How Many Logs?

[+R4D4R]

If I run a pocket query for all logs within some distance of some point, can I have it return all the logs of the caches or is it just limited to the first five?

[+Stunod]

It's just the LAST (most recent) five logs. Can't get any more.

Edited September 6, 2005 by Stunod

[+R4D4R]

Crud, thats unfortunate

[robertlipe]

Not to split hairs with armed chickens, but it's five logs plus your own logs. In most cases, the distinction doesn't matter, but we should probably spell that out for the young 'ens.

[+Stunod]

Not to split hairs with armed chickens, but it's five logs plus your own logs. In most cases, the distinction doesn't matter, but we should probably spell that out for the young 'ens.

Yes, my bad.

[+R4D4R]

either way :/

I was hoping to run a query once a month or so rather than once or twice a week to get the recent queries. I have a couple of terms that I am searching for so I can get a handle on a few different areas. Its going to take me longer with fewer returned logs, but I guess I can make it work out

[robertlipe]

More logs in the PQs is a frequently requested feature and has been for a long time. Aggregating PQs over time - like you're proposing - is a common workaround.

 

It's interesting that the reason given for limiting the number of logs is to reduce system load. Unfortunately, the current model somewhat encourages folks to do what you're doing and get 8 PQs instead of 1. I can't imagine that is a net win, but I'm just guessing.

 

I've been on hunts where the last six logs are essentially:

Aug 10. Harry - Found with Party of Five. Good thing we had Bill's coords.

Aug 10. Dick - Found with Party of Four. Good thing we had Bill's coords.

Aug 10. Tom - Found with Party of Four. Good thing we had Bill's coords.

Aug 10. Fred - Found with Party of Four. Good thing we had Bill's coords.

Aug 9. Waldo - Logging personal travel bug through.

Aug 9. Bill - Searched 500 foot circle around coords. Called placer. There's a typo on the cache page and the correct coords are a quarter mile away at XXXXXXX found quickly once I had the right coords.

[Jeremy]

It's interesting that the reason given for limiting the number of logs is to reduce system load. Unfortunately, the current model somewhat encourages folks to do what you're doing and get 8 PQs instead of 1. I can't imagine that is a net win, but I'm just guessing.

Call it the "super size me" phenomenon. The average user really doesn't need more than the last 5 logs but it is offered so what the heck. In this case "super size" has no downside (45 cents more, for example) so everyone ops for the maximum (20, all, or whatever). Add compounded interest and you have a compounded problem.

[robertlipe]

I'm familiar with the phenomenon you describe and it is indeed real. If there's a measureable computational cost, pass it on. ("Oh, you wanted the biggie fries? Those come bundled only with a sphygmomanometer, so that'll be an extra $74...")

 

Maybe Super Size PQ's count as two or maybe they return a max of 250 instead of 500 or something. Maybe they can only be scheduled to run on Frogday - you know, that eighth day of the week when your server can make one pass over the database without write locks, cache every page with the inner select on the logs done, store all the precomputed records (perhaps "include my logs" isn't an option here), and then just let the selection logic harvest the right ones for concatenation into the actual GPX.

 

More logs would also really help the "lonely cache club" kinds of projects, too.

 

A lot of us (and I can't claim that it's teh "average" user that's clamoring for it) would be delighted to have our "found by me" PQs - you know, the ones including archived caches - run only on Frogday.

 

I know there's programming involved. I've defended you against a lot of requested site changes, and I'm not "flipping" on you, but I think that studying this and considering the way that some folks use PQs would go a long way to reducing the number of PQs that are ordered up. (I'd wager a substantial number of PQs are ordered up right now solely for log aggregation.)

[Jeremy]

We're working on the "found by me" PQs now as we build in the Bookmark List PQ feature and attribute search features. I don't object to increasing the log number to 10 at least initially.

[robertlipe]

I think even ten logs would go a long way to solving the need to aggregate PQs. Thanx. It'll probably take some tub-thumping to get folks to pull such queries weekly instead of bi-weekly (or whatever) but that could be evangelized.

 

Come to think of it, there are times I'm scoping an area and I'd be happy with zero logs. No, .loc's won't cut it. At least in my case, if I had a knob to turn, I'd probably turn it down as much as I'd turn up. Only you will know if the power saving to you to put a dimmer in there instead of just using different bulb is worthwhile, but I thought I'd mention it.

 

Thanx for listening.

[+beejay&esskay]

I think even ten logs would go a long way to solving the need to aggregate PQs. Thanx. It'll probably take some tub-thumping to get folks to pull such queries weekly instead of bi-weekly (or whatever) but that could be evangelized.

I still see a problem here with convincing us that 10 logs means we should run our PQs less frequently...we still don't know if we will miss logs this way.

 

Now if we knew the logic was "10 [or 5] logs or 2 weeks of logs, whichever is more", then we would know that any "log harvesting" could reasonably be done every two weeks.

[+Hynr]

I find that when I need to look at past logs on my PDA I frequently go deeper than the last 5; I rarely go deeper than 10; I only go deeper than 20 if it is an event where I am trying to see if a particular person is going to be at an event to which I am traveling. I have that many to search because I don't discard logs once I have recieved them in a PQ. In the process I get many logs for many pages several times. It would indeed help me to be able to get more logs at once, but in the interest of efficiency, it would also be helpful to limit logs to only those that have dates since a particular date (e.g. the past 7 days). I am sure that with the ability to have at most 10 logs from only the last x days returned to me, my PQ results would be much smaller and I would have fewer holes in my dataset.

[+BigFurryMonster]

I don't need logs in PQ's. Can I get rid of them?