Jump to content

My Browser Builds (Part 1)


Recommended Posts

On 10/18/2018 at 2:10 PM, VistaLover said:

You don't specify, but if you're interested in an in-browser solution, I highly recommend the SaveFrom.net one, to which I have recently switched :):https://en.savefrom.net/user.php?vid=308#youtube

Scroll all the way down to Installation; the Mozilla Firefox extension is of the webextension format, so won't install in New Moon (does so in Serpent 52.9.0/55.0.0); but if you have Greasemonkey for Pale Moon installed, you can proceed to install their userscript from

https://download.sf-helper.com/chrome/helper.user.js

(Just load in a new tab and follow the prompt).

So, I installed the web extension in Basilisk and Greasemonkey and the sf-helper user script in NM. Both work fine. However....

... the web extension version prevents me from logging into Chase.com on Basilisk! I would never have figured that out, except I have the "official" version of Basilisk for the Win 7 side of my PC, and I didn't install the web extension there.

The user script version didn't block me from logging into Chase from NM 28, so I'll try it on the Win 7 side and see if it gives me the best of both worlds....

Edit: The script seems to work with Greasemonkey 4.1, and without interfering with me logging into Chase. So now I'll try it that way on the XP side. Edit 2: No joy. Greasemonkey 4.1 installs on @roytam1's Basilisk XP build, but doesn't seem to work for some reason.

So I guess the easiest solution on Basilisk XP is still to use the web extension, and just disable it when I want to log into Chase, and re-enable it when I want to download Youtube videos.

Edited by Mathwiz
Link to comment
Share on other sites


8 hours ago, roytam1 said:

you may need to set general.useragent.compatMode.firefox and general.useragent.compatMode.gecko to true in about:config

this trick won't work on MP28. it's not possible to toggle the value. why don't you set it to true by default ?

Link to comment
Share on other sites

25 minutes ago, caliber said:

this trick won't work on MP28. it's not possible to toggle the value. why don't you set it to true by default ?

New Moon uses different config options than Basilisk. In NM you have to set it in Preferences / Advanced / General / Compatibility:

On 10/22/2018 at 4:27 AM, zipandlips said:

yes yes I did everything!thank!:yes:

Screenshot_1.png

 

Link to comment
Share on other sites

8 hours ago, Mathwiz said:

Edit 2: No joy. Greasemonkey 4.1 installs on @roytam1's Basilisk XP build, but doesn't seem to work for some reason.

GM 4.1 is of the WebExtension format and is dependent on WE APIs not present inside Basilisk52/UXP (which, as stated numerous times, has only a limited subset of WE APIs compared to FxESR 52); if you want to install GM on Basilisk, go for the XUL (legacy) version 3.17 or the PM fork 3.31.4

If, OTOH, you insist on installing a userscript manager of the WE format, latest Violentmonkey 2.9.9 works fine with Basilisk (both 52/UXP + 55/moebius)... 

8 hours ago, Mathwiz said:

So I guess the easiest solution on Basilisk XP is still to use the web extension, and just disable it when I want to log into Chase, and re-enable it when I want to download Youtube videos.

The WE addon from SF.net has a feature that enables you to selectively disable/enable it per site; perhaps that's what you need (not tested, as I don't have an account with chase.com ;) ):

DFIH8dc.jpg

Regards

Edited by VistaLover
Link to comment
Share on other sites

3 hours ago, VistaLover said:

GM 4.1 is of the WebExtension format and is dependent on WE APIs not present inside Basilisk52/UXP (which, as stated numerous times, has only a limited subset of WE APIs compared to FxESR 52); if you want to install GM on Basilisk, go for the XUL (legacy) version 3.17 or the PM fork 3.31.4

But curiously, GM 4.1 does work (at least well enough to download Youtube videos) on the "official" Basilisk browser (version 52.9.2018.09.27); just not on @roytam1's XP builds (always say version 52.9.0). I'd think both versions would have the same WE APIs :unsure:

Nevertheless, since I reverted to the WE add-on for Basilisk on XP, I think this is the best solution: :cool:

3 hours ago, VistaLover said:

The WE addon from SF.net has a feature that enables you to selectively disable/enable it per site; perhaps that's what you need (not tested, as I don't have an account with chase.com ;) ):

I'll try it when I get back to my work PC on Monday and let you all know if it solved the problem.

