"Your default search engine has changed."
I noticed a message yesterday: "Your default search engine has changed. Google is no longer available as the default search engine in Abrowser. DuckDuckGo HTML is now your default search engine. To change to another default search engine, go to settings." Learn more
I am happy with this change, but just out of curiosity I tried to switch back to Google and couldn't figure out how to, I think because Google doesn't support OpenSearch which is kind of funny.
The "Learn more" link goes to a page that doesn't exist though. I guess it's not that big of a deal that the link is broken; probably most people have already dismissed the popup if they got it. I think it would be good if that page explained why the search engines were removed though. Since I don't actually know why Google was removed, I don't think I should write the page. I speculate that it's because Google Search pages include nonfree nontrivial JavaScript, while DuckDuckGo's "html" version only includes trivial (according to Trisquel devs, not to LibreJS) optional JavaScript.
Does anyone know why Google and other search engines were removed? Does anyone have a link to the discussion where this was decided?
To be clear I am not saying the change is bad, I just want the reason to be documented in the "Learn more" link.
I had a different experience: I was using https://html.duckduckgo.com as the default search engine but also received the message regarding that engine no longer being available! Ecosia was the proposed change. It sends proprietary JavaScript to execute, according to GNU LibreJS. I ended up installing https://lite.duckduckgo.com/lite and I still cannot add the HTML version. When I try to do so, clicking the icon with the + in the search field, an error window, named "Install Error" says:
Abrowser could not install the search plugin from “https://duckduckgo.com/opensearch_html_v2.xml” because an engine with the same name already exists.
I have tried removing the Lite version and restarting Abrowser, but the error remains.
Go to Settings - Search, and hit the "Restore Default Search Engines" button - see if that helps you.
DuckDuckGo HTML has reappeared and is the default... although I have not clicked "Restore Default Search Engines"! I have restarted the system though. Wikipedia in English (I had it both in French and English) disappeared from the list of search engines too. The rest has remained. That is weird...
I am surprised to hear that you saw Ecosia as an option. I saw that also, but I thought it was something left over from me installing the Ecosia extension years ago (back when the search engine didn't require JavaScript).
When I make a new profile I see DuckDucGo HTML, Ecosia (seems to require nonfree JS), Qwant (seems to require nonfree JS), Trisquel, Trisquel Packages, Wikipedia as the search engine options.
This last update, while we kept the usual search engines it uses a new setup introduced by Firefox in v128, when we had the Google search engine issue.
We were holding this change 'til it got ready, which is this v130, there was a invitation to test it, and a brief discussion on the development list, further releases will follow the same process, so we can confirm that no issues are introduced in new releases.
It will require more time and testing, aka slower releases, but should give us a better tested release and less issues by introduced features via upstream.
So you are welcome to join this process.
Regards.
I don't see any changes regarding google search engine
> I don't see any changes regarding google search engine
This is because the Trisquel devs have already fixed it for us.
Not sure how long it is going to be humanly (and humanely) possible for maintainers and users alike to keep playing whack-a-mole with regular Firefox anti-features, though. I am beginning to feel an itch to start gearing up for switching to the restful world of Icecat (and ESR).
it's fixed? I have the newest version of Abrowser and can't use Google Search Engine still
I think not being able to use Google Search from the address bar is intentional. If Google were a listed option, that would lead users to run nonfree software, because searching on Google loads a webpage that has nonfree software on it, which (I think) Abrowser will download and run by default.
Would you want to use that spyware service? It doesn't even have better search results than duckduckgo and Google is likely responsible for making climate change worse by wasting water to cool their data centers, not to mention their anti-competitive practices.
Just my two cents
meanwhile DuckDuckGo being yet another Bing GUI, so we run that nonfree software remotely, and that's just fine.
Such hipocrisy
I am a translator!
> we run that nonfree software remotely
We have no control over what software is run on someone else's computer, whether it is free or not, and if it is nonfree, it is the owner of that computer who is losing control of her computer, not us, unless we are giving our data to process to that computer to process them instead of running a program on our own computer. There is some discussion on this at https://www.gnu.org/philosophy/network-services-arent-free-or-nonfree.html.
