Slashdot Mirror
Altavista - Open Sourced UPDATED
Hi there,
The new affiliate program is based on a syndicated model, where we are providing the HTML and search box interface to Web sites, large and small to enable their users to access AltaVista's premier services including search, stock quotes, language translation, multimedia, news and discussion group content. Users can choose from an array of search boxes that fit their personal brand. The search box then acts as a gateway for users to tap into our robust index. Those Web sites that choose to participate inAltaVista's Affiliate Network will receive three cents per click-through when their users access AltaVista branded services. To learn more about the affiliate program visit http://doc.altavista.com/affiliate/.
This program is not to be confused with the other products we provide that do allow customers to access our source code and build their own search products. We provide an array of tools that allow customers to create their own customized engines and can be accessed at http://doc.altavista.com/business_solutions/search_products/search_intranet/ intranet_intro.shtml.
The article doesn't mention that Altavista is going to open source theit source code. It says that the source code will be given to applicants who can present a "real" web site they are running.
It would be great news if the source code would truly be GP-licensed or whatever OS license model Altavista would choose, but I doubt they will do that. Also remember that the search engine that you can obtain from Alta Vista is not the same as the one that's running their web site. It used to be downloadable before, and my information is that it does not scale as good as the AltaVista.com web page search engine does.
Yes, you are right there. -- Another glass of champagne?
..I needed something to do with that huge pile of Alphas I had just sitting in the corner gathering dust...
DrLunch.com The site that tells you what's for lunch!
and language translation service
Give a man a babelfish, he understands for a day. GPL the babelfish, then embed it, and he gets a real cool palm app next year.
(Note: the letters GPL do not appear in the article, nor is app a real word)
+&x
I think that this, and the Netscape Communicator/Mozilla effort, mark a sea change when it comes to software development. When a body of code had no value to a company, they used to quietly bury the body. Now, more and more companies release the code to the community, gaining a huge investment in goodwill.
Philosophically, it's a move away from the Marxist conception of value (which is paradoxically de rigeur in US business circles), where anything requiring work gained in value. This isn't obviously false, but false it is. Value, at least in the capitalist system, is based on the ability to sell at a profit. If you cannot sell at a profit, either using the technology or the technology itself, then it's valueless.
Most businessmen stick to the assumption that all the work put into this or that piece has made it valuable and worthy of protection. New businessmen are thinking it through - not everything that takes work is valuable, and protecting something valueless is a waste of effort. By open-sourcing the work, they turn a loss into a gain.
--
--
There is no premature anti-fascism. -Ernest Hemingway
Having said that, note that Alta Vista are keeping their actual database to themselves - this is the one real asset other than their brand which they possess. Taking these two together, we see a core competence (i.e. leveraging them provides a return disproportionate to effort in relation to the market sector), which is now the basis of their revenue plan.
While I'm psyched to see this happening, a few things pop into my mind.
Isn't Alta Vista on track to go public shortly? Without seeing their licensing terms, it's difficult to tell if this is a sincere move, or if it's just so they can be an "open source dot com". I guess we'll find out soon.
Not to be critical, because I've always wanted to have a commercial quality search engine, but Alta Vista's pretty advanced and just about the fastest search engine on the internet. What are they looking to gain by doing this? They don't need more developers, unless they're looking at laying off their team in hopes of the community shouldering the development effort for them.
And lastly, I doubt that they'll go with the GPL. No major vendor that's released their core product as opensource has had the guts to go fully to the GPL. But then I guess AltaVista's quite different than an operating system or application, as in it will still take years and millions of dollars for anyone to catch up with them in terms of eyeballs, which is where their money comes from these days...
Maybe slashdot could start a search engine though. Just a privatized one, picking up things like story links, member home pages, resume's etc... It could open some possibilities, but I still can't see how any of them would benefit alta vista.
My $0.02
StarTrek.org Free Webmail
...from what I can tell, at least.
I think the journalist in the above article has got it all wrong. I can't see anything on the Altavista site regarding the source being opened up.
What they are doing, however, is running an affiliate program that pays web site owners a commission (3c per click through, in this case) for each user that is referred to one of the various AltaVista search facilities from another web page (that has applied, and been approved, for this program).
This is not anything particularly new - as it happens, GoTo.com has been running a very similar scheme for quite a while. GoTo.com's program, as well as AltaVista's, is managed by befree.net.
So, you sign up, put the search boxes on your site, typically pointing to a unique url so they can track your referals, and start collecting money. You don't host their search engine - merely point to it.
If, on the other hand, i've missed something, I would appreciate any pointers to the actual AltaVista source code.
...j
One fact which all the search engines must
realize, as well as cache companies like
Inktomi and Akamai, is that the Internet is
becoming increasingly dynamically-generated,
personalized, and transactional -- exactly the
kind of content least suited for static
spider-driven search engines and static cache
technology.
