Let's see if I understand this correctly, Red Hat has now put a whole lot of open source / GPL software behind a paywall, where you have to pay $$ for a subscription in order to access the source code. Then if anyone uses that subscription to produce an Red Hat style distribution (ie: Red Hat minus any support services) their subscription will be cancelled... Is the above what I am reading here? : https://www.jeffgeerling.com/blog/2023/im-done-red-hat-enterprise-linux If this is true then it confirms my decision to use Debian GNU/Linux as for all its' faults (and it does have several) Debian GNU/Linux just doesn't pull this sort of @#$% .
On Mon, Jun 26, 2023 at 07:29:23PM -0400, Colin McGregor via talk wrote:
Let's see if I understand this correctly, Red Hat has now put a whole lot of open source / GPL software behind a paywall, where you have to pay $$ for a subscription in order to access the source code. Then if anyone uses that subscription to produce an Red Hat style distribution (ie: Red Hat minus any support services) their subscription will be cancelled... Is the above what I am reading here? :
https://www.jeffgeerling.com/blog/2023/im-done-red-hat-enterprise-linux
If this is true then it confirms my decision to use Debian GNU/Linux as for all its' faults (and it does have several) Debian GNU/Linux just doesn't pull this sort of @#$% .
Red Hat charges for access to their software and support. If you get their software that way, they of course are required to let you have the source code. But then they do in fact tell you that if you share that source code (which you are of course permitted to do by the license), they are allowed to cancel your contract and decide to never do business with you again. Well I think they say they can cancel your contract for any reason they want, but that sharing the source code would very likely make them do so. So similar to what the grsecurity people were trying to do, but at IBM scale instead. Red Hat is far from the company it was 30 years ago. Yeah I am happy I switched to debian 25 years ago because Red Hat's quality was so poor at the time. Debian having a better designed packaging system was a bonus. -- Len Sorensen
On 2023-06-27 08:19, Lennart Sorensen via talk wrote:
Let's see if I understand this correctly, Red Hat has now put a whole lot of open source / GPL software behind a paywall, where you have to pay $$ for a subscription in order to access the source code. Then if anyone uses that subscription to produce an Red Hat style distribution (ie: Red Hat minus any support services) their subscription will be cancelled... Is the above what I am reading here? :
https://www.jeffgeerling.com/blog/2023/im-done-red-hat-enterprise-linux
If this is true then it confirms my decision to use Debian GNU/Linux as for all its' faults (and it does have several) Debian GNU/Linux just doesn't pull this sort of @#$% . Red Hat charges for access to their software and support. If you get
On Mon, Jun 26, 2023 at 07:29:23PM -0400, Colin McGregor via talk wrote: their software that way, they of course are required to let you have the source code.
But then they do in fact tell you that if you share that source code I have looked a bit at various statements but nowhere near enough to comment authoritatively. Where do they actually lay out the restrictions on the use of the source code for their subscription?
Also curious. What is Canonical's(Ubuntu) take on cloning Ubuntu and selling it as a competing distribution?
(which you are of course permitted to do by the license), they are allowed to cancel your contract and decide to never do business with you again. Well I think they say they can cancel your contract for any reason they want, but that sharing the source code would very likely make them do so.
So similar to what the grsecurity people were trying to do, but at IBM scale instead. Red Hat is far from the company it was 30 years ago. Yeah I am happy I switched to debian 25 years ago because Red Hat's quality was so poor at the time. Debian having a better designed packaging system was a bonus.
