DuckDuckGo might be useful in the search list that comes by default in Web Browser. It uses primarily F/OSS and now will donate 10% of it's "gross revenue" to F/OSS projects (http://www.gabrielweinberg.com/blog/2010/11/help-me-start-a-foss-tithing-movement.html)
What do you think?
Since I have configured it as my default search engine, I am in favor of this idea! Nevertheless, you forgot to mention another plus point for DuckDuckGo: "By default, DuckDuckGo does not collect or share personal information."
I have used it for months now.
I think we should make it our default search engine.
I have used it for months now, too. However, Trisquel has a dilemma in this.
Trisquel is a distribution that is completely FOSS. By having Trisquel
endorse it, we could explicitly be condoning the use of non-FOSS
applications. This is an ethically murky situation for a fully FOSS
distro, in my opinion.
> I think we should make it our default search engine.
Hmm, I don't know. I'm pretty sure our default search engine right now is Google. While I don't have any data to back this up, I imagine the software behind DuckDuckGo is "freer" than that behind the Google search engine.
On the other hand, Trisquel is a fully free operating system. You're completely right in saying that we shouldn't be recommending the use of non-free software, even if it is a web service like it is in this case.
But on the other other hand (that's three hands, now), people need to do web searches. It's one of the primary things people do on the web. Is there actually a fully free search engine out there? I'd like to hear about it if there is.
Then again, Google is synonymous with web searching to many people. Providing them with something like DuckDuckGo, which they've probably never heard of, could be confusing and off-putting.
Like I said, I don't know. This certainly is a tricky situation.
I agree that web searches are necessary. I propose this as a solution.
Google and DDG both contribute to the open-source community (not necessarily
the FOSS community) with money and developers, so we shouldn't prefer one
over the other. I propose that we should make no search engine the
"default." Rather, we should do something like Google Chrome does when it
starts up and ask the user what search engine they want to use. Optionally,
we could include links to information about the various search engines. This
way, Trisquel is not seen as being preferential (the search engine listings
would be in randomized order) and the community has new search options.
As for a FOSS search engine, I don't know of any FOSS internet search
engines; the closest thing to it would be htdig.
On Thu, Nov 11, 2010 at 6:17 PM, <akirashinigami [at] gmail [dot] com> wrote:
> Google. While I don't have any data to back this up, I imagine the software
> behind DuckDuckGo is "freer" than that behind the Google search engine.
> On the other hand, Trisquel is a fully free operating system. You're
> completely right in saying that we shouldn't be recommending the use of
> non-free software, even if it is a web service like it is in this case.
> But on the other other hand (that's three hands, now), people need to do
> web searches. It's one of the primary things people do on the web. Is
> there actually a fully free search engine out there? I'd like to hear about
> it if there is.
> Then again, Google is synonymous with web searching to many people.
> Providing them with something like DuckDuckGo, which they've probably never
> heard of, could be confusing and off-putting.
> Like I said, I don't know. This certainly is a tricky situation.
Well, I only suggested putting DDG in the list of search engines, but making it default would be interesting.
As earlier said, Google the search engine is certainly not "freer" than DDG; neither is Yahoo, Bing, or the other search engines in the list right now. Yet DDG is partly based on Yahoo BOSS, which seems to not really be free software at all: http://info.yahoo.com/legal/us/yahoo/search/bosstos/bosstos-2317.html
Of course, Google/Bing/Yahoo are just as restrictive. And most of the code on GitHub for DDG seems to be under a free software license: https://github.com/duckduckgo
The owner of the site has said his intention is to make DDG 100% F/OSS... over a period of time. Of course, history shows that intentions, no matter how well promised, often don't turn out as *intended*. ;) (Personally I trust DDG, and like the owner and the DDG community, but who am I to speak for others?)
