Whether the pages are being dynamically generated or not is irrelevant. It's what's serving them out to your users and other visitors which counts.

UMSF.com pages may be being generated on the fly by IPB using PHP and SQL software, but those are not actually serving the pages to your visitors themselves. Instead they feed the pages they create to a webserving program like Apache or Microsoft's Internet Information Service (IIS) and it is they which send the pages out to the visitors. Since the webserving software will probably be keeping some kind of log of every page, image, or other file they send out (and also certain details--like IP addresses--of the computers requesting them), there will probably in effect be two sets of logs: those kept by IPB and those kept by your webserver.

The webserver's logs are the records which can be analyzed by programs like Analog.

Apache's weblog entries for the webpages, for example, are simply the URLs of the pages served. Weblog analyzers (which what I really should have called them) like Analog take those entries, match them up, and then use the results to produce tables & graphs. Dynamic pages make life tricky for Analog (dunno about its competitors) because its inclination is to treat differing URLs as separate pages, which means that if aren't careful you can end up with enormous tables and incomprehensible graphs. However, Analog also lets you use wildcards; and so long as you're canny when you're configuring it you ought to get acceptable results, especially if all you want is just a count of the total pages served minus the images.

The tricky part (also the more tedious part, because it tends to be an ongoing chore) is if you want to do more, like weed out the webbots from the humans visiting your site (assuming you allow the search engine bots through in the first place), or find out which are the most popular webbrowsers your visitors are using or which countries your site is most popular with, or the hit rate on a particular thread or group of threads. The latter in particular will require that you analyze what the URLs are your server is dishing out then use the wildcard and URL substitution features of a program like Analog to "herd" the URLs into groups for the analyzer to analyze. But there you're fortunate. UMSF.com's URLs are not that complex. (Grouping threads into forums would be less practical unless the server's logs keep some sort of indication as to which forum a thread belonged. The URLs as they stand do not.)

Of course, a weblog analyzer still needs access to the webserver logs; and if you're merely renting space on somebody else's machine access is probably out of the question.

======
Stephen

At the risk of being a wetblanket, UMSF.com is searchable by Google & other search engines. So unless your software weeds them out automatically I suspect a sizeable proportion of those 35,275,071 hits are due to search engine webspiders & other bots, not human visitors.

======
Stephen

A proportion, yes. Sizeable..not really. The search engines are intelligent enough to search through the 'lo-fi' version of the forum which has no imagery and so is one hit per page. Per visit - that automatically makes it 10% the size of a human. I actually have stats on the regularity and size of bot visits - but that's for another thread, another time.


Doug

Nice to hear it. However (and at the risk of getting OT even further ), that suggests one would therefore expect to be unable to search for UMSF.com's images using the Net's search engines.

That sounded like an hypothesis just waiting to be investigated, so I hurried off and put a few of the engines (those with an image search capability) through a test drive. The results I got were as follows:

===========================================
Google.com (using "site:unmannedspaceflight.com" as the search argument)
--960 webpages, 718 images
Altavista.com (using "host:unmannedspaceflight.com")
--7150 webpages, no images (but 2 videos)
Alltheweb.com (using "host:unmannedspaceflight.com")
--7090 webpages, no images (and 2 videos)
Search.netscape.com (using "unmannedspaceflight.com")
--1550 webpages, 41 images.
Search.yahoo.com (using "unmannedspaceflight.com")
--"About 11500" webpages, "about 96" images (and 2 videos)
Search.msn.com (using "site:unmannedspaceflight.com")
--106 webpages, no images
Search.ask.com (using "unmannedspaceflight.com")
--157 webpages, 5 images
===========================================

Conclusion: I guess some bots must be more intelligent than others.

======
Stephen

Where's Tesheiner? Normally the map's updated before I've even realised a drive has happened.

We had one tosol (823), a big one!

I don't know the exact figure as we had a site change (70 -> 71) but at the time of the change (very near - EDIT: about 1m from the end of the drive) we had 71m on the clock!

James

Putting us about here...

Click to view attachment

[quote name='jamescanvin' date='May 19 2006, 06:00 AM' post='54840']
Where's Tesheiner? Normally the map's updated before I've even realised a drive has happened.

He's probably arguing with rover's drivers why they didn't strictly follow his proposed route? Could only be the autonomous system that get a bug in it.

Well, the images were expected to be downlinked at 01:30 AM (my local time) and I was already snoring at that time.

