Known bugs (and workarounds, where possible) in the Observing Tool are described here. If you find a new problem, please submit the information via the Gemini HelpDesk and include any stack trace (by default Windows does not give a stack trace).

4 December 2009
Known bugs in 2010A.1.1.2:

• In the Windows version clicking on the Erase button in the Position Editor causes the OT to hang. Delete guide stars using the Target Component in the Program Editor
• In Repeat and Observe components the repeat (X) entry box does not properly save number that are typed in the box. Use the up and down arrow buttons to change the number of repeats. Or, type a number in the box and then click the up arrow, then the down arrow.

 

24 September 2009
GMOS and Michelle component bugs in 2009B.1.1.3:

• There is a bug related to the GMOS nod-and-shuffle tab that can prevent a program from being exported to XML. This is fixed in OT 2009B.2.1.1, unfortunately, this bug can appear during the update process. If it occurs there is a way to recover as long as you don't delete anything in the directory ~/.jsky/spdb. If you have a problem with a program then please submit a HelpDesk ticket and we will assist you.
• In echelle and medN modes, an entered central wavelength in the Michelle static component will often revert to the default of 10.5 microns. One way to help insure that the correct wavelength is used, even if the value in the static component appears wrong, is to add a Michelle instrument iterator to the top of the sequence and add a line with the central wavelength. All observers should be place underneath this iterator. This problem is fixed in the 2009B.2.1.1 version of the OT.
• The observing wavelength does not reset in multiple passes through a Michelle instrument iterator in which the filter is changing.The work-around is to add the Disperser (set to Mirror) to the Michelle iterator.

05 February 2007
Avispa 2.1 fetch attachment bug:
OnSolaris and certain Linux flavors (e.g. Fedora) there is a problem with the feature to fetch attachments. The files will appear to download but the files will not appear in the selected directory.  The problem is related to operating system restrictions on renaming files across file systems (disk partitions).  The easiest work around is to save all attachments to the /tmp directory and then manually move them tothe desired location.  The problem was fixed in the 2007B OT.

23 June 2006
Abeja 3.3 GNIRS and GMOS-S bugs:
In the GNIRS static component the number of coadds cannot be saved in the static component, the coadds must be set in a GNIRS sequence iterator. Also, changes to the PA in the static component cannot be saved. The PA can be changed by dragging the slit or IFU field in the Position Editor.  For GMOS-S, the visualization of the IFU nod& shuffle focal plain units in the Position Editor is incorrect andthe planned time does not appropriately account for the binning. All of these problems are fixed in Abeja version 3.5 released 30 June 2005.

29 June 2005
Palote 3.1 not compatible with Fedora Core 4 Linedistribution:
We have discovered that the Sun JRE included with the Palote_3.1 Linux distribution is not compatible with the Fedora Core 4 Linux distribution. The symptom is that the fetch/store fails immediately with a stack trace starting as follows:

ERROR [SwingWorker] 2005-06-28 18:16:28,748:  edu.gemini.oodb.OodbClientBase - Connection problemjava.net.SocketException: Invalid argument or cannot assign requested  address...with several more lines...

This is a known problem with the Fedora Core 4 people. The FC4 installation notes suggest that users upgrade their JRE to the latest 1.5_04 distribution from Sun at http://java.sun.com.We've tested this and confirmed that it fixes the fetch/store problem. After additional testing, we will make a special FC4 version availableon the Gemini ftp site.

 

9 January 2005
RA values considered to have units of hr:
The units for RA are assumed to be hr (hours). Values greater than 24 are'unwrapped' e.g. 243.7, intended to be in degrees, is interpreted as 10 revolutionsplus 3.7hr and is converted to an 'RA' of 3h 42m.

 

10 Dec 2004 - fixed in public re-release
Sequence not copied when modified observation merged:
After fetching a program and modifying observation sequences, the changes arenot correctly integrated when the program is stored to the Observing Database.This only affects Grillo-01.6, which was the initial public release availablefor about 12hr overnight on 9 December, and is fixed in subsequentreleases. 

 

4 Feb 2003 - fixed in river-10 internal release
Offset sequences not preserved when copies are made:
This occurs when an offset iterator is copied from one observation to another and then one of them is edited. The problem is that a new copy of the list is not being made; it is being shared between the copies. This bug also manifests as the apparent 'disappearance' ofoffset and nod-and-shuffle steps when exporting and re-importing saved (XMLfile) copies of the science program. 