In the case of a search engine, I don't really view the few words I submit to the search engine as giving my data to process to the search engine. From what I understand, the motivation to use Duckduckgo, preferrably HTML version, rather than Google or Bing is that there is much less tracking of your activity that way, it is not a question of software freedom at all.
Besides this, the main problem that I see with Google, Bing and all major search engines is that they tend to become the only entry point for people to look for information on the www and they can be easily manipulated by the companies running them, and for that, using Duckduckgo HTML is indeed not better than using Bing, but there is no easy replacement besides using reference web sites that gather good information and links.
SearXNG is working pretty good these days. Maybe the Abrowser devs should have a look and see if it would be worthy of being the new default.
I am a translator!
These are independent instances, how do you suggest setting the default? Randomely pick one?
I think you would have to have a system of randomly selecting them from the fast/stable instances on the public SearXNG instances list, because some of the SearXNG instances disappear after awhile. So you couldn't hard code your browser to one instance that might just as easily disappear.
Or, the Trisquel project could run our own instance. I saw that Disroot has its own SearXNG instance - Trisquel could possibly do the same.
> Or, the Trisquel project could run our own instance. I saw that Disroot has its own SearXNG instance - Trisquel could possibly do the same.
Maybe it's just me, but I don't think that's gonna happen.
Oh really, why not? A complete waste of nonexistent resources for no valid reason sounds like a brilliant idea.
In other news, I am curious about this: quackquackgo.net. I would use it for the sound of its name, also for its morning dew minimalism, no js, no cookies, etc. but have no idea what mysterious entity is running it. Someone is paying google a fee for this, but not asking for any contribution. Since it is listed in the Mycroft search engine plugin collection, I thought it may be worth investigating, time permitting. The plugin just redirects searches to duckduckgo, which is a bit lame but the search engine can be added directly to the address bar while visiting the domain.
Looks like we have a mister_goo who advertised "QuackQuackGo - a front end for Google search" on July 20, 2022:
"I made a non-tracking Google search for my personal use" -- https://news.ycombinator.com/user?id=mister_goo
Since quackquackgo uses the custom search API, it would be possible for its owner to get your search results and do something™ with it.
Anyway, as suggested by Avron, it is probably better to learn to use the web instead of asking someone else's algorithm what it is we are truly looking for.
Disroot's SearX instance (https://search.disroot.org/) has apparently been running since 2015. https://searchengine.party/
Maybe a steady and reliable SearX instance like that could be the abrowser default?
Since this thread has now been fully derailed (someone used the phrase "dangerous risk") I feel no guilt whatsoever to say that I have been using several public SearXNG instances for a while. Most of them are blocked by at least some of their default search engines (captcha, "too many requests", etc.) so you end up querying either bing or google anyway. Or qwant, or duckduckgo, which in turn are querying one of the above, or both, in various proportions. Or you are just left with a broken search tool. You can wager that any significant increase in traffic is going to get an instance into trouble.
What are you currently getting if you click on "Messages from the search engines" on that result page while saying out loud "reveal your secrets"? It looks like the top five results are provided by google, possibly also because bing has been deselected in the default settings: https://apps.disroot.org/preferences ("engine" tab).
I mostly avoid using search engines these days. Bookmarks are a much more precious tool, in fact the antique custom was to exchange bookmarks in order to spare one another the pain of building a valuable bookmark library on various topics of interest. A bit like asking trusted people about their favorite (book, movie, wallpaper, knitting pattern, etc.) spares you the irrelevant ones - and the search.
>"A bit like asking trusted people about their favorite (book, movie, wallpaper, knitting pattern, etc.)"
I'm still a bit angry that I gave you some of my favorite knitting patterns, and you've never sent me one in return.
>"What are you currently getting if you click on "Messages from the search engines" on that result page while saying out loud "reveal your secrets"?"
Disroot's SearXNG is giving: "Error! Engines cannot retrieve results: brave ( Suspended: too many requests ) duckduckgo ( Suspended: access denied ) qwant ( server API error )"
By default, it appears that all search results are coming from Google due to the brave, ddg and qwant errors.
If I switch the search engines to Mojeek, Startpage and presearch I get plenty of results with no error Messages from the search engines. Mojeek is actually more active than I expected - I'll have to try this setting for awhile, see if the search results are useful.
