AutoGuideBoxSetup command in spex.
   + slowguide's FullImage sents an image to BufA ( was BufD ).

July 10:
   + Installed new spex menus on guest accounts (BigDog/GuideDog).
     New menu items to launch XUI, DV and xgfit were added.
   + Added Slit, ndr, and slowcounts variable to spex cal_lamps save/restore command.
   + Changed GuideBox inputs on XUI to allow Center XY and wid/Hgt to be
     changed in separate prompts. The old way was: Upper-Left X,Y,Wid,Hgt.
   + Default NDR are 32.
   + Grating Menu Changes:
        LowRes renamed to LowRes60 ( step = 178150).
        New menu item is  LowRes15 ( step = 179350).
   + Recompiled spex for night time observing.

   Bill Vacca's email: Suggesting UI improvement for the guide modes.

17th-21th:
   + Mon: In irlab Manoa:
      fixed but in mylib's timeStr2sec() with didn't handle negative convertion
      if the hours are 0 (-00:xx:xx). Caused problems with parallatic angle with
      negative hour angle less than 1 hour.
     Recompiled spex for night time observing.
   + Tue took vaction.
   + Wed: In irlab Manoa:
     Fixed bug in 'AutoGuideBoxSetup' - Was using rotator angle, not position angle.
     Adjusted guideobox per Bill Vacca's email:
        New guideboxes for 0.8 slit position:
           ax = 215; ay= 215; awid=30; ahgt=30;
           bx = 216; by= 278; bwid=30; bhgt=30;
        New guideboxes for 0.5 slit position:
           ax = 215; ay= 212; awid=30; ahgt=30;
           bx = 216; by= 275; bwid=30; bhgt=30;
     Recompiled spex for night time observing.
   + Thur took vaction.

August 2000
   Spex was not scheduled due to repolishing of lens by manufacturer.
   Next spex run is Aug 28.

   + George install the fixed TPRO timerboard into bigdog. He stated it
     did sync to GPS time when the time signal was hooked up.
   + dv: draw guidebox line suing middle of pixel, not upper-left corner.

September 2000

  Per John Rayner's email:, the following was changed.
      + New autoGuideBox setting. (see email).
      + SXD pos changed from 719600 to 719750.
      + bm.set command works.

   Compiled and installed Spex version 0009 on the summit on 9.11.2000 for
   that nights observing.

   Work on changes for next version:
     + AutoGuideSetup command changes:
         - do not send beam switch offset command to TCS if GuideAB==A.
         - Sends center (x,y) of BoxA & B to DV's From(xy) and To(xy) on the TCS offset page.
         - Also updates the rotation angle - this cause the offset to be calculated.
     + Stop peeking into DSP RAM during movie modes (can cause acquisiton problems).
     + GPSTime ON is now the default.
     + show 'Macro in execute' on status window when running macros & beep when done.
     + added tcssim for debugging purposes.
     + Improved the estimated time remaining in a go. Previous didn't account for
       beam switch deadtime and obsmode AB.
     + Change FITS header:
          - deleted IDL line previously requested by Mike Cushing.
          - added calbox RPC3 informations (lamps & shutter).
     + Movie mode will now display the last image of each file on the local display.
     + Fixed sorting bug in Big.Movie mode when using subarrays.
     + recompiled all bigdog/guidedog/littledog source code.

   9.26 - Installed version v2000.09.26 as current version.

   We had a question of Itime & Efficiency. Here is my 000925.tony_email.txt
   describing a test using bigdog & my findings.

October 2000

   A SPeX occultation (001022.elliot.txt) from Jim Elliot/Amanda Bosh.

   - IO process has it own message queues so you should always be
     able to die using io task.
   - ParameterSave/ParameterRestore - add new parameter to list:
      guideab, filename.
   - Using slowcnt == 2 resulted in error - fixed.
   - check subarray bug in G.Movie mode - fixed.
   - Display cycle time of Movie files.
   - Improve the handling of xui message queue to reduced the 'not xui msgq' error.
     Also a new program 'delete_xui_msgq' is available on bigdog/guidedog to
     delete any allocated message queue. Use this rather that stopping/restarting IC.
   - lamp/shutter command have problems registering busy before the GO. - fixed.

An spex.0010 release was done to fixed the lamp/shutter problem. This was a pre-release
to version spex.0011.
   - fix bug with negative request for rotator ( reported by vacca ).
   - The command 'current/ic/tproShowTime' shows the Unix Time vs GPS time.
     If the GPS & System time is not sync, a warning message is display on the XUI.
   - Now we can support a 2nd DV for remote or simultaneous operations.
     The setup page has configuration option for dv1 & dv2. This also caused
     other widget to be moved (tcshost, ldhost to eng).
   - Movie mode - display the last frame of the movie (not first).
     check to make sure it working using subarrays.

November 2000

Spex version 0011 install as current on nov 1. Incorporates all changes listed above.

Minor fixed to version 0011
   + After INIT warm cal mirror should be IN & Dichroic OPEN.
      - This requires the cal mirror be initialize by moving toward the IN limit (pos=0).
   + Added log message for go.init to report some DSP variables.
   + Guide 'OffsetsTO' renamed to 'CorrectionTo'

Updated Spex version 0011 nov 11. Incorporates changes listed above.

December 2000

  No Changes - was national guard & vacation leave.