Site Map of Developer Resources

[Guest tohotom]

Forgive me if I am not seeing the forest for the trees - but I was wondering if there was a Site Map of the developer resources... I'm not talking about a site map to the directory structure - rather a single location one might look to access the various documents related to development stored on the DMOZ servers...

 

I have found many of the links spread out over several pages

 

e.g. http://rdf.dmoz.org/ or http://rdf.dmoz.org/rdf/Changes.html

or http://rdf.dmoz.org/rdf/tags.html etc.

 

- but I wanted to make sure that I wasn't missing any.

 

Also I was wondering if there is any documentation on the flat file system used by dmoz. I'm fairly certain that the RDF is not used as the basis of the dmoz site. The reason I ask is that I was wondering if anyone has attempted to create a true dmoz clone based on the dmoz content. I wouldn't mind giving it a try just for fun.

[bobrat]

I don't see how that is possible, the RDF only contains part of the DMOZ database, it does not contain the unreviewed sites, editor notes, and all kinds of other stuff. So it would not be possible to reconstruct the database from the RDF dump.

[Guest tohotom]

...it would not be possible to reconstruct the database from the RDF dump.

 

Quite right - you couldn't.

 

But that wasn't my question. The second part of my question was - is the flat file structure that DMOZ uses available (or at least documented) somewhere? Given the file structure - and a dump of the contents - one could conceiveably build a custom set of tools to edit its contents to mimic DMOZ.

 

Many people have already gone ahead and used the dumps to populate realtional databases and then built application to serve the data out. (in many cases mysql databases with perl and PHP serving out content)

 

I was curious whether anyone has tried to do what DMOZ does the same way as DMOZ does.

 

Now of course many might argure (and indeed have argued) "why bother? - the relational database model is the way to go - i don't know why DMOZ hasn't addopted it yet" I believe that is a valid question - BUT I am still interested in getting to know the guts of DMOZ and see if there are other applications of using the same system used to date.

 

One way to get to know it is copy it and then modify it - change it - learn from it and build something new out of it.

 

Thanks for the input. Sorry for the long winded clarification.

[windharp]

No, neither the sources nor the structure are available to the public. Since Netscape has bought the sources from the original developers, they hold the rights on them.

 

was curious whether anyone has tried to do what DMOZ does the same way as DMOZ does.

 

You already said it: Who should want that, except maybe some editors? Some senior editors who clearly remember the outages we had lately , surely would like the thought of an emergency system. Something we could use if the main system is down again. But since that is an enormous amount of work, nothing like this has been relaized by now.

 

Except from DMOZ, there are some smaller projects ( http://www.chefmoz.org / http://www.open-site.org/ / http://www.musicmoz.org/ ) which AFAIK use modified versions of the same system. But not with the same data of course.

[Guest tohotom]

Thanks for the info.

 

Seeing as Netscape is no stranger to open source I thought there might be access to the guts of DMOZ. I have taken a look at the source for the encyclopedia project but it is different even in its very nature (still not a bad starting point though).

 

Thanks again.

[totalxsive]

We use Berkeley DB as the database format, IIRC.

[Guest tohotom]

Sorry. Who are we? Who are we that use BerkelyDB?

 

<getting off topic here>

I've been wanting to play with BerkelyDBXML - any word on its ease of use?

</getting off topic here>

[totalxsive]

We = the ODP. All the directory data is in BerkeleyDB format, I believe. Sorry for the confusion.

[Guest tohotom]

Awesome. Thanks for the information.

[theseeker]

Actually, something I read from the founder once implied most of it was just flat files with a few BerkeleyDBs. Not that it's too important, as there's little difference.

 

The ODP software isn't open source, though chefmoz.org code is and that's also a Netscape project. Someday, many of us hope it will be, but there's appears to be a lot of red tape, and possibly some things I don't know in the way, so I wouldn't look for it to happen soon.