Strangely around the same time I switched to RedHat because I got tired of having to apply my own security patches to the kernel and applications because the distribution was shipping with largely unmodified sources. Like many things in life "your mileage may vary". -- Alvin Starr || land: (647)478-6285 Netvel Inc. || Cell: (416)806-0133 alvin@netvel.net ||
This Red Hat change concerns me. LONG: Some thoughts on what my "go to" distro pair should be. | From: Alvin Starr via talk <talk@gtalug.org> | On 2023-06-27 08:19, Lennart Sorensen via talk wrote: | > Yeah I am happy I switched to debian 25 years ago because Red Hat's | > quality was so poor at the time. Debian having a better designed | > packaging system was a bonus. | > | Strangely around the same time I switched to RedHat because I got tired of | having to apply my own security patches to the kernel and applications because | the distribution was shipping with largely unmodified sources. | Like many things in life "your mileage may vary". The RHEL / CentOS / clones drama is certainly unsettling. I don't think that Fedora is directly affected but it is hard to judge whether there will be secondary effects. One upcoming GTALUG talk will be from a Rocky Linux guy. That should be interesting. I've already been struggling with where I want to go for a stable system. Besides the drama, I just don't think that RHEL's pinning versions for 5 years is a good strategy. Backporting for that long feels like a wasted effort, prone to errors. Why do I care about the effort RH puts into backporting? - it creates a RH "moat": it prevents others from competing with them. Rocky, Alma, Oracle feel like clones, not creative competitors. That may be unfair to Oracle but that's not a company that I want a relationship with. - it is labour that feels wasted. Perhaps that labour could be used for more constructive purposes. On the other hand RH has added a lot to the community and does do a good job of beating back bugs. I do think that I need a pair of distros: one that is up to date, and one that is low-drama. If they are in the same family, that cuts down on redundant learning on my part. - CentOS + Fedora has been a good pair for me. TBH, CentOS has left me with technical debt: I get stuck on obsolete versions because the upgrade paths have been disrupted (twice!). Fedora release updates have been good for some years. For my workstation (desktop and laptop) use, I've been very happy with Fedora, but it sure has a firehose of updates. I don't think that it is affected directly by any of this. But if a lot of people migrate away from Red Hat stuff, it won't likely be good for Fedora. It feels as if RH steers the future of Linux by making so many contributions. - Ubuntu LTS + fresh Ubuntu has been pretty good. I've had more problems with package updates on Ubuntu than of Fedora, but it has been pretty good. Distro version upgrades have been good but not perfect in my modest experience. Canonical has repeatedly acted in ways that offend or scare me. So Ubuntu, although easy, feels like a potential trap. - debian Stable + Testing + Unstable. I don't have much experience with debian. I fear that the lack of full-time paid engineers might reduce the safety relative to RH (that could easily just be FUD). debian's goals are good by me. So: I'm thinking of switching to debian. I'd like to learn from others. How do you choose to solve these problems? Maybe some of them are non-problems. ================ Giles has a problem with needing a stable distro with a more recent FireFox. I suggested, against my preferences, that this might be a perfect use for Snaps/Flatpacks. I wonder if I should be using a stable distro everywhere but with containerized upgraded packages where they matter. I yet don't think so. The rest of my family uses Fedora on their workstations. But they hate applying updates (even when I do the work). They are way behind most of the time. Maybe a stable distro + a fresh FireFox would be best for them too. How many other packages would I need to have fresher-than-stable? - support for newer hardware - compilers etc. - more pain-points would be discovered. ================ A fundamental problem is that feature changes and bug fixes are usually mingled in upstream. In some cases, it is a false distinction. Few developers want to maintain a bunch of old releases. It is very hard for a distro to correctly separate these two, and yet that is required to maintain a stable distro.