Edit: Ugh. Chase has made some annoying changes to their Web site:

  1. network.http.sendRefererHeader must be set at the default of 2 now; otherwise you don't even get the login prompt!
  2. User agent must be a "classic" FF UA (no mention of Goanna or Basilisk), or you get their "your browser is out of date" nonsense and are redirected to Chase's mobile site. ("Official" Basilisk mentions both Gecko and Goanna, and both Firefox and Basilisk in its UA when FF compatibility mode is set to "true;" this is fine for most sites but not Chase's.)

Oh, well, the option to disable SF.net for the Chase website seems to work, at least on "official" Win 7 Basilisk; so I'm hopeful it will work on Roytam1's XP builds too. As I said, I'll find out Monday.

Edited by Mathwiz
Link to comment
Share on other sites

New build of basilisk/UXP for XP!

Test binary:
Win32 https://o.rths.cf/basilisk/basilisk52-g4.1.win32-git-20181027-4f59678ac-xpmod.7z
Win64 https://o.rths.cf/basilisk/basilisk52-g4.1.win64-git-20181027-4f59678ac-xpmod.7z

diff: https://o.rths.cf/basilisk/UXP-xp-gitdiff-20181027.7z

PM28XP build:
Win32 https://o.rths.cf/palemoon/palemoon-28.2.0a1.win32-git-20181027-4f59678ac-xpmod.7z
Win64 https://o.rths.cf/palemoon/palemoon-28.2.0a1.win64-git-20181027-4f59678ac-xpmod.7z

