I was talking to mtn-man and CharlieP at brunch and both mentioned having a problem with Plucker slowing down if they were "plucking" a large number of geocache pages produced by GPXSpinner (note the same problem can occur if GSAK is used to produce the pages or other tools that produce html from GPX). I had seen a potential solution so I thought I would post it here.

Lil Devil, the author of GPX Spinner, mentions the problem and potential solution in a groundspeak thread.

quote:

With that many waypoints, you'll notice that at first, the status window will be flying along. Then, about 1/2 to 2/3 through, it will slow waaay down, only showing one new line per second or so.

At this point, right-click in the status window and choose "clear all" and it will speed up again.

Now I assume that when everyone else says plucker they really mean plucker desktop. Which is the nice GUI interface most people use. Under the covers there is a real command line program which is the real plucker that the plucker desktop calls. I use it all the time and it never seems to slow down.

Remember "Real computer geeks use the command line."

It makes a big difference to select the spidering as depth-first.

You have to spider to at least 2 or you cannot read the hints, right? Spinner recommends a depth of 3.

My problem last night was trying to Pluck 2 GPX files at once. When I did them one at a time they were fine. Nice trick about clearing the window. Next time I process two large files at once I will try that.

quote:

Originally posted by KVOM:
It makes a big difference to select the spidering as depth-first.

Well in the same message that Lil Devil mentions using clear he says

quote:

From my tests, neither isn't really any slower. Breadth first may appear to be, based on the size of the queue, and all the pages it seems to have to skip over. Depth first does the same thing, just in a different order, so more pages get done up front, followed by more skipping at the end.

As long as you "clear all" once the display starts to slow down, breadth first is probably the best.

Me I don't know since I use a GSAK macro to filter then produce the HTML pages then call the command line plucker.

I would guess that the depth vs. breadth issue is dependent on how many "nearby" caches you choose to export. A breadth first cache will hit those links and likely need to suck in and cache almost every page before hitting many of the common links such as standard images.

I definitely found that deoth-first made a huge difference in timing as I normally export links to the 10 nearest caches.

I tried the "clear all trick" yesterday and it cut a 30 minute run down to about 2 minutes ... big difference. But I am also having a problem recently with Plucker locking up on certain files out of Spinner, so I may have to go back to Gpx2html anyway. LilDevil is looking at the problem, but he reports that 3 others have had the same problem and he was not able to re-create it. It does not look good.

Strange as since version 1.2.2 GPXSpinner has my code in it which removes some bad HTML that confuses Plucker. Do you know the GC number of the geocache causing the problem?.

quote:

Originally posted by CharlieP:
I tried the "clear all trick" yesterday and it cut a 30 minute run down to about 2 minutes ... big difference. But I am also having a problem recently with Plucker locking up on certain files out of Spinner, so I may have to go back to Gpx2html anyway. LilDevil is looking at the problem, but he reports that 3 others have had the same problem and he was not able to re-create it. It does not look good.

I've had this same problem. Could you post how you get around it? - I'm not familiar with Gpx2html.

quote:

Originally posted by AllenLacy:
Strange as since version 1.2.2 GPXSpinner has my code in it which removes some bad HTML that confuses Plucker. Do you know the GC number of the geocache causing the problem?.

I am running Spinner ver 2.0.6, does that have your code? Send me an email and let me know if you have a high speed connection ... if so I will send you both the gpx and html files that lock Plucker. I have not heard from lildevil so apparently he has not found a solution. I tried to count the files to see where it locked, and in both cases it was a file which had an external link, one to a TN geocachers assn., the other to a geocache rating site ... but I am not sure that is the problem.

quote:

Originally posted by mertat:
I've had this same problem. Could you post how you get around it? - I'm not familiar with Gpx2html.

Gpx2html is another routine similar to Spinner. It does not have all of the features of Spinner, but it runs faster and seems to have fewer problems. Download it here.

Fizzy Magic

I tried Gpx2html on the two gpx files that had caused the Plucker-lock with Spinner, and they both worked out of Gpx2html. Allen has identified the problem with one of the files and I am hopeful that LilDevil can modify Spinner to fix the problem. In the meantime I am back to Gpx2html.