My take: 1. I wouldn't go as far as "I saw this coming", but I have long suspected that the IBMification of Red Hat was far from complete -- layoffs, CentOS Stream, now this. RH employees that I know describe a hard shift in corporate culture. And I don't think they're done. It wouldn't surprise me the least if they even change the name of the product to "IBM Red Hat Enterprise Linux" or even just "IBM Linux". Now that they've effectively (and knowingly) destroyed the community goodwill that was formed over more than a decade of Linux Expos and Bob Young roadshows, I don't even see much added value in the RH brand to IBM; the Red Hat we've known for decades no longer exists. Come to think of it, the IBM that helped start LPI and championed Open Source against the SCO and Java assaults of a decade ago is also long gone. 2. IBM doesn't really give a damn about Alma and Rocky, they're just incidental casualties. The #1 and maybe only target of the subscription-wall action is IBM's longtime arch-enemy Oracle, which may now be forced to actually maintain its own distro and will no longer be able to claim bug-for-bug compatibility with RHEL (or whatever it will be called). They've calculated that the value of the harm this causes Oracle exceeds the lost value of community rejection. 3. This unfortunate momentum could be stopped (or at least slowed) by a Fedora developer revolt but I don't see that happening. 4. I see an opportunity for SUSE which maintains both an enterprise-Linux focus and good community relations. Are they up to it? As a longshot maybe even Oracle could try to seize the moment and try a charm offensive to attract a community... but that's unlikely considered its many burned bridges (Solaris, OpenOffice, Java) Evan Leibovitch, Toronto Canada @evanleibovitch / @el56 On Tue, Jun 27, 2023 at 11:23 AM D. Hugh Redelmeier via talk < talk@gtalug.org> wrote:
This Red Hat change concerns me.
LONG: Some thoughts on what my "go to" distro pair should be.
| From: Alvin Starr via talk <talk@gtalug.org>
| On 2023-06-27 08:19, Lennart Sorensen via talk wrote:
| > Yeah I am happy I switched to debian 25 years ago because Red Hat's | > quality was so poor at the time. Debian having a better designed | > packaging system was a bonus. | > | Strangely around the same time I switched to RedHat because I got tired of | having to apply my own security patches to the kernel and applications because | the distribution was shipping with largely unmodified sources. | Like many things in life "your mileage may vary".
The RHEL / CentOS / clones drama is certainly unsettling.
I don't think that Fedora is directly affected but it is hard to judge whether there will be secondary effects.
One upcoming GTALUG talk will be from a Rocky Linux guy. That should be interesting.
I've already been struggling with where I want to go for a stable system. Besides the drama, I just don't think that RHEL's pinning versions for 5 years is a good strategy. Backporting for that long feels like a wasted effort, prone to errors.
Why do I care about the effort RH puts into backporting?
- it creates a RH "moat": it prevents others from competing with them. Rocky, Alma, Oracle feel like clones, not creative competitors. That may be unfair to Oracle but that's not a company that I want a relationship with.
- it is labour that feels wasted. Perhaps that labour could be used for more constructive purposes.
On the other hand RH has added a lot to the community and does do a good job of beating back bugs.
I do think that I need a pair of distros: one that is up to date, and one that is low-drama. If they are in the same family, that cuts down on redundant learning on my part.
- CentOS + Fedora has been a good pair for me. TBH, CentOS has left me with technical debt: I get stuck on obsolete versions because the upgrade paths have been disrupted (twice!). Fedora release updates have been good for some years.
For my workstation (desktop and laptop) use, I've been very happy with Fedora, but it sure has a firehose of updates. I don't think that it is affected directly by any of this. But if a lot of people migrate away from Red Hat stuff, it won't likely be good for Fedora.
It feels as if RH steers the future of Linux by making so many contributions.
- Ubuntu LTS + fresh Ubuntu has been pretty good. I've had more problems with package updates on Ubuntu than of Fedora, but it has been pretty good. Distro version upgrades have been good but not perfect in my modest experience.
Canonical has repeatedly acted in ways that offend or scare me. So Ubuntu, although easy, feels like a potential trap.
- debian Stable + Testing + Unstable. I don't have much experience with debian. I fear that the lack of full-time paid engineers might reduce the safety relative to RH (that could easily just be FUD). debian's goals are good by me.
So: I'm thinking of switching to debian.
I'd like to learn from others. How do you choose to solve these problems? Maybe some of them are non-problems.
