New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Addressing default search engine privacy #88

Closed
RoxKilly opened this Issue Apr 24, 2017 · 23 comments

Comments

Projects
None yet
5 participants
@RoxKilly
Collaborator

RoxKilly commented Apr 24, 2017

One of the biggest privacy exposures users face on the web is definitely the history of their searches. Yahoo is Firefox's default search engine (Google was default until 2014). We all know about Yahoo's security and privacy history in recent years. In a word: deplorable.

What do you guys think of replacing Yahoo with a privacy friendly option such as DuckDuckGo which fetches Yahoo results behind the scenes? A javascript-free version is even available on AMO.

Alternatively, if you think users would rather see Google results, Startpage proxies the searches and stores no logs so Google can't track users or even know who is performing those searches.

If the default search engine cannot be set via user.js file, one possibility is to add a reference to private alternatives the way you make reference to addons that should be used to supplement the configs.

@Atavic

This comment has been minimized.

Show comment
Hide comment
@Atavic

Atavic Apr 24, 2017

Collaborator

I remove the search bar and use a local home page with links to search engines.

EDIT: I also remove favicons. The icons from searchbar can be used for tracking:

Collaborator

Atavic commented Apr 24, 2017

I remove the search bar and use a local home page with links to search engines.

EDIT: I also remove favicons. The icons from searchbar can be used for tracking:

@Thorin-Oakenpants

This comment has been minimized.

Show comment
Hide comment
@Thorin-Oakenpants

Thorin-Oakenpants Apr 25, 2017

Member

This. IMO, falls outside the scope of the project (via the user.js). For starters, setting the default engine requires it to be present, and different builds will have different default engines. It's also problematic from a language perspective. We have no right to impose/change someone's search engine. For those who have no idea that their search history per engine (bing, yahoo, google etc) is being collated (via a half dozen points, such as IP, SSL session ticket ids, cookies, wifi location data etc), well, 2000 called and want their village idiot back.

That said, if someone wants to add to the wiki with alternative search engines and tips and tricks, go for it. But it opens up a slippery slope of expanding beyond our mandate - what next, alternatives to gmail? A list of web email services? A list of hsts blocklists?

Will we then be expected to provide more and more OpSec info eg how to daisy chain VPNs, or use TOR over VPN vs VPN over TOR, or how to use tails, or how to buy a burner phone and avoid camera surveillance... or how to stick tiny pebbles in your shoes to hide your gait from biometric sensors

Look, I totally get that the search engine is a part of Firefox, and has controls. But who the hell are we to tell people what service(s) they can and can't use. I WILL not do this in the user.js (no arguments - at the very least it would be commented out, but I am totally against it period), but am open to a wiki page.

I am also open to listing these search engines from AMO (never used any myself, I used Add To Search Bar, which can be done manually anyway, so I actually have zero "installed" or default engines)

Member

Thorin-Oakenpants commented Apr 25, 2017

This. IMO, falls outside the scope of the project (via the user.js). For starters, setting the default engine requires it to be present, and different builds will have different default engines. It's also problematic from a language perspective. We have no right to impose/change someone's search engine. For those who have no idea that their search history per engine (bing, yahoo, google etc) is being collated (via a half dozen points, such as IP, SSL session ticket ids, cookies, wifi location data etc), well, 2000 called and want their village idiot back.

That said, if someone wants to add to the wiki with alternative search engines and tips and tricks, go for it. But it opens up a slippery slope of expanding beyond our mandate - what next, alternatives to gmail? A list of web email services? A list of hsts blocklists?

Will we then be expected to provide more and more OpSec info eg how to daisy chain VPNs, or use TOR over VPN vs VPN over TOR, or how to use tails, or how to buy a burner phone and avoid camera surveillance... or how to stick tiny pebbles in your shoes to hide your gait from biometric sensors

Look, I totally get that the search engine is a part of Firefox, and has controls. But who the hell are we to tell people what service(s) they can and can't use. I WILL not do this in the user.js (no arguments - at the very least it would be commented out, but I am totally against it period), but am open to a wiki page.

