LinuxQuestions.org - Requests for -current (14.2-->15.0)

- Slackware (https://www.linuxquestions.org/questions/slackware-14/)

- - Requests for -current (14.2-->15.0) (https://www.linuxquestions.org/questions/slackware-14/requests-for-current-14-2-15-0-a-4175620463/)

Quote:

Originally Posted by ljb643 (Post 6230093)

Followup re: geeqie, mplayer, ffmpeg, gimp and Samba/KRB5 problem:

The problem seems to be within Samba itself. So rebuilding those others will not help.

I re-installed samba-4.14.0-x86_64-1 and tried to start it:

Code:

# /etc/rc.d/rc.samba start

Starting Samba:  /usr/sbin/smbd -D

/usr/sbin/smbd: symbol lookup error: /usr/lib64/libkrb5samba-samba4.so: undefined symbol: krb5_set_default_tgs_ktypes, version krb5_3_MIT

                /usr/sbin/nmbd -D

/usr/sbin/nmbd: symbol lookup error: /usr/lib64/libkrb5samba-samba4.so: undefined symbol: krb5_set_default_tgs_ktypes, version krb5_3_MIT

geeqie etc. are not running because there is something wrong with the Samba libraries, at least on my system. I will look into this further.

Indeed, you can probably do a ldd /usr/lib64/libkrb5samba-samba4.so and may find it is referencing /lib64/libkrb5.so.3 instead of /usr/lib64/libkrb5.so.3

If reinstalling krb5 doesn't work, can you suggest if you have made any changes to your /etc/ld.so.conf ?

Andrew.

Quote:

Originally Posted by volkerdi (Post 6230092)

Try reinstalling the krb5 package. I suspect what may have happened is that aaa_libraries got installed after the updated krb5 package and replaced a couple of libraries with stale ones.

Yes that was it. Thanks. The krb5 package has some of the same libraries as the aaa_libraries package, but the ones in aaa_libraries won't run the new Samba-4.14. That cascades to failures of geeqie, mplayer, etc.

Quote:

Originally Posted by Ressy (Post 6229894)

so, you dont want issues reported here, you dont want people making comments on where this project is falling off, you want this to be a squeaky clean comments of praise only

any negative comments I've made are for lessons to be learned to improve slackware.

but since you dont want that, you dont want my bug reports and two of them are pretty serious, congrats you have one less person who gives a damn now.

If you look through this thread, there are plenty of people reporting bugs. Maybe you should read through how those bugs were presented compared to yours to see why you were getting the reactions you were.

You obviously don't belong as a -current user...

Code:

xap/mozilla-firefox-78.8.0esr-x86_64-4.txz: Rebuilt.

      Strip binaries and symlink duplicate binaries. Thanks to franzen.

      Add export MACH_USE_SYSTEM_PYTHON="1", upgrade build-deps/nodejs/,

      and compile with clang/clang++, fixing the use of this script to

      build more recent Firefox versions. Thanks to ponce.

Then, we can have also the latest Firefox 86.0.1 on /testing, please?

With all respect, while I truly understand that certainly shipping the Firefox for Enterprise makes sense for businesses, for home users like me, the usage of the latest Firefox permits to enjoy the state of art features and security and privacy measures published by Mozilla.

For example, the General Purpose Firefox 86.x provides separate cookies jars for every site, which is a fantastic good measure for protection of privacy for those like me who just surfs the wild Internet, instead of using portals on business Intranets.

Suggest to add wireguard tools.
https://git.zx2c4.com/wireguard-tools/about/

Quote:

Originally Posted by LuckyCyborg (Post 6230203)

Code:

xap/mozilla-firefox-78.8.0esr-x86_64-4.txz: Rebuilt.

      Strip binaries and symlink duplicate binaries. Thanks to franzen.

      Add export MACH_USE_SYSTEM_PYTHON="1", upgrade build-deps/nodejs/,

      and compile with clang/clang++, fixing the use of this script to

      build more recent Firefox versions. Thanks to ponce.

Then, we can have also the latest Firefox 86.0.1 on /testing, please?

I would have ask for extra :-)

I was to do testing.
Now instead revert, and broke all Mozilla app family, must build Firefox 86.0.1 in mainstream.
Also kde framework is release 5.80.0 , and any revert is broke this too mach.

Code:

Announcements

Saturday, 13 March 2021

KDE Ships Frameworks 5.80.0

KDE today announces the release of KDE Frameworks 5.80.0.



Thursday, 04 March 2021

20.12.3 Releases

Over 120 individual programs plus dozens of programmer libraries and feature plugins are released simultaneously as part of KDE's release service.



Tuesday, 02 March 2021

KDE Plasma 5.21.2, Bugfix Release for March

Today KDE releases a bugfix update to KDE Plasma 5.

https://kde.org

Friends, we build OS for developers and creation first, i mean Firefox developer edition is pretty idea.
Links:
Firefox ESR release cycle detail:
https://support.mozilla.org/en-US/kb...-release-cycle
Firefox Developer Edition:
https://www.mozilla.org/en-US/firefo...edium=referral

My child 10 y.old use Firefox 88.0 now.
I'm now use XDM and XFCE only. Is high stable, pretty and robust !
My web's:
http://7cyber.usluga.me
https://vk.com/7cyber
https://vk.com/dyabaru

Quote:

Originally Posted by LuckyCyborg (Post 6230203)

Code:

xap/mozilla-firefox-78.8.0esr-x86_64-4.txz: Rebuilt.