================
Giles has a problem with needing a stable distro with a more recent FireFox. I suggested, against my preferences, that this might be a perfect use for Snaps/Flatpacks.
I wonder if I should be using a stable distro everywhere but with containerized upgraded packages where they matter. I yet don't think so.
The rest of my family uses Fedora on their workstations. But they hate applying updates (even when I do the work). They are way behind most of the time. Maybe a stable distro + a fresh FireFox would be best for them too.
How many other packages would I need to have fresher-than-stable?
- support for newer hardware
- compilers etc.
- more pain-points would be discovered.
================
A fundamental problem is that feature changes and bug fixes are usually mingled in upstream. In some cases, it is a false distinction. Few developers want to maintain a bunch of old releases. It is very hard for a distro to correctly separate these two, and yet that is required to maintain a stable distro. --- Post to this mailing list talk@gtalug.org Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk
I was just reading about it - this is an interesting (depressing) take from Bradley Kuhn: https://sfconservancy.org/blog/2023/jun/23/rhel-gpl-analysis/ On Tue, 4 Jul 2023 at 06:40, Evan Leibovitch via talk <talk@gtalug.org> wrote:
My take:
I wouldn't go as far as "I saw this coming", but I have long suspected that the IBMification of Red Hat was far from complete -- layoffs, CentOS Stream, now this. RH employees that I know describe a hard shift in corporate culture. And I don't think they're done. It wouldn't surprise me the least if they even change the name of the product to "IBM Red Hat Enterprise Linux" or even just "IBM Linux". Now that they've effectively (and knowingly) destroyed the community goodwill that was formed over more than a decade of Linux Expos and Bob Young roadshows, I don't even see much added value in the RH brand to IBM; the Red Hat we've known for decades no longer exists. Come to think of it, the IBM that helped start LPI and championed Open Source against the SCO and Java assaults of a decade ago is also long gone.
IBM doesn't really give a damn about Alma and Rocky, they're just incidental casualties. The #1 and maybe only target of the subscription-wall action is IBM's longtime arch-enemy Oracle, which may now be forced to actually maintain its own distro and will no longer be able to claim bug-for-bug compatibility with RHEL (or whatever it will be called). They've calculated that the value of the harm this causes Oracle exceeds the lost value of community rejection.
This unfortunate momentum could be stopped (or at least slowed) by a Fedora developer revolt but I don't see that happening.
I see an opportunity for SUSE which maintains both an enterprise-Linux focus and good community relations. Are they up to it? As a longshot maybe even Oracle could try to seize the moment and try a charm offensive to attract a community... but that's unlikely considered its many burned bridges (Solaris, OpenOffice, Java)
Evan Leibovitch, Toronto Canada @evanleibovitch / @el56
On Tue, Jun 27, 2023 at 11:23 AM D. Hugh Redelmeier via talk <talk@gtalug.org> wrote:
This Red Hat change concerns me.
LONG: Some thoughts on what my "go to" distro pair should be.
| From: Alvin Starr via talk <talk@gtalug.org>
| On 2023-06-27 08:19, Lennart Sorensen via talk wrote:
| > Yeah I am happy I switched to debian 25 years ago because Red Hat's | > quality was so poor at the time. Debian having a better designed | > packaging system was a bonus. | > | Strangely around the same time I switched to RedHat because I got tired of | having to apply my own security patches to the kernel and applications because | the distribution was shipping with largely unmodified sources. | Like many things in life "your mileage may vary".
The RHEL / CentOS / clones drama is certainly unsettling.
I don't think that Fedora is directly affected but it is hard to judge whether there will be secondary effects.
One upcoming GTALUG talk will be from a Rocky Linux guy. That should be interesting.
I've already been struggling with where I want to go for a stable system. Besides the drama, I just don't think that RHEL's pinning versions for 5 years is a good strategy. Backporting for that long feels like a wasted effort, prone to errors.
