Jump to content

My Browser Builds (Part 1)


Recommended Posts

7 hours ago, roytam1 said:

it puts [object FormData] as payload POSTing to https://github.com/preview thats not a thing I can fix.

Many thanks for investigating this... Since, sadly, I'm not a coder, I found it hard to understand what you were saying there (:P), but after some further examination of my own I managed, at least, to identify what you were referring to ;) ...

The following is from New Moon 27 where, as reported, github comment preview does not work:

q8bliWb.jpg

... and the next one is from FirefoxESR 52, where github comment preview does work:

4zGM1ZO.jpg

It is unfortunate that this bug cannot be fixed in NM27/Tycho, but all is not lost; for the most part, Github is still usable there for an average user; for full compatibility with current Github, we have New Moon 28/UXP; plus, the other UXP browser, Serpent 52.9.0, can be made fully compatible with Github via a simple SSUAO, e.g.

general.useragent.override.github.com;Mozilla/5.0 (Windows NT 6.1; rv:52.0) Goanna/4.1 Basilisk/52.9.0

(no reference whatsoever to Gecko+Firefox seems to be enough...)

Likewise, in Basilisk (Serpent) 55/moebius I used:

general.useragent.override.github.com;Mozilla/5.0 (Windows NT 6.1; rv:53.0) Goanna/4.0 Basilisk/55.0.0

 

In FirefoxESR 52.9.0, which does not support SSUAOs, I used the HTTP Header Mangler v1.1.3 extension (WebExtension) with the rule:

# Restore GitHub JS functions in 52.0 =< Fx < 59.0
github.com
User-Agent= Mozilla/5.0 (Windows NT 6.1; rv:60.0) Gecko/20100101 Firefox/60.0

Many other solutions involving XUL and/or WE addons exist for FxESR 52; I just found that one to be quite elegant, plus it doesn't overcrowd the browser toolbar with yet another button... BTW, 59.0 is the lowest Firefox version (in the UA string) that does not cause Github to complain about an unsupported browser...

Two other browsers that are forked from FxESR 52, SeaMonkey 2.49.4 and Cyberfox 52.9.1, both support SSUAOs, so you can use the above Fx 60.0 UA string to restore GitHub on them... 

In closing, and while I'm not a Google Chrome person myself, I tried the last version of that browser that would run on Vista SP2, which is v50.0.2661.102; of course, GitHub would immediately flag Chrome 50 as an unsupported browser, but:

1. I installed User-Agent Switcher for Chrome extension; this one is capable of per-domain-UAOs

2. I configured the extension to spoof the Vivaldi (or Opera) browser on github.com; I first created a custom UA for Vivaldi (in the Opera group):

gN6evq9.jpg

and then used that in the extension's Permament Spoof list:

fE9gMHM.jpg

(Vivaldi 1.2.470.11[beta] is the maximum Vivaldi version - based on Chromium 50 - that would run on Vista)

I had first toyed with the idea of spoofing recent (> 65) Chrome versions to github, but that was not working; what really worked was spoofing either Vivaldi or Opera (and, not forgetting the original issue I reported here, "comment preview" does work in Chrome 50 !) ...

Though I've not verified this myself, I estimate a similar procedure can be used to restore GitHub in Google Chrome 49.0.2623.112, the last WinXP compatible version...

... I do hope all the above is found to be useful, at least by some of you here... ;)

Link to comment
Share on other sites


Some people decided to post links of my build into PM forum ( https://forum.palemoon.org/viewtopic.php?f=4&amp;t=20887 ), don't know if he want to introduce or alerting devs to take action. And Tobin is still same attitude.

So people here please don't mention my builds in their forum, that will be good for both sides.

(What Tobin said is totally "who fu*king cares" to me, so please don't argue on it with their devs.)

Link to comment
Share on other sites

1 hour ago, roytam1 said:

Some people decided to post links of my build into PM forum ( https://forum.palemoon.org/viewtopic.php?f=4&amp;t=20887 ), don't know if he want to introduce or alerting devs to take action

As a matter of fact, I do know @JoeyG :); he's a retired English teacher from the States, who currently lives in Germany with his wife and two beloved cats... I first met him in the mozillazine forums back in ca. 2012, when we were both Firefox Nightly testers ;) He was an avid Windows XP + Firefox enthusiast; after XP's EOL, he had to update many of his machines to Win7 and, if it hadn't been for aris's Classic Theme Restorer extension, he would have jumped off the Firefox bandwagon as soon as Australis was enforced upon us... :realmad: But the move to Firefox Quantum was a hard/bitter pill to swallow, so he ended up in Pale Moon Land...

Reading his posts over many years, I feel there was no ill intent on his part in posting @roytam1's blog link in the Moonchild forums; his was a genuine query for additional info regarding the "forks", posted in a subforum (General discussion) where such queries would be totally fine... But then again, "Matt A Tobin" is lurking there, hence the repercussions...