I am also open to listing these search engines from AMO (never used any myself, I used Add To Search Bar, which can be done manually anyway, so I actually have zero "installed" or default engines)

@RoxKilly

This comment has been minimized.

Show comment
Hide comment
@RoxKilly

RoxKilly Apr 25, 2017

Collaborator

But who the hell are we to tell people what service(s) they can and can't use

Im perplexed by this. I may be mistaken but didn't you disable the Pocket service at some point? Or Adobe's Encrypted Media DRM service? Or Firefox Hello?

This...falls outside the scope of the project. I WILL not do this in the user.js

OK. It just occurred to me that using the search bar is one way a lot of privacy gets exposed in FF's default configuration and I understood your objective to be to harden FF settings for privacy and security so I thought I should bring it up here. After all Search engine selection falls under the settings umbrella unlike "How to use a burner phone" or "How to stick pebbles in shoes to hide your gait". I don't think I went that far off the reservation. Thanks for reading and considering though.

Collaborator

RoxKilly commented Apr 25, 2017

But who the hell are we to tell people what service(s) they can and can't use

Im perplexed by this. I may be mistaken but didn't you disable the Pocket service at some point? Or Adobe's Encrypted Media DRM service? Or Firefox Hello?

This...falls outside the scope of the project. I WILL not do this in the user.js

OK. It just occurred to me that using the search bar is one way a lot of privacy gets exposed in FF's default configuration and I understood your objective to be to harden FF settings for privacy and security so I thought I should bring it up here. After all Search engine selection falls under the settings umbrella unlike "How to use a burner phone" or "How to stick pebbles in shoes to hide your gait". I don't think I went that far off the reservation. Thanks for reading and considering though.

@Thorin-Oakenpants

This comment has been minimized.

Show comment
Hide comment
@Thorin-Oakenpants

Thorin-Oakenpants Apr 25, 2017

Member

There's a slight BIG difference between removing services entirely (DRM) and dictating what service to use

Member

Thorin-Oakenpants commented Apr 25, 2017

There's a slight BIG difference between removing services entirely (DRM) and dictating what service to use

@earthlng

This comment has been minimized.

Show comment
Hide comment
@earthlng

earthlng Apr 25, 2017

Member

Hi, I agree with OP that search engines are a major problem. But I also agree with Pants that nowadays everyone should know about that. Heck, even my mother knows not to trust Google.
But since we also recommend Addons I think we can recommend some search engines, or even host some opensearch.xml files.
DDG is one of the built-in search engines and we could add the pref (commented out maybe) that sets DDG as the default: user_pref("browser.search.defaultenginename", "DuckDuckGo");

BUT the builtin engine also has some tracking parameters and a link for suggestions. So it's not ideal but still better than using the built-in yahoo engine (which also has those parameters)

  <Url type="text/html" method="get" template="https://duckduckgo.com/" rel="searchform">
    <Param name="q" value="{searchTerms}"/>
    <MozParam name="t" condition="purpose" purpose="contextmenu" value="ffcm"/>
    <MozParam name="t" condition="purpose" purpose="keyword"     value="ffab"/>
    <MozParam name="t" condition="purpose" purpose="searchbar"   value="ffsb"/>
    <MozParam name="t" condition="purpose" purpose="homepage"    value="ffhp"/>
    <MozParam name="t" condition="purpose" purpose="newtab"      value="ffnt"/>
  </Url>
  <Url type="application/x-suggestions+json" template="https://ac.duckduckgo.com/ac/">
    <Param name="q" value="{searchTerms}"/>
    <Param name="type" value="list"/>
  </Url>

We could sanitize that file and host it here, but it's not possible to overwrite an engine with the same name as the default ones, so we would also need to change the engine name.

