Jump to content

My Browser Builds (Part 1)


Recommended Posts

13 hours ago, luweitest said:

Video problem : https://v.youku.com/v_show/id_XMzE2ODg0MDAw.html

I can't play that video in Serpent 52, both normal mode and safe mode

I can reproduce with a new/clean Serpent 52 profile; after a few minutes of troubleshooting, it appears that the site does not like the default User-Agent string of Serpent; the fix is to change to Firefox compatibility mode; in "about:config", toggle "general.useragent.compatMode.firefox" to "true" and reload the page (it needs Adobe Flash NPAPI plugin):

A4f2zZY.jpg

In an actual dirty profile, make sure to disable any content blockers for that site (I had to do so with uB0) :whistle:

 

Link to comment
Share on other sites


18 hours ago, Mathwiz said:

Edit: The Tab Mix Plus add-on seems to cause issues with Moebius; in particular, the browser customization page is completely blank.

One problem with Serpent55/Moebius is that it advertises itself on AMO as being Firefox 55.0, when the moebius platform in reality is much more akin to Mozilla 53.0; so AMO will offer the installation of extensions originally meant for Fx55 compatibility; browsing AMO can be mitigated by issuing a SSUAO for it: 

general.useragent.override.addons.mozilla.org;Mozilla/5.0 (Windows NT 6.1; rv:53.0) Gecko/20100101 Firefox/53.0

But Serpent internally still claims to be Firefox 55.0, so - if automatic addon update is enabled - may even update existing extensions to their Fx55 compatible versions, possibly breaking things, especially where the GUI aspects of the browser are concerned... Internal extension update mechanism (from AMO) is controlled (among others) by the following two prefs:

extensions.update.background.url
extensions.update.url

In my own copy of St55, I edited the string %CURRENT_APP_VERSION% inside the value of both prefs to read 53.0.

Both above "fixes" were more to-the-point when AMO also included legacy extensions; now they only apply to the few WEs there that are still Fx53+ compatible...

One third, difficult to mitigate, issue is that Serpent 55 also reports itself as being Firefox 55 to individual extensions about to be installed/already installed. Some type of extensions have different internal code depending on the version (among the range of versions they support) of the browser they're installed in; so, if a certain addon has one set of internal code for Fx53 and another set of code (JS/CSS, etc.) for Fx55, when it is installed in Sepent55 the code for Fx55 will be active, whereas the "more" appropriate code would've been the one for Fx53.

I encountered such an issue when I tried to install my favourite Complete Theme (FTDeepDark by Stefano Rosselli) on Serpent 55; while the Fx55 compatible version on AMO (back in the day...) was v14.4, after installation many parts of the browser GUI were broken :(; then I downloaded and installed v14.3.1 (which AMO said was 53.0-55.0 compatible), but again, after installation, my problems were not rectified :no: - the theme was internally applying the v55 CSS code, instead of the also present 53.0 one... In the end, I had to modify the theme code myself to achieve full compatibility with Serpent 55 :yes:

I am not a user of TMP; on "caa:addon/tab-mix-plus/versions" I'm reading that the last version (0.5.5.0) supports Firefox 38.0 - 56.*; this is only an educated guess on my part, but I suspect TMP 0.5.5.0 is applying code to St55 written for Fx55 (breaking thus the customisation tab) when in fact it should've applied code for Fx53; in all fairness, this isn't an extension's fault (rather the app's misrepresenting itself...).

I would go for a previous version like 0.5.0.2 ("updated for Firefox 53"); also, it wouldn't hurt to try the latest version on BitBucket, 0.5.6.0 :)

Link to comment
Share on other sites

8 hours ago, Mathwiz said:

But - manually reinstall every add-on, user pref, and bookmark? Cumbersome/tiresome is an understatement! "Impractical" comes closer to the mark.

... I still stand firmly by the advice I posted on this! :P For further reference, people should really read Moonchild's words on the matter: 

Can I copy my profile between Firefox/Pale Moon/Basilisk?

Furthermore, things are not so "intimidating" as you make them sound; once you know what and how to selectively transfer from the old app's profile to the new browser profile, it's a matter of a few hours, possibly less on more simple profiles...