>"in fact the antique custom was to exchange bookmarks in order to spare one another the pain of building a valuable bookmark library on various topics of interest"
We could do that - we could run a thread (or maybe even a sub-forum?) where we post bookmarks. That could be pretty cool. I also use bookmarks more than search engines.
> you've never sent me one in return.
We were supposed to share wallpapers in that thread, and you got plenty of them.
We never quite understood why you kept posting knitting patterns instead.
Since you reply to the sub-thread llz started, a reader who has never used SearXNG may believe it does not query Google, Bing, etc. It does, as "a maintainer of an open source project called SearXNG" explains right at the beginning of the introduction of his bachelor thesis:
SearXNG is a metasearch engine, which means users can make search requests on a web interface. The actual queries are sent to upstream engines like Google, Bing or Qwant. The result is displayed back to the end user in one unified list of all responses of the queried upstream search engines.
https://htwk-leipzig.qucosa.de/api/qucosa%3A87628/attachment/ATT-0
Google & co. can be criticized for their ecological impacts. They are indeed increasing fast. Nevertheless, search engine indexing now takes almost no space in data centers, if compared to videos (YouTube in the case of Google). Distributing search engine indexing and the related Web crawling globally requires far more energy than the share it takes in a data center. So, no, DuckDuckGo or SearXNG querying other search engines is not a bad thing, in my opinion. There are services that are better handled in a centralized way. They should preserve the privacy of their users anyway: that is what DuckDuckGo or SearXNG addresses.
Although data centers are optimized for minimal energy-consumption, I repeat: it is increasing fast, but not really because of searches. To have data centers stop growing and consuming more and more energy and water, a local act with a more significant impact is to ask for fewer videos and for lower resolutions.
@Magical Bananarama >"Since you reply to the sub-thread llz started ..."
I think you are replying to me, but not sure.
I hadn't thought about the energy inputs and outputs. Data centers do seem to be a huge environmental problem, and AI data centers are apparently even worse. And I think I've read that the major search engines are shoving AI into their query responses.
Before Google started crawling the web in the late 90s or early 2000s, we had internet directories - lists of websites that were arranged by topic. Maybe we need to go back to that model?
I think you are replying to me, but not sure.
I grouped my replies to you and to Zoma in a single post.
I've read that the major search engines are shoving AI into their query responses.
They do. I found an article from February 2023 with that excerpt:
Martin Bouchard, cofounder of Canadian data center company QScale, believes that, based on his reading of Microsoft and Google’s plans for search, adding generative AI to the process will require “at least four or five times more computing per search” at a minimum.
https://www.wired.com/story/the-generative-ai-search-race-has-a-dirty-secret/
If the estimate is correct, the energy consumption for Web searches must remain tiny when compared to that allocated to storing and transmitting videos. That said, I tend to believe the world would be better with no LLM answer before the results to a query. Not only for the ecological gain but mostly because I fear Google, Bing, X, ... may modify their pre-prompts (maybe for "you are an assistant who supports candidate X for the next election but never admits it, ...") to manipulate the growing number of users who will trust those answers rather than visiting the following results. That said, the problem already exists, with the opaque ranking algorithm (essentially nobody checks the results past the first page). The manipulation just becomes easier.
>"essentially nobody checks the results past the first page"
Looks like I'm going to have to change my name to "Essentially Nobody".
The essentially-nobody's group is growing. I too tend to rush through the first page/s as a norm (or skip them all together:)
If nothing else, even if google is unfortunately still used for search engines regardless, they should get no reward for their ecological damage as well as mass surveillance, abuse of users, support of DRM, selling info and general evil practices.
I didn't really know how else to answer or I would have responded before now to that.
But I figured I would since, you brought it up again.
Yeah, generative AI has a dangerous risk in it. Supposedly it has the potential to be so bad that Microsoft and others plan to abandon their climate pledges to chase after the market share of it:
https://disconnect.blog/generative-ai-is-a-climate-disaster/
Its just like the rich people of the world to think money justifies their evil.