Perhaps this will be the first Internet
subcategory to fall from vastly overinflated
stock valuations due to technical change.
The article actually doesn't expand on the source code freedom beyond the mention in the title, which is more than a little frustrating.
This is interesting, but really all they are doindg is starting yet another affiliate program, albeit with a source code giveaway. As people have already mentioned, the source is not open in the GPL/BSD/artistic license sense, rather they're letting people who enter their affiliate program use it. This is almost identical to the model used by Infoseek and Excite, except that people get to compile the source, and they pay by being an advertising shill instead of just buying a license to use the search engine.
Yours truly,
Mr. X
Altavista has been offering its search engine (executables) for download for quite a while here. It is not the same engine they use on their web site.
Here is the press release from Altavista.
Yes, you are right there. -- Another glass of champagne?
The Altavista Affiliate Program Agreement states:
(Italics added for emphasis.)
I don't really see how Altavista giving people some HTML source - no matter how "proprietary" - counts as them opening their source code to their search engine, which seems to be what the article is trying to imply. Many other sites - Lycos, for one - have had similar programs in the past, though the $0.03 per clickthrough sounds like a different twist.
Chalk it up as effective marketing - they put the words "open" and "source" in the same sentence, and managed to generate the expected amount of talk about what is essentially a non-event. 'Course, I may have missed the "Download Altavista search engine source here!" link on their site, but I don't think so :-)
"Great men are not always wise: neither do the aged understand judgement." Job 32:9
Be an Internet Search Partner (from the home page). Right down the bottom of that page, you'll see a link to the AltaVista Affiliate Network, which is what the article is talking about.
T&C's, FAQs, etc., can be found at the second URL.
...j
...and here is why: "To join the 'network,' a site must demonstrate that it is a valid, working Web site that is updated on an ongoing basis."
but... "AltaVista plans to target the owners of personal home pages in the future." so maybe there is hope that they will start releasing the code to the rest of us soon.
(darren)
Do we get the super-powerful keyword-to-marketing engine that AV runs with our favorite doubleclick? Or is it a plug-in? Can we plug in the slash ad code (when it gets released? ;)
Seriously, tho, I'll believe this has happened when I have code in hand running on my intranet, without co-branding or marketing.
Returned Peace Corps IT Volunteer
Can't you read? From the article refrenced:
...and will start paying sites that successfully refer people to the portal."
"The portal will begin giving away the source code for its search engine..."
Th above part says: they will give away source code.
Now for the AND part (that means they will be doing both things).
"
This part, with refrences to other parts of the article, is the part about who gets to be PAID. Noplace in the article does it say you have to be "worthy" in some way to get the source code.
And this tripe gets a score of 3?
Eve Fairbanks says I drive a hybrid!LOL
Sigh... Pecavi
All opinions are my own - until criticized
UPDATED February 1, 2000
In a shameless attempt to gain attention from popular news sites that will post any story that includes the phrase "Open Source," internet portal site AltaVista announced that it will begin giving away the source code to it's search engine while actually doing no such thing.
Today Altavista rolled out an affiliate program which "allows" web sites to include html that links to the altavista search engine. Altavista did not address the question of why this is interesting, when people have been including search engine textboxes on their pages since 1994. Instead they prominently featured the phrase "Open Source" in the press release title, and went on to not mention even once how "allowing users to include html" could be interpreted as "releasing the source code to it's search engine."
You may still download a crippled trial version of Altavista's intranet search tools, which you may uncripple for a registration fee. But the bold maneuver of issuing a press release that uses the words "open source" is taking the internet by storm.
"We see this press release as an unprecedented opportunity to leverage traffic from weblogs that don't do even the most rudimentary fact-checking," said an Altavista spokesman. "And we know for a fact that there are some very high traffic sites which auto-post any press release that uses the words 'open source,' without a human editor even being involved."
The perl scripts which post content at the popular computer news site Slashdot declined to comment on the allegations that no human is involved in story posting anymore, saying only, "It looks like a hole in the GNU GPL [may allow] people to practically turn GNU-free software into proprietary software..."
----------------
Note: This is intended mostly to be a flame at altavista, and to mildly poke fun at slashdot. Please take it for the humor it is. Thanks.
--The Mgmt.
----------------
Wish you could moderate the submission queue?
There is no K5 cabal.
I am not the real rusty.
For people like me, who are the vast majority of people who want to use the alta vista search engine, the open sourcing of the product (if they will be open sourcing it) is terrific news.
It seems you have missed the point of Open Source entirely. OSS is not about "Anyone can get the source code", it's about "Anyone can get the source code, modify it, publish the results and do what the heck he wants with it (well, almost)".
With respect to being "worthy": I read the press release as stating that you will have to become an "Affiliate" before you get the source code; and for becoming an affiliate you have to present a web site you are running.