Member

earthlng commented Apr 25, 2017

Hi, I agree with OP that search engines are a major problem. But I also agree with Pants that nowadays everyone should know about that. Heck, even my mother knows not to trust Google.
But since we also recommend Addons I think we can recommend some search engines, or even host some opensearch.xml files.
DDG is one of the built-in search engines and we could add the pref (commented out maybe) that sets DDG as the default: user_pref("browser.search.defaultenginename", "DuckDuckGo");

BUT the builtin engine also has some tracking parameters and a link for suggestions. So it's not ideal but still better than using the built-in yahoo engine (which also has those parameters)

  <Url type="text/html" method="get" template="https://duckduckgo.com/" rel="searchform">
    <Param name="q" value="{searchTerms}"/>
    <MozParam name="t" condition="purpose" purpose="contextmenu" value="ffcm"/>
    <MozParam name="t" condition="purpose" purpose="keyword"     value="ffab"/>
    <MozParam name="t" condition="purpose" purpose="searchbar"   value="ffsb"/>
    <MozParam name="t" condition="purpose" purpose="homepage"    value="ffhp"/>
    <MozParam name="t" condition="purpose" purpose="newtab"      value="ffnt"/>
  </Url>
  <Url type="application/x-suggestions+json" template="https://ac.duckduckgo.com/ac/">
    <Param name="q" value="{searchTerms}"/>
    <Param name="type" value="list"/>
  </Url>

We could sanitize that file and host it here, but it's not possible to overwrite an engine with the same name as the default ones, so we would also need to change the engine name.

@Thorin-Oakenpants

This comment has been minimized.

Show comment
Hide comment
@Thorin-Oakenpants

Thorin-Oakenpants Apr 25, 2017

Member

Lets not get too involved with this. As I already said it's problematic in the user.js, and as I already said I am fine with some sort of wiki page (eg explain how to manually add sanitized engines, and/or point at some on AMO)

There is not only browser.search.defaultenginename there is also browser.search.defaultenginename.US, as well as browser.search.isUS. And then, how do we handle all the other locales/language builds? Has anyone looked at those? Do they need to be considered?

Disabling a service/feature/protocol (hello, webgl, DRM) is not the same as changing someone's choice of a service (which can't be disabled, at least not via the user.js) - so please don't compare them. IMO it is pointless adding anything to the user.js itself.

-slightly OT
As for the search box, I am currently using CTR to force the searchbox dropdown to itemize with an icon + text of the engine. With a default FF, it instead displays buttons only - I find this is terrible from a few angles (but I have until 57 to work it out) - when you select a button, it auto-searches, with CTR or old behaviour it does not. Nothing happens until I hit the enter key or the search icon. (I may have to do some more testing, its been a while)

example:

  • google engine: type in sans quotes "dwarf tossing", hit enter
  • change to wiki engine and blam! it auto searches wiki
  • what I wanted to do was simply change engines as I was about to search wiki for SOMETHING ELSE
  • even if the search field is empty, it still loads a search (eg google home search page)
  • I do not want to auto-clear the search field after searching, as I often need to search the same term on several sites

How do you make the default searchbar controls NOT DO ANYTHING unless you hit enter or click the search icon?

Member

Thorin-Oakenpants commented Apr 25, 2017

Lets not get too involved with this. As I already said it's problematic in the user.js, and as I already said I am fine with some sort of wiki page (eg explain how to manually add sanitized engines, and/or point at some on AMO)

There is not only browser.search.defaultenginename there is also browser.search.defaultenginename.US, as well as browser.search.isUS. And then, how do we handle all the other locales/language builds? Has anyone looked at those? Do they need to be considered?

Disabling a service/feature/protocol (hello, webgl, DRM) is not the same as changing someone's choice of a service (which can't be disabled, at least not via the user.js) - so please don't compare them. IMO it is pointless adding anything to the user.js itself.

-slightly OT
As for the search box, I am currently using CTR to force the searchbox dropdown to itemize with an icon + text of the engine. With a default FF, it instead displays buttons only - I find this is terrible from a few angles (but I have until 57 to work it out) - when you select a button, it auto-searches, with CTR or old behaviour it does not. Nothing happens until I hit the enter key or the search icon. (I may have to do some more testing, its been a while)

example:

  • google engine: type in sans quotes "dwarf tossing", hit enter
  • change to wiki engine and blam! it auto searches wiki
  • what I wanted to do was simply change engines as I was about to search wiki for SOMETHING ELSE
  • even if the search field is empty, it still loads a search (eg google home search page)
  • I do not want to auto-clear the search field after searching, as I often need to search the same term on several sites

How do you make the default searchbar controls NOT DO ANYTHING unless you hit enter or click the search icon?

