RIP DMOZ: The Open Directory Project Is Closing(searchengineland.com)
searchengineland.com
RIP DMOZ: The Open Directory Project Is Closing
http://searchengineland.com/rip-dmoz-open-directory-project-closing-270291
24 comments
I do agree that dmoz did a very terrible job at fulfilling the requests. That ultimately was the downfall and why so many didn't like DMOZ.
But there were way to get in, no matter what (unfortunately).
But there were way to get in, no matter what (unfortunately).
Exactly my experience as well.
the latest iteration of this idea seems to be the "awesome" lists on github [https://github.com/sindresorhus/awesome], which are a bit more decentralised and crowd-sourced while still being human-curated.
Exactly - the first time I looked at https://github.com/bayandin/awesome-awesomeness my immediate thought was that we're recreating Yahoo, or maybe publishing a webpage with all of your bookmarks is making a comeback.
I found DMOZ pretty well curated and responsive to requests... 15 years ago. If you tried to submit something recently though then, yeah, you were submitting to a zombie site.
Wikipedia, as it happens, curated some data illustrating the growth prior to ~2004-05 followed by complete stagnation; see the graph in this section: https://en.wikipedia.org/wiki/DMOZ#History
Wikipedia, as it happens, curated some data illustrating the growth prior to ~2004-05 followed by complete stagnation; see the graph in this section: https://en.wikipedia.org/wiki/DMOZ#History
Honestly, I'm surprised it took so long to get to this point. The directory has been declining in usefulness for the last decade or so to be honest.
And that's all because they can't seem to attract enough editors to keep it relevant. I mean, look at the games listing there, then drill down to consoles. Notice anything off?
Yeah, all the subcategories are from circa 2006. That area of the site is so poorly maintained that they don't even list categories for the 3DS, Vita, Wii U, Xbox One, PlayStation 4 or Nintendo Switch.
And that's the case with a lot of categories now. The TV section doesn't seem to have an easy to find category for Game of Thrones or the Walking Dead. The US presidents one doesn't have a category for Donald Trump (though it still has one for him as a candidate elsewhere). Basically, most of the site's categories are just woefully outdated, with the sites found in them not being much better.
The DMOZ has just been too infrequently updated and poorly maintained to really be useful for people nowadays.
So yeah, not surprising they had to call it quits. A site which doesn't have relevant information for most people just isn't worth keeping around.
And that's all because they can't seem to attract enough editors to keep it relevant. I mean, look at the games listing there, then drill down to consoles. Notice anything off?
Yeah, all the subcategories are from circa 2006. That area of the site is so poorly maintained that they don't even list categories for the 3DS, Vita, Wii U, Xbox One, PlayStation 4 or Nintendo Switch.
And that's the case with a lot of categories now. The TV section doesn't seem to have an easy to find category for Game of Thrones or the Walking Dead. The US presidents one doesn't have a category for Donald Trump (though it still has one for him as a candidate elsewhere). Basically, most of the site's categories are just woefully outdated, with the sites found in them not being much better.
The DMOZ has just been too infrequently updated and poorly maintained to really be useful for people nowadays.
So yeah, not surprising they had to call it quits. A site which doesn't have relevant information for most people just isn't worth keeping around.
I gave up on DMOZ when I repeatedly tried to volunteer to manage categories (that I happened to be a reasonably recognizable expert on), because they weren't being actively maintained, and it took months to get replies, if ever a reply arrived, at all.
That was many years ago, when it still had some relevancy to search engines. But, even then, it was a zombie. There were some outposts of activity, but it just wasn't a priority after all those acquisitions, and nobody felt enough ownership to maintain it. I don't know anything about the insider decisions or how it was managed or whatever, I just know that when I was enthusiastic about helping, I was ignored. I'd pretty much forgotten all about it and assumed it died years ago.
I like the idea, and I wish there were a good public data set like it. But, I guess since no one figured out how to make money with it, no one cared to make it work.