Since we're talking about migrating profiles between Mozilla-type applications, the ample Firefox-oriented literature found on line is still valid/applicable (with certain modifications, of course):

https://support.mozilla.org/en-US/kb/profiles-where-firefox-stores-user-data

https://support.mozilla.org/en-US/kb/back-and-restore-information-firefox-profiles

https://support.mozilla.org/en-US/kb/recovering-important-data-from-an-old-profile

http://kb.mozillazine.org/Transferring_data_to_a_new_profile_-_Firefox

Bookmarks, browsing history (download history included in that), sessionstore, custom search engines, login-credentials for sites and cookies are easily transferred in a matter of minutes.

Then give some time to configure the new browser's GUI to your liking; then some more time inside the browser's preferences tab/window to configure them accordingly (different browsers may - and do - have different settings...).

I'll agree that the most cumbersome part is re-installing needed extensions; but:

1. Several extensions provide native support for exporting their settings to files (.txt/.json/.ini etc. ), which you can then use to reconfigure them exactly the old way in a new installation (e.g. Classic Theme Restorer, CookieKeeper, uBlock Origin, to name a few...).

2. Some legacy extensions store their settings inside the actual profile folder, in various guises; uB0 stores its filters inside the extension-data directory, as an .sqlite database - Stylish/Stylem userstyles are stored in the form of a stylish.sqlite database in the root of the profile - GreaseMonkey userscripts are stored inside a gm_scripts directory, and so on... So all these are easily transferable!

3. The bulk of legacy extensions store their settings as "about:config" preferences; but these are actually stored inside file "pref.js" in the root of the profile. Using a text editor, you can open file "pref.js" from the old (backed-up) profile and selectively transfer extension settings to the new "pref.js" file (inside the new profile - with browser closed, of course...).

Speaking from past experience, I'd say the most elaborate profile migration for me was from official unstable Pale Moon 27.9.1a1 (on Tycho) to New Moon 28.0.0b2 (on UXP) by Roytam1; I did not migrate early on, because I wanted UXP's teething issues to be cured first :D ; the two platforms are oceans apart, so I opted for a "proper" transition; several old extensions had to be discarded for good, alternatives to be found, some had to be updated to later versions, etc. but it was all over in less than 3 hours... Some fine tuning was still needed during the week after the migration (e.g. I decided I had to quit trying to make FTDeepDark work and go for Dark Moon) but all-in-all I had the least of problems compared to people who chose the "easy" way out (and then sought help in the PM forums...). 

But of course, YMMV; if you took the quicker way and you're OK, so much the better! :P

Link to comment
Share on other sites

So guys, has anyone used exensions like Tamper Monkey or Custom Style Script to replace existing JS code of the site? Have you had problems with CSP headers during this? Please give an advice which extension is better for PM 27 and how to install it (maybe some editing of the manifest is needed). I will try to make GitHub working again in PM 27 then :)

Link to comment
Share on other sites

17 hours ago, VistaLover said:

I still stand firmly by the advice I posted on this! :P

Well, mostly, I was just confessing that I cheated, and got away with it :D; and opining that, if you're moving from FF 52.9 to Serpent/UXP, or from Serpent/UXP to Serpent/Moebius, that's a small enough jump that you'll probably get away with "cheating," as long as you know what might go wrong and how to fix it.

But I agree that yours is the right way to proceed, especially if you're making a more substantial jump (to NM, for instance). I just think it's ridiculous to expect most users to go through that. You don't get good at something like that unless you've done it several times, and most folks don't migrate to a new browser very often. So it's going to take most folks hours, and who wants to spend hours just migrating to a new Web browser? A lot of folks will just throw in the towel, stick with FF, and say goodbye to all their plug-ins and XUL add-ons.

Really, the "correct" procedure should be automated and included in the PM and Basilisk installers, but neither installer does that (and I certainly don't expect @roytam1 or @i430VX to!) I think MCP could get a lot more converts to their browsers if they would invest a little programming effort simplifying migration from FF. Of course, that would also imply not removing features that break existing FF installations, which they've already done at least twice (thrice if you count moving Basilisk away from AMO). So I have little hope they'll follow my advice.

On 4/3/2019 at 10:34 PM, Mathwiz said:

I had trouble with one legacy add-on: Enhanced Statusbar. Serpent/UXP had updated it to v2.1 and all was well, but Serpent/Moebius kept losing the enhanced status bar with that version! I downgraded to Enhanced Statusbar v2.0.5, and it's working fine in Moebius now.