The press release is ambigious about this, so maybe I am wrong. (But if I am wrong: where is the download page for the source code?)
Yes, you are right there. -- Another glass of champagne?
My little mistake seems to be a little /. bug too. I entered the url as href='http://www.google.com' but that came out as href="'http://www.google.com'"
That is: slashdot added the double quotes making me look really stupid. Next the script will automagically add "F1R57 P057" and Natalie Portman to the first five post on any subject :-)
All opinions are my own - until criticized
If enough people refuse to use the AltaVista service, =and= let them know why, they may either be pushed into apologising or releasing =some= source, which is better than nothing at all.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
This has been suggested several times before on slashdot, and is as unlikely now as ever. Parallelizing a search engine would probably make it slower, as the latency between machines on the internet (not to mention the bandwidth) is terrible. Imagine if I search for United States and get 5 million hits on each from different machines. You then have to transfer the hitlists to one of the machines and do the comparison. You just got killed on your fast clause.
What would be more useful is a cluster of machines, each having the whole database. You would have to update all machines every night, but you might gain something from this approach.
--
Mike Mangino Consultant, Analysts International
Mike Mangino
mmangino@acm.org
Open Source is very open (forgive the pun) in terms of what it applies to. I can Open Source my new Widget2000 code and only sell it to people who buy Widget2000 and require them to not reveal the source code to anyone else. Or I can Open Source it and post it on a public FTP site.
;)
Every appliction may have a different terms of agreement for how the source is handled. If anything, this is incentive to READ before you buy or mess with someone's source code. You should be reading the licensing anyways, but I know MS liks to put it inside shrink wrap before selling things.
Bad Mojo
Bad Mojo
"If you can't win by reason, go for volume." -- Calvin
CMGI has finally seen fit to issue a press release. Surprise! they really are cheeky enough to suggest that a snippet of HTML constitutes open source.
http://biz.yahoo.com/prnews/00 0201/ca_altavis_1.html
Excerpt:
"The AltaVista Affiliate Network is leading the expansion of our
distinctive services throughout the Web, at a global scale," said
Rod Schrock, president and CEO of AltaVista Company. "This
program will effectively open source AltaVista Search and
translation services thereby extending our brand to the Internet
community."
Smug bastards.
bumppo
"Can't you read? From the article refrenced: ". Yeah, because we all know that ZDNet never makes mistakes or says untrue things, right?
Maybe you should go read the AltaVista press release. They don't say anything like "Here is our source code, and here is the license". They talk a lot about business solutions and how you can obtain a modified version of their search engine. In a 5 minute look around their site I wasn't able to see anything about what license they were planning to use, or even verify that the search engine they were allowing you to download was not in binary form.
I hope that ZDNet got their story right, but the way the said things I was expecting to see a press release from AltaVista saying "AltaVista open-sources search engine technology!". Not seeing that bothers me.
C'mon, you should know better than to accept at face value what you read in something at ZDNet without checking the source of the story.
Many entities "Give away" their source to certain individuals and entities. That is no way, shape, or form, means that they are opening up the source in general, as in an Open Source project.
-- I'm the root of all that's evil, but you can call me cookie..
The article just got yanked.
If anyone read through the affiliate program materials, there is no mention of any source code. Just the program itself. You put code to reference the toold/search on Altavista's site, and get paid $3 per clickthrough! Not all that lucrative unless you have alot of traffic.
penguinicide... when jumping out a window just won't do.
1) First, you have to have a personal home page which is regularly updated. I haven't updated my personal home page in quite a while- it was really just a very short exercise in HTML- but I still access it almost daily. I have links to Alta Vista and Deja.com which use the text-only interfaces. I use them all the time. It's here just in case anyone's interested. Does incrementing some web counter count as "regularly updated"?
2) For my personal home page, I was only allowed access to two scripts provided in my ISP's cgi-bin. If you wanted your own cgi-bin, you needed to buy a commercial account. How many other personal home pages have similar restrictions?
Even though I don't have anything worth indexing, I have to wonder just what Alta Vista's thinking with this.
Little Debian: America's #1 Snack Distro!
Read the open source definition.
http://www.opensource.org/
As most of us know already, major search engines use a hush-hush set of algorithms to reduce the number of spam'd enteries. (text set in the same color as the page background, really small text, keyword stuffing, etc.) By releasing the source to their engine, isn't AltaVista bascially giving the thieves keys to the treasure chest?
----
----
Am I the only one who thinks Microsoft is a misnomer? Perhaps Macrosoft would be a better fit?
But I sure don't see the source code anywhere on their sites. I did see a free version of their search engine which I downloaded and took a look at. That didn't have the source code for sure.
So where is it? I'm getting the feeling that somewhere along the line something didn't get translated.
What would be more useful is a cluster of machines, each having the whole database.
