View previous topic :: View next topic |
Author |
Message |
AJM Apprentice
Joined: 25 Sep 2002 Posts: 189 Location: Aberdeen, Scotland
|
Posted: Tue Apr 18, 2017 6:32 pm Post subject: |
|
|
turtles wrote: | I too have run an ALSA only system for years.
It seems like the devs at Firefox are open to distros supporting ALSA:
[url]https://groups.google.com/forum/#!topic/mozilla.dev.platform/jRAqSTri66I[/url] |
A much better fix is just to purge Firefox from your system and emerge Pale Moon (there's an overlay.) Mozilla have made one arrogant, user-hostile change too many this time as far as I'm concerned.
The good news is that Pale Moon works perfectly well and you don't have to check each changelog to see what pointless non-feature you're going to have to disable this time! |
|
Back to top |
|
|
khayyam Watchman
Joined: 07 Jun 2012 Posts: 6227 Location: Room 101
|
Posted: Tue Apr 18, 2017 7:54 pm Post subject: |
|
|
AJM wrote: | A much better fix is just to purge Firefox from your system and emerge Pale Moon (there's an overlay.) |
AJM ... all of www-client/palemoon in the palemoon overlay are masked, and have been for some time.
AJM wrote: | The good news is that Pale Moon works perfectly well and you don't have to check each changelog to see what pointless non-feature you're going to have to disable this time! |
Of the ten builds I've attemped only two completed successfully, and then one of those was completely unusable and package.masked in the overlay the next day. Right now I'm running a version with known vunerabilities because there is no where to go, building the masked package(s) fails. So, that is far from "perfect", I expect to either go back to firecrap, or find some other browser.
best ... khay |
|
Back to top |
|
|
AJM Apprentice
Joined: 25 Sep 2002 Posts: 189 Location: Aberdeen, Scotland
|
Posted: Tue Apr 18, 2017 9:18 pm Post subject: |
|
|
khayyam wrote: |
Of the ten builds I've attemped only two completed successfully, and then one of those was completely unusable and package.masked in the overlay the next day. Right now I'm running a version with known vunerabilities because there is no where to go, building the masked package(s) fails. So, that is far from "perfect", I expect to either go back to firecrap, or find some other browser.
best ... khay |
Hmm... I did have problems building the latest release (required an older binutils which I'd removed since building the previous PM release, though I believe there's been a patch for this issue since.) Pale Moon has been 100% stable for me since I switched though, about a month ago.
What was your build error? |
|
Back to top |
|
|
khayyam Watchman
Joined: 07 Jun 2012 Posts: 6227 Location: Room 101
|
Posted: Tue Apr 18, 2017 9:35 pm Post subject: |
|
|
AJM wrote: | Hmm... I did have problems building the latest release (required an older binutils which I'd removed since building the previous PM release, though I believe there's been a patch for this issue since.) Pale Moon has been 100% stable for me since I switched though, about a month ago. |
AJM ... what do you mean by latest release? You mean 27.2.1-r1 ... which is package.mask'ed?
AJM wrote: | What was your build error? |
The build.log is 11M, so I won't pastebin. It dies linking ("dynamic relocations"), which sounds like it could be related to your "require[s] an older binutils".
build.log: | 8147:44.11 INPUT("StaticXULComponentsEnd/StaticXULComponentsEnd.o")
8147:44.11
8147:44.11 ../../build/unix/gold/ld: error: read-only segment has dynamic relocations
8147:44.11 /home/portage/www-client/palemoon-27.1.2/work/palemoon-27.1.2/media/libstagefright/binding/MoofParser.cpp:687: error: undefined reference to 'GetDemuxerLog()'
8147:44.11 /home/portage/www-client/palemoon-27.1.2/work/palemoon-27.1.2/media/libstagefright/binding/MoofParser.cpp:694: error: undefined reference to 'GetDemuxerLog()'
8147:44.11 /home/portage/www-client/palemoon-27.1.2/work/palemoon-27.1.2/media/libstagefright/binding/MoofParser.cpp:556: error: undefined reference to 'GetDemuxerLog()'
8147:44.11 /home/portage/www-client/palemoon-27.1.2/work/palemoon-27.1.2/media/libstagefright/binding/MoofParser.cpp:564: error: undefined reference to 'GetDemuxerLog()'
8147:44.11 collect2: error: ld returned 1 exit status |
That is a after 5 and a half hours of being the only single running process with the machine completely maxed out (and I mean completely maxed).
best ... khay |
|
Back to top |
|
|
AJM Apprentice
Joined: 25 Sep 2002 Posts: 189 Location: Aberdeen, Scotland
|
Posted: Tue Apr 18, 2017 10:06 pm Post subject: |
|
|
khayyam wrote: | AJM ... what do you mean by latest release? You mean 27.2.1-r1 ... which is package.mask'ed? |
I'm currently using pale moon 27.2.1... yes it's masked, but works perfectly for me.