 

20 January 2003 - fixed in river-10 internal release
Image quality observing condition not loaded correctly:
From semester 2003A, the previous image quality observing conditions (20%, 50%,80% and "any") were changed to (20%, 70%, 85% and "any").Some science program or Phase I proposal created with OT or PIT from 2002B orearlier and imported into later versions of OT are not correctly translated,e.g. 80% is interpreted as "any". The image quality observingconditions should be checked and changed manually.

 

11 July 2002 - fixed in OT river release of January 2003
Offset iterator steps are defined in wrong axis forrectangular grid:
In the telescope offset iterator, a rectangular grid of steps can be createdusing the grid pattern tab. The number of steps specified in the paxis (the upper button) incorrectly produces that number of steps in the qaxis (and vice versa). This can be seen by inspecting the iteration table orvisually by viewing the offsets in the position editor. The initial offsetand spacing parameters are handled correctly. 

The pop-up tip (when the cursor is held over the step boxes) incorrectlylabels them as RA and dec. The relation between the [p,q] coordinatesystem and (RA, dec) depends on the defined position angle on the sky. 

 

17 April 2002 - fixed in OT limberlost release of 10 July 2002
GMOS OIWFS does not correspond to current hardwareconfiguration:
As it has an asymmetric configuration relative to the optical axis, the GMOSon-instrument wavefront sensor's patrol field will change when the instrument ismounted on the up-looking and any side-looking port. (The effective patrol fieldis flipped about the x-axis by reflection in the science fold mirror). Thusalthough the same guide star may in principle be used regardless of port, theinstrument may need to be rotated to acquire that star and the OIWFS probe armwill vignette a different part of the field. 

The current public OT release (OT2002A.11) assumes that GMOS is on theup-looking port; in fact the instrument is now mounted on a side-looking port.PIs should continue to define their programs using the public OT; their GeminiContact Scientist will notify them of any problems. 

 

5 February 2002 - fixed in OT river release of January 2003 (use thePosition Editor...Catalog...Proxy Server menu)
Guide star catalogue or digital sky survey searchesfail because user machine is behind proxy server:
If the user's computer connects to the outside world via an HTTP proxy serverthen PIT catalogue assistant searches, PIT proposal submission (UK, Canada and Australia)and OT guide star or image searches will fail, often with the error exception occurred while retrieving the data:java.net.ConnectException: Connection refused. This problem arisesbecause PIT and OT currently do not support sites with an HTTP proxy(support is being added for the next public releases of PIT andOT).

There is a workaround for those using PIT and OT under solaris or linuxoperating systems. (If you are using Windows, please contact us as there mayalso be a solution). The Java Virtual Machine (JVM) on Solaris and Linux allows setting two system properties at runtime to indicate that HTTP access should go through an HTTP proxy server.

The two system properties are:

• http.proxyHost -> The host name of the proxy machine
• http.proxyPort -> The port used by the proxy host

By setting the site specific values for these two properties before starting PIT and OT, the access for posting the proposal and gathering WFS stars will go through the indicated proxy. You may find the name of  your proxy by looking in your web browser preferences.

A modified command for starting PIT would look like thefollowing (where the -D allows setting the system property on the command line):

 exec java ${JFLAGS} -Dhttp.proxyHost="proxy host" -Dhttp.proxyPort="port" edu.gemini.phase1.tool.app.PIT 

An example PIT.sh script is available thatincludes two new variables (called PROXYHOST and PROXYPORT) and can be editedwith your local information. Similar modifications should work for the OT.

 

23 October 2001 - only an issue for semester 2001B
Bugs in Phase I - to - Phase II skeleton conversion.In the translation of approved Phase I proposals to the XML skeletons for PhaseII several errors were introduced: 

• In some Science Programs no instrument component (NIRI or GMOS) was included in each observation, in others the instrument was included but with default parameters. In both cases the PIs should define the instrument configuration required for their observations.
• The observing condition constraints often were incorrectly transcribed. The observing conditions included within the Phase I proposal summary element of the Phase II XML file are correct, but the conditions within each observation are wrong. PIs should check and, if necessary, correct the observation conditions. 

 

3 May 2001
In (first) public release OT1.0.0

• Help-page tutorials are those from old OT
• When WFS tags are deleted in the position editor they are correctly deleted from the (target component) target list but not from the offset iterator table