Just wanted to clarify that Extended Statusbar is a legacy (XUL) add-on, not a WE one. So both version 2.05 and 2.1 should be compatible with FF 52, NM 28, and Serpent - even with Moebius, there's no chance of version-dependent code being misled by Moebius's reported version 55. Unfortunately, that means it's no longer on AMO. v2.0.5 is availiable in the Classic Add-Ons Archive, but FF 52.9 users will need to download v2.1 from the PM or Basilisk add-on site (or from GitHub: https://github.com/kustodian/extended-statusbar, although I don't see any .xpi files there; don't know if you can just download the .zip and rename, changing the extension to .xpi, or if you need to compile something). Also, those versions probably aren't signed, so you'll need to flip the pref (forgot which one) that lets unsigned add-ons be installed (and it'll show in about:addons with a warning) in FF 52.9.

I think v2.1 has a bug. I had similar trouble adding it to NM 28, although I was able to fix it by toggling a new pref, from Web Protocols to All Protocols:

Capture5.PNG.446c91016c0f6af1a1d932f66edd5d99.PNG (Web protocols is the default and is supposed to mimic the v2.0.5 behavior, so it doesn't appear on, say, about: pages.)

Link to comment
Share on other sites

New build of Serpent/UXP for XP!

Test binary:
Win32 https://o.rths.cf/basilisk/basilisk52-g4.1.win32-git-20190406-cc07da9cb-xpmod.7z
Win64 https://o.rths.cf/basilisk/basilisk52-g4.1.win64-git-20190406-cc07da9cb-xpmod.7z

source code that is comparable to my current working tree is available here: https://github.com/roytam1/UXP/commits/custom

NM28XP build:
Win32 https://o.rths.cf/palemoon/palemoon-28.5.0a1.win32-git-20190406-cc07da9cb-xpmod.7z
Win64 https://o.rths.cf/palemoon/palemoon-28.5.0a1.win64-git-20190406-cc07da9cb-xpmod.7z

Official repo changes since my last build:
- Update SQLite to 3.27.2 (60d420b0a)
- Issue #187: Remove solaris conditional code. (25779d371)
- Issue #187: Remove solaris 1st party code OS checks. (6b968b13d)
- Remove 18 yo OS/2 and AIX bustage workaround. (1bf6a5c7d)
- Remove AIX VisualAge C++ compiler bug workaround. (49a3f5a53)
- Remove AIX 1st party code OS checks, part 1 (5d409fdca)
- Remove some HPUX leftovers. (b9a8bca64)
- Remove AIX 1st party code OS checks, part 2 (a2786c051)
- Unhook CR exception handler. (bfc977280)
- Remove crashreporter toolkit files. (ff2f287f8)
- Only draw a border on the tab bar when on bottom (520f99601)
- Bug 1360343 - ensure maskSurface is not null before dereference, since it can be null because of OOM or gfx device reset. r=dvander (71c031832)
- Merge pull request #1033 from Lootyhoof/issue-1025 (352121c2a)
- add a nullptr check in nsSVGUtils::PaintFrameWithEffects (00baf2836)
- Merge pull request #1035 from kn-yami/bug1360343 (a04fdfd89)
- Zap the robots. (6e4da38d2)
- CotPM update (cc07da9cb)

My changes since my last build:
- Revert "Issue #756 - Remove Contextual Identity from DevTools" (393112bc2)
- Revert "Issue #756 - Remove Contextual Identity from UnifiedComplete" (979a96c49)
- Revert "Issue #756 - Remove Contextual Identity from Basilisk" (a10f2e94a)
- Revert "Issue #756 - Remove contextual identity from BackgroundPageThumbs.jsm" (f097325b9)
- Revert "Issue #756 - Remove Contextual Identity References from Pale Moon" (4d4a6085d)
- nss: update nss to hg rev e5e10a46b9ad with vc2013 hackfix (dcdc5d70e)

Link to comment
Share on other sites

New build of BOC/UXP for XP!

