DuckDuckGo Wants Answers to Linux Questions
fossforce.com
fossforce.com
Chrome supports using different search engines directly.
In my Chrome, searching youtube.com is just "you<tab>query<return>", the en.wikipedia.org is "en<tab>query<return>", Google Maps is "maps..." etc.
The nice thing is that it picks up various search engines automatically. Useful for work etc.
And if you want to use DuckDuckGo occasionally, it's just a "duck<tab>" away.
(And Google is my default search engine.)
Also, doesn't !g just open a Google search with the specified query?
I like instant answers, but the DDG guys need to learn how to play nicely in this space. They are a search engine, not a content provider. One of the key issues I think sites have with Google at the moment is that if they start pulling in their data then people don't actually visit their page.
Wikipedia is a good example - it relies on contributions from the general public, and if people just get aggregated data packaged on the search engine that's great for the searcher, but not very good for the site. StackOverflow is another site that I could well see affected.
I'm not sure how this could be solved, but there needs to be a way for sites that rely on community engagement to get active participants, otherwise the unintentional consequences of this may be that they are adversely affected - or they might start competing more directly... Which is what the WMF tried to do, in the most ridiculously crack-handed way possible.
I'm starting to use DDG, you guys (I'm assuming you are from DDG?) have improved and you are more nimble and transparent. I'm hoping this big of feedback might lead to some thoughts and with a bit of creativity some sort of mechanism could be put into place to solve this.
In regards to scraping SO answers, I don't see the problem. In a classic search engine, the title of the page is listed, with an excerpt of the text on the page immediately below it.
In DDG, IIRC, the SO instant answers display the exact same information, but just format it better. Same with the wikibooks answers (those are the only instant answers I can remember triggering recently). What's wrong with that?
Have a look at the pull request here:
https://github.com/duckduckgo/zeroclickinfo-goodies/pull/296...
They are only following on the industry standard for this feature. Google, Bing, Yahoo all do this already. DDG adding this feature (but offering to let people correct the entries and open-sourcing the answers) is just achieving feature parity, not diving into new territory.
If they start tracking me to achieve "feature parity", that's sort of one less feature I can do without!
For cheat sheets, we either write the information from scratch (as in the tennis cheat sheet) or rely on existing key->value data such as keyboard shortcuts or command parameters. Again, we link to the source in both cases although our cheat sheets often end up featuring improvements or fixes, added by either the author or the community at a later date. They become living, open source resources and are also exposed in our API: https://duckduckgo.com/api
You're right - we're not a content provider, but so far we've found people appreciate saving even a click or a page load to find what they're searching for.
My only point is that all search engines tend to cannibalise the ones they index to a certain degree when they aggregate or present results from their own website. It doesn't make it wrong to do so, or even unethical. I just think it would be nice if some sort of mechanism could be worked out to encourage people to visit and contribute more to the sites that generate the data being used.
DDG makes it quite plain where the data is coming from, and is very open and transparent with what you guys are doing, so please just take what I wrote as a comment and maybe a request, but certainly not a criticism.
Edit: I just reread my comment and I expressed myself rather poorly I think. The sentence "DDG need to learn how to play nicely..." was not at all what I was trying to convey. DDG do play nicely with others, far more so than anyone else. I apologise for that phrasing, that was unintentionally rude.
I didn't take your comment as rude at all and you make valid points. We're continually looking at how best to serve our users, contributors and the wider web while building trust and protecting privacy, and such feedback helps us do that.
https://duck.co/ia/view/linux_cheat_sheet https://duck.co/ia/view/linux_error
To actually see them in action, here are two example queries:
https://duckduckgo.com/?q=linux%20cheat%20sheet&ia=cheatshee... https://duckduckgo.com/?q=linux%20enoent
It would be great if Linux experts could check for errors and suggest improvements. This can be done via the "Create issue" button in each overview page. It's also possible to create new Instant Answers or submit PRs for improvements. We have a list of Instant Answers we're looking for, and links to our docs, here: https://github.com/duckduckgo/duckduckgo/wiki/Programming-IA...
I see you've already been doing this (thank you!) but hopefully explaining it here clears up confusion for others.
Contributing was a pleasure :-)