Last updated: 29 January 2001

Autostar Updater 2.3 Bug List

From: rick@pinefields.com (Richard B. Emerson)

This is a description of problems encountered with Autostar Updater version A2.3 ("A2.3"). There are also some comments about Autostar firmware 2.1Ek ("2.1Ek") but this document does not primarily address problems with this firmware except as they relate to issues with A2.3. Throughout this document are recommendations (marked "RECOMMENDATION: " at the top of the paragraph) on how to address the problems raised in the ITEM: paragraphs.

The platform used for the procedures described below is an IBM ThinkPad 770ED running Win98, patched to date, 128M RAM and 5G hard drive – approx. 1G free space. All downloads are direct from Meade and not third party sites except where specifically noted. The author of this document, Richard B. Emerson at rick@pinefields.com.

ITEM: Autostar Updater version A2.1 ("A2.1") was removed from the host platform before installing A2.3 which was obtained from the Meade web site. All registry keys applying to A2.1 were also removed. On completion of installation, my Autostar was connected via a third party serial cable (supplied by ScopeTronix). A "safe load" as described in the text file supplied with A2.3 was done to force the new firmware into the Autostar. I did not load any objects (comets, asteroids, or satellites) or tours not supplied in the A2.3 distribution. At the end of the load, I used the Autostar utilities to determine there was 37.4K free in the Autostar RAM.

I initialized the Autostar by entering a custom location (Lansdale, PA 40-15N, 075-18W, -5 hours relative to UTC), disabled the Sun and start up displays (Sun warning and “Getting Started”), and set the display brightness and contrast.

I then re-started A2.3, now with expert mode enabled by setting the EXPERT key to 1, and set the time of day from my platform. That worked (as confirmed by using MODE to display local time of day and LST), however there is no indication from A2.3 that the event ran properly.

RECOMMENDATION: A "Time set successfully" prompt should be displayed. There is further discussion of the date/time setting function below.

Also, the Initialize, Version, and Date/Time functions should be available to users. None of these items is likely to damage an Autostar load and can be helpful during loading objects. Presently the Delete Data function also does not confirm successful completion of its task; a suitable message should be displayed. The Delete Data button is capable of causing problems and should be protected with "Really remove all data?" warnings and "Type the word 'YES' to continue" locks. In some cases, where there are too many objects in an Autostar, it really is best to delete all data sets and start over. For that reason the Delete Data function should not be completely concealed from users.

ITEM: There is no discussion of when any data set or ephemeredes is replaced in or removed from an Autostar. It's unclear whether uploading ephemeredes from A2.3 overwrites all data sets in an Autostar, only replaces existing items and writes new items, or only replaces new items. For example, assume an Autostar has elements for Mir stored in it and a new TLE is installed in the A2.3 library. On doing an ephemeredes upload, is the exiting data set in the Autostar replaced with new data? I assume this is the case but there is no clear statement of how the data replacement process works. Mir will be de- orbited around 6 March and this presents another issue for maintaining Autostar data sets: how are individual or groups of objects removed from Autostar memory?

RECOMMENDATION: Clarify the rules of data maintenance in Autostar and Autostar Updater documentation. Add the ability to delete both single objects and groups of objects if this capability is not present.

ITEM: On querying the Autostar for version information, I got back this ROM ID: |A|21|Ek Dec 12 1992 @ 10:32:38 This is a minor point, but note the error in the date. RECOMMENDATION: The ROM version date should have the proper date.

ITEM: Exiting the expert mode panel, I attempted to upload the current (that is, supplied by the Meade web site) ephemeredes for all objects. The comet and asteroid data collection process did not produce error messages but the satellite data collection stopped with a message saying "Can't get NASA Special Interest Group (SIG) from registry". I selected Interesting from the group list in the display and hit OK; there were no further error messages.

Either the selection of a SIG should be part of setting up for data collection or a default value should be present in the PC's registry, loaded during A2.3 installation. Additionally, the satellite SIG selection process should allow selection of multiple SIG's instead of the current limit of only one SIG.

There are two problems with tour uploading. First, there seems to be no way to select multiple tours, for uploading, from the Meade tour source; the present scheme allows for the choice of only one tour. Second, selecting a tour of objects seen in the spring skies seems to point to a Messier tour and there is a warning the Messier tour is already present.

RECOMMENDATION: The tour selection process should allow the selection of multiple tours without restarting the entire ephemeredes collection process for each additional tour. The apparent mismatch of items seen in the Meade tour source site should be resolved as appropriate.

Also, the Meade ephemeredes web pages are generally in need of updating. For example, there are still references to ST-99, a Space Shuttle mission that has been over for a long time.

ITEM: With the ephemeredes collection process completed with the errors or exceptions noted above, I proceeded to upload the new data to my Autostar. As none of the exceptions seemed to be fatal, I felt safe in doing so.

The comet and asteroid ephemeredes uploaded without comment. The "Available [space]" display, however, began to show values that made little or no sense. The three or four digit values counted back through zero and then were replaced with either eight digit numbers or negative seven digit numbers. It appears that the counters are kept as long integers that have rolled over past zero. How valid the displays are during upload is, in my opinion, anyone's guess.

The satellite ephemeredes upload process appeared to be working well until the following message appeared in an error message display: "Failed to write from file c:\program files\meade\Autostar\autostar update\Ephemerides\LibSatellite.ROM at record 49". There is no choice but to acknowledge the message with "OK". At that point the upload seemed to upload tours (as indicated by a progress bar, listing tour names). At the end of the process, the number of items in the library matched in all cases except the satellite count that stopped at 49 instead of over 130 satellites.