A *big* drive, indeed. I tried first to compare the new pics with those taken on sol 821, no luck. But then making a polar projection is usually the solution to the puzzle.
Click to view attachment

My updated route map.
Click to view attachment

A *big* drive, indeed. I tried first to compare the new pics with those taken on sol 821, no luck. But then making a polar projection is usually the solution to the puzzle.

When one can correlate ground picts to so many details from orbital picts of MGS (even if the one I'm talking about is realy good ), I can't realy imagine what MRO's will look like.
Rover's HG antena will surely be visible on some and we can even hope to get one of the sundial...just for Doug.

Edit : I don't think they'll actualy USE the calibration target to make colored picts

MRO won't be quite THAT good.

http://www.unmannedspaceflight.com/doug_im...hirise_oppy.jpg

Is an optimistic simulation of HiRISE looking at Oppy as it would have appeared at Purgatory

Doug

[quote name='djellison' date='May 19 2006, 11:42 AM' post='54884']
MRO won't be quite THAT good.

Do you mean even when using the "compensation move" (sorry can't remember the exact words) they use sometimes on MGS ?

MC can step in here if I'm wrong, but I believe the way in which HiRISE works - the CPROTO technique that MGS uses just won't be possible (or even worthwhile) with HiRISE.

Doug

Doug:

I think he posted that specific observation some time ago in response to a similar discussion.

Bob Shaw

how deep is corner crater? Do you know? And how hight is a rim of crater? Can it be good scenary?

With this last giant leap Oppy has arrived the area covered by MGS image S0500863 (I tend to get the codes mixed, so correct me if needed), meaning that it's now possible to plot Oppy's route on a 3-D route map, either crosseyed or anaglyph.

I'll come back later and attach a an xeyed example of Oppys' Sol 823 location.

(Edit) Came back at last. The image suffered for unknown reason in the process but still retains enough of its three dimensions to be worth looking. I rather indulged myself in the work, driving Oppy as far as to the bottom of Victoria. Note that Oppy's route is *really* three-dimensional.

Corner Crater is less than 600m away, I think will be reached within the end of June...

Nice work but you made a huge mistake (I'm surprised Climber didn't notice). You placed the "Beacon" on the NEAR rim of Victoria, and we all know it's on the FAR rim.

it was marswiggle's pic that convinced me to join the NEAR team....

Great to see your work method Tesheiner, thanks a lot!

(BTW, I lived in Alcalá de Henares for six months in the mid-nineties, fun to see that you hail from there)

[quote name='ElkGroveDan' date='May 21 2006, 04:09 AM' post='55121']
Nice work but you made a huge mistake (I'm surprised Climber didn't notice). You placed the "Beacon" on the NEAR rim of Victoria, and we all know it's on the FAR rim.

Could not connect over the week-end...Thanks ElkGroveDan, we'd better run our own version. Never take what is on the Internet for granted

Edit : I also take the opportunity to fully correct the pictures Sol XXX should read SOL 935

Small world, isn't it?

http://en.wikipedia.org/wiki/Alcala_de_Henares

Route map updated to sol 828.

Click to view attachment

Corner crater is now at the middle of the travel to the near rim of Victoria.

What a great coincidence that the ripples are oriented in just about the best direction possible!

I reckon that with the sol 828 drive we are for the first time closer to the rim of Victoria than the rim of Erebus. We are still maybe two drives away from being closer to the centre of Victoria than the centre of Erebus.



Small celebration for 100th post.

[quote name='Joffan' date='May 24 2006, 10:10 PM' post='55615']
Small celebration for 100th post.

Welcome to the club! You're just in time to celebrate the 1500 sols over garanty cumulated by both rovers.

I'm writing up several weeks' worth of "Q and A" segments for Planetary Radio and someone asked me "Has Opportunity driven out of the landing ellipse yet?" I see here that that happened around the middle of March, but that's a bit of a short answer for the radio show -- any suggestions about what other fun facts I could put in my answer? What was the straight-line distance from Eagle to the ellipse exit point? What was the wheel odometry at that point? How far has she driven since then? Maybe I can follow that up with some facts about what awaits at Victoria...hmmm...

--Emily

An important part of that story is precisely where Oppy plopped down in the landing elipse. What were the dimensions of the elipse and how far from the centre, and in what direction (Southeast as I recall). In other words how far off the centreline did Oppy land, and then how far south did it have to drive to exit the elipse. Comparions with where Spirit set down within its own elipse would add something too.