Why do I care about the effort RH puts into backporting?
- it creates a RH "moat": it prevents others from competing with them. Rocky, Alma, Oracle feel like clones, not creative competitors. That may be unfair to Oracle but that's not a company that I want a relationship with.
- it is labour that feels wasted. Perhaps that labour could be used for more constructive purposes.
On the other hand RH has added a lot to the community and does do a good job of beating back bugs.
I do think that I need a pair of distros: one that is up to date, and one that is low-drama. If they are in the same family, that cuts down on redundant learning on my part.
- CentOS + Fedora has been a good pair for me. TBH, CentOS has left me with technical debt: I get stuck on obsolete versions because the upgrade paths have been disrupted (twice!). Fedora release updates have been good for some years.
For my workstation (desktop and laptop) use, I've been very happy with Fedora, but it sure has a firehose of updates. I don't think that it is affected directly by any of this. But if a lot of people migrate away from Red Hat stuff, it won't likely be good for Fedora.
It feels as if RH steers the future of Linux by making so many contributions.
- Ubuntu LTS + fresh Ubuntu has been pretty good. I've had more problems with package updates on Ubuntu than of Fedora, but it has been pretty good. Distro version upgrades have been good but not perfect in my modest experience.
Canonical has repeatedly acted in ways that offend or scare me. So Ubuntu, although easy, feels like a potential trap.
- debian Stable + Testing + Unstable. I don't have much experience with debian. I fear that the lack of full-time paid engineers might reduce the safety relative to RH (that could easily just be FUD). debian's goals are good by me.
So: I'm thinking of switching to debian.
I'd like to learn from others. How do you choose to solve these problems? Maybe some of them are non-problems.
================
Giles has a problem with needing a stable distro with a more recent FireFox. I suggested, against my preferences, that this might be a perfect use for Snaps/Flatpacks.
I wonder if I should be using a stable distro everywhere but with containerized upgraded packages where they matter. I yet don't think so.
The rest of my family uses Fedora on their workstations. But they hate applying updates (even when I do the work). They are way behind most of the time. Maybe a stable distro + a fresh FireFox would be best for them too.
How many other packages would I need to have fresher-than-stable?
- support for newer hardware
- compilers etc.
- more pain-points would be discovered.
================
A fundamental problem is that feature changes and bug fixes are usually mingled in upstream. In some cases, it is a false distinction. Few developers want to maintain a bunch of old releases. It is very hard for a distro to correctly separate these two, and yet that is required to maintain a stable distro. --- Post to this mailing list talk@gtalug.org Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk
--- Post to this mailing list talk@gtalug.org Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk
-- Giles https://www.gilesorr.com/ gilesorr@gmail.com
Evan Leibovitch via talk wrote on 2023-07-04 03:39:
My take:
2. IBM doesn't really give a damn about Alma and Rocky, they're just incidental casualties. The #1 and maybe only target of the subscription-wall action is IBM's longtime arch-enemy Oracle, which may now be forced to actually maintain its own distro
This was one of my first thoughts too - Oracle will no longer be able to "leech" off of RedHat, which is rather delicious considering Oracle's reputation.
4. I see an opportunity for SUSE which maintains both an enterprise-Linux focus and good community relations. Are they up to it? As a longshot maybe even Oracle could try to seize the moment and try a charm offensive to attract a community... but that's unlikely considered its many burned bridges (Solaris, OpenOffice, Java)
That's an interesting idea - the rise of SUSE / OpenSuse. I kinda like the idea - I don't have personal experience with it, but it seems well-regarded. Seems the whole English speaking world (of podcasters and YouTubers) has a bit of a blind spot to SUSE and they probably deserve more attention. Good call, hope it pans out. As for Oracle, they have free / cheap cloud hosting for tiny instances, so maybe they'll try a charm offensive, however even with a free tiny-VPS, they don't seem to have much mind share. I can't really see this. Charm & Oracle don't belong in the same sentence. rb
On Thu, Jul 6, 2023 at 7:04 AM BCLUG via talk <talk@gtalug.org> wrote:
Evan Leibovitch via talk wrote on 2023-07-04 03:39:
4. I see an opportunity for SUSE which maintains both an enterprise-Linux focus and good community relations. Are they up to it? As a longshot maybe even Oracle could try to seize the moment and try a charm offensive to attract a community... but that's unlikely considered its many burned bridges (Solaris, OpenOffice, Java)
That's an interesting idea - the rise of SUSE / OpenSuse. I kinda like the idea - I don't have personal experience with it, but it seems well-regarded.