      Strip binaries and symlink duplicate binaries. Thanks to franzen.

      Add export MACH_USE_SYSTEM_PYTHON="1", upgrade build-deps/nodejs/,

      and compile with clang/clang++, fixing the use of this script to

      build more recent Firefox versions. Thanks to ponce.

I am testing from today :
https://github.com/marco70-ms/slackb...ozilla-firefox

Quote:

Originally Posted by LuckyCyborg (Post 6230203)

[code]

Then, we can have also the latest Firefox 86.0.1 on /testing, please?

probably not a good idea, I'm using the FF from mozilla.org installed as a user under /usr/local/firefox/
which works ok, but every 2 and a half days it wants to update itself, which would make it pretty hard to followif it's a system package.....
just my 0,02$
-rasp

Quote:

Originally Posted by rasp (Post 6230228)

What I asked for is just an alternate package - it is not about a shocking new addition on Slackware.

Also, IF you do not know yet, it's empirically confirmed that our BDFL knows a way to convince Firefox to NOT try to update itself while it's built as "system package" and so we already have Firefox for Enterprise 78.x packaged on Slackware -current.

Just my 0,02₽ (Russian Ruble)

Code:

Sun Mar 14 03:24:31 UTC 2021

xap/mozilla-firefox-78.8.0esr-x86_64-4.txz: Rebuilt.

      Strip binaries and symlink duplicate binaries. Thanks to franzen.

      Add export MACH_USE_SYSTEM_PYTHON="1", upgrade build-deps/nodejs/,

      and compile with clang/clang++, fixing the use of this script to

      build more recent Firefox versions. Thanks to ponce.

I just finished building mozilla-firefox-86.0.1-x86_64-4cgs.txz using the build from the source tree. Copied the source mozilla-firefox directory over to my build tree, removed the ESR tarballs, dropped in the 86.0.1 tarball and ran the SlackBuild and here I am posting from Firefox 86.0.1

I remember when the ESR version was introduced. Looking though old ChangeLog.txt files it looks like it was 13.37. We went back and forth for a while between Firefox and FirefoxESR in -current. I used to just drop the latest Firefox tarball in to the source and build it. Eventually that became harder and hard to get the compile to succeed so I just ran with the ESR version. I'm hoping now, based on this successful compile I will be able to do the same again.

Who needs extra or testing when I can just build it.

:hattip: Pat and ponce!

Quote:

Originally Posted by chrisretusn (Post 6230300)

Code:

Sun Mar 14 03:24:31 UTC 2021

xap/mozilla-firefox-78.8.0esr-x86_64-4.txz: Rebuilt.

      Strip binaries and symlink duplicate binaries. Thanks to franzen.

      Add export MACH_USE_SYSTEM_PYTHON="1", upgrade build-deps/nodejs/,

      and compile with clang/clang++, fixing the use of this script to

      build more recent Firefox versions. Thanks to ponce.

NOT everybody is capable of building it.

Right now I own over twelve older boxes capable to run Slackware -current, with CPUs ranging from Intel Core2 Duo to i3-3220 (and similar AM2 or AM3 AMD CPUs) and memory of 4GB or 8GB. And a main computer with 16GB RAM and an AMD Athlon x4 605e.

Before requesting the addition of current Firefox on /testing, I've discovered empirically that no one of those boxes other than the main box is capable to build Firefox.

Yes, I build myself the latest Firefox since long time, lurking after various notes leaved by @Ponce on this forum and others experiences.

BUT, this times I tried to understand how enabling the system NSS and NSPR interacts with an otherwise known build like was on Rust 15, and that's WHY I did multiple slightly diffferent builds on several boxes. All of them, excluding the main box, failed miserably by ending on the swap hell. And I was stubborn enough to try on all of them.

So, one needs at least 16GB RAM for building a modern Firefox.

That's a quite high stack for today, as I have many friends who uses computers with 4GB or 8GB RAM.

And a box with 8GB memory is considered a quite decent one in many places of the World.

Just because YOU (or me, or some others) have a computer capable to build a modern Firefox, this does NOT make a rule that also everybody can. From my real life experience, I would say that the majority can't do this task.

Quote:

Originally Posted by chrisretusn (Post 6230300)

:hattip: Pat and ponce!

Indeed!

@LuckyCyborg: swapinzram is your friend.

Quote:

Originally Posted by Didier Spaier (Post 6230335)

@LuckyCyborg: swapinzram is your friend.

NOT when on a box with 8GB RAM, the associated 8GB SWAP is used almost integrally. ;)

Believe or NOT, I known about swapping on RAM and I use also techniques like compressed SWAP caches.

BTW, usually I add to kernel command line as standard:

Code:

zswap.enabled=1 zswap.compressor=lzo zswap.max_pool_percent=20 zswap.zpool=z3fold

You are right 8GB of RAM is not enough.
For me, the compilation took about 30-40 minutes.
Most of the programs it compiles go through the ram.
Too bad the disk.

My fstab mod:

Code:

tmpfs /tmp tmpfs defaults,noatime,mode=1777 0 0 ## SSD

sysctl.conf:

Code:

vm.swappiness=1