That was many years ago, when it still had some relevancy to search engines. But, even then, it was a zombie. There were some outposts of activity, but it just wasn't a priority after all those acquisitions, and nobody felt enough ownership to maintain it. I don't know anything about the insider decisions or how it was managed or whatever, I just know that when I was enthusiastic about helping, I was ignored. I'd pretty much forgotten all about it and assumed it died years ago.
I like the idea, and I wish there were a good public data set like it. But, I guess since no one figured out how to make money with it, no one cared to make it work.
For those who still see value in a curated directory, please help me combine efforts of people willing to contribute to a DMOZ revival by filling out this form: https://goo.gl/forms/0cFa3bXLGoo12M6l2
Such a pity. I have always used DMOZ as a source for seeding any search engine I am playing around with as you get a good breadth of sites to start crawling.
Half tempted to create an open clone to keep it going. Any suggestions for a name would be appreciated.
EDIT - Ordered freemoz.org and will attempt to get somthing going ASAP
Half tempted to create an open clone to keep it going. Any suggestions for a name would be appreciated.
EDIT - Ordered freemoz.org and will attempt to get somthing going ASAP
The blekko slashtag dataset is a lot better, just sayin: https://github.com/blekko/slashtag-data
(the founders of dmoz were the founders of blekko)
(the founders of dmoz were the founders of blekko)
Looks like a good data source to seed with.
What’s sad here is that we’re now in an era where AI and other computers are now categorizing and approving websites. It’s no longer a volunteer-editor driven web world. Now, only a search engine algorithm decides whether or not a site is trusted. It used to be that a human approved a site, and a search engine trusted that human's opinion. No longer.
Personally I think it's a good thing. No opinions involved and everyone is treated equally.
Also sad from the AI perspective, this was a great, public human-labeled dataset.
It's funny, right now there are two comments:
>> What’s sad here is that we’re now in an era where AI and other computers are now categorizing and approving websites.
and
>> Good riddance! It was crap anyway.
I kind of agree with both of them. It does make me sad to see DMOZ go, but at the same time, maybe, just maybe it's ok to admit algorithms do a better job of this stuff. As a librarian, I want to think we humans are better at categorizing and cataloging everything.
>> What’s sad here is that we’re now in an era where AI and other computers are now categorizing and approving websites.
and
>> Good riddance! It was crap anyway.
I kind of agree with both of them. It does make me sad to see DMOZ go, but at the same time, maybe, just maybe it's ok to admit algorithms do a better job of this stuff. As a librarian, I want to think we humans are better at categorizing and cataloging everything.
I don't think DMOZ was the best example of a well curated collection though. It might have lived had it some direction and actual upkeep.
Last I looked at it, half the outbound links were dead, which could have been managed automatically.
Credible anecdotes say reasonable new submissions were routinely ignored.
There are curated niche lists, many on GitHub, that are better than Google for certain things.
Last I looked at it, half the outbound links were dead, which could have been managed automatically.
Credible anecdotes say reasonable new submissions were routinely ignored.
There are curated niche lists, many on GitHub, that are better than Google for certain things.
Apparently search engines, like Google, think that computers do a better job than humans. I still like the human element, even though it may not be as efficient.
Search engines are a specific type of tool where algorithms just make more sense. Between the amount of data that needs to be processed, and the relatively low impact of being incorrect. There are a lot of places Google wants to apply algorithms (like customer service, censorship products, and driving cars) where I do not believe it's warranted to replace humans with software.
On March 13, 2017 DMOZ should just approve all submissions. Especially those that have been waiting 7+ years to get their site approved.
No, that's a terrible idea. The vast majority of unapproved suggestions are likely to be crap (or just outright spam).
(And that's assuming that the site will stay up, in any form, after the published "closure" date. For all we know, it may simply go away.)
(And that's assuming that the site will stay up, in any form, after the published "closure" date. For all we know, it may simply go away.)
I was trying to be snarky with that comment. Ha
Contrast that with Google, who often have content of my sites returned in search results, within days of it being published—without me having had to lift a finger.
The idea of "human curation" is fine in principle but, as sites like DMOZ, Wikipedia, StackOverflow, etc show, it very often ends up being monopolised by a few self-important nobodies seeing it as a way to wield some authority in life.