Update Time Stamp

According to the RDF samples, there is not a "site last updated/validated" timestamp, yet it seems that the official RDF documenation allows such a date. It would seem that this could be implimented for "newly touched" entries only, and would be extremely usefull within 6 to 12 months.

This would allow users of the RDF data (if not direct users of dmoz) to be able to judge the summary of a site (and likelyhood of accuracy) by how old the listing update is. I initially felt this would be most usefull at ChefMoz.org, but since have come to think it would be equally usefull directly in the dmoz.

Hmm, some problems. A site listed 4 years ago could still be valid, while a site listed just yesterday might not exist today.

Additionally, sites tend to be listed once and then the entry never touched again until either a content change is noticed needing an update to the description, or a spelling error in the original entry is found, or the site is moved around due to a category reorganisation, and so on.

And, how would you handle a site that is listed in two categories, one where it has been listed for 4 years, and another where it was only listed for the very first time just last month? Which date would you include on each entry? If the date is the last edited per category date then the date really doesn't mean much about the up-to-date-ness of the site entries, as both are up to date, even though one might show a very old date.

I think GAVollink has a good point though. That might not be a bad thing to include in the RDF, considering we do include the last update date of categories. Also, what I do know of the system makes me believe it wouldn't be impossible to implement.

However, there are currently several technical changes underway that must be finished before new features will be considered and there are a number of new features that editors need even more than that, so it may be some time before anything like this would be considered, much less implemented.

When there comes a time where I think it might be considered, though, I'll suggest it myself.

giz,
Yes, I would consider the date/time (both in terms of the RDF standard - as well as the dmoz use of the standard) to be a per-category, per-entry stamp. After-all, as the editor guidelines clearly point out, multiple category listings should be the exception. Your point about a 4 year old listing being more-valid than a recent listing may be true. The invalid listing (through abuse and automated link checks) is more likely to be weeded out quickly than a 'stale' listing though.

Stale is to say that a site that has not garnered a second look from the editor for 4 years maybe has not changed itself in 4 years (maybe it has). It certainly wouldn't be something that would need to be visibly listed, but could be very usefull as another sort/ranking option by the downstream RDF data users (search engine sites), not to mention something the the editors themselves may find comes in handy.

theseeker,
You are certainly right that there are more pressing matters, but I'm very glad to see that my suggestion was not summarily dismissed (though I won't personally feel insulted if it never happens - it's the nature of software). Either way, I thank you for your support.

All,
Any idea of a place where I could suggest the same for the specific ChefMoz RDF data? I think it would be even more useful to end-users for the ChefMoz data, but I cannot find a similar forum or suggestion box for the ChefMoz site.

Thank you for your time.

to be a per-category, per-entry stamp.

Click to expand...

Well, from this point on I think this has to be discussed internally. Only thing I want to add: I don't like it, if people from the outside are enabled to do detailed profiling about the work in special areas (aka of special editors). I would not mind if a per-category stamp was included in the RDF (it is visble on the pages anyway), but would object to a per-entry stamp, either in the RDF or on the pages.

(Since it is somehow related Same goes for the already proposed "New Entry" flag - in my eyes a site that is new is in no way better than one that is already listed. If someone really wants this flag, he is invited to regularly download the RDF and create it himself by comparing them.

Stale is to say that a site that has not garnered a second look from the editor for 4 years

Click to expand...

You won't see that anyway. A site that has been looked at and not changed (becuase there is nothing to change) would not receive a new timestamp. And in my experience this is the majority of sites that I rereview

that my suggestion was not summarily dismissed

Click to expand...

We only do that to two types of suggestions:
1) Those that were already dicussed over and ober. ["Will there be review notification e-mail?" - "No. Maybe in some distant future."]
2) Proposals that are contrary to our policies. ["I have a cool idea how ODP could benefit from taking money from [...]!" - "*Sigh* We had a poll that voted the fact that ODP is not powered by money on #1 of the best features list. AOL/Netscape released a note stating that they will keep it free."

Stale is to say that a site that has not garnered a second look from the editor for 4 years maybe has not changed itself in 4 years

Click to expand...

Just wanted to comment that a site that has not been updated in 4 years is not necessarily stale. There are many, many webpages that provide good detailed information on a topic which is still as current and appropriate today as when it was written many years ago. Age of information alone is not an indicator of worth.