If you have a VStar bug to report, please feel free to post here.
You may want to check:
- that no-one has posted the same report here;
- the GitHub bug tracker to see whether the bug has already been captured. Ultimately, this is where bug reports posted here will migrate to.
VStar has been updated again to fix this. Please try launching VStar and you should see that it is now version 2.21.1 and that the B-V plugin works.
In preparing for a new session of my VStar course I load many different data files in one vstar session. After something more than 20 or 30 file loads I get an out heap memory overflow (something like that). I didn't take a screen shot. VStar crashes and I have to end it with task manager and restart. It has happened 3 times in the last 3 days. I know this is unusually high usage of the program but it appears to have a memory leak. It is easy to me to avoid if I remember to simply close the program every few hours and restart it. The tricky part of that starement is "If I remember ... ." This is probably something you already know about but I didn't see it mentioned in this forum.
Thanks for your bug report. The short answer is that this is a known problem that has been addressed in some ways but I completely agree, it is not yet a solved problem and needs more attention.
There is an old ticket for this but I've just created a new one:
There is at least one simple way to improve the scenario you are faced with: increase the amount of memory available to the Java Virtual Machine when VStar is running, or "heap" size. This is mentioned in the ticket above. I can say more here or contact you via email about this.
I think VStar has some memory leaks. After sequentially loading analyzing and closing a number of data files, annecdotally it seems to be more than 20 or so, some with a few thousand observations, others with a few hundred, VStar seems to run out of memory as though it wasn't fully emptying its memory space when each data file is closed.
A screenshot of the error message when this occurred when starting a DCDFT on the next file
Today I've been working on and running a script that loads a few hundred datasets of at least 1000 observations per dataset, does period analysis, creates and saves a phase plot and saves this and light curve PNGs.
I'll have to compare notes with you in email on what you're loading compared to this etc.
I purchased aa new Mac laptop and get his message when trying to run VStar:
"VStar" can't be opened because Apple cannot check ot for malicious software.
The software needs to be updated. Contact the developer for more information.
I am using the latest (2.21.2) VStar download
Anyone have any ideas?
Apologies for the delay.
I'm running macOS Catalina. Hw about you?
Are you using WebStart or the dmg downloaded from here:
I'm guessing the latter. If so, in Apple Preferences -> Security & Privacy -> General, does "Allow apps downloaded from:" have "App Store" or "App Store and identified developers" selected?
What version of Java? I don't think that's the issue here, but just for comparison.
!. Running the dmg download V 2.21.2
2. Running Catalina 10.15.7
3. Running Java 8 update 261
4. App store and identified developers is selected
I dug out of a closet an HP laptop. I put Java and VSTAR on it. No problem getting VSTAR to come up. BUT! I installed the AAVSO Upload File plugin and when I try to use an Upload File i get the error:
Obsevation Read Error: No observations for specified period
I'm sure I am doing something wrong, but darned if I know what! A few years back I used VSTAR extensively. Thanks for your support and patience.
The error on HP you see may be due to not first selecting the plugin from the list in the file dialog, but using the standard simple/download plugin. You can switch back to the old way of having obs source plugins in File menu via plugin prefs to check.
On the Mac, have you tried ctrl-clicking the VStar file once unpacked from the DMG and then clicking Open? I wish I could see a screen shot of what you are seeing. Perhaps we could Skype if we still can't figure it out.
HP Prob: Mia Culpa. Wouldnt you know it, the upload file I chose for a test was without a check star! The field is very sparse. Upon choosing another upload file, all was well
Mac: Let me try a few things today. If I dont get anywhere, we can go to the next level.
again , thanks for your help and patience,
Do you do Zoom? I could send you a Zoom link and share the VSTAR screen with you.
There is also Teamviewer. With it you could take remote control of my Mac
Zoom is fine. Perhaps we can start there and go to Teamviewer if necessary.
What days/times are good for you? We can then match that up with my timezone (UTC+10.5).
I am currently UTC-4. On November 1 we will move to UTC-5.