Quote: |
build.log: | 8147:44.11 INPUT("StaticXULComponentsEnd/StaticXULComponentsEnd.o")
8147:44.11
8147:44.11 ../../build/unix/gold/ld: error: read-only segment has dynamic relocations
|
That is a after 5 and a half hours of being the only single running process with the machine completely maxed out (and I mean completely maxed).
best ... khay |
Yep, that looks like the same error I had. Building with sys-devel/binutils-2.25.1-r1 should work around it, although I have a feeling I saw patches to fix it too - maybe on the pale moon forum?
I'm spoiled with my "new" PC (actually built mostly free second hand bits) , Core i5 2500K and SSD... build times are fairly insignificant now - though having used Gentoo a fair bit on a Raspberry Pi recently I feel your pain! |
|
Back to top |
|
|
turtles Veteran
Joined: 31 Dec 2004 Posts: 1657
|
Posted: Sat Apr 22, 2017 4:02 am Post subject: |
|
|
AJM wrote: | turtles wrote: | I too have run an ALSA only system for years.
It seems like the devs at Firefox are open to distros supporting ALSA:
[url]https://groups.google.com/forum/#!topic/mozilla.dev.platform/jRAqSTri66I[/url] |
A much better fix is just to purge Firefox from your system and emerge Pale Moon (there's an overlay.) Mozilla have made one arrogant, user-hostile change too many this time as far as I'm concerned.
The good news is that Pale Moon works perfectly well and you don't have to check each changelog to see what pointless non-feature you're going to have to disable this time! |
Yeah FF is going the "lets write yet another programming language" route, scary ...
I'll try it out however FF is no systemd and I have to run all web browsers possible for work related tasks so i will always need ff.
Also I noticed netflicks 'just works' TM now in firefox so I can ditch wine and silver light.
It also is not really hard to update the old ALSA code in they refuse to do and the ALSA project is interested in helping so it will probably get done. _________________ Donate to Gentoo |
|
Back to top |
|
|
Anon-E-moose Watchman
Joined: 23 May 2008 Posts: 6098 Location: Dallas area
|
Posted: Sat Apr 22, 2017 10:50 am Post subject: |
|
|
Another thing coming along for FF is the switch to gtk3.
Maybe someone will fork the 45 or so series and keep alsa, gtk2 and just address security concerns, instead of trying to "enhance" our browsing experience with stuff that no one, outside of ff devs wants. _________________ PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland |
|
Back to top |
|
|
tld Veteran
Joined: 09 Dec 2003 Posts: 1816
|
Posted: Sat Apr 22, 2017 2:37 pm Post subject: |
|
|
Anon-E-moose wrote: | Another thing coming along for FF is the switch to gtk3.
Maybe someone will fork the 45 or so series and keep alsa, gtk2 and just address security concerns, instead of trying to "enhance" our browsing experience with stuff that no one, outside of ff devs wants. | The gtk3 part bothers me even more than the ALSA support. Unreal. It's hard to imagine that the original goal of FF was supposedly to be light weight. To call the FF devs self important ass-hats would be an insult to self important ass-hats frankly. They just don't seem to get that it really is just a browser. Just sad.
Maybe I'm wrong, but I don't get a good impression of palemoon (even reading here frankly), or I'd go that route.