There are already 100% free search engines, but most are incomplete, older, and require downloading their software. Others are just the API for a search engine. ( http://en.wikipedia.org/wiki/Category:Free_search_engine_software )
Having a window that pops up and asks what search engine you want to use (and allowing you to change the choice later) would be great. However, this still could be considered as indirectly promoting non-free software. After all, isn't Canonical's inclusion of Adobe Flash and Air in its repository considered "promotion of non-free software"? ( http://www.gnu.org/distros/common-distros.html#Ubuntu ) True, Flash is directly non-free and the search plugins aren't, but Canonical's decision was indirectly promoting it.
Besides, if you had the choice between a brand you're familiar with and some other brand that you've never heard of, which would you take?
This issue brings up a lot of ethical questions: Is having Google or another non-free site the default search promoting non-free software? What about "open-source" sites? What about part-free software sites? Is indirectly using non-free software the same as directly using it? What about sites that contribute to F/OSS projects? Google has contributed to F/OSS, but they are also the guys with Chrome and privacy concerns.
Perhaps it's a necessary evil that isn't so evil? Or maybe we need to make a DDG alternative?
Distributing, under a Free Software License, every software that DuckDuckGo runs, would give the user the freedom to have its own search engine. Well, that would be nice but I do not think every user should crawl the Web! The Web search engine should be regarded as an advanced Web page rather than a software. It is not SaaS.
However it raises another important concern: privacy. That is why I prefer DuckDuckGo to to Google and think that DuckDuckGo should encourage the user to take consciousness of the privacy issue (large companies that knows everything about you by storing every query you make; potentially selling them to spammers; potentially helping dictatorships to identify political opponents; etc).
Distributing, under a Free Software License, every software that DuckDuckGo runs, would give the user the freedom to have its own search engine. Well, that would be nice but I do not think every user should crawl the Web!
True, but I don't think that's the point. The point of free software is basically to have the ability to modify, redistribute, and use without worry of consent, patents, etc, along with source code availability in those same conditions. The ability to know what you're running is just an added benefit.
The Web search engine should be regarded as an advanced Web page rather than a software. It is not SaaS.
Isn't that like judging a book by it's cover? The web page is just the graphical front-end to the engine. All the algorithms that find what you're looking for are not put in the source code of the web page.
Don't forget freedom 0: the freedom to run the software. This is actually the most fundamental freedom granted by Free softwares. Indeed, the other freedom are irrelevant without freedom 0: if you don't run your own copy of the software, you just do not know what code is executed even if this software is distributed under a Free software license! Indeed, by using a, so-called, service on the Internet you were not distributed the software beforehand. It is not your copy and the owner of the service may modify the service as she wishes. If this service is a Free software, this is even guaranteed by freedom 1!
To fight against this threat (SaaS), every user should understand the danger of letting her computing in the hand (well, in the CPU) of other people. For instance they should prefer LibreOffice (or any other Free software word processor) and refuse Google Docs. However, as I wrote in my last post, crawling the Web is not your computing. That is why it should not be regarded as SaaS.
That is RMS's vision too: http://balance.fsf.org/video/lp2010-sat-rms+awards.ogv (this authoritative argument will certainly make you more thoroughly consider my point ;-) )
EDIT: I took again a look at this video and the question whether search engines are SaaS was actually asked (at 37'30''). Richard Stallman answered:
"I don't think that search engines are softwares as a service because they are mainly doing is they are gathering gigantic amounts of data from all around the Net and then we ask to look up in their data. And that's not doing your own computing unless you imagine you are going to spide the whole Net. And each of us is going to spide the whole Net. That's not gonna work."
I am not a native English speaker and I may have misunderstood some words (the audio quality of the video is not perfect).
No, you are correct. The FSF does not consider search engines SaaS (http://www.gnu.org/philosophy/who-does-that-server-really-serve.html)
See the section on "Distinguishing SaaS from Other Network Services"
"Services such as search engines collect data from around the web and let you examine it. Looking through their collection of data isn't your own computing in the usual sense—you didn't provide that collection—so using such a service to search the web is not SaaS. (However, using someone else's search engine to implement a search facility for your own site is SaaS.)"