1 hour ago, roytam1 said:

And Tobin is still same attitude... (redacted) (What Tobin said is totally "who fu*king cares" to me

The mere mention of "XP/Vista/fork" is enough to ignite Matt and make him explode all over the place :realmad:; we all know that by now; I have taught myself to be immune of his insults and stay calm about it; in fact, "who fu*king cares" what he thinks... Roy's hard efforts are valued immensely by all those who get to use his builds! :thumbup

As for @JoeyG, perhaps @roytam1 can invite him via PM into this wonderful community, where members here are more tolerant of people's freedom of choice to use/keep using the OS they most feel comfortable with - or, in some cases, keep using the OS they are stuck with (i.e. it is impossible/impractical to move to a newer OS)!

Edited by VistaLover
excess whitespace
Link to comment
Share on other sites

I can't resist myself anymore .

Mr. Ashole Tobin said

"They are unbranded builds so there isn't anything that can be done to stop them.. But they should make it more clear that they are unofficial builds. I suggest users do NOT use builds from a modified codebase maintained by lazy "developers" who cut corners and do not properly account for issues targeting NT 5x. Additionally, any user of these builds is not entitled to support from us."

Dear sir Tobin the moron, what is the benefit of replacing GetTickCount with GetTickCount64?

What wrong in not supporting h264? What heck with putting some random import function which has nothing to do ?

If Roy and Fedor is lazy who hell are you people keeping million of bug and outdated code base . Our code base is more upto date , more feature rich , more optimised than your Nt6.x function crap .

As for your homework , Mr as***** Tobin sir , please kindly benchmark a same application with nt6.x API and sameone with nt5.x . One with nt5.x will perform way better.

These guys are just  some childish folks who think something not compatible with nt5.x is superior . Biggest example is palemoon and notepad3. This proves how well Microsoft brain wash silly folks to allow them execute their legendary policy of "embrace extend extinguish"

I apologize in advance for any typo . I am writing from a android phone.

 

 

Link to comment
Share on other sites

1 hour ago, Dibya said:

I can't resist myself anymore

keep calm and ignore Tobin :)

1 hour ago, Dibya said:

what is the benefit of replacing GetTickCount with GetTickCount64

to prevent overflow of ticks after 49 days? but since browser will not be that stable after several hours of heavy use so it is non-issue here :)

Link to comment
Share on other sites

28 minutes ago, roytam1 said:

keep calm and ignore Tobin :)

to prevent overflow of ticks after 49 days? but since browser will not be that stable after several hours of heavy use so it is non-issue here :)

I will stay calm. Who keeps browser open for 49days?

Link to comment
Share on other sites

1 hour ago, Dibya said:

I will stay calm. Who keeps browser open for 49days?

I do. %) But those were the good times, long forgotten... When browsers were stable and webpages were less resource-hungry.

Btw, wasn't that tobin guy banned previously? Also, what would happen when that counter gets overflown?

Link to comment
Share on other sites

On 11/15/2018 at 6:02 PM, roytam1 said:

it puts [object FormData] as payload POSTing to https://github.com/preview
thats not a thing I can fix.

Apologies for coming back to this, but I'm not the type of person that easily gives up ;):P

I conducted further tests on FirefoxESR 38.8.0, MozillaESR 38.0 being the platform that Tycho is forked from, and, as expected, I can reproduce there the same bug I reported in New Moon 27... Searching to find the first Firefox version in which this bug was fixed (cause we already know it works in FxESR 52, hence UXP platform), I was lucky to discover that the issue is fixed in Firefox 39.0:

W5cuCpv.jpg

Given that the ESR branch rarely contains any major feature enhancements (except for security+stability fixes applied to the Fx major version it is forked from), it is safe to assume that the code that fixes this is found in the changelog Fx-38.0...Fx-39.0

Searching on the net for "Fetch API + Request Payload + FormData", I found the following references:

https://developer.mozilla.org/en-US/docs/Web/API/FormData/Using_FormData_Objects

https://hacks.mozilla.org/2015/03/this-api-is-so-fetching/

Quote

As of this writing, the Fetch API is available in Firefox 39 (currently Nightly) and Chrome 42 (currently dev)

and, most interestingly,

https://hacks.mozilla.org/2015/03/using-the-firefox-devtools-to-debug-fetch-on-github/

which mentions Bugzilla #1143857 ; As said, I'm not a coder, but, perhaps, is this what you need to apply to Tycho to fix the GitHub Comment Preview bug in New Moon 27 (provided fetch API is enabled) ?

Thanks for your excellent user support, BTW :thumbup

 

Link to comment
Share on other sites

What's annoying to me is the fact that those who rail endlessly against XP almost never mention that it is  being updated and it's not a secret since people have been taking advantage of the PosReady 2009 hack for years. In the last four months I've downloaded and installed at least 22 security patches alone for this older system. Not only that but you can still get the latest virus programs, flash player, and thanks to this project an updated browser.

It is an old OS and probably has holes nonetheless but it's not like it isn't getting support. I'm usually on something newer but millions use XP regularly for whatever reason and it works fine, always has.