I am retired so it might be easier for me to match your days. Given a few days notice I can probably match a day you choose. 03:00 UTC until 11:00 UTC would not work for me
I am trialling the Gaia DR2 Photometry plugin. First, I displayed the light curve for RR Ori from 2456698-2458198 JD, corresponding closely to a time span during which Gaia DR2 has published results for RR Ori.
When I try to add Gaia DR2 photometry to the light curve, the following error is displayed:
"New Star From Observation Source Read Error
Entries in the Gaia Load Parameters dialog were:
Gaia source_id 3349083629043128320. This was obtained from the Basic search on the Gaia Archive web page, with the target resolved by SIMBAD.
The error occurs whether or not one or both of the boxes "Transform Gaia passbands to V, R I?" and "Add to current?" are checked.
Since RR Ori/Gaia DR2 3349083629043128320 works just fine for me I guess we need to dig deeper.
Let me just make sure of a couple points on recreating this. Your earlier post mentioned a JD range. Are you entering a Vela Filter for JD range in the Gaia Load Paramters dialog? Also, do you get this error if you attempt to load this Gaia data immediately after starting VStar?
The other thing we should check is versions. Can you let me know what the top line of VStar's Help->About window says? (eg. VStar 2.21.2 (git revision 1cbfe07)) Next, please open File->Preferences, click the Plug-ins tab and let me know what the last part of the URL path says (eg. vstar-plugins-2.21.2).
Finally, have you been able to get any Gaia data to load?
I should have thought to look first at my version of VStar. I'll do that when I get a chance and get back to you. If I need to update my version, I'll do that, try again and let you know the result.
If I have the latest version, I'll look at your other questions.
I am running VStar 2.21.2 (git revision lcbfe07).
"Are you entering a Vela Filter for JD range in the Gaia Load Paramters dialog?" I don't know what this means.
"Also, do you get this error if you attempt to load this Gaia data immediately after starting VStar?" The error occurs on any attempt to load Gaia data, irrespective of time since starting VStar.
"Next, please open File->Preferences, click the Plug-ins tab and let me know what the last part of the URL path says (eg. vstar-plugins-2.21.2)." The path is https://www.aavso.org/sites/default/files/vstar-plugins/vstar-plugins-2…. The "load Pulins" box is checked.
And last, no I have not been able to load any Gaia data.
With that I can recreate the error. Here is what I think will fix your problem.
From VStar open Tool->Plugin Manager. It takes a few seconds to open.
Scroll down to the Gaia DR2 plugin. Highlight it. Click Delete. Close VStar with the red button that appears and open it again.
Open the plug-ins preference pane and change the last part of that URL from 2.20.0 to 2.21.2.
Open the Plug-in Manger again. Scroll down to the Gaia plugin which will now be near the bottom of the list. Highlight it. Click Install. Click the red button to shutdown VStar and start it once more. The Gaia plug-in should now work for you.
Let me know how that goes.
Following the recent website upgrade, I have downloaded VStar to my Mac (macOS Big Sur), but it doesn’t seem to want to download the plug-ins.
When I click on Plug-in Manager in the Tools menu, it says “Error reading remote plug-in information”.
I note that in the Preferences / Plug-Ins Tab, the location base URL is set to https://www.aavso.org/sites/default/files/vstar-plugins/vstar-plugins-2.21.2. Is this correct?
I cannot locate the base file on the website at all.
Regards: Paul (YPFA)
There is a problem, yes. Thanks for reporting it. HQ is aware and looking into it.
That URL is correct.
There is a workaround:
- Download the latest plugins from GitHub
- Unzip it
- Change VStar Preferences to point to the unzipped folder via "Select Local Directory"
- You'll see something like "file:///Users/david/Downloads/vstar-plugins" as the URL once selected
- Open Tools -> Plug-in Manager
- You'll notice that this loads faster as well!
See also page 109 of the user manual.