Seems the whole English speaking world (of podcasters and YouTubers) has a bit of a blind spot to SUSE and they probably deserve more attention.
Good call, hope it pans out.
This path just got WAY more interesting. SUSE has committed to creating and maintaining an open source fork of RHEL: https://news.itsfoss.com/suse-rhel-fork/ - Evan
On Tue, Jun 27, 2023 at 10:22 AM D. Hugh Redelmeier via talk <talk@gtalug.org> wrote:
This Red Hat change concerns me. snip
It feels as if RH steers the future of Linux by making so many contributions.
- Ubuntu LTS + fresh Ubuntu has been pretty good. I've had more problems with package updates on Ubuntu than of Fedora, but it has been pretty good. Distro version upgrades have been good but not perfect in my modest experience.
Canonical has repeatedly acted in ways that offend or scare me. So Ubuntu, although easy, feels like a potential trap.
Found this to be true - - - - spent a lot of time a number of years ago looking into LXD - - - the snap environment is one that NEVER will be seen here again if I can help it. Canonical is, imo, desperately looking for ways to monetize their brand - - - - - someone's just have to have more $$$$$$$$$.
- debian Stable + Testing + Unstable. I don't have much experience with debian. I fear that the lack of full-time paid engineers might reduce the safety relative to RH (that could easily just be FUD). debian's goals are good by me.
So: I'm thinking of switching to debian.
I was with Debian for over 10 years - - - have now switched to Devuan - - - thereby getting rid of another item of 'control'.
I'd like to learn from others. How do you choose to solve these problems? Maybe some of them are non-problems.
The problem is that the small encroachments don't tend to isolate - - they tend to grow - - - somewhat like microbes! (With similar results in my experience!).
================
Giles has a problem with needing a stable distro with a more recent FireFox. I suggested, against my preferences, that this might be a perfect use for Snaps/Flatpacks.
Snaps are a system controlling adventure - - - be aware of this BEFORE starting down that road. Haven't used flatpacks. I find that browsers are needing updates almost on a daily basis. Am wondering if there is a way of reducing the 'encroachment' of the nefarious bits of cruft adhering to all browsers (at least as far as I see).
I wonder if I should be using a stable distro everywhere but with containerized upgraded packages where they matter. I yet don't think so.
The rest of my family uses Fedora on their workstations. But they hate applying updates (even when I do the work). They are way behind most of the time. Maybe a stable distro + a fresh FireFox would be best for them too.
I've already been informed that if I weren't doing updates windows would be applied post haste - - - something about the devil they know!!!! (Work systems are all M$ Win!)
How many other packages would I need to have fresher-than-stable?
- support for newer hardware
- compilers etc.
- more pain-points would be discovered.
================
A fundamental problem is that feature changes and bug fixes are usually mingled in upstream. In some cases, it is a false distinction. Few developers want to maintain a bunch of old releases. It is very hard for a distro to correctly separate these two, and yet that is required to maintain a stable distro.
