<br><br><div class="gmail_quote">On Thu, May 15, 2008 at 4:14 PM, Paul Vixie <<a href="mailto:vixie@isc.org">vixie@isc.org</a>> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
"I love the smell of napalm in the morning...smells like...victory."<br>
<div class="Ih2E3d"><br>
<a href="mailto:jwales@wikia.com">jwales@wikia.com</a> (Jimmy Wales) writes:<br>
<br>
> Paul Vixie wrote:<br>
</div>> > 1. why is ISC's the only backend? ...<br>
<div class="Ih2E3d">><br>
> We strongly support this notion of backend syndication, and we are<br>
> hopeful that as the infrastructure and protocols mature, we will get more<br>
> and more people working with us on this.<br>
<br>
</div>working with which "us"? is it wikia's remit to recruit new co-backends,<br>
to add flooding features to the backend so that crawlers can bulk-share<br>
their findings with other crawlers? if so, please be as transparent as<br>
possible about it, up to and including writing "open letters" to prospects<br>
which are cc'd to search-l, and doing an RFC-like public review on search-l<br>
of the protocols and mechanisms that will be used at the backend-backend<br>
layer. this request follows from my observation that everybody is excited<br>
about search syndication, but the voices we most need to hear belong to<br>
people who don't want to be "along for the ride" with wikia as sole driver.<br>
<br>
(note, i am NOT saying that being a sole driver was wikia's intent, only<br>
that if somebody had other reasons to question this effort, they could use<br>
any lack of transparency as support for their assertion that wikia was the<br>
sole driver and the rest of us were along for the ride, untrue though it is.)<br>
<br>
the first step on this path would seem to be an exhaustive list of other<br>
search crawlers/indexers, with a confidence rating of each, such that we'd<br>
all know who the players are and we'd all know how likely it was that each<br>
of them would want to be part of a coordinated universal search syndicate.<br>
(for example, i'd assign confidence level "0%" to google.)<br>
<br>
> > 3. who is driving the syndication model? ...<br>
<div class="Ih2E3d">><br>
> We just transitioned our New York office to fulltime work on the search<br>
> project, and Dan Lewis is being put fulltime on the task of community<br>
> outreach: answering critics, beating the drum, and doing the detailed<br>
> work of working with inbound inquiries from potential partners who are<br>
> already interested, outreach to potential partners who are not yet<br>
> interested, etc.<br>
<br>
</div>i hope i get to meet dan, in my ISC role. i hope everybody on search-l gets<br>
to meet dan or at least hear from dan, in our roles as interested supporters<br>
in the syndicated search model that wikia is championing here.</blockquote><div> </div><div>well what cant you tell from his user page (<a href="http://www.wikia.com/wiki/User:DNL">http://www.wikia.com/wiki/User:DNL</a>) :-p i have to agree with this, that it would be nice to talk/hear from all the staff on the project, maybe a nice new thread, telling us who they are and what they do etc, cos if im honest i have no idea who "david" is. <br>
</div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>
> > 4. what else is jer working on? ...<br>
<div class="Ih2E3d">><br>
> Jer is fulltime on search, as are several others. Dennis, Seth,<br>
> Jeffrey, David, Aaron, Dan... I feel that I am forgetting someone.<br>
><br>
> We are prepared to ramp up our commitment as we start to get traction,<br>
> as well. At the present time, every time I ask the team what we need to<br>
> buy, they say "not yet, we are coding". :)<br>
<br>
</div>at the backend, i want to suggest that even though we have another thousand<br>
or so donated servers in our warehouse, they are power-inefficent anti-green,<br>
and it might be good to cap the power utilization at 20kW (where it is now)<br>
and eventually buy more modern 1U's as a way to get more computrons here. we<br>
could throw another 10kW at it during a transition, to avoid downtime, but in<br>
the long run we need all Pentium III chips to move to museums, or landfills,<br>
and starting pretty soon, Pentium 4 chips as well.</blockquote><div>is there nothing else that could be done with those servers? leaving them in a warehouse seems wierd to me, you could always mail them to ppl :-p<br> </div>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>
<br>
> > 5. who else is working on this, outside of wikia? ...<br>
<div class="Ih2E3d">><br>
> Strong preference that we get lots of people coding on a fully open<br>
> system, as they like it. I think so far we have not done a great job of<br>
> outreach, but then again, we have not had everything in place to get<br>
> people oriented and started.<br>
><br>
> Also, we view ourselves as a "good neighbor" part of the existing Nutch<br>