Test binary:
MailNews Win32 https://o.rths.cf/boc-uxp/mailnews.win32-20190406-325e148-uxp-cc07da9cb-xpmod.7z
Browser-only Suite Win32 https://o.rths.cf/boc-uxp/bnavigator.win32-20190406-325e148-uxp-cc07da9cb-xpmod.7z

source patch (excluding UXP): https://o.rths.cf/boc-uxp/boc-uxp-src-xpmod-20190223.7z

No official repo changes since my last build.

For UXP changes please see above.

Link to comment
Share on other sites

New build of post-deprecated Serpent/moebius for XP!
* Notice: This repo will not be built on regular schedule, and changes are experimental as usual.
** Current moebius patch level should be on par with 52.9, but some security patches can not be applied/ported due to source milestone differences between versions.

Test binary:
Win32 http://o.rths.cf/basilisk/basilisk55-win32-git-20190406-4d70836fa-xpmod.7z
Win64 http://o.rths.cf/basilisk/basilisk55-win64-git-20190406-4d70836fa-xpmod.7z

repo: https://github.com/roytam1/basilisk55

Repo changes:
- nss: update nss to hg rev e5e10a46b9ad with vc2013 hackfix (b8c68d7f9)
- imported mozilla upstream changes: bug1468217, bug1436241, bug1464039, bug1458048, bug1408695, bug1423649 (45bc84221)
- port most of XP related hunks from iceweaselXP-53 (17c64c5c7)
- ported change from iceweaselXP-53: Restored eme-adobe plugin support for Windows XP systems. (8dc3c7f3e)
- fixed inverted logic in rev 17c64c5c71e12338934b7ae4ebfac2daf54efca9 (4d70836fa)

Link to comment
Share on other sites

New regular/weekly KM-Goanna release:
https://o.rths.cf/kmeleon/KM76.1.1-Goanna-20190406.7z

Changelog:

Out-of-tree changes:
* update Goanna3 to git d86d78c8b..4ed6f4941:
- import change from rmottola/Arctic-Fox:
 - further pointer style tweak for patching (90943bf7a)
 - Bug 1137910 part 1. Don't lose track of the original 'this' object in THIS_SAVEDFRAME, so we can actually do things based on the principal of the object we're working with. (0c6bf89dc)
 - Bug 1137910 part 2. Give chrome callers that are indirectly (e.g. via nsIStackFrame) poking at content-captured stacks the content view of the stack. (9b9d8981d)
 - more pointer style to pach easier (210ef0c78)
 - even more pointer style (8dad567c6)
 - Bug 1130576 - Rename JSPropertyOp -> JSGetterOp and JSStrictPropertyOp -> JSSetterOp. (0714aa290)
 - Bug 1083359 - Part 1 - Add the asyncCause and asyncParent properties to the native SavedFrame object. (1c99ec1bc)
 - more pointerstyle (21148a28d)
 - Bug 1135827 - Don't root more times than necessary in js::SavedStacks::insertFrames. (904461ae0)
 - comment fix (9401372c6)
 - Bug 1083359 - Part 2 - Allow C++ code to provide an async stack when calling a JS function. (e5ce833fa)
 - Bug 1133081, part 1 - Switch from js::PropDesc to JSPropertyDescriptor for all users of js::StandardDefineProperty (mainly Object.defineProperty/Properties and the corresponding Debugger.Object methods). (4754703df)
 - pointer style (58e1befc2)
 - Bug 1133094 - Object.defineProperty() on scripted proxy incorrectly sets {[[Configurable]]: true} if it's missing. (c4da7e6c6)
 - Bug 1133081, part 2 - Switch from js::PropDesc to JSPropertyDescriptor for js::StandardDefineProperty implementation. (07efac725)
 - Bug 1133081, part 3 - Switch from js::PropDesc to JSPropertyDescriptor for more odds and ends. (119b5711e)
 - more pointer style (b2c2a8383)
 - Bug 1133294 - Object.getOwnPropertyDescriptor(obj, key) must always return either undefined or a complete property descriptor, even if obj is a proxy. (39bc723a7)
 - Bug 1133081, part 4 - Reimplement the remaining PropDesc methods and delete PropDesc. (3b447fb7b)
 - Bug 1130537 - Reimplement js::SetPropertyIgnoringNamedGetter to follow ES6 draft rev 32 9.1.9 [[Set]]. (7a681465a) (e1ef2cd0b)
- import change from rmottola/Arctic-Fox: Bug 1133081, part 5 - Remove non-asserting PropertyDescriptor accessors in favor of the new PropDesc-inspired asserting accessors. (9da3866fa)
 import change from tenfourfox: #399: null check in FindNearestCommonAncestorFrame from PM 978 (23f303460) (d83c58e95)
- nss: update nss to hg rev e5e10a46b9ad with vc2013 hackfix (c532d18da)
- import change from rmottola/Arctic-Fox:
 - some pointerstyle (5eafc4481)
 - even more pointer style (8b1c94dd0)
 - Bug 1138984 - Allow 'new' template objects to have dynamic slots (31cff4565) (593d1c056)
- import change from rmottola/Arctic-Fox:
 - revert apparently PaleMoon specific change to CreateLazyScriptsForCompartment, so later FF patches apply smoothly (ed89197ec)
 - Bug 1140428 - Warn when __noSuchMethod__ is used. (976c93809)
 - Bug 1133759 - Always shrink GC buffers at the end of a shrinking GC, even if we don't compact (5d9fadcfa)
 - pointerstyle tweaks again (26426837a)
 - Bug 1125101 - Incrementalise compacting GC by zones (9badc2329)
 - Bug 1136597 - Mark all relevant cross compartment debugger edges when sweeping. (Adapted) (964373773)
 - some pointerstyle updates (1ff2d78b7)
 - more pointer style (be38e9b5c)
 - Bug 1135100 - Don't bother to check if things we don't relocate have been forwarded (8da001057)
 - Bug 1135100 - Remove Cell* versions of various GC methods forcing callers to use correct type (a1d09d1fa)
 - pointer style (6f7d4091e)
 - Bug 1137573 - OdinMonkey: Alignment Mask Analysis (fd86305f3)
 and removed addTelemetry hunk from rev 976c93809 (4ed6f4941)