RECOMMENDATION: There should be a prediction of “resources used” which prevents the selection of too many objects for the space available in the Autostar. It appears, from later experiments, that this feature is already present in the tour editor function of A2.1.

Restarting my Autostar, I then examined the object lists for comets, asteroids, and satellites without incident. When I selected Guided Tour, however, line 2 of the display was blank and, at that point, nothing would move the Autostar out of Guided Tour mode. It was frozen and only powering down the telescope and restarting Autostar cleared the problem. Returning to Guided Tour produced the same effect.

On reconnecting the Autostar to A2.3, the "In Handbox" display showed 49 satellites loaded and only 1 tour which, during data collection from the Autostar, displayed a name made up of nonsense characters and not the expected "Tonight's Best".

I used the expert mode panel to delete all object and tour data from the Autostar, exited A2.3 and deleted all LibXxx files for all objects and tours. I restarted A2.3 and did a software upload which installed zero comets, 58 asteroids, 42 satellites, and 3 tours. There were no error messages and the three tours displayed properly.

ITEM: It appears that, in assembling the memory images for upload, there is a problem with detecting exhausting memory resources. Although an error message is generated ("Failed to write...") during the upload process, it appears the memory image being uploaded is defective and there is no termination of the upload. As a result, a defective memory image causes problems when attempting to use anything uploaded "downstream" from that error point.

RECOMMENDATION: A2.3 should be able to at least detect memory resource exhaustion as soon as it occurs and not during upload. A2.3 should not be allowed to continue an upload that installs a significantly defective memory image.

ITEM: Two other problems were encountered in using the expert panel. First, the Properties button produces a GPF every time it's clicked. This function works, albeit poorly, with A2.1.

RECOMMENDATION: If the Properties button is no longer functional, it should be removed or a code stub and a "Not supported" message should be used to prevent the GPF caused, I assume, by defective or "under development" code.

ITEM: Using the Date/Time button causes the Autostar to initialize in some cases. As Autostar initialization resets its clock to the last known date and 20:00:00 local time, the uploaded date and time are lost.

RECOMMENDATION: Since A2.3 knows when an initialization is triggered by A2.3, it should either disallow sending date and time if Autostar is not ready to accept the data or warn that the data was lost during the forced initialization.

ITEM: There are a number of issues which, while not affecting the Autostar directly, diminish the overall Autostar Updater / Autostar package's usefulness.

There is no way to handle groups within a single object. For example, assume that the satellite ephemeredes include a number of rocket boosters. There is no way to manually mark all of these objects and delete them at once. They must be deleted one at a time, through a time-consuming process. A related problem is the inability to mark a number of library items for inclusion in an ephemeredes update. Again, each object must be marked individually through a multi-step process.

RECOMMENDATION: The editors should be improved to allow a way to process groups of objects at one time. Standard Windows SHIFT-click, CTRL- click, and click-and-drag procedures should be supported in editing the object list. (The drag-and-drop feature does allow adding files; this feature appears to work correctly with tours but, because of problems with adding objects, was not tested with TLE's and comet and asteroid files).

ITEM: If all object ROM files are deleted, the object editors only have the object "<ADD NEW>" in a list. Closing the editor produces an error message, saying <ADD NEW> is an invalid name because names cannot begin with "<".

RECOMMENDATION: The editor code should recognize <ADD NEW> as a special case and proceed accordingly.

ITEM: The 2.1Ek ROM images do not include Polaris as an alignment star although it is present in 2.0h's alignment star list. While I understand the issues surrounding using Polaris to align a scope, having it in the list is useful in several instances including helping a new user to get a good idea of where Polaris is in the sky. Since north orientation is important in aligning in even AltAz mode, knowing where Polaris is has more than minor value. Removing Polaris from the alignment list is a very unfortunate decision.

RECOMMENDATION: Include Polaris in the alignment star list and include a note in the firmware documentation stating why aligning to Polaris can produce aiming errors over time.

ITEM: There is no way to operate any Autostar Updater if an Autostar is not connected. The advantage of being able to operate Updater without an Autostar is that it's possible to examine library contents and test for resource use or exhaustion (see above recommendations) without altering Autostar contents.

RECOMMENDATION: Allow Updater to inspect library contents without requiring an Autostar to be connected.

NOTE: I cannot stress enough that the problem which begins with the "Failed to write..." message effectively makes the Autostar unreliable in that, should the user scroll to Guided Tour mode, the Autostar will stop working. This is a serious problem that cannot be worked around. While it's possible to view some of the recommendations listed above as "cosmetic changes", Autostar Updater, as seen in A2.3, is a program that is essentially unusable but is presented in the Meade web site as production quality software. Moreover, because of A2.3's many problems, there is the very real possibility that customers, assuming the software is serviceable, will use it with very adverse consequences which include rendering an Autostar controller unusable. This reflects poorly on Meade and its commitment to its customers. I strongly urge that Autostar Updater be given the attention it so badly needs.

-# 30 #-

Return to the top of this page.

Go back to the Autostar Information page.

Go back to the ETX Home Page.

Copyright ©1999-2001 Michael L. Weasner / etx@me.com
Submittals are Copyright © 2001 by the Submitter
URL = http://www.weasner.com/etx/autostar/as_au23buglist.html