In VStar, if I use "View > Simple filter": ObsCode equal to <myObsCode>, I can mark my own observations with the "Filtered" color. In the Preferences (Series Size) I set the size of (e.g. VIS, Johnson V ...) to a value greater than that for "Filtered". So I can see my own observations with a (smaller) blueish dot in front of the original observation. So far so good. But for Cousins R the Filtered dot is BEHIND the observation point, so you don't see it.
By the way: Is it possible to mark own observations with a cross, like in Light Curve Generator LGC v1?
PS: Preview (for this forum) doesn't work (Windows 10): The website encountered an unexpected error. Please try again later.
This is a pre-defined ordering. It's a good point though. Filtered observations should be in front I suppose, i.e. the last to be rendered. Or the rendering should be user definable.
There is no user defined shape for own/filtered/other series observations implemented yet.
I've created tickets for both:
Re: Preview and Windows 10, that's something to feed back on a different forum, e.g. https://www.aavso.org/forum/4999
I was trying to analyze some TESS data and I couldn't.
I got the following error message in a window called "Observation Source Error":
access denied ("java.lang.RuntimePermission" "accessDeclaredMembers")
I had been working normally a couple of weeks ago.
I tried on my Laptop and on my PC, both under Windows 10.
I have Oracle Java v8 291-b10 installed.
Any idea or fix?
Do you use JavaWebStart or a stand-alone installation?
What VStar version do you use? We released 2.21.3 this month.
Ideally, under Windows, use VStarWinSetup-2.21.3.zip to install. The installer contains all plugins. See also
Yes, JavaWebStart provides the latest VStar core, however the plugins should be updated too (via plug-in manager from within VStar). Stand-alone installation is better anyway, IMHO.
I installed it and when I tried to run it I got a "Java vortual machine launcher" window with the message:
"Error: could not create the Java Virtual Machine.
Error: A Fatal exception has occured. Program will exit."
I went to the Debug Mode and I got this:
Launching Java.exe with parameters:
Invalid maximum heap size: -Xmx8130m
The specified size exceeds the maximum representable size.
Error: Could not create the Java Virtual Machine.
Error: A Fatal exception has occured. Program will exit.
In most cases, this is caused by 32bit Java under 64bit Windows.
You can check if Java is 32 or 64 bit using the attached tiny Java program (javatest.class). Download javatest.zip , unpack it into VStar folder, run Command Prompt, change directory to VStar, and type:
Look at the line 'sun.arch.data.model'. If Java is 64bit, the value is 64, otherwise, 32 (for 32bit Java)
If you used Windows Installer and launched VStar via VStar.exe, try to comment out memory-related parameters in VStar.ini (see https://github.com/AAVSO/VStar/wiki/Installation-Recipes#how-to-install-vstar-under-windows , "Troubleshooting memory problems" section)
I was trying to install a plug-in but was unable to do so because the Plug-in Manager could not access the remote plug-in data.
I have logged this as an issue in SourceForge
I hope it is an easy fix, as I would like to install the ASAS file loader plugin for a project I'm working on.
Cameron (MCJC) :-)
I tried to reproduce this which I did successfully. I ran the webstart version of vstar from by Win10 machine. This resulted in vstar 2.21.3 running. I checked the URL for the prefs and found https://www.aavso.org/sites/default/files/vstar-plugins/vstar-plugins-2.21.2 set. When I then ran the plugin manager from the tool menu I got an error dialog with the text "Error reading remote plug-in information.".
I went back to the prefs and changed the trailing "2" in the URL to "3" and applied that. Then the plugin manager worked as correctly - delete, add, etc.
So the URL to set is "https://www.aavso.org/sites/default/files/vstar-plugins/vstar-plugins-2.21.3".
I hope this works for you.
It's a while since I've used plug-ins so I clicked Tools>Plug-in Manager to update.
An error dialogue "Plug-in Manager: Error reading remote plug-in information" displayed.
I've Googled this error, and there are relevant conversations on the Forum, but I can't see a clear solution.
First of all, it's would be good to update VStar to the newest version. This is because some plug-ins are related to VStar core functionality that may change.
(It only matters for offline installation, if you use the Jawa WebStart option, you always get the newest stable release)
Then please check (in Preferences->Plug-ins) the "Plug-in location base URL". It must end with the current VStar version number. If in doubt, click "Set defaults".
Then press OK and try to use the Plug-in Manager again. See also https://github.com/AAVSO/VStar/wiki/Installation-Recipes#plug-ins
I solved the problem by making sure I had the latest version of VStar, but because of previous comments elsewhere I downloaded from GitHub.
All good now.
I was running Vstar just fine a few months ago. Started it up today and I can't see any local files on my hard drive. I've spent about three hours going through everything I know, looking at all the permissions I can find, and nothing. I updated Java (Java 8 update 311), VStar (2.21.3). I'm on a Mac running Catalina V 10.15.7. Next step is to update to Big Sur, whenever I have time to deal with a 12GB download.
VStar will download from the AAVSO database, just won't open local .csv etc. files because none show up in the window when I select File>New star from file...
I worked google for a while, I did manage to get VStar to install plug-ins by following instructions I found here, namely resetting to default to make sure I pointed to the right web address.
Paul York reported what appears to be the same bug:
I've added a note there re: your Java and Mac OS X versions.
So, it was fine with VStar 2.21.2?
Are you starting VStar with WebStart or a downloaded release?
I am using Mac OS X Catalina 10.15.4 and don't see this problem. Odd.
If anyone is experiencing this problem, it would be good to know.
I tried both the webstart and downloading from github. Same either way.
Yes, it was fine with 2.21.2. I'm pretty convinced it's on my end, and not VStar. Just need to find the right setting to tweak.At least I think!
In VStar I am trying to use the "B-V" function, switching to view Observations, then download the resulting B-V data to my laptop.
I can get to the B-V observations, noting that the B-V data for a given run is between the V and B observation from that run.
However, when I save the resulting CSV data onto my computer, I see that some columns are not downloaded. For example, in VStar column B is "Calendar Date", but in the downloaded file Column B is V magnitude.
Another item of note is that when I just look at the raw B and V observations in VStar, one column is "Transformed", so I know whether the data was transformed. But when viewing or downloading the B-V view, the transformed column is missing.
Am I missing something here or are these bugs??
The Calendar Date is derived from the time column value and displayed as opposed to being part of the initially loaded data. When saved as AAVSO Download format, only time is saved, since a. calendar date is derivable, b. AAVSO Download format does not contain calendar date. Neither does Simple format (the type specified in the save observation dialog). Does that make sense?
The question of whether other fields should be added to B-V observations, such as Transformed, comparison star fields, CMag, KMag, charts etc, is an interesting one. Clearly comparison star fields (CMag, KMag) are problematic because they are specific to the original B and V observations. The time value is the mean of the B and V observation times. The magnitude is of course B-V and no uncertainty is recorded. I suppose that mean uncertainty could be used, but what does this really mean for a B-V observation? These are after all, synthetic observations. The observer code or codes (hopefully the same!) are recorded to aid connection to the original B and V observations, other than simply the mean time.
So, from my perspective, this aspect of the current state of the B-V plugin is a feature rather than a bug.
However, I invite further discussion, of course.
There is an unrelated bug I found in the course of thinking about this (https://github.com/AAVSO/VStar/issues/246) that will be fixed in the next release: the name column is not set for B-V observations, which is a requirement for AAVSO Download format files to be loaded. Therefore, B-V observations cannot currently be loaded after being saved. Saving as Simple format instead is a current workaround.
I see your point…
I see your point. Thank you.
the B-V view and download is something I just happened upon a few days ago. I was/am interested in making a B-V vs V chart for a particular star, and in viewing the observations in VStar 1st noticed the B-V in File pulldown.
For my use, the most useful format would have B-V, V and B magnitudes all in one row, but I can make Excel do that for me. It would be easier if the V row always was above the B-V row, but I understand why that can't occur as some observers image B before V...
All in all this is a really nice function to have and is making my analysis quicker! Thank you.