Tom |
|
Back to top |
|
|
asturm Developer
Joined: 05 Apr 2007 Posts: 8936
|
Posted: Sat Apr 22, 2017 2:45 pm Post subject: |
|
|
In what way is gtk+-3 heavier than gtk+-2? Bear in mind that the latter is being phased out everywhere. |
|
Back to top |
|
|
Ant P. Watchman
Joined: 18 Apr 2009 Posts: 6920
|
Posted: Sat Apr 22, 2017 11:34 pm Post subject: |
|
|
asturm wrote: | In what way is gtk+-3 heavier than gtk+-2? Bear in mind that the latter is being phased out everywhere. |
The maintenance of having to update/hack themes every time they break in a point release? If Gtk3 is unstable enough to make KDE developers ragequit over oxygen-gtk3, what chance do users have? |
|
Back to top |
|
|
fcl Tux's lil' helper
Joined: 31 Dec 2016 Posts: 77
|
Posted: Sun Apr 23, 2017 5:18 pm Post subject: |
|
|
I'm no fan of Gnome but to be fair gtk3 theming api has been stable for quite a long time. It took way too long for them to stabilize though. I hope they learned from this and do things better with gtk4 *shudders* |
|
Back to top |
|
|
Anon-E-moose Watchman
Joined: 23 May 2008 Posts: 6098 Location: Dallas area
|
Posted: Sun Apr 23, 2017 10:50 pm Post subject: |
|
|
The other question to ask is how long will gtk3 have their focus since they've already got their eyes on gtk4? _________________ PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland |
|
Back to top |
|
|
asturm Developer
Joined: 05 Apr 2007 Posts: 8936
|
Posted: Sun Apr 23, 2017 11:02 pm Post subject: |
|
|
fcl wrote: | I'm no fan of Gnome but to be fair gtk3 theming api has been stable for quite a long time. |
3.20 ain't that long ago, that broke basically everything for breeze-gtk.
However:
Ant P. wrote: | asturm wrote: | In what way is gtk+-3 heavier than gtk+-2? Bear in mind that the latter is being phased out everywhere. |
The maintenance of having to update/hack themes every time they break in a point release? If Gtk3 is unstable enough to make KDE developers ragequit over oxygen-gtk3, what chance do users have? |
Apparently they consider gtk+-3 'finished' now, so this shouldn't happen anymore. But somehow I don't think tld had the theme breakage in mind when calling it 'heavy'. |
|
Back to top |
|
|
mv Watchman
Joined: 20 Apr 2005 Posts: 6747
|
Posted: Mon Apr 24, 2017 7:07 am Post subject: |
|
|
asturm wrote: | But somehow I don't think tld had the theme breakage in mind when calling it 'heavy'. |
Perhaps tlld had the atk-bridge stuff in mind which forces in several deps, including dbus.
Fortunately, BSD maintain a patch which makes this part optional. A corresponding ebuild (which also uses optionally minimal debugging information) is in the mv overlay. |
|
Back to top |
|
|
mv Watchman
Joined: 20 Apr 2005 Posts: 6747
|
Posted: Mon Apr 24, 2017 7:42 am Post subject: |
|
|
tld wrote: | I don't get a good impression of palemoon (even reading here frankly) |
Can you be more precise?
Maybe you have to distinguish between palemoon - the project - and palemoon - the ebuild: I think all criticism mentioned here referred to the ebuild.
In fact, the ebuild from the palemoon overlay is unusually restrictive: Allowing only compiler and library versions which work on the maintainer's machine. In some sense this is understandable, but for some users like me it is simply too restrictive.
Note that in the octopus overlay also a palemoon overlay is maintained which is much more liberal and allows a fine-tuning of system libs. After my last conflicts due to the locked hunspell version, I swtiched to that ebuild. The generated palemoon is rock stable. |
|
Back to top |
|
|
tld Veteran
Joined: 09 Dec 2003 Posts: 1816
|
Posted: Mon Apr 24, 2017 5:19 pm Post subject: |
|
|
mv: Yea, I think most of what I read about palemoon that seemed questionable was related to the ebuild. I'm on an old x86 machine, and I am in fact able to compile FF, though it takes time. I seem to read a lot about the palemoon compile really tanking the system...not sure on all that. I've also read some criticisms (on soylentnews.org I believe) where some questioned the palemoon project altogether, as being too dependent on upstream for patches, and otherwise being questionable. I have no clue if there was any validity to any of that.
On the gtk3 thing, you're correct: That wants to pull in crap I don't use including dbus.
I wouldn't put it past the folks involved to, at some point, try and make some gtk version dependent on systemd...though I don't think they'd get away with that. Then again, I wouldn't have thought they'd get away with the Gnome dependency, and they have.