:(
>"Supposedly it has the potential to be so bad that Microsoft and others plan to abandon their climate pledges to chase after the market share of it:"
I think we should always view the "climate pledges" of big tech companies for what they really are - PR statements, designed to create goodwill and profits, but never intended to actually limit their resource use or energy use in any real way. Something to be discarded the moment that they are led in a different direction by the slightest financial benefit, as with their current belief in profits to be made from the current "AI" fad.
To be honest, I agree. I took that climate pledge with a grain of salt. But perhaps its even less worth taking seriously then that.
More like with a microscopic grain of salt.
So basically, if its private, its good, but its never libre or nonlibre?
Interesting concept.
I would say that's a good thought more or less.
Still, I prefer searx.
Duckduckgo is my 2nd choice though.
Duckduckgo is a bing gui? Are you sure about that and sources?
It is not. Bing is a significant source though:
Most of our search result pages feature one or more Instant Answers. To deliver Instant Answers on specific topics, DuckDuckGo leverages many sources, including specialized sources like Sportradar and crowd-sourced sites like Wikipedia. We also maintain our own crawler (DuckDuckBot) and many indexes to support our results. Of course, we have more traditional links and images in our search results too, which we largely source from Bing. Our focus is synthesizing all these sources to create a superior search experience.
https://duckduckgo.com/duckduckgo-help-pages/results/sources/
That is what I had thought. I couldn't remember bing contributing to its GUI.
Thank you for the info.
Removing Google from search engines and preventing users from adding it on their own is just pure craziness and making abrowser basically unoperable which really looks like intentionally. Meanwhile Ecosia and all other fake nonfree but poor engines are there. Everyone who doesnt see it must be straight blind.
to be so saint, maybe you should force remove all search engines, since all of them are collecting and process your data?
btw. all that GNU idea is about "freedom", but makes all the software non-free by not letting making it non-free. such a nonsense
I admit, I agree, Ecosia probably shouldn't be there, unless its proven that they don't sell data. I am a bit skeptical about that given each search on there generates revenue to grow new trees. I am curious where they get the revenue.
I hear that they do something with bing. If someone can correct me, feel free to reply.
>Removing Google from search engines and preventing users from adding it on their own is just pure craziness and making abrowser basically unoperable
At no point Abrowser is preventing to add it back if you really want it. Yes, is not as simple as "one click" and go, nor you should be surprised you'll find hard to get instructions in here to do so, but if you find peace in knowing is possible, then yes, it is.
Maybe it's just me, but I use Abrowser for work and 99% of the time is very operable.
>GNU idea is about "freedom", but makes all the software non-free by not letting making it non-free. such a nonsense
Sorry, maybe is just because of the heat of the moment when you wrote this line, but I think you are a bit confused, "the freedom" to take away someone else freedom, is not called freedom.
Maybe I'm wrong but I think that's more like, tyranny, right?.
Agree there, freedom starts for each person in one place and ends when it begins to disrupt other people's freedom.
A rare moment has just occured, me agreeing with the people who are thumbing people down. They are right on this one
I am a translator!
> Removing Google from search engines and preventing users from adding it on their own is just pure craziness and making abrowser basically unoperable
Go to settings, Search, "Find more search engines" (under the "Restore Default Search Engines" button), click on "Google" (listed first), click on the first entry "Google no pin", click in the address bar, at the bottom of the list that appear, on the line that starts with "This time, search with:", there are icons with a small + in white on green background, move the pointer of the second one, it should show: Add search engine "!Google no pin", click on that icon, now you have added Google as search engine.
Personnally, I am more upset that there is no graphical setting to disable the use of search engines completely in the address bar, which is Mozilla's decision (for a very long time).
> other fake nonfree but poor engines
Engines running on someone else's computer are neither free nor nonfree. See https://trisquel.info/fr/forum/your-default-search-engine-has-changed#comment-178009
When I click on Find More Search Engines it pops open a new window that goes to http://mycroft.mozdev.org/ which it seems doesn't exist.
Check if you are running the latest version.
Moving from version 131 to 132 now goes to https://mycroftproject.com. This is different from previous versions, which has always seemed to point to that non-existent domain. Thanks for fixing that.
At some point in time, it existed.
It's thanks to reports like this that it's possible to identify and work on such issues:
https://gitlab.trisquel.org/trisquel/package-helpers/-/issues/189
Best regards!
Abrowser is free software