@earthlng

This comment has been minimized.

Show comment
Hide comment
@earthlng

earthlng Apr 25, 2017

Member

And then, how do we handle all the other locales/language builds?

Yes, that does seem to be a problem. The searchplugins in the omni.ja are in a "en-US" subfolder so they are likely localized and DuckDuckGo is maybe called PooPooHoo in Thailand or something ;)

How do you make the default searchbar controls NOT DO ANYTHING unless you hit enter or click the search icon?

good question - no idea. I basically only use the searchbar for the indicator when an opensearch.xml is detected. Otherwise I copy/paste the search term and search directly in the urlbar with "s dwarf tossing" for startpage and then "x dwarf tossing" for ixquick etc. Definitely not ideal but I never bothered to search for a better way to do it.

Member

earthlng commented Apr 25, 2017

And then, how do we handle all the other locales/language builds?

Yes, that does seem to be a problem. The searchplugins in the omni.ja are in a "en-US" subfolder so they are likely localized and DuckDuckGo is maybe called PooPooHoo in Thailand or something ;)

How do you make the default searchbar controls NOT DO ANYTHING unless you hit enter or click the search icon?

good question - no idea. I basically only use the searchbar for the indicator when an opensearch.xml is detected. Otherwise I copy/paste the search term and search directly in the urlbar with "s dwarf tossing" for startpage and then "x dwarf tossing" for ixquick etc. Definitely not ideal but I never bothered to search for a better way to do it.

@Thorin-Oakenpants

This comment has been minimized.

Show comment
Hide comment
@Thorin-Oakenpants

Thorin-Oakenpants Apr 25, 2017

Member

good question - no idea

We have until 57, and from some early mockups it looks like the list MAY be back (there are some definite u-turns on UIX)

Member

Thorin-Oakenpants commented Apr 25, 2017

good question - no idea

We have until 57, and from some early mockups it looks like the list MAY be back (there are some definite u-turns on UIX)

@Thorin-Oakenpants Thorin-Oakenpants referenced this issue May 1, 2017

Open

sticky: wiki stiki #65

3 of 6 tasks complete
@Thorin-Oakenpants

This comment has been minimized.

Show comment
Hide comment
@Thorin-Oakenpants

Thorin-Oakenpants May 1, 2017

Member

I added this to the sticky wiki as an item to keep track of. When or if someone drafts up a wiki article (it's just github markup you can use in here), and we finalize it, then I can add the pref to the user.js with a note about the engine name having to exist, etc, AND link to the wiki article

wiki ideas:

  • list of search engines
  • list of AMO search engines
  • list of sanitized code
  • how to manually add these engines
  • make sure we build in keywords
  • maybe provide a zip bundle for profile\searchplugins
Member

Thorin-Oakenpants commented May 1, 2017

I added this to the sticky wiki as an item to keep track of. When or if someone drafts up a wiki article (it's just github markup you can use in here), and we finalize it, then I can add the pref to the user.js with a note about the engine name having to exist, etc, AND link to the wiki article

wiki ideas:

  • list of search engines
  • list of AMO search engines
  • list of sanitized code
  • how to manually add these engines
  • make sure we build in keywords
  • maybe provide a zip bundle for profile\searchplugins
@atomGit

This comment has been minimized.