It's almost funny how some of these anti-XP internet freaks go on a rampage, as if they have something to lose. No matter what forum their heads act like they're about to explode at the mere mention of XP. The end of support for PosReady is next April so it will require extra locking down but I wonder how many malware/virus trolls even bother with it anymore, they probably moved on to Windows 10.   

Link to comment
Share on other sites

4 hours ago, VistaLover said:

Apologies for coming back to this, but I'm not the type of person that easily gives up ;):P

I conducted further tests on FirefoxESR 38.8.0, MozillaESR 38.0 being the platform that Tycho is forked from, and, as expected, I can reproduce there the same bug I reported in New Moon 27... Searching to find the first Firefox version in which this bug was fixed (cause we already know it works in FxESR 52, hence UXP platform), I was lucky to discover that the issue is fixed in Firefox 39.0:

Given that the ESR branch rarely contains any major feature enhancements (except for security+stability fixes applied to the Fx major version it is forked from), it is safe to assume that the code that fixes this is found in the changelog Fx-38.0...Fx-39.0

Searching on the net for "Fetch API + Request Payload + FormData", I found the following references:

https://developer.mozilla.org/en-US/docs/Web/API/FormData/Using_FormData_Objects

https://hacks.mozilla.org/2015/03/this-api-is-so-fetching/

and, most interestingly,

https://hacks.mozilla.org/2015/03/using-the-firefox-devtools-to-debug-fetch-on-github/

which mentions Bugzilla #1143857 ; As said, I'm not a coder, but, perhaps, is this what you need to apply to Tycho to fix the GitHub Comment Preview bug in New Moon 27 (provided fetch API is enabled) ?

Thanks for your excellent user support, BTW :thumbup

 

as I stated before, some fixes of fetch API need to be backported to PM27. I'll have a closer look to the changes later to see if they can be ported.

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-20181117-c94825c86-xpmod.7z
Win64 https://o.rths.cf/basilisk/basilisk52-g4.1.win64-git-20181117-c94825c86-xpmod.7z

diff: https://o.rths.cf/basilisk/UXP-xp-gitdiff-20181110.7z

PM28XP build:
Win32 https://o.rths.cf/palemoon/palemoon-28.2.0a1.win32-git-20181117-c94825c86-xpmod.7z
Win64 https://o.rths.cf/palemoon/palemoon-28.2.0a1.win64-git-20181117-c94825c86-xpmod.7z

Official repo changes since my last build:
- Don't build internal updater by default and set default update channel to an inactive one. (9d2d304a6)
- [PALEMOON] Use generic application icon for external applications in about:feeds (c8d90e332)
- Revert "[PALEMOON] Use generic application icon for external applications in about:feeds" (e3f4fe530)
- [PALEMOON] Use generic application icon for external applications in about:feeds (81c27113a)
- Revert "[PALEMOON] Focus and select the first non-collapsed text element in the bookmark properties dialog" (0cded9cf5)
- Merge branch 'master' of https://github.com/MoonchildProductions/UXP (3438b20bd)
- Issue #861 - The bookmark properties window should focus on the Name Picker text box on init (b199e3aa0)
- [Pale Moon] Sync notification widget's messageImage should inherit the 'type' attribute (46a5c51bf)
- [PALEMOON] Re-apply removal of unnecessary code (Findbar is no longer in the gBrowser) (16e5a6bc2)
- Merge pull request #869 from FranklinDM/sync_notification-work (3c342ab20)
- Split the content viewer marking function in CCUncollectable out, so a document-based marking function can be used separately. (2db00388a)
- Merge branch 'master' of https://github.com/MoonchildProductions/UXP (d8e3bde97)
- Fix build bustage. (a748fde4c)
- Tell the cycle collector about nsScriptLoadRequest::mElement. (c94825c86)

My changes since my last build:
- update TwemojiMozilla.ttf to Twemoji-v11.2.0 base

Link to comment
Share on other sites

New Palemoon 27 Build!
* This build is beyond official 27.9.4 build.

32bit https://o.rths.cf/palemoon/palemoon-27.9.1a1.win32-git-20181117-a6bf60af2-xpmod.7z
32bit SSE https://o.rths.cf/palemoon/palemoon-27.9.1a1.win32-git-20181117-a6bf60af2-xpmod-sse.7z
32bit noSSE https://o.rths.cf/palemoon/palemoon-27.9.1a1.win32-git-20181117-a6bf60af2-xpmod-ia32.7z

64bit https://o.rths.cf/palemoon/palemoon-27.9.1a1.win64-git-20181117-a6bf60af2-xpmod.7z

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

repo changes since my last build:
- ported mozilla upstream changes:
 - Nikhil Marathe - Bug 1139665 - Check content load policy in FetchDriver. r=baku (25ded6c22f64)
 - Nikhil Marathe - Bug 1143857 - Add FormData serialize support to Fetch API. r=ehsan (1ff28a6f0b77)
 and enable fetch API in pref. (a6bf60af2)

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...