And I meant to add that, sorry, I don't know the numbers. The JPL web site shows the elipses here, without the actual landing location

http://marsrovers.jpl.nasa.gov/gallery/pre.../20040124a.html

Here are some FunFacts i'd like to know (sorry i dont have answer but i think some of this info already lurks in these threads):
- altitude gain/loss between Eagle and Victoria
- depth of excursion into Endurance crater
- relate these measurements for earthly familarity, ie., soccer field lengths or how long it would take to walk at a human pace (or whatever sounds fun: snail's pace, swim, speed of sound or dust devil on mars,...).
- how small the MER ellipses are in comparison to Pathfinder, viking,...
...and lots lots more!

I too would be curious as to how the elevation changes from Endurance to Erebus to Victoria. That is, how many metres will we have moved up-section from Endurance to Erebus to Victoria? And when the rover drove inside Endurance it examined a 7 meter vertical section. How much of a section are they hoping to see at Victoria?

.....By the way, Emily, loved your article on the Titan radar results. Very intriguing geology, to say the least.

I'd like to know, too, but I suspect that the truth is that we don't know. We don't know what the geometry of the beds are (short of "essentailly flat-lying") and without soem sort of stratigraphic marker we can't figure out whee we are in the section. Hopefully we'll find something that we saw at Endurance to something that we will see at Corner or Victoria...

--Bill

And I, Emily, would be fascinated to learn how much verbiage and bandwidth UMSFers have expended in fervid, global cyber-debate on:
1. Where the landing ellipse should go?
2. Whether the line should be thin, fat or fuzzy?
3. When we will arrive at Victoria?
4. How we will know when we get there?
5. Which side of it we can see?
6. Which side of it the Beacon is on?
7. Whether the above questions have been answered in a fair and democratic manner?
Someday all this will be answered in depth in a Tom Wolfe novel, which will be praised as his "funniest since The Right Stuff." But you can use it for now.

28 more meters tosol, according to the tracking data.

How wrong is that this time Tesh?

Updated route map.

Click to view attachment



I wouldn't call it "wrong" but simply "different". The map may have errors and the different methods to estimate a rover's position too; all those things add to the probability of finding different results.

How diplomatic!

Interesting question.

Here is a plot of change in drive number vs drive distance for all the changes in position listed in the tracking database since sol 781 (start of site 68):

Click to view attachment

I was quite surprised to see there is quite a strong correlation for a lot of the points, i'm guessing that indicates the 'drive so far then check for slippage' parameter. Also note the high black points from recent drives. It looks like they may be driving slightly more aggressively now the ripples arn't so big.

Also note all those points with large drive number changes but very small (even zero) drive distances - don't understand that.

Zooming in on that graph:

Click to view attachment

Shows the small changes in drive number which correspond to auto-nav drives. Very little correlation.

Note there are even some points with no drive number change but some movement, as well as points with no movement but a number change - now I'm confused!

Moving quickly on. Here is the same plot but with the total distance and drive number change for each sol:

Click to view attachment

Which is much more sane - giving this (very rough) rule of thumb: For a drive increment of 36 (which is for example xxDx to xxEx etc.) the rover will have moved approximately 6 meters. Of course a change in driving stratagy/aggressiveness could change all this overnight.

Here ends this pointless exersice, make driving distance estimates from drive numbers at you peril.

James

Thought I'd share this which I wizzed over to Emily last night...

"
Right ….

I started with this…

http://www.msss.com/mars_images/moc/2004/01/24/



That ellipse is about 87km long, 11km wide – I’m not sure which ellipse that is ( final targeted ellipse after final TCM, or the launch ellipse…it might be worth pinging someone at MER HQ to see if they know exactly where that one came from – but it’s the one we used on the forum )



Going from an estimated centre of that ellipse - the actual landing is, by my maths, 23km downtrack, and 0.4km right of the centre line – with the MER entry trajectory coming from the left.



Using this…

http://www.planetary.org/image/oppo-stooke-map.jpg



(I can’t WAIT for HiRISE imagery)





And this

http://www.unmannedspaceflight.com/index.p...pe=post&id=4907



I would estimate that the southern edge of the landing ellipse was between 4200 and 4400 metres south of the landing site, and crossed at some point between 760 and 773



Sol 760 Odom was 6682m, sol 773 Odom was 7249m….I wouldn’t want to try and drag in either of those sol numbers or distances driven – we did our best overlaying that MSSS image ( the larger 10m res one is available at the bottom of the page linked up the top of this email ) onto the route maps that people have done – but we’re not really sure – so that’s the sort of ‘max range’ based on what we have. There might be an MER person who knows more accurately."