Tom |
|
Back to top |
|
|
Tony0945 Watchman
Joined: 25 Jul 2006 Posts: 5127 Location: Illinois, USA
|
Posted: Mon Apr 24, 2017 8:03 pm Post subject: |
|
|
Here is my own ebuild for palemoon-bin which works fine. Yes, unfortunately, upstream is hostile to Gentoo. Code: | # Copyright 2016 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
EAPI=5
inherit eutils
DESCRIPTION="ebuild for palemoon browser binary install"
HOMEPAGE="http://www.palemoon.org"
SRC_URI="http://linux.palemoon.org/files/${PV}/palemoon-${PV}.en-US.linux-x86_64.tar.bz2"
LICENSE="Palemoon-BRL"
SLOT="0"
KEYWORDS="~amd64"
RESTRICT="mirror"
IUSE="X "
DEPEND=""
#all runtime dependencies are contained in downloaded tarball
RDEPEND="${DEPEND}"
S="${WORKDIR}"
src_install() {
mkdir "${S}"/palemoon-bin
cp -R "${S}"/palemoon/* "${S}"/palemoon-bin
# files will be installed to /opt rather than /usr/bin
insinto /opt
# the untarred files are in work/palemoon
# this command copies them from there to /opt/palemoon-bin (not /opt as you might think)
doins -r "${S}"/palemoon-bin
dodir /usr/bin
dosym /opt/palemoon-bin/palemoon-bin /usr/bin/palemoon-bin
# following cribbed from firefox-bin ebuild
local size sizes icon_path icon name
sizes="16 32 48"
icon_path="${S}/palemoon-bin/browser/chrome/icons/default"
icon="${PN}"
name="Palemoon"
# Install icons and .desktop for menu entry
for size in ${sizes}; do
insinto "/usr/share/icons/hicolor/${size}x${size}/apps"
newins "${icon_path}/default${size}.png" "${icon}.png" || die
done
# The 128x128 icon has a different name
insinto "/usr/share/icons/hicolor/128x128/apps"
newins "${S}/palemoon-bin/browser/icons/mozicon128.png" "${icon}.png" || die
# Install a 48x48 icon into /usr/share/pixmaps for legacy DEs
newicon "${S}"/palemoon-bin/browser/chrome/icons/default/default48.png ${PN}.png
domenu "${FILESDIR}"/icon/${PN}.desktop
}
pkg_postinst()
{
chmod +x /opt/palemoon-bin/palemoon-bin
} | The license is NOT approved by upstream. They insist that the license text which I copied from the webpage use the exact fonts they use, which is impossible in a text file.
I've used this for at least a year. When a new version comes out, I just copy the ebuild to the new number and run "repoman manifest".
I was using firefox-bin anyway. Too much of a PITA to compile for no observable benefit. |
|
Back to top |
|
|
khayyam Watchman
Joined: 07 Jun 2012 Posts: 6227 Location: Room 101
|
Posted: Mon Apr 24, 2017 10:31 pm Post subject: |
|
|
mv wrote: | Note that in the octopus overlay also a palemoon overlay is maintained which is much more liberal and allows a fine-tuning of system libs. After my last conflicts due to the locked hunspell version, I swtiched to that ebuild. The generated palemoon is rock stable. |
mv ... thanks for that information, having added the octopus overlay my next though was, will this link with =sys-devel/binutils-2.26.1, or am I similarly required to downgrade to 2.25.1-r1 as per AJM's report?
best ... khay |
|
Back to top |
|
|
mv Watchman
Joined: 20 Apr 2005 Posts: 6747
|
Posted: Tue Apr 25, 2017 7:52 am Post subject: |
|
|
khayyam wrote: | will this link with =sys-devel/binutils-2.26.1, or am I similarly required to downgrade to 2.25.1-r1 as per AJM's report? |
I didn't understand which problems you had. I am using gcc:6 and binutils-2.27 since ages - I don't have older versions on my system - and had always compiled palemoon only with this: For the ebuild from the palemoon overlay, I had applied some patches to avoid the forced test for the gcc version, and I had also patched out system-cairo, because this is the only system library which caused frequent crashes for me. Finally, I had also to patch out atk-bridge stuff when trying gtk:3, because my gtk3 does not provide atk-bride.
For the ebuild from octopus, I did not apply any patches (except for atk-bridge), used all system libs with the exception of system-cairo and system-sqlite (the latter had some underlinking issue with ICU so that palemoon did not install with it).