* Notice: the changelog above may not always applicable to XULRunner code which K-Meleon uses.

A goanna3 source tree that has kmeleon adaption patch applied is available here: https://github.com/roytam1/palemoon27/tree/kmeleon76

Edited by roytam1
Link to comment
Share on other sites

New New Moon 27 Build!

32bit https://o.rths.cf/palemoon/palemoon-27.9.6.win32-git-20190406-4ed6f4941-xpmod.7z
32bit SSE https://o.rths.cf/palemoon/palemoon-27.9.6.win32-git-20190406-4ed6f4941-xpmod-sse.7z
32bit noSSE https://o.rths.cf/palemoon/palemoon-27.9.6.win32-git-20190406-4ed6f4941-xpmod-ia32.7z

64bit https://o.rths.cf/palemoon/palemoon-27.9.6.win64-git-20190406-4ed6f4941-xpmod.7z

source repo: https://github.com/roytam1/palemoon27

repo changes since my last build:
- import change from rmottola/Arctic-Fox:
 - further pointer style tweak for patching (90943bf7a)
 - Bug 1137910 part 1. Don't lose track of the original 'this' object in THIS_SAVEDFRAME, so we can actually do things based on the principal of the object we're working with. (0c6bf89dc)
 - Bug 1137910 part 2. Give chrome callers that are indirectly (e.g. via nsIStackFrame) poking at content-captured stacks the content view of the stack. (9b9d8981d)
 - more pointer style to pach easier (210ef0c78)
 - even more pointer style (8dad567c6)
 - Bug 1130576 - Rename JSPropertyOp -> JSGetterOp and JSStrictPropertyOp -> JSSetterOp. (0714aa290)
 - Bug 1083359 - Part 1 - Add the asyncCause and asyncParent properties to the native SavedFrame object. (1c99ec1bc)
 - more pointerstyle (21148a28d)
 - Bug 1135827 - Don't root more times than necessary in js::SavedStacks::insertFrames. (904461ae0)
 - comment fix (9401372c6)
 - Bug 1083359 - Part 2 - Allow C++ code to provide an async stack when calling a JS function. (e5ce833fa)
 - Bug 1133081, part 1 - Switch from js::PropDesc to JSPropertyDescriptor for all users of js::StandardDefineProperty (mainly Object.defineProperty/Properties and the corresponding Debugger.Object methods). (4754703df)
 - pointer style (58e1befc2)
 - Bug 1133094 - Object.defineProperty() on scripted proxy incorrectly sets {[[Configurable]]: true} if it's missing. (c4da7e6c6)
 - Bug 1133081, part 2 - Switch from js::PropDesc to JSPropertyDescriptor for js::StandardDefineProperty implementation. (07efac725)
 - Bug 1133081, part 3 - Switch from js::PropDesc to JSPropertyDescriptor for more odds and ends. (119b5711e)
 - more pointer style (b2c2a8383)
 - Bug 1133294 - Object.getOwnPropertyDescriptor(obj, key) must always return either undefined or a complete property descriptor, even if obj is a proxy. (39bc723a7)
 - Bug 1133081, part 4 - Reimplement the remaining PropDesc methods and delete PropDesc. (3b447fb7b)
 - Bug 1130537 - Reimplement js::SetPropertyIgnoringNamedGetter to follow ES6 draft rev 32 9.1.9 [[Set]]. (7a681465a) (e1ef2cd0b)
- import change from rmottola/Arctic-Fox: Bug 1133081, part 5 - Remove non-asserting PropertyDescriptor accessors in favor of the new PropDesc-inspired asserting accessors. (9da3866fa)