Official repo changes since my last build:
- remove named function syntax in addonManager.js (0d97c6ada)
- remove named function syntax in AddonManager.jsm (06086f82f)
- remove named function syntax in amWebInstallListener.js (d218160bf)
- remove named function syntax in LightweightThemeManager.jsm (dd574ec7b)
- make style of amInstallTrigger.js more consistent (6667a10d3)
- Issue #795 - Make nsIUserInfo component optional (19b6f2d9b)
- Port the libvpx mozbuild generator to aom (07c17b6b9)
- Import aom library (68569dee1)
- Generate build description for libaom (a66e91651)
- Add --enable-av1 configure switch (99ed16064)
- Add AOMDecoder (9aea199da)
- Add AOMDecoder to AgnosticDecoderModule (3e0443e4c)
- Recognize AV1 in WebMDemuxer (ee2cb65e2)
- Add missing includes to WebMDemuxer (e5545e10a)
- Add aom to the list of thirdparty code (e8480f8a7)
- [aom] Remove 32-bit Mac OS build config (08619db14)
- [aom] Enable YASM (f45278950)
- Make AOMDecoder actually build (26dc168e3)
- Remove aom_codec_stream_info_t sz field references (1036d1fa0)
- Add av1 to MediaSource.isTypeSupported (0cc51bc10)
- Update aom to slightly newer commit ID (df9477dfa)
- Fix typo (build bustage) (4b5e22956)
- Add missing includes to FFmpegLibWrapper (dddc2aa9e)
- [av1] Clean up duplicate filenames check (afda5e384)
- [aom] Add x86-win32-gcc config (857b86f25)
- Make aom_config.asm match upstream (ecdaf7930)
- [aom] Filter out CONFIG_EXT_PARTITION_TYPES (4121d7571)
- [aom] Remove unused option (77887af9c)
- Update aom to commit id f5bdeac22930ff4c6b219be49c843db35970b918 (7369c7d7a)
- Export aom_config.h (21598d1bd)
- [aom] Fix stream info peeking (e394e2049)
- [aom] Resample high bit depth frames (4653be960)
- [aom] Don't resample 8-bit images (0c98b7165)
- [aom] Fix win32 debug build (01eb57073)
- Update aom to commit id e87fb2378f01103d5d6e477a4ef6892dc714e614 (ec910d814)
- [av1] Fix build issues (464c3130f)
- Fix aom compile errors with VS2015 (125aff11b)
- Add Python cmake parser (39e842a83)
- Add generate_sources_mozbuild.py (5c0c6c73a)
- Update generate_sources_mozbuild.sh (b425400e9)
- Add aom_version.h (c410f04e8)
- Add support for SSE 4.2 to libaom moz.build (14805f6dd)
- Update aom to v1.0.0 (bbcc64772)
- Do not build aomstats unless examples are enabled (ab5b4462a)
- [aom] Fixup moz.build (4f63fc3bd)
- Updates to AOMDecoder for aom v1.0.0 (ee1300453)
- Add missing : to libaom moz.build (9469bc196)
- [aom] No longer necessary to run lint_config.sh anymore (a4d3c59dc)
- Update libaom to rev b25610052a1398032320008d69b51d2da94f5928 (b8df135c9)
- [aom] Only build stats code if examples are enabled (ba5a1ea60)
- Update libaom vendor scripts to account for upstream changes (41fbdea45)
- Update libaom to commit ID 1e227d41f0616de9548a673a83a21ef990b62591 (d2499ead9)
- Remove superfluous message manager property (f91ab88b9)
- Merge pull request #838 from FranklinDM/notification-work (714ccb6f8)
- Put AV1 codec behind a pref (6ddf66542)
- [css-grid] Don't shrink-wrap the inline size when we have an available size when measuring block size (88671bcc3)
- [css-grid] Reftests for 88671bcc3633af5a5e34715518e0804ca892fac7 (7110f93fb)
- Merge pull request #841 from JustOff/PR_css_grid_shrink_wrap (54deaf443)
- Merge pull request #839 from kn-yami/issue516 (38173d9e4)
- fix naming of getStartupChanges function (ca4b2c502)
- Merge pull request #843 from kn-yami/issue516 (a9fbfba2a)
- align Basilisk sidebar context menu behaviour with Pale Moon (50050aac0)
- [PALEMOON] Add filter query to about:permissions (0d7afa78f)
- [PALEMOON] Open about:permissions when notification settings are requested (4c488cb7f)
- Fix canPlayType/isTypeSupported for AV1 content (e5b30fc95)
- Downsample av1 images unconditionally (070c2cb24)
- Use larger stack for media decoder threads (47a01617e)
- Fix reference to undefined property warnings in toolkit's browser.xml. (61a05da01)
- Add support for AV1 in MP4 (29f718ef7)
- Ensure we correctly parse the finalized codec string for av1 (23013dda6)
- backport m-c bug 1333174 - Don't use NS_ENSURE_SUCCESS at nsIOService:793 (dbac7d691)
- Merge pull request #844 from kn-yami/issue768 (1d3233c49)
- Improve graph edge resolution code. (980b7e4da)
- Code style improvements (no bug). (b4e78255c)
- Revert "Add support for AV1 in MP4" (192199b03)
- Show hidden left border of search button on about:newtab page (1b5066040)
- Basilisk: Trivial fixes for Mac build (d9c5e2b80)
- Merge pull request #849 from adeshkp/basilisk-mac-build (70c214344)
- Merge pull request #848 from adeshkp/search-btn-border (e4836ec73)
- backport m-c bug 1333172 - Avoid 1100 warnings by replacing NS_ENSURE_SUCCESS in nsNetUtilInlines.h:180 (e4cc27a0e)
- Merge pull request #847 from g4jc/debugspam (81acc4099)
- Merge pull request #850 from trav90/add-av1-support (d1a35c3fa)
- Merge pull request #796 from g4jc/nsuserinfo (f85840a05)
- Merge pull request #845 from JustOff/PR_notifications_open_settings (d07a4f2ba)
- Revert "Make nsIUserInfo component optional with --disable-userinfo at build time" (afe69140a)
- Merge pull request #851 from MoonchildProductions/revert-796-nsuserinfo (f71c04d81)
- Bug 1489785: Remove a workaround for gcc (6db23ae65)
- Bug 1489785: Remove a workaround for gcc (065f6f9e5)
- Merge pull request #852 from wicknix/master (f5b13e91d)
- Issue #795 - Make nsIUserInfo component optional (6a2ff38f0)
- Merge pull request #854 from g4jc/nsuserinfo (4f59678ac)