Yes, of course I agree that search engines aren't SaaS. But that wasn't my point. What I was trying to say was that when code from some search engine is released, say for simple indexing of your own hard drive or another purpose, and this code is proprietary or "open-source" (the bad open-source), your freedoms are restricted if you use it. If only some of the code is released, free or proprietary, the code enters into the "gray" area. It's part-free, so it's not all free. But search engines aren't SaaS, so does it even matter?
Here's how I see it: If you're a developer, and you use code released by a search engine and it's proprietary, then your program is obviously non-free.
But if you're a user, and you use the output of a proprietary (or part-proprietary) search engine or program, you may or may not be using proprietary code. Why? If the link the engine gave you is non-free or sells non-free programs, then the output is non-free. Of course, it is not the engine's job to filter, unless you specifically ask it to do so.
Think of it this way: If you went up to a stranger and asked him directions to your friend's house, does that make your friend a stranger? No it doesn't, except if your "friend" really is a stranger or has strangers at his house.
This does not, however, excuse developers from using non-free code in search engines. Users aren't guilty however, unless they use this non-free code locally!
Let me stress though that I believe a free software search engine (as I believe DDG will become) is an excellent idea, and should be encouraged. I don't think all people should index the web (as not all users of Trisquel should make their own OS,) but I do think that this code should be available for the public to use, study, modify, redistribute, and re-release with the source code available under those same terms.
BTW, I am in no way against DDG. Great features, community, and privacy make it a great search engine. (Of course, I'm open to all ideas and alternatives!) +1 for making it default! :-)
I think everyone here agrees on that:
* one should use Free software to do its own computing (a search engine on her website is part of it) unless she does not value her freedom;
* one cannot impose her view on someone's else computing (a search engine on someone else website is part of it and so is a feature that would filter the output).
I agree that Trisquel should encourage the participation to the Free software community. In this way, a company that develop and distribute softwares under the terms of Free licenses should be preferred. DuckDuckGo does so.
However, the freedoms to study and modify only apply to the code you received. The one the company runs (and that you use from your Web browser) may be different (freedom 1 guarantees this right to the company). You cannot know therefore you cannot study the search engine you use from your Web browser. And, of course, you cannot modify this service either since you do not own it. As a consequence, the licenses (or absence of license!) of the search engines in the list of Trisquel's Web browser are irrelevant.
This is great if he follows through. A FOSS internet search engine of
DDG's caliber would certainly change the web.
However, I still do not believe Trisquel should endorse something only
to find out that it doesn't become FOSS due to reason A, or only parts
of it are open-source due to reason B. Were Trisquel to prematurely
endorse it, it could be quite a mishap should this happen.
> DDG has clarified his intentions (pro-F/OSS.)
Do you consider that Trisquel currently endorses Google?
* If yes, in what way is it better to endorse Google (number 1 for spying on the Web users) instead of DuckDuckGo (that has a real policy to preserve its users' privacy)?
* If no, I do not get your point.
Furthermore, if DuckDuckGo publishes all its developments under a Free license, I still do not get in what way it would change anything for the Web browser user: they do not run their own copy therefore cannot profit from the four freedoms. They only use DuckDuckGo's copy. Again, we are talking about the list of search engines in Trisquel's Web browser and not about a search engine that would be in the repository to crawl your own data (the latter definitely has to be free).
And if might add that DDG does not encourage people to use proprietary software locally, whereas Google, with Chrome and Google Earth, does.
> * If yes, in what way is it better to endorse Google (number 1 for spying on
> the Web users) instead of DuckDuckGo (that has a real policy to preserve its
> users' privacy)?
> * If no, I do not get your point.
I'm with you. Currently (Trisquel 4.0), Web Browser comes with search support for the following:
- Creative Commons
Only the ones marked in bold can in any sense be regarded as respecting users' freedom. That's only 3/8 == 37.5% of the supplied search engines, not even including the default engine.
If anything, I'd suggest cutting the number of search engines to the following:
- DuckDuckGo (SSL)
- Creative Commons
3/5 == 60% free, and DDG moving in that direction. Maybe throw in Yahoo if you expect anybody to actually use it regularly (I haven't in years, except occasionally on a whim).