 import change from tenfourfox: #399: null check in FindNearestCommonAncestorFrame from PM 978 (23f303460) (d83c58e95)
- nss: update nss to hg rev e5e10a46b9ad with vc2013 hackfix (c532d18da)
- import change from rmottola/Arctic-Fox:
 - some pointerstyle (5eafc4481)
 - even more pointer style (8b1c94dd0)
 - Bug 1138984 - Allow 'new' template objects to have dynamic slots (31cff4565) (593d1c056)
- import change from rmottola/Arctic-Fox:
 - revert apparently PaleMoon specific change to CreateLazyScriptsForCompartment, so later FF patches apply smoothly (ed89197ec)
 - Bug 1140428 - Warn when __noSuchMethod__ is used. (976c93809)
 - Bug 1133759 - Always shrink GC buffers at the end of a shrinking GC, even if we don't compact (5d9fadcfa)
 - pointerstyle tweaks again (26426837a)
 - Bug 1125101 - Incrementalise compacting GC by zones (9badc2329)
 - Bug 1136597 - Mark all relevant cross compartment debugger edges when sweeping. (Adapted) (964373773)
 - some pointerstyle updates (1ff2d78b7)
 - more pointer style (be38e9b5c)
 - Bug 1135100 - Don't bother to check if things we don't relocate have been forwarded (8da001057)
 - Bug 1135100 - Remove Cell* versions of various GC methods forcing callers to use correct type (a1d09d1fa)
 - pointer style (6f7d4091e)
 - Bug 1137573 - OdinMonkey: Alignment Mask Analysis (fd86305f3)
 and removed addTelemetry hunk from rev 976c93809 (4ed6f4941)

Edited by roytam1
Link to comment
Share on other sites

My server's file has been updated to support the latest builds...

Updating the installer itself is not necessary this time, but I would recommend it as some changes happened:

CHANGE LOG:

- Support for New Moon SSE and noSSE browsers to be installed.

         Note: I have no way to tell if the installer itself works on PCs without SSE as I have none with 2k/xp on them
         Note x2: These are the only PM27 builds currently supported are the SSE/noSSE, because I didn't want the installer to get "crowded"

-Possible Bug fixed with the win2k version of the installer

Another note: for now, the installer versions will continue to be separate as putting them together would seem to require a complete rework of the installer.

As always, the installers are available here:
Windows XP and later Version
Win2k Version

 

Link to comment
Share on other sites

3 hours ago, someguy25 said:

bug.jpg

Seems like a bug after i upgraded to the latest version :dubbio:

seems to be a bug in: https://github.com/MoonchildProductions/UXP/commit/00baf283622b47ad7926c6e62364854d3dfbc00a

EDIT: fix confirmed by reverting this commit

EDIT2: all today's UXP based binaries are inplace updated.

Edited by roytam1
Link to comment
Share on other sites

Thanks to Roy for restoring container tabs in UXP, it saves me the trouble of opening another browser but on a side note their temporary removal did get me to try out the moebius version and now I've got two nice running browsers to play around with. --:)

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...