Closed
Bug 820582
Opened 12 years ago
Closed 11 years ago
Add private browsing package.json flag
Categories
(Add-on SDK Graveyard :: General, defect, P1)
Add-on SDK Graveyard
General
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: evold, Assigned: evold)
References
Details
Attachments
(1 file)
No description provided.
Assignee | ||
Updated•12 years ago
|
Summary: Add private-browsing opt-in to package.json → Add private-browsing opt-in key to package.json
Assignee | ||
Comment 2•12 years ago
|
||
I am thinking about adding a private-browsing opt-in key to package.json which would mean modules like context-menu, widget, page-mod, and probably more modules would all need to be disabled on pb enabled windows by default, unless the developer opts-in to be used on those windows. Can I get your thoughts on this Dave, Irakli, and anyone else?
Assignee: nobody → evold
Flags: needinfo?(zer0)
Flags: needinfo?(rFobic)
Flags: needinfo?(jgriffiths)
Flags: needinfo?(dtownsend+bugmail)
Would it be easier to require the addon to require("private-browsing") than to require adding things to package.json?
Would it be easier to require the addon to require("private-browsing") than to require adding things to package.json?
Comment 5•12 years ago
|
||
As I stated on the list, I'm concerned that this flag doesn't give the user much choice as to whether an add-on should work or not in PB mode.
Flags: needinfo?(zer0)
Flags: needinfo?(jgriffiths)
Comment 6•12 years ago
|
||
(In reply to Erik Vold [:erikvold] [:ztatic] from comment #2) > I am thinking about adding a private-browsing opt-in key to package.json > which would mean modules like context-menu, widget, page-mod, and probably > more modules would all need to be disabled on pb enabled windows by default, > unless the developer opts-in to be used on those windows. > > Can I get your thoughts on this Dave, Irakli, and anyone else? Why those modules? By themselves they do nothing that is against the rules of private browsing. It'd be better to disabling things like the prefs modules which actually have the ability to store private data, but I'm not sure that is a great idea either. AMO has a policy that add-ons be required to properly monitor PB state and behave accordingly, I think we should rely on that rather than trying to put in safeguards that aren't actually protecting the thing we need to protect.
Flags: needinfo?(dtownsend+bugmail)
Assignee | ||
Comment 7•12 years ago
|
||
(In reply to Dave Townsend (:Mossop) from comment #6) > (In reply to Erik Vold [:erikvold] [:ztatic] from comment #2) > > I am thinking about adding a private-browsing opt-in key to package.json > > which would mean modules like context-menu, widget, page-mod, and probably > > more modules would all need to be disabled on pb enabled windows by default, > > unless the developer opts-in to be used on those windows. > > > > Can I get your thoughts on this Dave, Irakli, and anyone else? > > Why those modules? By themselves they do nothing that is against the rules > of private browsing. It'd be better to disabling things like the prefs > modules which actually have the ability to store private data, but I'm not > sure that is a great idea either. > > AMO has a policy that add-ons be required to properly monitor PB state and > behave accordingly, I think we should rely on that rather than trying to put > in safeguards that aren't actually protecting the thing we need to protect. Perhaps I was too concerned with avoiding add-on breakage, that is the only reason I had for bringing up this safeguard which looks more and more like a rabbit hole now. I'm ok with simply changing the behavior of the panel module and reaching out to developers that use it too.
Flags: needinfo?(rFobic)
Priority: -- → P1
Updated•12 years ago
|
Summary: Add private-browsing opt-in key to package.json → Add private browsing preference by default to package.json that is then exposed to the user.
Comment 8•12 years ago
|
||
As per the discussion in triage today, we were thinking a default pref could be added to package.json for pwpb opt-in.
Assignee | ||
Updated•12 years ago
|
Summary: Add private browsing preference by default to package.json that is then exposed to the user. → Add private browsing simple-pref
Assignee | ||
Updated•11 years ago
|
No longer blocks: sdk-pwpb
Summary: Add private browsing simple-pref → Add private browsing package.json flag
Assignee | ||
Comment 9•11 years ago
|
||
Pointer to Github pull-request
Assignee | ||
Updated•11 years ago
|
Attachment #711204 -
Flags: review?(rFobic)
Assignee | ||
Updated•11 years ago
|
Attachment #711204 -
Flags: review?(poirot.alex)
Comment 10•11 years ago
|
||
Comment on attachment 711204 [details] Pointer to Github pull request: https://github.com/mozilla/addon-sdk/pull/763 Please note: There were bunch of pull requests containing each other, so comments may be little spread out.