[/size]

When you think about it - they got a bit of a downrange distance on both of them due to the atmospheric conditions, but they were both VERY close to centre line which is the real gauge of their accuracy in targetting...400m or so for Opportunity is utterly astonishing.







[size="2"]Doug

The Oppster has been really cleaning out the attic in recent updates. In the past, this has signified the commencement of a major advancement campaign. Can we expect a serious attempt to add mileage to the odometer?

So you could argue that since it was a 456 million km journey they hit the target to a precision of around 1 part in a billion.

I do a chart in my talks that shows the still of Dan's animation at bridle cut.....400,000,000,000 metres above the airbags, 6.3m below

Doug

Mmm.. I noticed all the past images being offloaded but I hadn't associated it with the prospect of an aggressive driving campaign... I hope you're right!

So do I but I think they might be preparing for the software update, and that might throw a complete spanner in the works for both the arrival at Victoria pool and the "Where is the Beacon" soccer match.

An updated route map after sol 833 "drive".

Click to view attachment

Edited: Corrected wrong sol number.

Deviate from Tesheiner's route at your peril

It looks like there is no way to avoid going over some of the dunes between here and Corner Crater. Hopefully, when Oppy is unstuck again, a better more careful route can be derived. We don't need to make it to VC until Sol 933 anyway

Your concern for Oppy's welfare is deeply moving, Eric.

Yes. I was so moved that I looked around for something or someone to hug. The nearest victim was a pair of red/cyan glasses.

I wanted to post some recent observations. This might better belong in one of the other threads, but since it relates to my interest in observing Opportunity's movements, I'll drop it here.

We've recently seen Opportunity's path move into a region for which decent 3D orbital imagery is available. I have long been awaiting this. It would have been nice to have such coverage available for the whole route map. Although the parallax on the plains is minimal, it seems that we can observe some of the topography. It appears to me that Opportunity has recently decended into an ancient crater whose circular outline is apparent on Tesheiner's maps. The decent can be seen in recent anaglyphs from Opportunity. It appears to me that the crater's basin is directly in our path, and I wonder if the aeolian sediment in this crater is more treacherous than elsewhere, and responsible for the recent wheel embedding hazard.

The attached file is an anaglyph of orbital imagery for this area. It's rough, but it has 4 yellow dots corresponding to the last four locations of the rover on Tesh's last map. It might have had five locations, but I did not distinguish between his last two.

Anaglyphs are definitely a more practical way to see maps in 3D, compared to my Xeyed version previously. Thanks for sending this image. Will save me from the risk of getting my eyes permanently crossed due to my working methods, and several other people from some splitting headaches if you instead of me keep sending these maps every now and then.
--
And yes, there seem to be a noticeable depression and some even bigger-looking dunes ahead of Oppy. I think there's no going there now and she must be backed and driven to west to follow the 'original' paved path.

James recently helped me out on a question I had about site drive number changes and their correlation to driving distances.



Oppy's current predicament got me thinking about this again. It seems to me someone somewhere (in JPL) must be deciding what the site location is going to be called. I can't believe the naming is a random process and therefore it seems logical to tie it to distance.

The first attempt to extract Oppy from the latest sand trap has resulted in a drive site change from 71GB to 71GG, a change of 5 units on the fourth character. Perhaps this equates to 5 cm, but they would have commanded a much longer drive to achieve this.

I'm now wondering if the site drive location is more related to commanded driving distances rather than actual drive distances achieved. This would make some sense, as they probably have to decide on the site number for file naming before the drive takes place, images are taken and then sent to Earth.

Here continues this pointless exercise - but think of the benefits if we could actually work out precise driving distances from the site numbers.

Castor

No - it really is just a drive number. Currently, because they use regular visidom slip checks, each instance of that will trigger a new site number - and I believe they do these every 6 or so metres..ish.

Of course, if you do a lot of tiny short drives with lots of imaging in an attempt to say, get out of a dune, then you'll crank up the drive number quite quickly. I think they may have done 4 intermedite imaging sequences and one end imaging sequence with the first drive-out-of-dune sequence, which would = an increase in drive number of 5.

It's not related to distance driven, just number of drives done.

Doug