I tend to run in debian's equivalent of 'testing'. Found over the last more than 10 years that that was a reasonable compromise to stability and currentness. HTH
A couple of points following on to 01bigtenor's reply: According to Wikipedia: "Devuan is a fork of the Debian Linux distribution that uses sysvinit, runit or OpenRC instead of systemd. Devuan aims to avoid 'lock-in' by projects like systemd and aims to maintain compatibility with other init systems to avoid detaching Linux from other Unix systems." When systemd was initially a horrible mess, I was very much in favour of this. But for the last several years systemd has been rock-steady, and once you wrap your head around the basics, it's a LOT easier to use than maintaining those damn /etc/rc.N/ folders. At least that's been my experience. As for Debian stable vs. testing: 20 years ago Debian testing was great ... about a decade ago I found it to be noticeably less stable, and the constant volume of updates got on my nerves. So I switched to stable. Even if the stability of testing has improved, I don't want to put up with the updates. If I need something newer, I can "pin" a package, getting it either from testing or backports. This is admittedly a PITA to set up, and I hardly ever use it because of that ... but I _have_ used it, and it proved reasonably reliable in practice. Although getting your head around pinning weights is ... nasty. On Tue, 4 Jul 2023 at 08:24, o1bigtenor via talk <talk@gtalug.org> wrote:
On Tue, Jun 27, 2023 at 10:22 AM D. Hugh Redelmeier via talk <talk@gtalug.org> wrote:
This Red Hat change concerns me. snip
It feels as if RH steers the future of Linux by making so many contributions.
- Ubuntu LTS + fresh Ubuntu has been pretty good. I've had more problems with package updates on Ubuntu than of Fedora, but it has been pretty good. Distro version upgrades have been good but not perfect in my modest experience.
Canonical has repeatedly acted in ways that offend or scare me. So Ubuntu, although easy, feels like a potential trap.
Found this to be true - - - - spent a lot of time a number of years ago looking into LXD - - - the snap environment is one that NEVER will be seen here again if I can help it.
Canonical is, imo, desperately looking for ways to monetize their brand - - - - - someone's just have to have more $$$$$$$$$.
- debian Stable + Testing + Unstable. I don't have much experience with debian. I fear that the lack of full-time paid engineers might reduce the safety relative to RH (that could easily just be FUD). debian's goals are good by me.
So: I'm thinking of switching to debian.
I was with Debian for over 10 years - - - have now switched to Devuan - - - thereby getting rid of another item of 'control'.
I'd like to learn from others. How do you choose to solve these problems? Maybe some of them are non-problems.
The problem is that the small encroachments don't tend to isolate - - they tend to grow - - - somewhat like microbes! (With similar results in my experience!).
================
Giles has a problem with needing a stable distro with a more recent FireFox. I suggested, against my preferences, that this might be a perfect use for Snaps/Flatpacks.
Snaps are a system controlling adventure - - - be aware of this BEFORE starting down that road. Haven't used flatpacks.
I find that browsers are needing updates almost on a daily basis. Am wondering if there is a way of reducing the 'encroachment' of the nefarious bits of cruft adhering to all browsers (at least as far as I see).
I wonder if I should be using a stable distro everywhere but with containerized upgraded packages where they matter. I yet don't think so.
The rest of my family uses Fedora on their workstations. But they hate applying updates (even when I do the work). They are way behind most of the time. Maybe a stable distro + a fresh FireFox would be best for them too.
I've already been informed that if I weren't doing updates windows would be applied post haste - - - something about the devil they know!!!! (Work systems are all M$ Win!)
How many other packages would I need to have fresher-than-stable?
- support for newer hardware
- compilers etc.
- more pain-points would be discovered.
================
A fundamental problem is that feature changes and bug fixes are usually mingled in upstream. In some cases, it is a false distinction. Few developers want to maintain a bunch of old releases. It is very hard for a distro to correctly separate these two, and yet that is required to maintain a stable distro.
I tend to run in debian's equivalent of 'testing'. Found over the last more than 10 years that that was a reasonable compromise to stability and currentness.