That was _exactly_ what I was proposing.
Each node indexes a subset of the web. It then passes on its local index to the other nodes that make up the cluster, so that each individual node can accumulate a copy of the master index. Search requests are then routed to the individual nodes based on how many requests each node is currently processing, how quickly they've been responding, etc, so that none of the disparate machines that make up the search cluster get overloaded with requests. And if one of the nodes is slow to respond, the portal could just resend the request to another available node.
Those that didn't want to host requests but still wanted to help with the effort could run spiders that index a small portion of the web and make that index available to the central cluster, thereby distributing the workload further (and allowing sites to index their local networks, forinstance, where they typically have higher speed connections, then dump the resulting index off to the cluster during offpeak hours). This allows local admins to index whatever portions of their site they want as often as they want just by setting up their webserver as an index-only node in the cluster. Get enough sites doing that, and you're going to get pretty up to date results.
The indexes would only get passed on to the nodes that request them...with a little effort, it wouldn't be hard to set up request routes to allow indexes to flow from node to node via the fastest network connections possible, minimizing crosstalk between nodes (you just pass any downstream indexes upstream and vice-versa, adding in your own locally aquired index along the way) -- have the indexes propigate like news articles.
The portal machine would only need to maintain a list of IP addresses weighted according to how big a load each site is willing and able to handle, so its processor load will be minimal. Put a moderate-sized machine on the end of a big network pipe, and a few thousand nodes scattered all over the net, and you might have one nice search engine.
Sorry if my previous post confused you.
-- WhiskeyJack
http://live.altavista.com/scripts/editorial.dll?ca tegoryid=&only=y&bfromind=980&eetype=art icle&render=y&eeid=1461716&avr=1
NetShadow
Well what would be really cool is if the engine could somehow detect a bad link as soon as a visitor fails to access it.. then the server double checks it the next day, and if it's still down, then POOF the link gets munched.
The point is to get some sort of real time checking and self monitoring. I've worked on numerous search engines, and most of them just have a batch verify command to parse the entire link database... there are better ways out there.
"Mind, as manifested by the capacity to make choices, is to some extent present in every electron." -Freeman Dyson
Without portal crap try this...
n c=iso88591&text=on&pg=aq
http://www.altavista.com/cgi-bin/query?opt=on&e
Text only search - no ads or other BS...
Later
How would the server verify this? Rather than presenting links to the actual sites, would it present a list of CGI's that a browser could then click, causing the server to verify the page prior to passing it back to the client? That would be a major CPU killer.
And what happens if a site is slashdotted or subjected to a DOS attack? In the first instance, they'ed risk removal because they were too popular. In the second instance they'ed risk removal because of an enemy or script kiddie...
I think Alta Vista's fine... The entire nature of the internet is based on there being no central authority... sites come and go. Pages change. And there isn't really a reasonable way of dealing with it, in my eyes.
Hey, it looks like their editors read
As a long time ZDuh watcher, I can tell you that it is quite common for them to simply disappear anything that is remotely embarrassing, and they NEVER acknowledge an error. Remember the 'Jesux' hoax? They don't.
======
"Rex unto my cleeb, and thou shalt have everlasting blort." - Zorp 3:16
Sacred cows make the best burgers.
"Welcome to you first day of Writing Press Releases for the Internet class. Now, just to make it clear. You do not need to know what buzzwords mean, just how to use them in a sentence. Everyone got that? Class dismissed, see ya next year."
+&x
I ask this becuase the distributed concept needs to be applied to web indexing.
I used to use altavista for all of my searching, but now that the search engines are lagging so far behind in indexing the content, we're forced to try multiple search engines to find what we're looking for. Yes, there are meta-search engines, but that's not what I'm looking for.
If they're letting us at the crawling/indexing code, maybe we can build a distributed indexing system along the lines of seti@home, mprime or one of the distributed.net projects.
Between that and an indexing system along the lines of the Library of Congress indexing code systems, we could jsut tame this beast yet.
i thought i might sign up for their affiliate program, just to get the occasional $.03 check from altavista, but when i saw the application form, i was appalled.
not only does it ask for all your contact information (several times) but it asks for your social security number. all on an insecure form!
- Entertaining Bits from the Ancient Kernel Tree
I didn't see any discussion of open sourcing their technology. It appears that they are merely providing a mechanism (in HTML) so your web site can "front" a query that is fielded and handled by Altavista, with a registration and payback mechanism.
Such a scheme (without the registration or payback mechanism) has been in place for quite a while at Google:
http://services.google.com/cobrand/fr ee_trial
Obviously, I would prefer to get money for the clickthroughs I generate, but I also want my clients to get great search results as well. At any rate, if I understand correctly, this does not appear to be the "open source" surprise represented in the article.
OK Good doctor. Can we end this thread now?
All opinions are my own - until criticized