Attachment #711204 -
Flags: review?(rFobic)
Attachment #711204 -
Flags: review?(poirot.alex)
Attachment #711204 -
Flags: review-
Assignee | ||
Updated•11 years ago
|
Attachment #711204 -
Flags: review- → review?(rFobic)
Assignee | ||
Comment 11•11 years ago
|
||
I added docs and got the other comments I believe, ready for another round now Irakli.
Assignee | ||
Comment 12•11 years ago
|
||
(In reply to Irakli Gozilalishvili [:irakli] [:gozala] [@gozala] from bug 839746 comment #4) > Maybe property name can be made more descriptive though, like > `supports-private-browsing: true, { supports: { private-browsing: true } }` > or something, I'll let native ENG speakers bike-shed / decide on proper > naming. How about `{ permissions: [ "private-browsing" ] }` like chrome extensions: http://developer.chrome.com/extensions/declare_permissions.html
Comment 13•11 years ago
|
||
(In reply to Erik Vold [:erikvold] [:ztatic] from comment #12) > (In reply to Irakli Gozilalishvili [:irakli] [:gozala] [@gozala] from bug > 839746 comment #4) > > > Maybe property name can be made more descriptive though, like > > `supports-private-browsing: true, { supports: { private-browsing: true } }` > > or something, I'll let native ENG speakers bike-shed / decide on proper > > naming. > > How about `{ permissions: [ "private-browsing" ] }` like chrome extensions: > > http://developer.chrome.com/extensions/declare_permissions.html Maybe, I don't know to be honest, I don't in particular mind with previous option, it would be just great if you and wbamberg, mossop could figure correct naming.
Comment 14•11 years ago
|
||
Comment on attachment 711204 [details] Pointer to Github pull request: https://github.com/mozilla/addon-sdk/pull/763 Looks good, just make sure to use full require paths as pointed out in review, and would be great if you can settle with Will / Mossop on property name before landing it.
Attachment #711204 -
Flags: review?(rFobic) → review+
Comment 15•11 years ago
|
||
On the second thought we should probably follow the ff os lead here and copy their approach with app.manifest that is also similar to chrome extensions that you've suggested: https://developer.mozilla.org/en-US/docs/Apps/Manifest?redirectlocale=en-US&redirectslug=OpenWebApps%2FThe_Manifest#permissions
Comment 16•11 years ago
|
||
If we want to follow openwebapp manifest, (I'd vote for this), it would be something like this: { "id": "my-addon-id", "permissions": { "private-browsing": {} } } It isn't just an array like chrome as we can pass optional arguments for each permission. A common one is `access` property with read, write and readwrite possible values.
Assignee | ||
Comment 17•11 years ago
|
||
I'm using the "permissions" property now, that and https://github.com/mozilla/addon-sdk/pull/763/files#diff-0 are the only changes, mind taking a quick review of that before I merge Irakli?
Flags: needinfo?(rFobic)
Comment 18•11 years ago
|
||
Commits pushed to master at https://github.com/mozilla/addon-sdk https://github.com/mozilla/addon-sdk/commit/c2db6ffe65523b657f4e7c5a25da8a3bf8b47018 bug 820582: fix for tests that used loader w/out metadata https://github.com/mozilla/addon-sdk/commit/b2ec37edb05d7b789ad780815ef88e44220a8b44 bug 820582: review updates, self module now exports isPrivateBrowsingSupported, and there is a test that this is true when it should be https://github.com/mozilla/addon-sdk/commit/5b4fbc80d7c9f82122390ea5c0f54f5b308bccdf Bug 820582: address review comment on diff in lib/sdk/loader/cuddlefish.js https://github.com/mozilla/addon-sdk/commit/9fcbce9c79356c857f2c2830b0f384a72d229039 Bug 820582 adding baisc docs, more to come https://github.com/mozilla/addon-sdk/commit/36d6971858670711017ced10d16168940c0b97e8 Bug 820582 - updating package.json to use permissions property https://github.com/mozilla/addon-sdk/commit/7018afebad65d5529db30099bffa56e4a657b78b Bug 820582: freezing metadata and metadata.permissions in bootstrap.js file now and no longer need to do this in the loader; freezing should only be done for add-on metadata https://github.com/mozilla/addon-sdk/commit/9ded485783703a2447b46719bf60d60a64ba2c4c bug 820582: updating package.json docs to reflect the new permissions key https://github.com/mozilla/addon-sdk/commit/d1f3d5136fc2de16e2b8eae04dd8acebeafde42b Merge pull request #763 from erikvold/820582 Fix Bug 820582 adding private-browsing flag to package.json r=@Gozala
Updated•11 years ago
|
Status: NEW → RESOLVED
Closed: 11 years ago
Resolution: --- → FIXED
Assignee | ||
Updated•11 years ago
|
Blocks: sdk-pwpb-fx21
You need to log in
before you can comment on or make changes to this bug.
Description
•