Show comment
Hide comment
@atomGit

atomGit May 14, 2017

Collaborator

re:

  • list of sanitized code
  • how to manually add these engines

i use the XML Search Engines Exporter/Importer add-on to export the engines, both default and those i add, and then sanitize them and dump them in /searchplugins folder (you also have to delete the search.json.mozlz4 file and restart the browser to get them to register)

btw, one of the best privacy-centric meta-search engines i've found is searx - it uses, or can use, a whopping pile of 3rd party engines and is very customizable - rather than overloading searx.me however, please consider using one of the many other privately hosted sear instances

unfortunately, other than the big boys, there seem to be very few alternative indexes that actually crawl the www looking for content - DDG, Startpage, searx, etc., all pull their results from Google, Bing, Yahoo, etc., which means their results are also open to censorship

YaCe is a notable exception in that it is an indexer and also a distributed P2P search engine, but it's index is still very small - you can find a demo here

Collaborator

atomGit commented May 14, 2017

re:

  • list of sanitized code
  • how to manually add these engines

i use the XML Search Engines Exporter/Importer add-on to export the engines, both default and those i add, and then sanitize them and dump them in /searchplugins folder (you also have to delete the search.json.mozlz4 file and restart the browser to get them to register)

btw, one of the best privacy-centric meta-search engines i've found is searx - it uses, or can use, a whopping pile of 3rd party engines and is very customizable - rather than overloading searx.me however, please consider using one of the many other privately hosted sear instances

unfortunately, other than the big boys, there seem to be very few alternative indexes that actually crawl the www looking for content - DDG, Startpage, searx, etc., all pull their results from Google, Bing, Yahoo, etc., which means their results are also open to censorship

YaCe is a notable exception in that it is an indexer and also a distributed P2P search engine, but it's index is still very small - you can find a demo here

@atomGit

This comment has been minimized.

Show comment
Hide comment
@atomGit

atomGit May 15, 2017

Collaborator

I added this to the sticky wiki as an item to keep track of. When or if someone drafts up a wiki article (it's just github markup you can use in here), and we finalize it, then I can add the pref to the user.js with a note about the engine name having to exist, etc, AND link to the wiki article

wiki ideas:

  • list of search engines
  • list of AMO search engines
  • list of sanitized code
  • how to manually add these engines
  • make sure we build in keywords
  • maybe provide a zip bundle for profile\searchplugins

i'd like to contribute to this if you want but i don't know how - is the wiki done through pull requests or do i just comment here?

i can lay out the basics regarding how to add and sanitize search engines and list some engines i think are good as far as results and respecting privacy

Collaborator

atomGit commented May 15, 2017

I added this to the sticky wiki as an item to keep track of. When or if someone drafts up a wiki article (it's just github markup you can use in here), and we finalize it, then I can add the pref to the user.js with a note about the engine name having to exist, etc, AND link to the wiki article

wiki ideas:

  • list of search engines
  • list of AMO search engines
  • list of sanitized code
  • how to manually add these engines
  • make sure we build in keywords
  • maybe provide a zip bundle for profile\searchplugins

i'd like to contribute to this if you want but i don't know how - is the wiki done through pull requests or do i just comment here?

i can lay out the basics regarding how to add and sanitize search engines and list some engines i think are good as far as results and respecting privacy

@Thorin-Oakenpants

This comment has been minimized.

Show comment
Hide comment
@Thorin-Oakenpants

Thorin-Oakenpants May 15, 2017

Member

You could create a wiki page in your repo - or paste a draft in here. Or you could do an article on your 12bytes blog, and we'll reproduce it here, and credit you and link back to 12bytes - or we can all collaborate together to come up with a three/four page search wiki section (and you could repost em at 12bytes as a joint author). All a wiki is, is basic markup. I forgot to add to that itemized list that we need to make sure we have unique base64 icons (eg google vs google images etc - yes I still think we should provide sanitized google ones).

