opencaching.com

Team BMW-Biker · Joined: 01 Oct 2003 Posts: 209

Hello everybody,

i think we must first decide, how the structure must look like.

Here is my idea:
http://www.dietz.homeip.net/geocache/opencaching/ideas001/generalstructure.htm

Please post your ideas here and i will make a new mind-map.
We can discuse about all ... please do it.

I will post in a few time an idea for a replication-circle and an idea for the server-user-areas ...

Thanks for reading and discusing.
Oli from the Team BMW-Biker

Team BMW-Biker · Joined: 01 Oct 2003 Posts: 209

ok, here are the other "first-ideas":
http://www.dietz.homeip.net/geocache/opencaching/ideas001/replicationcircle.htm
and
http://www.dietz.homeip.net/geocache/opencaching/ideas001/serveruserareas.htm

Happy braining, Oli[/url]

Team BMW-Biker · Joined: 01 Oct 2003 Posts: 209

Hello again,

here are some more ideas

... about Database-Structure:
http://www.dietz.homeip.net/geocache/opencaching/ideas001/databasestructure.htm
... about a Replication Protokoll:
http://www.dietz.homeip.net/geocache/opencaching/ideas001/replicationprotokol.htm

... Oli

Vinnie · Joined: 23 Sep 2003 Posts: 71 Location: Cologne, Germany

Hi Oli,
great (and useful) work!

Some remarks from here:

1. I do not like these ugly guids. I prefer "readable" guids. For the caches I proposed something consisting of coordinates and time in another thread.

2. Concerning the general structure and replication: (You probably know) my proposal is to use a usenet-like network based on NNTP as "data backbone". In my opinion this solves many many problems in data distribution and replication and that we should not reinvent the wheel. But that has been discussed over and over already. We (Raven and me) have just setup a 2-node NNTP network and we have started exploring the whole thing. Reports will follow Smile

3. Our XML should be an extension of the gc.com extended GPX format, such that all software built for gc.com-GPX will accept our data as well. You may look for further ideas on XML elements in another thread where this issue was discussed and in the Navicache DTD.
Legal issues (concerning the gc.com-GPX) have been discussed, but I am very convinced that this can be no problem

Team BMW-Biker · Joined: 01 Oct 2003 Posts: 209

Hi Vinnie,

Team BMW-Biker · Joined: 01 Oct 2003 Posts: 209

Hi,

I've take some changes ...
http://www.dietz.homeip.net/geocache/opencaching/ideas002/changes.htm

... Oli

raven · Joined: 29 Sep 2003 Posts: 84 Location: Bielefeld, Germany

yep, thats a really useful ressource you built there.
here's the point in which i differ from your ideas, or maybe just don't understand them clearly ;)

- full nodes vs. regional servers
your system uses "server-user-areas", where each server/node is the "owner" of the data from users from his geographical area. i imagine a system where the data is totally the same on all nodes and no node is the "owner" of some data this has several reasons:
- it's failsafe, if one node goes down for whatever reasons, the users can instantly switch to another node and still have all their caches, logs etc. available to them. ok, i looked a bit closer at your proposal and saw that you also want to do a full replication. but if all data is available on all nodes, why do you need one node to be the "owner" of the data?
- the nodes don't need to be for a special region. there should be the possibility of several nodes that are in the same geographical region. if opencaching.com wants "their" node to cater to geocachers around the world (like gc.com and navicache do now), they are free to do so. if geocaching.de/opencaching.de wants their frontend and features to be the best thing for german-speaking geocachers, they can do so, too. but still, all caches are available from all these sites.
- the user gets the freedom to choose what node he wants to use depending on the features its frontend offers etc., not based on what node is in his region.
if opencaching.com will run a node in the U.S. and navicache maybe (that is only a speculation!) switches to being an opencaching-node, too, that's no problem. competition is a good thing and the user is free to choose whatever node he wants to use (i think i'm repeating myself...).

the most important thing is the bold question up there, i'm a bit confused. i don't really understand why you make that regional distinction, yet you say, that all nodes should have access to all data... i don't think the scheme of "server-user-areas" isn't really neccessary, of course a node can make a frontend that is specifically for users of their region, but that doesn't have anything to do with the data itself... please explain why you think it is important for making replication easier to have an owner of a dataset. if i create a cache on one node, but update it on another, which node would be the owner then?
_________________
Quoth the raven, "Nevermore"

Team BMW-Biker · Joined: 01 Oct 2003 Posts: 209

Hello,

thank you very much for critical feedback !

Renegade Knight · Joined: 15 Sep 2003 Posts: 151

One thing to keep in mind is that as opencaching comes into being that several things must be balanced.

This is a partial list. More things will come along as we develop this organization.

1) Putting data into the hands of geocachers.
2) Protecting the data from people who don’t need it or who would cause other problems.
3) Landowner concerns. Part of why we are doing this is to become self regulating rather than having landowenrs ban geocaching as a nuisance or make the rules so restrictive that they might as well have banned it.
4) Finance the site. Many members make a site inexpensive for everyone who uses it. There does need to be a site and it does have to make enough money to pay for bandwidth, servers, continued development, and a myriad of other things that will come up. The site has to give the members a reason to come here, or they won’t.

The solution to one problem can prevent another problem from being solved. The site needs to address the needs of the entire community by looking at solutions until the one that does the best overall job is decided on.

If we pick out a distributed network and learn that it prevents us from accomplishing something else that is important that we didn’t think of until down the road, all that work will be scrapped and we will start over.

The site can be built using a conservative model that allows the most future flexibility. By doing that we get a site up ASAP and yet as the organization grows it can continue to solve problems as they come along. Since the organization will be owned by geocachers it will now be the community as a whole that directs the long run vision.

The question is what is a conservative model that allows the most flexablity to solve problems as they come along. The goal is to not have to scrap the work we do and start over should a problem come along that needed a different site model to solve.

SpinWebby · Joined: 12 Oct 2003 Posts: 1 Location: Des Moines, Iowa

First, I must admit, I haven't done alot of reading here as yet, but RSS may be an option. Its a combination of news reader and email using XML. I use FeedDemon to read my messages. I'll go explore Navicache for now, but thought I'd mention it in case no one has yet.
_________________
=======================
SpinWebby
Great Plains Geocaching
www.gpgeocaching.com
=======================