From now --enable-av1 option is compiled in. you can enable AV1 decoding by toggling "media.av1.enabled" in about:config.
Test passed from site https://demo.bitmovin.com/public/firefox/av1/ but still not recognized by youtube.
Testing in my main machine (i5-3450), 720p 500kbps AV1 will cause laggy playback in test site above.

Edited by roytam1
Link to comment
Share on other sites

New build of Firefox 45ESR SSE:

test binary: https://o.rths.cf/gpc/files1.rt/firefox-45.9.12-20181027-b317f782f-win32-sse.7z

repo: https://github.com/roytam1/mozilla45esr

Changes since my last build:
- import changes from tenfourfox:
 - #525: pref off for FPR10 (6912ea34a)
 - #526: update certs and pins (8a7210bac)
 - #526: M1492823 M1483699 M1496340 M1492524 (f818695f0) (0f0a4f2f2)
- bump revision to 45.9.12 (3d4f729ec)
- update NSS to rev 9dd8d5e2156e with nss339-vc2013.diff applied (c00989e51)
- backout #error from M1499198 to fix build (b317f782f)

Link to comment
Share on other sites

Didn't work, but it was a good suggestion roytam1.

If you google, you'll find a long list of new and old posts about browsers not rendering reCAPTCHA. I don't
know why Chrome is only browser that renders reCAPTCHA consistently. I must have performed fifty searches.
Never found anything that actually worked.

I'm not sure if this exists, but is there an app that creates a log file about all browser activity after clicking on
a reCAPTCHA link? If there is such an app, I could post the results here at MSFN. Maybe one of you guys
would see the problem in the log file.

Edited by HoppaLong
Link to comment
Share on other sites

3 hours ago, Dibya said:

@roytam1 please can you compile Latest Firefox without TLS limitation https://docs.microsoft.com/en-us/cpp/build/reference/zc-threadsafeinit-thread-safe-local-static-initialization?view=vs-2017

Then it might work with wrapper library based on wine

as stated before, there is exclusive use of Win7 API, and Rust side issues.

Link to comment
Share on other sites

10 hours ago, caliber said:

thanks @Mathwiz. we need here more people like you who post snapshots for doomies.

what is the difference or advantage of the firefox compatibility over the Gecko one ?

I'm not positive, but I believe "Gecko Compatibility" replaces "Goanna/4.1" with "Gecko/20100101" in the user agent string (for the benefit of sites that don't know what the "Goanna" rendering engine is).

"Firefox Compatibility" does that too, but it also replaces "Pale Moon/28.2" with "Firefox/52.9." Again for the benefit of sites that don't know about (or don't like) Pale Moon.

About the only advantage of "Gecko Compatibility" over "Firefox Compatibility" would be that sites that do know about Pale Moon won't report version 28.2 as "out of date," but may report Firefox 52.9 as out of date. In my experience though, you're less likely to run into problems using "Firefox Compatibility" than using "Gecko Compatibility." (But as always, YMMV.)

Link to comment
Share on other sites

15 hours ago, HoppaLong said:

I don't know why Chrome is only browser that renders reCAPTCHA consistently.

I don't know this for a fact, but I'll hazard a guess: Google keeps changing reCAPTCHA code and only tests it thoroughly with their own browser, Chrome. Maybe also the latest version of FF, if we're lucky. But Pale Moon and Basilisk, both based on an older FF version, 52.9? Fuhgeddaboudit.

My guess is, they're trying to stay ahead of the bots, so it's not totally nefarious; but they also have no incentive not to break older browsers. After all, it just might drive more users to Chrome in the process. What a lucky happenstance for them! After all, FF will keep up. Eventually Pale Moon and Basilisk will probably catch up too, but by then Google will probably have changed something else and broken them yet again.

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...