HTH --- Post to this mailing list talk@gtalug.org Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk
-- Giles https://www.gilesorr.com/ gilesorr@gmail.com
| From: Giles Orr via talk <talk@gtalug.org> Thank you and others for telling us about your experience. | If I need something newer, I can "pin" a | package, getting it either from testing or backports. This is | admittedly a PITA to set up, and I hardly ever use it because of that | ... but I _have_ used it, and it proved reasonably reliable in | practice. Although getting your head around pinning weights is ... | nasty. Can you use that to get the newer FireFox that you need?
On Tue, Jul 4, 2023 at 10:01 PM D. Hugh Redelmeier via talk <talk@gtalug.org> wrote:
| From: Giles Orr via talk <talk@gtalug.org>
Thank you and others for telling us about your experience.
| If I need something newer, I can "pin" a | package, getting it either from testing or backports. This is | admittedly a PITA to set up, and I hardly ever use it because of that | ... but I _have_ used it, and it proved reasonably reliable in | practice. Although getting your head around pinning weights is ... | nasty.
Can you use that to get the newer FireFox that you need?
I've been using Firefox-esr for years from debian testing. Seems to work reasonably. At present its my predominant use browser. HTH
On Tue, 4 Jul 2023 at 23:01, D. Hugh Redelmeier via talk <talk@gtalug.org> wrote:
| From: Giles Orr via talk <talk@gtalug.org>
Thank you and others for telling us about your experience.
| If I need something newer, I can "pin" a | package, getting it either from testing or backports. This is | admittedly a PITA to set up, and I hardly ever use it because of that | ... but I _have_ used it, and it proved reasonably reliable in | practice. Although getting your head around pinning weights is ... | nasty.
Can you use that to get the newer FireFox that you need?
Uhhh ... I had never considered it, which seems more than a little blind. I haven't used pinning in a couple years, and when I did it was usually to get a piece of software that wasn't in the current distro (but was in the upcoming release). However, on closer examination (I used this page: https://www.debian.org/distrib/packages - you should become close friends with this if using Debian) "testing" also includes Firefox ESR, so the answer would appear to be no. Makes sense in a depressing way: Debian wants a _stable_ version for stable, and testing is for testing things that will become the stable distro ... Huh - testing is called "trixie." I always wonder which toy is coming next ... -- Giles https://www.gilesorr.com/ gilesorr@gmail.com
On 2023-07-05 08:57, Giles Orr via talk wrote:
... Makes sense in a depressing way: Debian wants a _stable_ version for stable, and testing is for testing things that will become the stable distro ...
Huh - testing is called "trixie." I always wonder which toy is coming next ...
Debian testing runs in a cycle with their releases. Just after the preceding one (bookworm) came out, it opened to a lot of heavy-lifting upgrades and you're more likely to see glitches. Then they start calling code freezes and stabilize things when they get closer to releasing the next one (trixie) in about two years and you'll see fine-tuning of bugs in the pre-release era and testing will be a lot more stable for any innocent users, then comes the grand release announcement and the next new cycle. Anthony
Giles Orr via talk wrote on 2023-07-04 18:09:
But for the last several years systemd has been rock-steady, and once you wrap your head around the basics, it's a LOT easier to use than maintaining those damn /etc/rc.N/ folders. At least that's been my experience.
My experience too - systemd is fine and is much easier to maintain, better documented, and has more features. And, since it's used in all the major Linux distros, I feel reasonably comfortable that it's thoroughly tested. If I reject systemd via a "boutique" distro, what else am I giving up? Devuan - is the team there large enough to keep up with the ever-changing security issues, features, etc? If systemd were anathema to someone, shouldn't they move to BSD, where a large and experienced team exists? Thought I'd add my 2c rb
participants (9)
-
Alvin Starr -
Anthony de Boer -
BCLUG -
Colin McGregor -
D. Hugh Redelmeier -
Evan Leibovitch -
Giles Orr -
Lennart Sorensen -
o1bigtenor