Collaborators with "write" access I think can edit the wiki. Otherwise users can fork the repo, and do a mockup in their fork.

eg sample draft

### Search Engines

1. AMO
- [DuckDuckGo](https://blahblah.com/) - **h**ttps://blahblah.com

I think we should decide the structure. I would like to store all the sanitzied xml files unzipped in a folder, but list them in the wiki. So we could have something like

🔸 Search Engines

  • 4.1 Into (written by @RoxKilly), list of engines (make a new topic and list all engines), their AMO, and a sanitized XML (stored in searchplugins root folder here in repo)
  • 4.2 How to maintain/edit (eg rename them, change the base64 icon, change keyword, how to not clash with existing installed engine names)
  • 4.3 How to use the search bar without unintentionally leaking/searching (note, I use CTR which restores the "list" rather than just icons - without CTR I cannot properly control searching when changing engines - any touch of the icon to change engines makes a new search, even if the field is null), eg tips on key word searching etc

We could easily get page 4.1's list of engines done, then decide on keywords, then create sanitized XMLs and upload them all, and then do a 4.2 on how to add/edit/maintain them.

Member

Thorin-Oakenpants commented May 15, 2017

You could create a wiki page in your repo - or paste a draft in here. Or you could do an article on your 12bytes blog, and we'll reproduce it here, and credit you and link back to 12bytes - or we can all collaborate together to come up with a three/four page search wiki section (and you could repost em at 12bytes as a joint author). All a wiki is, is basic markup. I forgot to add to that itemized list that we need to make sure we have unique base64 icons (eg google vs google images etc - yes I still think we should provide sanitized google ones).

Collaborators with "write" access I think can edit the wiki. Otherwise users can fork the repo, and do a mockup in their fork.

eg sample draft

### Search Engines

1. AMO
- [DuckDuckGo](https://blahblah.com/) - **h**ttps://blahblah.com

I think we should decide the structure. I would like to store all the sanitzied xml files unzipped in a folder, but list them in the wiki. So we could have something like

🔸 Search Engines

  • 4.1 Into (written by @RoxKilly), list of engines (make a new topic and list all engines), their AMO, and a sanitized XML (stored in searchplugins root folder here in repo)
  • 4.2 How to maintain/edit (eg rename them, change the base64 icon, change keyword, how to not clash with existing installed engine names)
  • 4.3 How to use the search bar without unintentionally leaking/searching (note, I use CTR which restores the "list" rather than just icons - without CTR I cannot properly control searching when changing engines - any touch of the icon to change engines makes a new search, even if the field is null), eg tips on key word searching etc

We could easily get page 4.1's list of engines done, then decide on keywords, then create sanitized XMLs and upload them all, and then do a 4.2 on how to add/edit/maintain them.

@Thorin-Oakenpants

This comment has been minimized.

Show comment
Hide comment
@Thorin-Oakenpants

Thorin-Oakenpants May 15, 2017

Member

Step 1: see #118 for the initial page 4.1 list of search engines with info on them, page links
Step 2: write a wee intro on page 4.1
Step 3: add in links to AMO
Step 4: decide keywords, build sanitized XUL versions for all of them (upload to root subfolder on repo)
Step 5: write 4.2 - how to remove all your installed engines, add all our sanitized ones, then pick and choose what to keep and change the order
Step 6. write 4.3 - how to edit/create xml engines (eg site specific) and how to make a base64 icon etc
Step 7. have a beer [Assigned: earthlng]

Member

Thorin-Oakenpants commented May 15, 2017

Step 1: see #118 for the initial page 4.1 list of search engines with info on them, page links
Step 2: write a wee intro on page 4.1
Step 3: add in links to AMO
Step 4: decide keywords, build sanitized XUL versions for all of them (upload to root subfolder on repo)
Step 5: write 4.2 - how to remove all your installed engines, add all our sanitized ones, then pick and choose what to keep and change the order
Step 6. write 4.3 - how to edit/create xml engines (eg site specific) and how to make a base64 icon etc
Step 7. have a beer [Assigned: earthlng]

@earthlng

This comment has been minimized.

Show comment
Hide comment
@earthlng

earthlng May 15, 2017

Member

I would volunteer to do step 7

then decide on keywords

afaik keywords can not be set in the xml file and need to be done manually in search options.
They are then stored in the search.json.mozlz4. That's also one of the reasons why I prefer to update or add search engines manually via the console, instead of deleting and re-creating search.json.mozlz4.
The other reasons are that the default engines get un-hidden and my custom SE order is also lost.