You could add OpenStreetMap for one more bold string... and you could remove Google for one less non-bold. :-)
eBay and Amazon are very large corporations and have done certain things over the years that deserve criticism, but their inclusion is not really a freedom issue. It is possibly a privacy issue, though.
I use the Duck Duck Go search engine for some time now and if I may add in it's favor, let's not neglect the "!Bang" option, an option that lets the user search with other search engines right from the DuckDuckGo interface.
That makes, in this thread, five Trisquel users using DuckDuckGo and in favor of setting it as the default Web search engine for the distribution. I thought I was the only one! :-)
Talking about the default list of search engines, I believe that OpenStreetMap's search is not part of it, is it? That would be another useful addition. OpenStreetMap is the Wikipedia of the maps (content distributed under the CC-BY-SA 2.0 license).
I agree with OpenStreetMap's search being included if it is free
software, which I believe it is. We would need to determine that beforehand.
> in favor of setting it as the default Web search engine for the
> distribution. I thought I was the only one! :-)
> Talking about the default list of search engines, I believe that
> OpenStreetMap's search is not part of it, is it? That would be another
> useful addition. OpenStreetMap is the Wikipedia of the maps (content
> distributed under the CC-BY-SA 2.0 license).
The maps are Free (Creative Commons BY-SA 2.0). I do not know about the search engine. However, as I explained earlier, even if it is not, it does not restrict my freedom as a simple Trisquel user (it would does as an OpenStreetMap contributor though)! As a consequence, I do not think Trisquel needs to determine whether the included search engines are Free.
OpenStreetMap contributors may have develop their own search engine and do not distribute it (hence no license attached to it). There is no problem with that. Neither for them (they have control over their computing since THEY developed the search engine), nor for mere users (searching OpenStreetMap data is NOT doing their computing).
See my above comments wrt F/OSS search engines.
OSM is currently licensed under CC-BY-SA, but they're changing it to their own:
It appears to be a free-license:
I also believe their software is free software, though I'm not sure.
The section "What license is being changed?" in the page pointed by the first link you gave gives an answer about the softwares OpenStreetMap uses:
"It does not cover software and software source code, which are usually but not always GPL (GNU Public License)."
However, again, this point should be irrelevant to Trisquel users (hence to the inclusion of the search engine in its Web browser). Only OpenStreetMap contributors should care.
Is there any official decision with respect to the Web browser's list of search engines in the next Trisquel? Do we need to upload the idea of inserting that of OpenStreetMap and DuckDuckGo (as the default engine?) via a feature request on this page?
Ixquick looks decent at first glance but they push proprietary browser plugins to their users (see http://us2.ixquick.com/eng/download-ixquick-plugin.html). It's also rather suspicious that they operate with so many names, including generic ones like "startpage."
Scroogle OTOH is rather neat and I've used it before. I seem to remember them publishing the source code somewhere, but can't remember if it was under a free license or not. Regardless, since Scroogle depends on Google, that means Scroogle is vulnerable to Google trying to lock them out, as Google has tried to do many times.
How about something like Trisearch.com? ;) Let's show them all how to do it right. Take the zing out of the bing.
Trisquel is about building a 100% Free operating system. That is a lot of work already. Developing a search browser is a completely different task. It has little to do with Free software (from the user's point of view, it is more about accessing a Web site) and I do not see why Trisquel would be able to do a better job than DuckDuckGo. Why competing when you can collaborate?
FYI, DuckDuckGo already primarily bases its search results off of Bing's, with a lot of fine-tuning and info gathered from other engines, including its own crawler bot.
It is always better to contribute to an existing project that fulfills your goals (in DDG's case, a pain-free search engine that doesn't track you) than to branch off into a new one.
I hear you, I just know sometimes if you want it done right...
Plus, I am thinking of ways to get attention to Trisquel. Perhaps that is a long shot high resource undertaking. But I can't help but wishing to see everything having a 100% libre offering.