Both versions run rather stable, though that from octopus appears to be even more stable (I don't know why: It's the same sources, but perhaps some more system libs are used, e.g. for encryption):
The version from the palemoon overlay crashed reproduciible when llogging into google mail, the version from octopus does not (or only very rarely) in this case.
In some sense this crash is "to be expected": Some old firefox versions had suffered from exactly the same problem.
But apart from this (which probably is related to google talk plugin and video conversion), my experience with both palemoons is very positive (and also some bugs I reported had been sanely solved).
The gtk:3 variant loks ugly (e.g. scrollbars do not work), and unless you install the adawait theme, most icons are jjust dummys.
However, gtk:3 support is not yet claimed to be ready by upstream. gtk:2 works nice. |
|
Back to top |
|
|
tld Veteran
Joined: 09 Dec 2003 Posts: 1816
|
Posted: Tue Apr 25, 2017 5:54 pm Post subject: |
|
|
Tony0945 wrote: | The license is NOT approved by upstream. They insist that the license text which I copied from the webpage use the exact fonts they use, which is impossible in a text file. | Hahah...OMG. That may actually be the most dickish thing I've ever heard in my life. Amazing. |
|
Back to top |
|
|
Ant P. Watchman
Joined: 18 Apr 2009 Posts: 6920
|
Posted: Tue Apr 25, 2017 7:58 pm Post subject: |
|
|
tld wrote: | Tony0945 wrote: | The license is NOT approved by upstream. They insist that the license text which I copied from the webpage use the exact fonts they use, which is impossible in a text file. | Hahah...OMG. That may actually be the most dickish thing I've ever heard in my life. Amazing. |
I think the (now thankfully dead) ion3 WM holds that crown. BS license, and the author actively went around harassing distros that tried to package it. |
|
Back to top |
|
|
Zucca Moderator
Joined: 14 Jun 2007 Posts: 3345 Location: Rasi, Finland
|
Posted: Tue Apr 25, 2017 10:12 pm Post subject: |
|
|
tld wrote: | Tony0945 wrote: | The license is NOT approved by upstream. They insist that the license text which I copied from the webpage use the exact fonts they use, which is impossible in a text file. | Hahah...OMG. That may actually be the most dickish thing I've ever heard in my life. Amazing. | ... I wonder it that's even legal in some places... |
|
Back to top |
|
|
Zucca Moderator
Joined: 14 Jun 2007 Posts: 3345 Location: Rasi, Finland
|
Posted: Tue Apr 25, 2017 10:19 pm Post subject: |
|
|
Ant P. wrote: | I think the (now thankfully dead) ion3 WM holds that crown. BS license, and the author actively went around harassing distros that tried to package it. | The developer is a shame spot. I cannot believe he's a Finn.
I apologize for his behaviour. ;P |
|
Back to top |
|
|
AJM Apprentice
Joined: 25 Sep 2002 Posts: 189 Location: Aberdeen, Scotland
|
Posted: Tue Apr 25, 2017 10:39 pm Post subject: |
|
|
mv wrote: |
I didn't understand which problems you had. I am using gcc:6 and binutils-2.27 since ages - I don't have older versions on my system |
This is why you didn't see the problem, IIRC it only occurs with older versions of GCC coupled with newer versions of binutils. I'm on Code: | (Gentoo 4.9.4 p1.0, pie-0.6.4) |
|
|
Back to top |
|
|
khayyam Watchman
Joined: 07 Jun 2012 Posts: 6227 Location: Room 101
|
Posted: Tue Apr 25, 2017 11:07 pm Post subject: |
|
|
Zucca wrote: | tld wrote: | Tony0945 wrote: | The license is NOT approved by upstream. They insist that the license text which I copied from the webpage use the exact fonts they use, which is impossible in a text file. |
Hahah...OMG. That may actually be the most dickish thing I've ever heard in my life. Amazing. |
... I wonder it that's even legal in some places... |
Zucca ... it's not a recognised legal convention anywhere, a contract isn't the font, or paper, it's the content. A contract could state something like "void if not signed with the blood of a virgin", but any court would exclude this as a valid criteria for the contract being applicable. So too with a licence, the licence is not the font, and if this is a stipulation, it would no doubt be excluded as applicable.
best ... khay |
|
Back to top |
|
|
|