Member

earthlng commented May 15, 2017

I would volunteer to do step 7

then decide on keywords

afaik keywords can not be set in the xml file and need to be done manually in search options.
They are then stored in the search.json.mozlz4. That's also one of the reasons why I prefer to update or add search engines manually via the console, instead of deleting and re-creating search.json.mozlz4.
The other reasons are that the default engines get un-hidden and my custom SE order is also lost.

@Thorin-Oakenpants

This comment has been minimized.

Show comment
Hide comment
@Thorin-Oakenpants

Thorin-Oakenpants May 15, 2017

Member

^^ My take is we provide a zip or individually ALL the xml files (base64 icons inside). Instructions on say wiki page 4.2

  • Open FF, delete ALL your search engines
  • Close FF, stick in ALL the search engine XML files you want in the dir
  • Open FF, go to search options, change the order to hearts content
  • ^^ might need to also assign keywords

Wiki page 4.3 would be how to make your own - eg site specific ones. Site Specific search engines should be the first choice, but we ain't gonna list all of those. I frequently use about a dozen non general search engines, just to stay off google (and no, the other engines suck at finding what i am after).

Member

Thorin-Oakenpants commented May 15, 2017

^^ My take is we provide a zip or individually ALL the xml files (base64 icons inside). Instructions on say wiki page 4.2

  • Open FF, delete ALL your search engines
  • Close FF, stick in ALL the search engine XML files you want in the dir
  • Open FF, go to search options, change the order to hearts content
  • ^^ might need to also assign keywords

Wiki page 4.3 would be how to make your own - eg site specific ones. Site Specific search engines should be the first choice, but we ain't gonna list all of those. I frequently use about a dozen non general search engines, just to stay off google (and no, the other engines suck at finding what i am after).

@earthlng

This comment has been minimized.

Show comment
Hide comment
@earthlng

earthlng May 15, 2017

Member

afaik it doesn't work that way. Once the search.json.mozlz4 exists, the searchplugins folder is ignored.

Member

earthlng commented May 15, 2017

afaik it doesn't work that way. Once the search.json.mozlz4 exists, the searchplugins folder is ignored.

@Thorin-Oakenpants

This comment has been minimized.

Show comment
Hide comment
@Thorin-Oakenpants

Thorin-Oakenpants May 15, 2017

Member

I'll have to play. I migrated all my engines that way before.

Member

Thorin-Oakenpants commented May 15, 2017

I'll have to play. I migrated all my engines that way before.

@earthlng

This comment has been minimized.

Show comment
Hide comment
@earthlng

earthlng May 15, 2017

Member

yes, it used to work but not anymore. Also if you change/edit an xml file in searchplugins it won't update it.
You need to delete search.json.mozlz4 or manually re-add the SE (or with an addon I assume)

Member

earthlng commented May 15, 2017

yes, it used to work but not anymore. Also if you change/edit an xml file in searchplugins it won't update it.
You need to delete search.json.mozlz4 or manually re-add the SE (or with an addon I assume)

@atomGit

This comment has been minimized.

Show comment
Hide comment
@atomGit

atomGit May 16, 2017

Collaborator

afaik it doesn't work that way. Once the search.json.mozlz4 exists, the searchplugins folder is ignored.

and ...

yes, it used to work but not anymore. Also if you change/edit an xml file in searchplugins it won't update it.

my understanding is that the searchplugins folder isn't ignored entirely - if you add an engine and delete search.json.mozlz4, the new engine becomes available in your search drop-down - i think what's happening is the new engine is imported into search.json.mozlz4, but i'm not entirely sure - for example, i added the Trove engine and the Searx engine as xml files, but only Trove shows up in search.json.mozlz4 - why this is, i don't know, but both are available from the search bar

as far as having to delete search.json.mozlz4, i believe this was the case before the move to this file (i forget the file name of the old one ... searchengines.json or something?) - even with the old file, you still had to delete it to refresh the available engines in the search bar

... and i think you're going to need some help with step 7 - luckily i'm available to collaborate regarding this issue :)

