Never mind, I extracted to desktop and it appears to be working.

Edit: It appears the issue was not with vista but with the values for Xmx and DmaxPanBytes. I tweaked it to work at the following:

-vmargs
-Xmx1398m
-DmaxPanBytes=1100m

If I raise the Xmx to 1399 it gives me the error. I'll leave the DmaxPanBytes at 1100 till I find a panorama that needs more then 1100MB of memory.

BTW, is this a known issue?

I don't that's an issue as such; I think those numbers are simply too high.

Yes, I guess it's generally known that if you set the Xmx too high, some Java implementations just don't handle that. Also, if you set DmaxPanBytes higher than than the amount of OpenGL textures that can be handled at one time by your graphics hardware, then you will see performance problems. IIRC, the application is delivered with reasonable defaults.

Could it possibly be an issue with 64 bit Vista? Anyone else running MMB on 64 bit?

I run it on XP64 - no problems at all.

Okay I think I got it. I needed extract MMB to the desktop instead of the "program files" folder in order for it to update from 2.0.2 to 2.0.5.

I am running Vista 64bit and MMB used to work fine for me. And i absolutely love the panoramas viewer in MMB, it did just about everything i could ask for. But now it just wont draw any panoramas. When i try to make a viewer with panorama the cursor spins for a moment like its trying, and then just seems to give up and window remains blank. otherwise everything else about MMB still works just fine. and ive got plenty 4GB memory installed in this intel Q9300 2.5Ghz box with GeForce 9800GT.

i hadnt changed anything about the "MidnightMarsBrowser.ini" file from the defaults, but after it stopped working, i upped it a bit as i found on some part of the MMB site in an attempt to get it to work but no go. i also tried upping the setting to the values given above by 'briv1016' and they dont help either. i get the same error and have to lower it a bit extra to avoid that. whatever i set it to, its a no go for the panoramas.

I've tried all sorts of things like re-downloading MMB and reinstalling the latest Java versions.
I've tried both MMB 2.02 and 2.05 (since some panorama code was reworked between the verions) but they run into the same problem. ive even tried downloading and running a completely separate instance of MMB and then downloading all new content to see if it would work in that pristine microcosm, but no go.. this and a litany of other sweet little gotchas have prodded me into regretting my choice of Vista64.

has anyone else run into this problem?
are there any sort of tests that i can do to get an error log from MMB or some other module that might plug into the same panorama functions as MMB (if thats how it works) to see whats broken, as it doesnt seem to be MMB itself but some dependency that is crippled. im not a programmer but am handy enough, and would love to get this feature working again.

Did you try extracting to desktop instead of Program Files?

..in fact its never been in the program files directory. instead, it stopped working in place at
C:\Users\atomoid\Pictures\MidnightMarsBrowser
so i tried placing/running the unzipped folder at C:\Users\atomoid\Desktop\MidnightMarsBrowser 2 as well (had to re-choose the 'MMB Workspace'), but no granola..

Why was it in the pictures file?

i placed it there at first since it seemed a good place to keep a ton of images rather than the documents directory, and the MMB program doesn't hold any gripe about where its located, just as long as it remains in its original folder structure that it unzips to.

i also tried updating my Nvidia drivers to see if that helped. didnt. so its still systems non-go for launch.

the fact that i can download a fresh copy of MMB and unzip it and download all new content and the the panorama feature still doesnt work makes me conclude nothing must be wrong with MMB itself, but then what else could be preventing MMB panos from working? how does it render the panos? does anyone know if there is some sort of directX component or rendering engine outside of the MMB folder structure that is a dependency?

Do you have the latest java?

yup, did that, auto-updates too, hmmmm...

I don't know if every instance of MMB produces the same folder structure on every single computer, but I was curious and just checked the size of MMB on my computer and realized we're about to pass a milestone of sorts: as of this moment my MMB folder contains 19,996 subfolders. One more sol and we'll be over 20,000! In another sol or two we should turn over to 360,000 files. All of which is eating up 32.3 Gigs on my hard drive. Oy.

I'm not complaining, though, because to do so would be to complain about the rovers' longevity. I hope to be forced someday to upgrade my hard drive in order to keep holding all that lovely rover image data.

You are such a lightweight

91.4 GB on disk (90,627,082,502 bytes) for 407,487 items


I have 50 gig of IMG's though. That's probably cheating. Just deleting them now.

Ha, yeah, I've only downloaded the JPGs.

Weird, I have the same 32.3GB but have a whopping 399,016 items listed.

Perhaps you have more windows thumbnail files.

Yes, I guess OSX is counting the .DS_Store (or whatever it's called) files, which adds the extra 20,000 (1 per directory) that I have over Emily.

Please forgive me if this has already been discussed as a possible feature in MMB, but are there any plans for MMB to create a Autopano Pro Giga project file (.pano) for panoramas? This would be a neat feature to implement into any future releases.

I know there is support for PTgui http://www.unmannedspaceflight.com/index.p...ost&p=68219.

Read this thread and decide for yourself - particularly mhowards posts regarding development.

It doesn't hurt to ask, but yeah, I'm pretty busy and not working on MMB anymore.

I guess in theory MMB is open source so somebody else could work on it, but in practice that's never happened AFAIK.

I was just throwing another idea out there for anyone considering future work on MMB. I am not very experienced with the java language, but I might take a gander at the source.

Something that has bugged me for a while. I'm using MMB 1.5.2, and I have not done any updates for a while (I know, shame on me...). When I do run the default update for the exploratorium it says that that I have not done an update since June 2008 (that doesn't sound right?) but then proceeds to say there are no new images.

And for JPL, it only finds some very recent sol images even though I should have at least a couple of month's worth of sols missing.

Any hints as to why it is not finding anything for the exploratorium and almost nothing for JPL?

Thanks,

Airbag.

Checking Spirit raw images from Exploratorium. Last download time is 24-Jun-2008 07:00
Reading http://nasa.exploratorium.edu/mars/spirit/forward_hazcam/... scanning...
Reading http://nasa.exploratorium.edu/mars/spirit/rear_hazcam/... scanning...
Reading http://nasa.exploratorium.edu/mars/spirit/navcam/... scanning...
Reading http://nasa.exploratorium.edu/mars/spirit/pancam/... scanning...
Reading http://nasa.exploratorium.edu/mars/spirit/micro_imager/... scanning...
There are 0 new Spirit images at Exploratorium.

[added later: MMB 2.0 does seem to be updating just fine from exploratorium, so maybe not a real issue then]

IIRC there was a change in the way the images were labeled, MMB2.0 was updated to deal with the change but version one was not.

Just a thought...would MMB be eligible to enter this Mars competition?
http://beamartian.jpl.nasa.gov/pic/coop-league#/Details

Has this been mentioned before? I think it's worth investigating.

Huh. I poked around on the competition website for a few minutes and, to be honest, I'm not sure I understand the goal of the contest. It's kind of a strange experience to hear someone talking about some new "MER Dataset API" when we've been using the MER data for like six years now. But I'm guessing it's for students, and if I were supposed to know about it I would. Anyway, it seems to presume certain things like a Microsoft development environment, which (to be honest) I couldn't have much less time for.

To answer the question previous to that: yes, MMB version 1.x is dead now. It no longer works with changes to Exploratorium, although I imagine it might still work against the JPL site.

I notice MMB produces anaglyphs from all possible combinations of left and right images. While good in theory, eventually producing a complete image when the final image is downloaded; with the recent downloading of up to 16 versions of the same image, in practice it can lead to 200+ copies of the same image. Would it be possible to quickly update the software to shut this feature off?

Sorry; not going to happen. You can go into the workspace file hierarchy and delete the files in question, then choose the "Rebuild Image Indexes" menu option in the program. The directories with the anaglyphs are 1/na, 1/pa, 1/ra, 1/fa, 2/na, etc. (1 or 2 for rover number, na for navcam anaglyph, pa for Pancam anaglpyh, and so on).

Would combining recently updated "ImageMetadata_Archive" and "ImageMetadata_Update" (currently blank) files be a good proxy for the total number of mission images when you adjust for M1, 2, 3, etc versions of the same image?

Why not back them up, try it yourself, and report back. The message MH has given regarding how much time he has to work on MMB couldn't be any stronger.

I was just wondering because it’s approaching 250,000 rows and it's kind of a milestone if that’s how many images have been taken and downlinked.

Double check for yourself here : http://pancam.astro.cornell.edu/merweb/mer_edr_counter.pl

Ok, I’m getting three different numbers depending where I look. The MER website which I know counts multiples has the count at 265,077 images. The Cornell link you gave has the “Total Images” (ESF+EFF+EDN) at 258,228 images. And the MMB metadata method I described earlier has the total at 249,135 images.

Now right off the bat we can get rid of the MER website total because we know it includes multiple versions of the same image. Now if we take the difference of the Cornell total and the MMB total we get 9,093 images. I thought that maybe the difference is due to onboard data but Spirit and Opportunity only have 1,621 and 2,319 onboard images respectively and a good portion of those have at least a partial version on the ground.

In the end I’m not sure where the disparity lies. I think it’s safe to say that we either have already pasted or are about to pass the 250,000 unique image milestone. Yay!

BTW, thanks for the link.

Hello,

I have a few days getting the following error message when I try downloading new images.



Would you happen to anyone else?. mhoward Any solution?

Exploratorium computer says "no"...

( see this post )

Ok, thak you.

Hello,

Since the failure occurred at Exploratorium, the latest images from Opportunity is there are from 26 May. I tried to download new images via JPL, but today I've managed to download images from beyond of sol 2231.

On the official website of the mission, JPL site, the latest RAW images are of sol 2273.

mhoward, can you make a patch that allows downloading from this or else we will no longer have the latest images with MMB?.

A long time ago that you do not include updates to MMB, from what I understand you've put aside their development, however, as the missions continue to be active would be nice if you would give a minimum of support to continue using this great piece of software. At least until the end of the mission.

A greeting.

Hi Dig,

The images you mentioned are being downloaded. I think what you are observing is that I haven't updated the metadata which drives the panorama feature for a couple weeks. I've been away from home; that situation has now been rectified.

I think people probably have forgotten - or never realized - that the Panorama feature in MMB depends on data updates which I put on my own server, and which get automatically downloaded by MMB. I have been posting those updates quietly, regularly, for something like five years now. It might be nice if people realize that a "minimum of support" is already going into this project.

As I have mentioned several times, I do not have time to make code changes to MMB anymore. When I say that, I am speaking quite literally. If Exploratorium "gives up the ghost" (dies) or if JPL changes its image posting format, then that will mark the end of this little technology experiment, at least in its current incarnation. It is probably best to keep in mind that I am not NASA and do not have NASA's resources.

I will make changes to the Google code project page to try to make this more clear. The current closed status of the project is currently only mentioned in the "Supporting the Project" section, so I guess I can see where people don't read that far down.

Regards.

The 3D Panorama feature doesn't work anymore on Nvidia Forceware 258.96. (at least on the GTX460) It shows the panorama for a few seconds, then nothing....

EDIT: strange. When running MMB as administrator it works, kind of. Very choppy. The card doesn't throttle to full 3D frequencies. MMB uses OpenGL hardware acceleration, doesn't it?

Quick solution for you....





...use different drivers.

Now that this thread is alive, I'll post a question I have since months ago...

When MMB checks and finds new "metadata" it always detects that the "ImageMetadata_archive" file has to be updated. That's a big one, currently at 24MB, at take its time to be downloaded. Is there really new data on that file every time metadata is updated, therefore the need to download it, or may it be a bug?

Quick answer for you..... No.
These drivers are the only ones for the gtx460. It's a brand new card. There should be new drivers out any day now, but I doubt those will fix this problem. I hope it's not a unique 'feature' of this card.

Tesheiner: It shouldn't be doing that, and it's not doing it here (I don't think). Try manually deleting the file (smells like a permissions issue). And/or check your Java version.

OWW: As far as I remember (which is increasingly less far), it just uses Java functionality to access OpenGL. Not much I can do about version incompatibilities at this point. See if any other Java 3D apps are working.

I don't know any other java3d programs. Could you point me to one?

Mars24 for starters. I think. Suggestions welcome.

I moved all metadata files to a backup directory and updating it right now. I will tell here what happens after updating it once again.

It's still doing the same thing.
I then checked my current Java version and it was 6.17 running in Windows XP. After updating it to 6.21 I tried to enforce a metadata update by deleting the homepage.html and latest2.txt files and the result is the same.


All "*_archive" files are dated 06Apr2010 but curiously some of them are always downloaded and some others are not updated and "same last modified time" is displayed.

I've just reset the archive metadata online; not sure if that will help, but it's worth a try. It might only help randomly, since what you're describing sounds suspiciously like a Java version peculiarity.

Well, now the results I get are consistent; all files (archive and update) were detected as new, downloaded, and I see them dated today Sep 24th around 17:20 (GMT+1). Let's wait some days and I'll tell what happens on the next update which should be the "interesting" one.