> project: Dennis is a Nutch committer who is starting to work on a set of<br>
> ideas he is calling "Nutch 2.0".<br>
<br>
</div>domain names are brands, and <a href="http://swlabs.org" target="_blank">swlabs.org</a> isn't a meaningful one for an SVN<br>
used by this project. i suggest moving it to <a href="http://svn.search.wikia.com" target="_blank">svn.search.wikia.com</a> or<br>
<a href="http://svn.search.isc.org" target="_blank">svn.search.isc.org</a> or similar, and adding a bug tracker, or installing a<br>
sourceforge instance, or similar. a wiki by itself does not a community<br>
make.<br>
</blockquote><div><br>this is one thing that we are working on atm. we are setting up a vanilla drupal instance and are going to get it all sorted out with the svn, bug tracker, blog etc etc, and hoepfully that will be up in the next few weeks :-D<br>
</div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>
> > 6. where are the mini-articles stored? ...<br>
<div class="Ih2E3d">><br>
> It's all GFDL, and we make available database dumps. We would have to<br>
> consider a "flood" of incoming data from a community/editorial point of<br>
> view, but totally welcome it, and are totally committed to sharing<br>
> everything extremely liberally.<br>
<br>
</div>that's good to hear. it'll be even better to have a PoC. who here can run<br>
a <a href="http://non-wikia.com" target="_blank">non-wikia.com</a> social search backend, which can consume these database dumps<br>
and allow local editing/authorship of mini-articles, so that the editorial<br>
considerations of this kind of flooding can be learned while the monster is<br>
still in its larval stage? note that if it's free and open, ISC could host<br>
it, and i'd even be willing to throw a couple more P-III's and P4's on the<br>
stove for it, as long as nobody at ISC has to do any sysadmin on them.</blockquote><div><br>maybe use the <a href="http://swlabs.org">swlabs.org</a> for this?<br> <br></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>
<br>
> > 7. given that the idea of "taking on google" is silly, ...<br>
<div class="Ih2E3d">><br>
> I think this is a really great question. :)<br>
><br>
> One of the things I have been arguing is that we are no threat to google<br>
> even if we are wildly successful at "making search part of the internet<br>
> infrastructure" as you put it...<br>
><br>
> Google's brand is tied up with search, but Google's business is not<br>
> searhc, per se, but the matching of advertisements to user actions and<br>
> intentions online. The threat to google is not an open source<br>
> alternative that helps 1,000 small competitors to flourish, but a single<br>
> large proprietary competitor (Powerset?) that captures enough market<br>
> share to take away the advertising marketplace.<br>
><br>
> 1,000 small competitors are much more likely to simply partner with<br>
> Google for ad revenues, because buyers go where the sellers are, and<br>
> sellers go where the buyers are.<br>
<br>
</div>you've gotta write a whitepaper to that effect. not just an e-mail message<br>
and not a wiki article, but an honest to gods high end marketing hit piece<br>
which you roll up and brandish at audiences when you speak about wikia<br>
search. the noncredibility of the early claims about "taking on google" is<br>
the biggest weakness wikia search has got.<br>
<div class="Ih2E3d"><br>
> > 8. has anybody reached out to yahoo and microsoft to see if they'd like to<br>
> > join this effort or at least sponsor it, since as #2 and #3 in internet<br>
> > search today, they're the ones with the most to gain if we change the game.<br>
> > and if nobody's doing this now, and i did it, what would wikia say about<br>
> > sharing the sponsorship burden with other players, perhaps larger players?<br>
><br>
> We have done some of this, and would be eager to support you if you want<br>
> to help us with it. We can talk privately about the status of current<br>
> talks but there is nothing to report and nothing likely to happen right<br>
> away... but there are a lot of interested parties in the industry.<br>
<br>
</div>ok, have dan call me, i understand that strategic relationships like this are<br>
covered by NDA during negotiations, and that i need to know what irons are<br>
in which fires before i go poking around. note that more whitepapers about<br>
the design and goals of all this would help me in that kind of outreach.<br>
--<br>
<font color="#888888">Paul Vixie<br>
</font><div><div></div><div class="Wj3C7c">_______________________________________________<br>
Wikia Search mailing list<br>
<a href="http://alpha.search.wikia.com/" target="_blank">http://alpha.search.wikia.com/</a><br>
Change options or unsubscribe: <a href="http://lists.wikia.com/mailman/options/search-l" target="_blank">http://lists.wikia.com/mailman/options/search-l</a><br>
</div></div></blockquote></div><br>