Collaborator

atomGit commented May 16, 2017

afaik it doesn't work that way. Once the search.json.mozlz4 exists, the searchplugins folder is ignored.

and ...

yes, it used to work but not anymore. Also if you change/edit an xml file in searchplugins it won't update it.

my understanding is that the searchplugins folder isn't ignored entirely - if you add an engine and delete search.json.mozlz4, the new engine becomes available in your search drop-down - i think what's happening is the new engine is imported into search.json.mozlz4, but i'm not entirely sure - for example, i added the Trove engine and the Searx engine as xml files, but only Trove shows up in search.json.mozlz4 - why this is, i don't know, but both are available from the search bar

as far as having to delete search.json.mozlz4, i believe this was the case before the move to this file (i forget the file name of the old one ... searchengines.json or something?) - even with the old file, you still had to delete it to refresh the available engines in the search bar

... and i think you're going to need some help with step 7 - luckily i'm available to collaborate regarding this issue :)

@earthlng

This comment has been minimized.

Show comment
Hide comment
@earthlng

earthlng May 16, 2017

Member

@atomGit can you remove the image shenanigans in your article, please?
We link to it from our Readme and it's not great that I can't even see the images without enabling JS and allow the redirect on sync.1dmp.io, or use a custom userscript to fix it in-place. Ideally host the images on your own domain and use the img src path instead of data-src, so it works without Javascript. IMHO it's a bit paradoxical to write an article about blocking/controlling 3rd party resources + JS while using those exact same practices. I'm only asking to change it for that one article. If you keep doing it that way on all your other articles I don't care.

edit: holy shit man, interesting articles btw

Member

earthlng commented May 16, 2017

@atomGit can you remove the image shenanigans in your article, please?
We link to it from our Readme and it's not great that I can't even see the images without enabling JS and allow the redirect on sync.1dmp.io, or use a custom userscript to fix it in-place. Ideally host the images on your own domain and use the img src path instead of data-src, so it works without Javascript. IMHO it's a bit paradoxical to write an article about blocking/controlling 3rd party resources + JS while using those exact same practices. I'm only asking to change it for that one article. If you keep doing it that way on all your other articles I don't care.

edit: holy shit man, interesting articles btw

@atomGit

This comment has been minimized.

Show comment
Hide comment
@atomGit

atomGit May 16, 2017

Collaborator

i thought that lazy-load plugin i was using worked when JS was disabled - it didn't, so i just switched to another one

as for self-hosting the images, that's something i intended to do and then forgot about it - i don't have a lot of time to fool around, but i'll get that done at some point

edit: holy shit man ...

lol - yeah, that's something i've been studying for a pretty long while now - try the quiz if you didn't already

Collaborator

atomGit commented May 16, 2017

i thought that lazy-load plugin i was using worked when JS was disabled - it didn't, so i just switched to another one

as for self-hosting the images, that's something i intended to do and then forgot about it - i don't have a lot of time to fool around, but i'll get that done at some point

edit: holy shit man ...

lol - yeah, that's something i've been studying for a pretty long while now - try the quiz if you didn't already

@earthlng

This comment has been minimized.

Show comment
Hide comment
@earthlng

earthlng May 16, 2017

Member

self-hosting the images ... i'll get that done at some point

Thanks mate

try the quiz if you didn't already

my lawyer tells me not to reply to this ;) (not really but you know what I mean)

Member

earthlng commented May 16, 2017

self-hosting the images ... i'll get that done at some point

Thanks mate

try the quiz if you didn't already

my lawyer tells me not to reply to this ;) (not really but you know what I mean)

@earthlng earthlng referenced this issue Jun 21, 2017

Closed

ToDo: diffs FF54-FF55 #144

24 of 24 tasks complete
@Thorin-Oakenpants

This comment has been minimized.

Show comment
Hide comment
@Thorin-Oakenpants

Thorin-Oakenpants Sep 9, 2017

Member

#118 can handle the task for doing the wiki pages - closing this

Member

Thorin-Oakenpants commented Sep 9, 2017

#118 can handle the task for doing the wiki pages - closing this

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment