|
|
Log in / Subscribe / Register

Klumpp: A few words about the future of the Limba project

[Posted June 18, 2016 by n8willis]

Those concerned about the proliferation of application-packaging formats will soon have one fewer to worry about. At his blog, Matthias Klumpp announces that he intends to scale back his work on Limba, the cross-distribution application-packaging format he has developed as an extension of the ideas in the earlier Listaller. The decision comes on the heels of discussions with Flatpak developer Alexander Larsson, since the two projects overlap in many respects: "Alex and I had very productive discussions, and except for the modularity issue, we were pretty much on the same page in every other aspect regarding the sandboxing and app-distribution matters."

Given that he has several other active projects in development, Klumpp has decided to throttle back on Limba, although he will continue to hack on it "as a research project" and sees several opportunities where it might still fit into vendor-independent software distribution down the road. "This is good news for all the people out there using the Tanglu Linux distribution, AppStream-metadata-consuming services, PackageKit on Debian, etc. – those will receive more attention," Klumpp concludes.

to post comments

this is sad

Posted Jun 18, 2016 7:29 UTC (Sat) by HelloWorld (guest, #56129) [Link] (86 responses)

I had hoped that distributions would eventually stop distributing applications and rely on cross-distro packages instead, but I suspect with flatpak that just isn't going to happen. Its dependency concept (of having one single dependency, the “runtime”, and bundling everything else) is too primitive, one will invariably end up with wasted space and security vulnerabilities in poorly-maintained packages. So enough advantages for distro-specific packages will remain for them not to go away. systemd fixed many other inconsistencies, but it looks like this one will be with us for years to come.

this is sad

Posted Jun 18, 2016 9:59 UTC (Sat) by khim (subscriber, #9252) [Link] (60 responses)

Why is that such a big deal? Dependency model could be changed later if needed.

As I've said years ago: Linux distros are on borrowed time and clocks are tickling.

Android move to desktop is slow (Google just released first development version of Android support for first proper laptop device…), thus Linux distros still have time: my 3 years optimistic estimate was clearly wrong, 5-7 years look more and more likely.

But remember that even mainstream press, slow as clueless as usual, have realized what's happening… it's a race: will Linux distros be able to provide at least some incentive for ISVs before they will all be rendered irrelevant? We'll see. Interesting times, interesting times indeed.

P.S. Of course GNU/Linux will survive as “hobbyists project” even if this attempt will fail (and some even claim that is what they desire—similarly to KolibriOS, AROS or Haiku), that's not about that, that's about making it actually relevant for “Joe Average”… will it happen or not? Linux distros wasted sooo much time, will they be able to finally get their act together?

this is sad

Posted Jun 18, 2016 10:49 UTC (Sat) by Wol (subscriber, #4433) [Link] (50 responses)

The problem with your world view is that Google Play is a bloated monster ...

We have two Android mobile phones in our house that are now pretty much useless, because they are caught in the following vicious circle ...

Phone 1 has updates switched off, so nothing works because all the internet interfaces have been upgraded ...

Phone 2 has no apps (that we actually WANT, as opposed to foisted on us), because the apps we want need an updated Google Play, and in order to update Google Play we need to delete the apps to make space!

So in your world view everyone will be forced to use the distro called "Android desktop" aka "Google Play". Thanks, but no thanks. How many apps are foisted onto my phone? Twenty? Thirty? How many of those apps do I actually use? One? Two? That's one hell of a vulnerability attack surface THAT SHOULDN'T EVEN BE THERE!

Cheers,
Wol

this is sad

Posted Jun 18, 2016 13:40 UTC (Sat) by khim (subscriber, #9252) [Link] (49 responses)

So in your world view everyone will be forced to use the distro called "Android desktop" aka "Google Play".

Not necessarily. There are still iOS, macOS and Windows. These would still be supported, too (for about a decade, at least). Just GNU/Linux would be dropped.

There are about 1% of users who use GNU/Linux today on desktop and many more—on server. But if you think that these are popular because it's FOSS then you are sorely mistaken: GNU/Linux is used because it's cheap, not because it's open. Android is free, too, thus it could easily fill that niche.

The only question is: when and if it'll become usable enough.

How many apps are foisted onto my phone? Twenty? Thirty? How many of those apps do I actually use? One? Two? That's one hell of a vulnerability attack surface THAT SHOULDN'T EVEN BE THERE!

Welcome to the real world. The OSes which people actually use are all made like this. Well, ChromeOS is kinda-sorta exception, but it clearly lost to Android and Google basically admits that now, too.

It's not question if whether your want it or not, it's just how things are always done in the “real world”. Even when Linux enjoyed brief popularity on Netbooks it was preinstalled in the exact same fashion.

Of course they only had various sets of bloatware and no support from ISVs thus they quickly faded away, but it's naive to expect that you would be able to just go and pick exactly what you want: hardware companies need to make a living somehow.

this is sad

Posted Jun 19, 2016 1:20 UTC (Sun) by drag (guest, #31333) [Link] (4 responses)

The solution to bloatware is to glet devices you can install your own oses/firmware on.

This requires not only software but requires respectful hardware. This is one of the classic ironic things about the Linux Desktop... We had the benefit of Microsoft using its clout to force standards on the x86 platform. a this meant Linux had a relatively easy time in PCs.

Android is freaking awesome FLOSS OS, but you won't benefit from it if you depend on stock firewares on leased/contracted phones.

this is sad

Posted Jun 20, 2016 13:43 UTC (Mon) by clump (subscriber, #27801) [Link] (3 responses)

Android is freaking awesome FLOSS OS, but you won't benefit from it if you depend on stock firewares on leased/contracted phones.
Android is open source, certainly, but "awesome?" Does Android have a particularly vibrant, collaborative open source community behind it? Is code not tossed over the fence when one entity considers it done? Are security updates made available in a reasonable amount of time to users?

While you can buy phones that make it simpler to change firmware, you're going to be installing firmware downstream of Android where the above problems now matter much more to you.

this is sad

Posted Jun 20, 2016 14:10 UTC (Mon) by drag (guest, #31333) [Link] (2 responses)

> Android is open source, certainly, but "awesome?"

In terms of being able to meet the needs of end users and mobile devices it is a massive improvement over any other FLOSS system out there. With Wayland and container-based packaging 'regular' Linux is coming close, but it's till going to be years before it's up to were Android is. Trying to retrofit modern features on aging code base is going to be more difficult then writing something modern from scratch.

> While you can buy phones that make it simpler to change firmware, you're going to be installing firmware downstream of Android where the above problems now matter much more to you.

Any other FLOSS solution in it's position will face the same problems. People with proprietary drivers, closed firmware, FCC rules, patent lawsuits, losers trying to get revenge through copyright lawsuits, vendors seeking revenue streams through advertising, leasing locked-down phones at massive discounts with obnoxious contracts, and nasty pre-installed applications, etc etc. That is the price you pay for being successful.

Standardization in x86 platform benefits regular Linux users significantly and this standardization is due to Microsoft. People wanting to use FLOSS on other platforms are going need to be very careful in what they spend their money on. Even then they are going to be burned occasionally unless they find a company that specifically caters to this segment.

this is sad

Posted Jun 20, 2016 15:05 UTC (Mon) by clump (subscriber, #27801) [Link]

In terms of being able to meet the needs of end users and mobile devices it is a massive improvement over any other FLOSS system out there. With Wayland and container-based packaging 'regular' Linux is coming close, but it's till going to be years before it's up to were Android is. Trying to retrofit modern features on aging code base is going to be more difficult then writing something modern from scratch.
Except for little concerns like security, lock-in, and having a healthy ecosystem I agree. Wayland and containers are completely irrelevant here.

Any other FLOSS solution in it's position will face the same problems. People with proprietary drivers, closed firmware, FCC rules, patent lawsuits, losers trying to get revenge through copyright lawsuits, vendors seeking revenue streams through advertising, leasing locked-down phones at massive discounts with obnoxious contracts, and nasty pre-installed applications, etc etc. That is the price you pay for being successful.
The situation for everything you've described is far better for users in the desktop x86 world than the mobile world. The solution is not to give up and proclaim problems aren't fixable.

this is sad

Posted Jun 20, 2016 16:30 UTC (Mon) by anselm (subscriber, #2796) [Link]

With Wayland and container-based packaging 'regular' Linux is coming close, but it's till going to be years before it's up to were Android is.

Sailfish OS (being a smartphone Linux that is based on, among other things, Wayland, Qt, and systemd) is actually not at all bad. As a base system it is rock solid and does what it is supposed to do. It's certainly not “years before it's up to where Android is”. Granted, some of the preloaded productivity apps could use a few more features, but even stock Android is, in many cases, nothing to write home about in that area. And my Sailfish-OS based phone has other advantages, too – having to charge it only once every three days when the Android phone I used to have lasted barely 24 hours on one charge is a definite plus in my book. Oh, and proper multitasking, too.

this is sad

Posted Jun 19, 2016 20:36 UTC (Sun) by callegar (guest, #16148) [Link] (41 responses)

The other way round, I'd say that GNU/Linux is not very successful on the desktop because it is not cheap enough for Average Joe. On day zero it costs exactly like another mainstream OS that comes ready installed on the new PC. From that day on, it costs him more time because the hardware is less supported and you never know if the scanner, camera, printer, TV card, USB video dongle, CDMA dongle, video capture dongle, home surveillance camera, will work seamless, will work at all, or will require you hours to make it work. Even shopping for electronic equipment is a time consuming exercise when you are on Linux. Ultimately, time is the most precious asset we have and Joe Average does not use Linux on the desktop because it is far more (time)expensive for him. Distros did never and will never have enough weight to fix this. Google has.

this is sad

Posted Jun 19, 2016 21:38 UTC (Sun) by khim (subscriber, #9252) [Link] (36 responses)

From that day on, it costs him more time because the hardware is less supported and you never know if the scanner, camera, printer, TV card, USB video dongle, CDMA dongle, video capture dongle, home surveillance camera, will work seamless, will work at all, or will require you hours to make it work.

What you said is true for macOS, too. And it's more expensive in purely monetary terms, too. Still it's vastly more popular than Linux.

Distros did never and will never have enough weight to fix this.

Not WRT hardware, no.

Google has.

Not really. Chromebooks are even worse there than Windows - and yet they occupy 7 out of 20 most popular slots on amazon.

Yes, time is the most precious asset - but that does not mean that distributions couldn't do anything about it. They could, they just never held that as #1 priority. Other considerations dominated. And the most probematic was always the inability to create portable installable package. They fought tooth and nail to make life of their users miserable.

Crazy distro makers rants miss the point so completely I don't even know what to say. When I read “Put out the sources with a makefile—the maintainers can handle the rest (heck, the makefile is optional, we can write one)” I weep.

Most software packages out there don't have developers or maintainers. They are built by some freelancer who was paid some lump sum. Once. Then they are used for years. There are no developers, there are no upstream and there are no one who could “Put out the sources with a makefile”!

Is it good? Is it bad? No, that's just how thing are done in “real world”. When you want to install an oven or a bathhouse you pay some lump sum to someone knows how to do that—and then use it while it works. Any maintenance is something you try to avoid and you certainly wouldn't expect to see the need to rebuild an oven when you replace a carpet! Well, you could damage an oven when you replace a carpet, but that's not something you expect, rather that's “something that happens when thing go bad™”. That's how software is handled, too. This works will all popular OSes, GNU/Linux was the only exception. Now it finally joins the club.

this is sad

Posted Jun 20, 2016 0:32 UTC (Mon) by anselm (subscriber, #2796) [Link] (29 responses)

What you said is true for macOS, too. And it's more expensive in purely monetary terms, too. Still it's vastly more popular than Linux.

The important difference here is that for the vast majority of users, Windows or MacOS come preinstalled with the machine. This means that keeping them is essentially “free” while replacing them with Linux carries a considerable cost right there, and most users aren't prepared to pay that cost.

Consequently, as long as that cost is not removed, Linux could have the greatest app store in the world but people would still buy Macs or Windows PCs – not because they are necessarily the best machine for them, but because Macs and PCs today are generally reasonably good, and changing the operating system on one's machine is a hassle that people can just as easily do without.

this is sad

Posted Jun 20, 2016 3:45 UTC (Mon) by bronson (subscriber, #4806) [Link] (28 responses)

It's easy enough to get Linux preinstalled too. The reason so few people are interested in the Linux desktop runs deeper than that.

this is sad

Posted Jun 20, 2016 8:07 UTC (Mon) by anselm (subscriber, #2796) [Link] (7 responses)

It's easy enough to get Linux preinstalled too.

Here in Germany it isn't – you can't walk into the local white-box electronics store and buy a Linux machine “over the counter” the way you can buy a Windows PC. There are no glitzy high-street shops that sell Linux PCs the way the Apple Store sells Macs. Amazon offers me exactly one computer with pre-installed Linux, and that's a low-spec machine with Linux Mint that has some pretty bad reviews. With diligence you can probably find some specialist outfits that sell reasonable pre-installed Linux PCs on the Internet, but that's not how most people buy their computers. I don't consider that “easy enough”.

The reason so few people are interested in the Linux desktop runs deeper than that.

I support a few friends (non-technical people) who are running Linux on their desktops. They're happy with it – they enjoy the performance, the reliability, and the fact that they don't need to worry about malware. I've been doing this for almost 20 years now and I don't buy the idea that there is something inherent about Linux that turns people off once they have it running properly – at least not more so than with the other popular operating systems.

this is sad

Posted Jun 20, 2016 8:47 UTC (Mon) by Cyberax (✭ supporter ✭, #52523) [Link] (6 responses)

> Here in Germany it isn't – you can't walk into the local white-box electronics store and buy a Linux machine “over the counter” the way you can buy a Windows PC.
Computers (especially laptops) with Linux were popular in Russia back in middle 2000-s. For a very simple reason - they didn't include the price of Windows license. Consequently, the first question of buyers of these computers was: "Can I install Windows on it?"

Pre-installed Linux desktops were tried many times but they have never caught up.

> I support a few friends (non-technical people) who are running Linux on their desktops. They're happy with it – they enjoy the performance, the reliability, and the fact that they don't need to worry about malware.
Are they using anything that's outside of Chromebooks' capabilities?

this is sad

Posted Jun 20, 2016 9:06 UTC (Mon) by anselm (subscriber, #2796) [Link] (5 responses)

Pre-installed Linux desktops were tried many times but they have never caught up.

All the pre-installed Linux desktops I've seen were low-spec machines with weird Linux distributions that were built in a slap-dash way. Even the mainstream vendors that experimented with pre-installed Linux often didn't do a great job on it. No wonder people don't like it. I have yet to meet someone who was presented with a well-installed Linux on good hardware who wanted to replace it with (of all things) Windows.

The real problem is that vendors don't see a lot of gain in providing good pre-installed Linux, because customers don't seem to want it. Customers don't want Linux because they've heard it is different from Windows (which they usually know) and nobody they know has it. And nobody they know has Linux because it isn't sold pre-installed in white-box stores alongside pre-installed Windows PCs and Macs, because vendors don't offer good pre-installed Linux machines. Catch-22.

Are they using anything that's outside of Chromebooks' capabilities?

How would I know? I don't look over their shoulders all the time, and I have no idea what Chromebooks can actually do. For example, can Chromebooks handle multiple displays? Do Chromebooks run Digikam, The GIMP, and Kdenlive? These are things some of my friends use as a matter of course.

this is sad

Posted Jun 20, 2016 13:37 UTC (Mon) by johannbg (guest, #65743) [Link]

My experience too and there was no price difference in laptops ( from those few available with linux pre-installed ) here and those with windows. Then there was a lack of applications compatible ( or just feature compatible ) to what is available on windows outside the standard applications ( email, browser, office simple editor and calculator etc ) as well as hardware incompatibilities and no easy way to do a simple firmware upgrade on the device from vendors and so on.

Those novice end users ( which includes developers ) that atleast switched to Fedora for a brief time to try linux and did not have a joyful experience in doing so left the linux ecosystem entirely ( the complaint from developers not being able to use Fedora was they spent more time setting up their development environment and use manage it than hacking and novice end user did not like the instability in the ui design in Gnome 2 and 3. ). Those end users did not try alternative desktop environment like the KDE version of Fedora ( The entire Fedora brand suffered at the cost of one desktop environment ) or other distributions like Ubuntu or Debian or Mint since from their perspective they had tried "Linux" and it did not work for them and they did just want to go back to the environment they where familiar with ( Windows,Apple ). Arguably it would be better for if those Desktop Environment would release their own release then users could stay with/away from GnomeOS and try kdeEOS, ChromeOS etc. as in those desktop environment brands would suffer the loss ( or success ) of end users directly themselves but it really does not matter since linux desktop will never materialize for novice end user ( outside potentially Android ) when the ecosystem remains so fragmented but only be usable for tinkeres and administrators which I think everyone is aware of except those that are developing desktop environment.

this is sad

Posted Jun 20, 2016 17:35 UTC (Mon) by Cyberax (✭ supporter ✭, #52523) [Link] (3 responses)

> All the pre-installed Linux desktops I've seen were low-spec machines with weird Linux distributions that were built in a slap-dash way.
I've seen all kinds.

> Even the mainstream vendors that experimented with pre-installed Linux often didn't do a great job on it.
Dell is still selling Linux laptops and they actually tried to make them nice (for example, they wrote DKMS for Debian/Ubuntu).

Ultimately, making a Linux PC is not complicated. The fact that it has never attracted more than a trace interest is notable.

this is sad

Posted Jun 21, 2016 15:28 UTC (Tue) by krake (guest, #55996) [Link] (2 responses)

> I've seen all kinds.

Then you are in luck.
I've never seen any other then maybe Netbooks in physical stores and online stores still fail to offer Linux as an OS option.

> Dell is still selling Linux laptops

They started recently with single models for which reports often say that things don't actually work and official blogs telling about workarounds or making promises of updates.
They have yet to actually offer Linux as an option on their machines for customers who buy computers based on requirements.

> The fact that it has never attracted more than a trace interest is notable.

We don't know, since there has never been any attempt in offering Linux as a pre-install option from any of the major vendors.
Only specialized vendors such as ZaReason or System76 do and they don't have widespread presence and only a limited range of models.

So far, whenever I needed to buy a new laptop, none of the vendors who's products met the requirements closely enough to get into the final consideration was able to offer the device without OS, let alone with Linux.

It is great if that is different in the place where you are from.

this is sad

Posted Jun 21, 2016 21:49 UTC (Tue) by Cyberax (✭ supporter ✭, #52523) [Link]

> They started recently with single models for which reports often say that things don't actually work and official blogs telling about workarounds or making promises of updates.
You mean, 8 years ago is "recently"? And their models are reasonable.

> So far, whenever I needed to buy a new laptop, none of the vendors who's products met the requirements closely enough to get into the final consideration was able to offer the device without OS, let alone with Linux.
Not anymore, after anti-piracy laws got stricter.

"Reports"

Posted Jun 21, 2016 22:31 UTC (Tue) by zlynx (guest, #2285) [Link]

Going by "reports", my report is that the 2013 Dell XPS 13 Developer with Ubuntu worked really well for me. Another "report" is that I am right now using a 2015 Dell m3800, which was going to be a Developer model but I didn't want to wait so I bought the Windows 8.1 version and installed Fedora on it. It also works great.

> We don't know, since there has never been any attempt in offering Linux as a pre-install option from any of the major vendors.

Dell seems pretty major to me.

And they have the small laptop (XPS 13) and the big laptop (m3800) with Linux. I am having a hard time imagining requirements that can't be met by one or the other. Maybe price. Neither one is particularly cheap. Optical drives, wired Ethernet, or old weird ports are handled by adapters.

this is sad

Posted Jun 20, 2016 10:46 UTC (Mon) by krake (guest, #55996) [Link] (19 responses)

> It's easy enough to get Linux preinstalled too.

This might depend on country, but I don't think that is true in general.

Aside from what anselm already said about not being available in physical stores, I haven't come across a single vendor's online configurator that would have Linux as one of the OS options.

At best you can get a single specific model with Linux, which you have to find first.

Again, these online shops tend to be artifically country specific, but every time I was shopping for a new laptop over the past 15 years I've never seen Linux as one of the OS options on any significant range of models on either HP's, Sony's, Dell's or Lenovos's configurators in Austria or Germany.

this is sad

Posted Jun 20, 2016 17:35 UTC (Mon) by bronson (subscriber, #4806) [Link] (18 responses)

Here's one example: http://www.dell.com/learn/us/en/555/campaigns/xps-linux-l... It's not quite as bleak as you're saying. Still, I think we can agree that the demand for preinstalled machines is vanishingly small?

I'm just trying to show that the lack of demand for Linux on the Desktop has very little to do with who installs it.

this is sad

Posted Jun 21, 2016 8:21 UTC (Tue) by krake (guest, #55996) [Link] (17 responses)

Well, this is one of the special model options I was referring about, not a general availability of choosing Linux as the operating system in the model configurator.

Also, in the case of Dell, they had very strange artificial limitations on who they would sell this to.
E.g. for a long time they would have such a special model in their German store and not in their Austrian one and deny any shipping from Germany to Austria as well.
How that was legal under the terms of the EU common market is beyond me.

So from my point of view we are far, far away from "easy enough".

> Still, I think we can agree that the demand for preinstalled machines is vanishingly small?

I don't understand how you can say that.
The staggering majority of people buy preinstalled machines, I only know a handful of people who have installed e.g. a different Windows than they bought a PC with.

Also, if the demand for preinstalled machines were vanishingly small as you think, wouldn't that imply that the majority of systems would be offered and bought without anything preinstalled?

this is sad

Posted Jun 21, 2016 9:53 UTC (Tue) by nye (guest, #51576) [Link] (16 responses)

>> Still, I think we can agree that the demand for preinstalled machines is vanishingly small?
>I don't understand how you can say that.
>The staggering majority of people buy preinstalled machines, I only know a handful of people who have installed e.g. a different Windows than they bought a PC with.

I think the GP meant preinstalled *with Linux*.

this is sad

Posted Jun 21, 2016 10:50 UTC (Tue) by anselm (subscriber, #2796) [Link] (12 responses)

We can't really say a lot about demand for preinstalled Linux, other than that if there is demand it is apparently not communicated well enough to computer manufacturers, or computer manufacturers do not think it is worth their while to respond to whatever demand there is.

I don't think there's active “demand” for Windows or MacOS in any real sense, either, to a point where people walk into a computer store and ask for any computer as long as it comes with Windows, and not one of those nasty Macs. Few people buy a computer from Dell or Acer or Lenovo specifically because it comes with Windows – instead they buy a computer because in the first instance they want a computer, and whatever operating system it comes with is mostly incidental. Conversely, many people buy a Mac because they like the sleek hardware and the way it is marketed as the creative person's computer (as opposed to the boring office-supply type computer that everybody else is selling) and possibly also the operating system, not specifically because they're in love with MacOS and the Mac is (unfortunately) the only computer it will run on. How little many people care about what operating system their computer is running is also evidenced by the fact that they tend not to upgrade the operating system that their computer came with – note how Microsoft needs to practically drag many people kicking and screaming into allowing Windows 10 to be installed on their machines.

Linux's problem in this setup is simply that few people know about it to the point where they want it on their computer, and of the people who do, most are happy to install it themselves (if only so they get to pick the particular flavour of Linux they end up with). Therefore there is little incentive for manufacturers to go to the considerable trouble and expense of actually spec'ing a computer that contains only Linux-friendly hardware, getting that built, getting a version of Linux with the right drivers together, putting that on the computers (where otherwise the hard disks would come from the hard disk factory with Windows already on them), training sales reps and service and support staff, and so on. Next comes the uphill battle of convincing people that they actually want this particular machine with preinstalled Ubuntu when of the prospective clientele 20% would really prefer Debian and 30% would really prefer Fedora, and those wouldn't touch a Ubuntu preinstall with a long pole as a matter of principle – they could of course install whatever they liked after the fact but then why bother to buy preinstalled? (There is the added complication that due to all this extra work the Linux machine will probably end up being more expensive than the one with Windows, which for a good-quality Linux preinstall wouldn't be a huge problem but would p*ss off all those people who think Linux, being much-hyped “free software”, ought to be free, as in beer, and so the Linux PC ought to be cheaper than its Windows equivalent. – The latter theory leads us to Cyberax's Russia where people would buy a “Linux” PC because it was cheaper than one that came with Windows, and then install a copy of Windows that fell off the back of a truck somewhere.)

It's reasonable to assume that the demand for Linux would be a lot greater if (a) people weren't preconditioned to expect Windows, (b) people were actually in a position to choose between Windows and Linux versions of the same basic hardware, on a spectrum of low-end to high-end machines, and (c) some entity proselytised for Linux to the same degree that Microsoft proselytises for Windows. This seems to work for Playstation vs. Xbox and various other market segments, but unfortunately it is not how the PC world rolls – not because Linux isn't good enough, but because Microsoft had a ten-year headstart in the late 1980s and early 1990s where it was basically the only game in town, and now it is difficult to dislodge. (In the game-console market, Microsoft started as a complete non-entity but managed to butt in only because they didn't need to make money on the Xbox.)

this is sad

Posted Jun 21, 2016 11:45 UTC (Tue) by ballombe (subscriber, #9523) [Link] (1 responses)

It is worth than that. My university procurement rules means that I have the choice between 3 laptops designed to run windows that are completely linux-hostile or to insist on using Mac OS and somehow be waived the procurement rules and be allowed to buy any Apple laptop.

So the users that are running linux on the lab desktop, get a poor experience with linux on their windows laptop and next time pick a Macbook.

this is sad

Posted Jun 21, 2016 15:33 UTC (Tue) by krake (guest, #55996) [Link]

If the Windows laptop options are significantly Linux-hostile, it might be worth checking if the allowed Apple laptops are less Linux-hostile.

While Apple hardware can also be difficult, it might actually be easier depending on how bad the Windows laptop hardware is

this is sad

Posted Jun 21, 2016 11:45 UTC (Tue) by hitmark (guest, #34609) [Link] (9 responses)

The Windows 10 thing has less to do with fear of upgrading, and more to do with people confusing telemetry with spying.

this is sad

Posted Jun 21, 2016 15:44 UTC (Tue) by Wol (subscriber, #4433) [Link] (8 responses)

And if I have to pay for internet access by the megabyte?

WHY do I have to have a computer that *insists* on phoning home and racking up bills, when I see absolutely no use whatsoever for the (lack of) value generated.

Cheers,
Wol

this is sad

Posted Jun 21, 2016 16:17 UTC (Tue) by nye (guest, #51576) [Link] (7 responses)

>WHY do I have to have a computer that *insists* on phoning home and racking up bills, when I see absolutely no use whatsoever for the (lack of) value generated.

If you have no use for those features, you might consider turning them off? As a Linux user, I'm sure you're used to the idea of changing the default settings to ones that you prefer.

[There are certain bits of telemetry that can't be turned off if you're using a Windows Insider build, since the entire purpose of those is to gather feedback.]

this is sad

Posted Jun 21, 2016 21:32 UTC (Tue) by Cyberax (✭ supporter ✭, #52523) [Link] (6 responses)

You can not turn off updates in Windows 10. Let me repeat, you can NOT turn off updates. It can be done through undocumented registry tweaks, but not through user interface.

Windows 10 tries to guess if you're on a metered connection by checking your IP address and doing a tracert (bypassing active VPN, but what's a security violation between friends?) to see if you're inside a mobile network. It usually works but sometimes it doesn't, for example when you're using VPN from your 4G-enabled OpenWRT box.

this is sad

Posted Jun 22, 2016 6:09 UTC (Wed) by zdzichu (subscriber, #17118) [Link] (1 responses)

I'm sure they're first using more standard method, like watching for ANDROID_METERED (and equivalents for other OSes) stuffed into DHCP options.

this is sad

Posted Jun 22, 2016 6:12 UTC (Wed) by Cyberax (✭ supporter ✭, #52523) [Link]

Probably. Except I set up this access point myself on OpenWRT. And then a couple of Windows VMs easily ate my month's traffic allowance.

Grr..

BTW, thanks for the tip, I'll add this option to my config.

this is sad

Posted Jun 22, 2016 10:45 UTC (Wed) by nye (guest, #51576) [Link] (3 responses)

>You can not [easily] turn off updates in Windows 10.

Good. It shouldn't be possible for a clueless user to shoot themselves in the foot by doing something that's such a monumentally bad idea, but the option is available for people who seriously know what they're doing.

>Let me repeat, you can NOT turn off updates. It can be done through undocumented registry tweaks, but not through user interface.

It can be done via Group Policy, which is the standard UI for administrative settings that aren't appropriate for most users. For the home edition, I believe you can't use Group Policy but have to edit the registry directly - Group Policy is to a large extent a graphical front end for the registry that explains the settings and their available options (actually this is the case for much of the contents of Settings and the Control Panel). So using the home edition, you'd need to find documentation online on how to do this.

>Windows 10 tries to guess if you're on a metered connection by checking your IP address and doing a tracert (bypassing active VPN, but what's a security violation between friends?) to see if you're inside a mobile network

No, the option to choose whether a connection is metered is up to the user. By default it's a toggle that can only be set for wireless networks, though there are ways to set it for a wired connection that aren't exposed to the UI (and I think I remember reading that it was becoming standard in the next release, but I could be wrong about that). I agree that it's silly to make the assumption that wired connections won't be metered, but that's a very separate issue.

You might be thinking of Windows 10 Mobile? I've not used that so I don't know how it differs, but I can't find any reference to the behaviour you describe.

I find it so enervating seeing the endless array of baseless FUD on LWN. I wish I could just let go and accept that most of the things people say here are clueless and misleading and that's the way it will always be, but some part of me can't let go of the idea that if just one person can be prevented from unknowingly seeing and believing some falsehood then it will be worth the effort.

this is sad

Posted Jun 22, 2016 17:15 UTC (Wed) by Cyberax (✭ supporter ✭, #52523) [Link] (2 responses)

> It can be done via Group Policy, which is the standard UI for administrative settings that aren't appropriate for most users.
GP now requires Enterprise versions that is not even sold to mere users.

> I find it so enervating seeing the endless array of baseless FUD on LWN.
What, unturnable updates are FUD?

Nope, Win10 is now firmly in "crapware and spyware" land.

this is sad

Posted Jun 23, 2016 10:33 UTC (Thu) by nye (guest, #51576) [Link]

>GP now requires Enterprise versions that is not even sold to mere users.

Yet again, this is not true. Where did you get this idea? Even a cursory Google search would have disproved this.

>What, unturnable updates are FUD?

You keep saying things that are blatantly untrue on a topic you clearly have absolutely no knowledge of at all.
Would you behave this way on a forum where there's a reasonable chance of finding knowledgeable people, or do you just do it here because you expect there won't be anyone with enough experience to call you on your bullshit?

Are there in fact any topics you do know anything about? I've read numerous posts from you on topics that I know little about, that seem informative and with no reason to disbelieve them. Now though I'm wondering if they just seem that way because I don't know enough about the subject to recognise when somebody is making up nonsense, so everything you say has to be treated as probably untrue unless carefully verified.

this is sad

Posted Jun 23, 2016 17:13 UTC (Thu) by zlynx (guest, #2285) [Link]

Windows 10 updates cannot be ignored forever. This is true.

If you have Pro or Enterprise they can be deferred. They can be scheduled.

They cannot be ignored.

This is a FEATURE and I support it totally. I never want to see another Windows 7 laptop used for meeting room video conferences and presentations with 66 pending Critical and Required updates.

If you don't want them installing automatically in the middle of your work then PAY ATTENTION to the popup dialog and say when you want them done. No, you can't keep pushing it off four hours at a time, and if you try they will install anyway. That's on you.

If you don't want updates and you claim you don't need them then don't connect a Windows system to the internet. Now you're safe from updates and security holes.

this is sad

Posted Jun 21, 2016 11:13 UTC (Tue) by krake (guest, #55996) [Link] (2 responses)

> I think the GP meant preinstalled *with Linux*.

I considered that but then the sentence would make even less sense.

How would be agree that the demand for preinstalled Linuxsystems is vanishingly small if there are no suitable options?

Would you agree that the demand for breathable air is vanishingly small based on only medical, diving and space equipment vendors selling it?

Or agree that there is no demand for colored cars based on Ford selling their Model-T only in black?

this is sad

Posted Jun 21, 2016 15:51 UTC (Tue) by Wol (subscriber, #4433) [Link] (1 responses)

"I'm getting really fed up telling all these people that there's no demand for linux pre-installs. You're the tenth person I've had to tell today!"

If you tell demand by recording sales, of course there's no demand if nobody can find what they're looking for!

Cheers,
Wol

this is sad

Posted Jun 21, 2016 16:55 UTC (Tue) by krake (guest, #55996) [Link]

Exactly!
Or if you make it impossible to order even if you find it, like in the case of Dell in Europe.

this is sad

Posted Jun 20, 2016 9:21 UTC (Mon) by callegar (guest, #16148) [Link] (1 responses)

> What you said is true for macOS, too. And it's more expensive in purely monetary terms, too. Still it's vastly more popular than Linux.

I would not say so. If you are on a Mac, then, most of the time you end up buying hardware "for the Mac". Does it cost more than equivalent hardware for PCs? Yes, it does. But then it just works and typically it pleases your eye too.
I never see hardware that is designed to play nicely with Linux and immediately recognizable as such. Even when some piece of hardware carries the penguin logo, most of the time it is a pain. Typically, it comes with some binary only kernel module that works with a single version of the kernel which is never the one you are on and that certainly will not be the one you'll be on in one year. Even things that could be easily done in user space in a future proof way, typically are not or are entangled with GUIs that do not play nice (or coherently) with the rest of your system.

> Not really. Chromebooks are even worse there than Windows - and yet they occupy 7 out of 20 most popular slots on amazon.

Yes, because at least the hardware that is /inside/ the chromebook just works and will keep working when ChromeOS gets updates. With linux it is not like that. Even if I buy a PC with linux preinstalled, I can not be sure of what is going to happen and what is going to break and how much time it will require to repair when I upgrade the OS. Even on a 2008 workstation that I have and that carries a major brand suspend and resume is broken every odd OS upgrade. On laptops you need to be prepared to always have glitches with the graphics card.

For the rest, I see your point, even if I do not 100% agree with it. I think that there are portions of software for which the distro provided packaging will always be better (where integration and coupling between various parts can make a difference) and some items where easing the ISV packaging in a cross-distro fashion can be important (commercial software, large packages that are sufficiently autonomous, things that carry their own package manager like texlive, etc.).

Incidentally, if Android succeeds on the desktop, I fear it to be a bit like Windows 3.1 was in the days of its maximum glory: sort of unstable (all my phones already are, because apps can bring the OS to its knees easily since they are never designed with system constraints in mind), cheap and preinstalled, used by everyone because everyone uses it (and you cannot do differently), not really loved by anyone.

this is sad

Posted Jun 21, 2016 12:08 UTC (Tue) by hitmark (guest, #34609) [Link]

One may also want to consider software.

Once one have had a MacOS or Windows computer for a year or so one has built up a large cache of software that one wants to continue using.

Never mind that i swear Bill Gates himself has gone on record saying he would rather see people pirate MS products than contemplate alternatives.

MS didn't even implement more than a serial number until Windows XP, and then supposedly because otherwise the BSA would have revoked their membership.

this is sad

Posted Jun 20, 2016 10:21 UTC (Mon) by jond (subscriber, #37669) [Link]

What you said is true for macOS, too.
Not so much, these days.

this is sad

Posted Jun 20, 2016 10:32 UTC (Mon) by krake (guest, #55996) [Link] (2 responses)

> What you said is true for macOS, too. And it's more expensive in purely monetary terms, too. Still it's vastly more popular than Linux.

It might a similar situation from the technical point of view but not from the user point of view.

With macOS or Mac computers or Apple in general there isn't an assumption that random things will work.
It is even more or less the other way around, i.e. people will assume it does not work unless the thing in question is either an Apple device as well or specifically blessed by Apple.

People who have switched from Windows to Mac do not assume their old computer will work with macOS, they specifically buy a new one that has it pre-installed.
They will be annoyed if some of their peripherals don't work anymore but buy new ones that do instead of switching back.

So what callegar said is effectively not true for macOS

this is sad

Posted Jun 21, 2016 8:32 UTC (Tue) by hitmark (guest, #34609) [Link] (1 responses)

And that may well indicate a big obstacle.

Most non-techie computer users equate hardware with software.

We have probably all snickered to the tech support tale of "what OS are you using? Dell". Basically most people treat a computer like they treat a TV or similar, is a sealed whole and not a sum of parts.

So if you were to put two computers of the same brand side by side, one running Window and one running Linux, customers may well get confused as they equate the brand with Window out of habit.

Similarly you have HTC used the same look of watch widget for their phones since the Windows PocketPC days.

this is sad

Posted Jun 21, 2016 15:44 UTC (Tue) by krake (guest, #55996) [Link]

Indeed.

Amplified by the continued wrong/misleading/restrictive use of the term "PC" to only mean a Windows PC, when in fact an Apple PC is as much a PC, or one running Linux, etc.

this is sad

Posted Jun 20, 2016 5:48 UTC (Mon) by nhippi (subscriber, #34640) [Link] (3 responses)

> you never know if the scanner, camera, printer, TV card, USB video dongle, CDMA dongle, video capture dongle, home surveillance camera, will work seamless, will work at all, or will require you hours to make it work.

ChromeOS supports none of that yet has beating legacy Linux desktop in popularity. People say they want hardware X and software Y on Linux, yet they happily choose platforms where neither is true (iOS, Mac OS, ChromeOS). I'm making two conclusions

- people prefer a clear "not supported" over "git clone the snapshot, build, tinker variables and it somewhat works"
- uncanny valley effect. When libreoffice looks almost like office but has many stubble differences, people get incredibly frustrated. When you tell them there is no direct copy of office, they find some weird way to workaround ala https://xkcd.com/763/

this is sad

Posted Jun 20, 2016 11:16 UTC (Mon) by khim (subscriber, #9252) [Link]

- people prefer a clear "not supported" over "git clone the snapshot, build, tinker variables and it somewhat works"
For very obvious reason, too: most “Joe Averages” have no idea what git even means (well, they know the meaning: git means git is “a stupid or worthless person”—but how could that help to fix camera driver? they have no idea). Thus for them it's “it does not work” vs “you don't have enough magic points to make it work”. Second is much worse because when you know that something does not work you are at least in the clear, when you don't have enough “magic points” then you have this piece of hardware which is there and tempting… but even so it's uselsss.

this is sad

Posted Jun 21, 2016 12:11 UTC (Tue) by hitmark (guest, #34609) [Link] (1 responses)

Thats because Google have them marketed specifically as Chromebooks. Not as PC (hello Windows) or Mac (hello MacOS) but Chromebooks (hello ChromeOS).

They are effectively marketed more like a phone or TV or some other "device". Not as a PC built from standardized parts (even though they are internally just another laptop PC).

this is sad

Posted Jun 21, 2016 15:45 UTC (Tue) by krake (guest, #55996) [Link]

Very good point!

this is sad

Posted Jun 20, 2016 10:36 UTC (Mon) by krake (guest, #55996) [Link]

> But if you think that these are popular because it's FOSS then you are sorely mistaken: GNU/Linux is used because it's cheap, not because it's open.

I don't think it can be reduced to being cheap and not being open.

Without it being open it would not have spread beyond desktop as no third party would have been able to adapt it to any other needs.

As far as I can tell being open and collaboratively developed were the major contributing factors to having it spread so widely across the device spectrum and not be confined to the original, single use case.

this is sad

Posted Jun 20, 2016 13:30 UTC (Mon) by clump (subscriber, #27801) [Link]

There are about 1% of users who use GNU/Linux today on desktop and many more—on server. But if you think that these are popular because it's FOSS then you are sorely mistaken: GNU/Linux is used because it's cheap, not because it's open. Android is free, too, thus it could easily fill that niche.
And yet paid Linux is doing quite well on servers, as you mention. Dell's Developer Edition/Linux laptops have seen multiple generations. Linux is used for many reasons, and by your own reasoning, "cheap" does not explain why.

There's no argument that the "desktop" is changing. The reason seems to be that many people who barely needed PC's bought them anyway, while in modern times many personal computing uses can be handled by mobile devices.

Mobile devices are abysmal for security, content creation, and lock-in.

this is sad

Posted Jun 19, 2016 13:13 UTC (Sun) by rqosa (subscriber, #24136) [Link] (7 responses)

> about making [GNU/Linux distros] actually relevant for “Joe Average”… will it happen or not?

But are any of the current GNU/Linux distros actually aimed at "Joe Average desktop user"? There's Fedora Workstation, there's Ubuntu… and of course ChromeOS is a Linux distribution too. These are probably the only ones actually competing for this market segment in the first place.

> it's a race: will Red Hat and/or Canonical's desktop Linux distros be able to provide at least some incentive for ISVs before they will all be rendered irrelevant Google's desktop Linux distro (whether it be Android, ChromeOS, or some fusion of the two) wins out in that market segment instead?

That's really all that's at stake in this competition. Meanwhile, other disctributions (RHEL/CentOS/enterprise distributions in general, Arch/Gentoo/hobbyist distributions in general, etc.) will continue to address the needs of their current user bases.

Incidentally, the fact that ChromeOS now can run Android apps just proves what I was saying in my posts in the previous thread that you linked to here: soon, "the only remaining difference between "Android" and "a Linux distribution" will be that "Android" has the proprietary Google stuff preinstalled", and consequently there will cease to be any significant difference between community-developed "traditional" Linux distributions (Arch, Debian, etc.) and community-developed Android derivatives (CyanogenMod, Replicant, etc.).

And finally: bundling/sandboxing tools such as Flatpak are a good thing to have available for certain purposes. For example, it's been said here before that "For any software that [users] do care about (whether that's a browser, email sever, PCB designer, compiler, etc) where they want the latest and greatest function, they want that (and only that) to change" (emphasis mine). So, once Flatpak et al. have matured, I expect that it will be common for Linux desktop users to have the majority of software on their computer be from native distribution packages, and then add a few standalone "application bundles" (e.g. for a bleeding-edge version of some specific program, for proprietary games, etc.) on top of the base system.

this is sad

Posted Jun 19, 2016 22:01 UTC (Sun) by khim (subscriber, #9252) [Link] (6 responses)

But are any of the current GNU/Linux distros actually aimed at "Joe Average desktop user"?

At least Ubuntu is aimed at “Joe Average”—but even then it took many years till they've gotten support for things like Steam (and even then there are still compatibility issues even with supported games).

There's Fedora Workstation, there's Ubuntu… and of course ChromeOS is a Linux distribution too. These are probably the only ones actually competing for this market segment in the first place.
And these are the only ones which provide a way to develop software for them, too. Of course ChromeOS was always crazily limited by what it could do and Fedora/Ubuntu had no stable SDK, but now situation is changing: ChromeOS have finally gotten support for a decent development platform, Fedora/Ubuntu are creating Flatpak and Snap.

It's just funny: Google was incredibly slow, but Canonical and RedHat wasted so much time that these solutions become usable at about the same time…

Meanwhile, other disctributions (RHEL/CentOS/enterprise distributions in general, Arch/Gentoo/hobbyist distributions in general, etc.) will continue to address the needs of their current user bases.

But for how long? If Android will render Fedora and/or Ubuntu irrelevant then what will make it impossible to move it to server? And if server payments will dry up then how long RHEL/CentOS/enterprise distributions in general will be around? And if they will go away then who would fund for the development of software which is then packaged by Arch/Gentoo/hobbyist distributions in general?

This may all sound really far-fetched today, but the whole thing really looks like a repeat of UNIX wars and eventual replacement of “UNIX world” with “GNU/Linux world”… only this time around GNU/Linux world would be replaced with Android…

Sure, GNU/Linux will survive… somewhere on the outskirts… but in what capacity?

this is sad

Posted Jun 20, 2016 1:28 UTC (Mon) by rqosa (subscriber, #24136) [Link] (5 responses)

> If Android will render Fedora and/or Ubuntu irrelevant then what will make it impossible to move it to server?

If some variety of Android does eventually get used on servers, then it will just be used to run the same kinds of server software (RDBMSes, web-application servers, etc.) that Linux runs today.

> And if server payments will dry up then how long RHEL/CentOS/enterprise distributions in general will be around?

There's nothing stopping Red Hat / Oracle / other enterprise-Linux-distro maker from offering an "enterprise / server-oriented Android distribution" if there ever gets to be customer demand for anything like that. (Having access to the Google Play app store is totally irrelevant for the server market.)

> And if they will go away then who would fund for the development of software which is then packaged by Arch/Gentoo/hobbyist distributions in general?

Apart from core system infrastructure (the kernel, glibc, GCC, etc.) and actual enterprise-server software, most of the current hobbyist-oriented desktop-Unix software ecosystem (e.g. desktop apps for GNOME/KDE/etc.) isn't funded by enterprise-Linux-distribution companies anyway.

> only this time around GNU/Linux world would be replaced with Android…

Not everyone wants to use a stock Android system, hence the existence of CyanogenMod et al. And once Android moves to the desktop, there's nothing stopping Arch/Debian/Gentoo/etc. from becoming "third-party Android distributions" like CyanogenMod is. At that point, there's no technical impediment to running Android software on GNU/Linux — the only remaining impediment is getting access to the Google Play app store, and they can work around that in the same way that CyanogenMod does (i.e. back up the preinstalled stuff from stock Android during the installation process).

(Plus, who knows, maybe Google will eventually make the Android app store accessible in a way similar to how the Chrome app store is available on Chrome-for-Linux today. And if desktop-Android actually gets a majority of the desktop market-share without Google having opened up the app store that way, then it's time for an antitrust lawsuit.)

It's inherent in the nature of FLOSS that, once it's (legally) possible to develop a hobbyist-oriented OS distribution, someone will. That's why projects like CyanogenMod and Arch Linux exist today and will continue to exist, without needing to have a large portion of the general public using them.

this is sad

Posted Jun 20, 2016 7:46 UTC (Mon) by alankila (guest, #47141) [Link] (4 responses)

> If some variety of Android does eventually get used on servers, then it will just be used to run the same kinds of server software (RDBMSes, web-application servers, etc.) that Linux runs today.

This is probably true. The key thing for me is the cost and reliability of the hardware. I think that at first, a server like this wouldn't probably be running in a rack in server room somewhere, but it would be more like some RPi-sized box installed in the customer premises that happens to run Android instead of some traditional Linux distribution. These small and cheap devices are becoming powerful enough to suffice to run operations and act as localized hubs that can even control hardware via the GPIO things. And they have very attractive price tags.

this is sad

Posted Jun 20, 2016 8:54 UTC (Mon) by Cyberax (✭ supporter ✭, #52523) [Link] (3 responses)

> This is probably true. The key thing for me is the cost and reliability of the hardware.
Nobody really cares about hardware reliability for commodity servers. As long as it doesn't die every day, you just migrate to another hardware in case of problems.

> I think that at first, a server like this wouldn't probably be running in a rack in server room somewhere, but it would be more like some RPi-sized box installed in the customer premises that happens to run Android instead of some traditional Linux distribution.
Why? The current direction is away from customers' premises to consolidated server farms. And it makes little difference for AWS or Google Compute what OS you're running inside a VM.

this is sad

Posted Jun 21, 2016 23:44 UTC (Tue) by alankila (guest, #47141) [Link] (2 responses)

I think we just have very different kind of deployments in mind. Cloud-style crap is only possible if your networking is reliable. If it isn't, you want locally deployed servers again.

this is sad

Posted Jun 22, 2016 1:02 UTC (Wed) by Cyberax (✭ supporter ✭, #52523) [Link] (1 responses)

At this point in time, it's usually cheaper and easier to upgrade your network connection. Sure, lots of places on Earth still don't have reliable Internet connectivity, but would you place your office there?

this is sad

Posted Jun 23, 2016 0:25 UTC (Thu) by Wol (subscriber, #4433) [Link]

> Sure, lots of places on Earth still don't have reliable Internet connectivity, but would you place your office there?

You mean, like London?

I wish my internet connection was reliable (things like CG-NAT and bufferbloat don't help ...)

For a "necessary utility", the availability of reliable internet is extremely patchy. Fortunately, the government has decided that "broadband for all" is a necessity - until a couple of months ago the phone companies' push for "superfast broadband" quite explicitly ignored the 10% of UK households that don't have broadband at all, it was initially intended as a pure upgrade of existing broadband, not rolling broadband out to the "have nots".

Cheers,
Wol

this is sad

Posted Jun 20, 2016 17:08 UTC (Mon) by rahvin (guest, #16953) [Link]

>As I've said years ago: Linux distros are on borrowed time and clocks are tickling.

Your timetable is wrong because your assumption is wrong. What's happening isn't an entire market shifting, but a fragmentation of a market that was serving participants poorly. The prior market we all lived under is that computing resources came in one generic variety (desktop/laptop) that could do everything. What we are moving to is fragmenting out the content consumption users into a separate hardware/software paradigm targeted specifically to consumption which does other things very poorly. Consumers of content don't need mice and keyboards. They don't need much input capability at all.

There is certainly a portion of the residential consumer market that doesn't need general purpose computing and will be quite happy only with a device targeted at consumption. Personally I expect a least a 50% decline in PC sales that will remain permanent. There is also a secondary issue in that computing power has now exceeded most people's needs (This may or may not change with a killer app like VR). The result being that those people who still need the large general processing of a computer/laptop are refreshing their hardware at a much slower rate, moving from 3-5 years to 5-8 years. This is mostly because PC's have stopped improving, for the most part you can't tell a computer built yesterday from computer built 5 years ago just by using it.

I believe the idea that PC/Laptops are going away is an overestimation of these two effects. There are far too many computing uses that need general PC's and mice/keyboards that are not going away. Most of the people with phones/tablets and even chromebooks still have general PC's they use, they just don't use them for the tasks the other is good for, such as content consumption. This is one of the benefits of fragmentation, as you end up with more total sales overall, but the bigger problem for PC's has been the stagnation in processing power and the lack of software that needs it. Unless this changes the PC refresh cycle is going to keep stretching out and will result in real harm to the computer market. We've already seen a significant price increase at the high end PC due to this.

this is sad

Posted Jun 18, 2016 13:01 UTC (Sat) by NAR (subscriber, #1313) [Link] (3 responses)

"security vulnerabilities in poorly-maintained packages."

Do I care if there's a "security vulnerability" in the DVD authoring software that I've just used? No. Do I care if there's a "security vulnerability" in the CD copying software I've just used? No. Do I want them to break because a Firefox upgrade pulls in upgrades with incompatible changes in some shared library that these applications use? Hell no. I rather have 10 versions of the same DLL on my system if the 10 applications using them work than hoping every time I upgrade something something else breaks.

false dichotomy

Posted Jun 18, 2016 13:28 UTC (Sat) by HelloWorld (guest, #56129) [Link] (2 responses)

It's entirely possible to have the best of both worlds nowadays, just because library dependencies can be shared doesn't mean they have to. Flatpak otoh forces libraries to be bundled.

false dichotomy

Posted Jun 18, 2016 15:55 UTC (Sat) by drag (guest, #31333) [Link] (1 responses)

Flatpak forces bundling of libraries that are not part of the runtime. Runtimes get shared. I expect as it matures you could combine runtimes. Whatever ostree supports.

false dichotomy

Posted Jun 18, 2016 20:56 UTC (Sat) by alexl (guest, #19068) [Link]

ostree deduplicates/shares any identical files. Of course thats not the same as one being updated when the other is...

this is sad

Posted Jun 18, 2016 16:01 UTC (Sat) by josh (subscriber, #17465) [Link] (20 responses)

> I had hoped that distributions would eventually stop distributing applications and rely on cross-distro packages instead

This will never happen as long as distributions have different policies. The value of Debian isn't in the .deb format, it's in Debian Policy.

this is sad

Posted Jun 18, 2016 18:51 UTC (Sat) by amacater (subscriber, #790) [Link] (19 responses)

The value of Debian is that it cares about dependencies and dependency chains - and licences. Policy and developer practice helps reinforce this.

Ubuntu's Snap format and Flatpak and the others that seek to produce "all dependencies in one place" are fantastic if you want e.g. Firefox on a phone and that's all you want.

Anywhere else, they seem large, bloated and don't serve to produce a better ecosystem once you've got several of them around.

Fundamentally - you've got to understand what your program needs.

this is sad

Posted Jun 19, 2016 2:58 UTC (Sun) by drag (guest, #31333) [Link] (4 responses)

If you look you'll see that they are not substantially larger then their rpm/deb equivalents if done correctly. Recently they re-did the Libreoffice snap that took the size down from ~1GiB down to 280-ish MiB or so. That is only 70-ish megs larger then equivalent rpm or deb files and even then those don't include Java, which is necessary to unlock all the Libreoffice features.

this is sad

Posted Jun 19, 2016 8:56 UTC (Sun) by lsl (subscriber, #86508) [Link]

As it is, the flatpak variants are also strictly less useful. With a standard LO installation, I can easily use its filters from a script to convert weird document formats to something more usable. I can just install a JRE to unlock the remaining features. Does the latter work with the Flatpak? What would I need to go through to achieve the former?

LibreOffice is already the kind of software that's a pretty good match for what Flatpak can do. The utility of many other programs will be much more severely restricted when decoupled from the system.

Interaction with the system is a desired feature of most software. These are not smartphone apps. It's also the hard part of any sandboxing effort. Annoying the user with thousands of file picker dialogs is not going to cut it except for the most trivial of programs.

this is sad

Posted Jun 19, 2016 8:59 UTC (Sun) by nim-nim (subscriber, #34454) [Link] (1 responses)

They're comparing pathological cases, apps which have been decades in writing, with lots of specific private code, and which have been forced to sanitize their requirements due to years of distro pressures.

That's pretty much the ideal case for flatpack or snap (I suspect just taking apache openoffice instead of libreoffice would be quite less rosy).

Even there they have ~ 30% of wastage.

That's the usual "trade dev convenience for everyone else's" and "cargo cult windows/apple/android to win market share¹" (the only change is who is being cargo culted, usually via themes of other superficial things)

It will probably end with the usual "dev does not understand why no one follows him".

I just hope they won't damage too much the platform in the meantime.

¹ Years of efforts poured in the Linux desktop, with lavish use of distro gifts and exceptions, have produced little in terms of market share. Meanwhile the people who focused on doing their own thing instead of aping the mistakes of others have actually found their way to people homes (in appliances, in owncloud, etc). With things like Office 365 and skype Microsoft is actually on the way to the network desktop that could have been done years ago Linux side if people had not focused on burying its *nix network heritage and copying every half-hassed network-ignorant Microsoft or Apple tech they could find out.

this is sad

Posted Jun 19, 2016 9:46 UTC (Sun) by halla (subscriber, #14185) [Link]

"trade dev convenience for everyone else's"

It has nothing to do with "dev convenenience" -- I would prefer not to have to make binaries, it's a timesink. And "everyone else's" only applies if by "everyone" you mean "everyone but end users".

this is sad

Posted Jun 19, 2016 21:03 UTC (Sun) by callegar (guest, #16148) [Link]

Doesn't this just prove that you can bundle all libraries in DEBs exactly as the snap packages do? This comparison is against the LibO provided DEBs, that bundle a lot of stuff. If you compare with the Debian provided DEBs, I think you get 280MB vs ~100MB, which is almost 200% larger.
Increasing size is not such a big issue for the disk (you can get a bigger one), it is an issue for your cache memory.

this is sad

Posted Jun 20, 2016 0:47 UTC (Mon) by zblaxell (subscriber, #26385) [Link] (13 responses)

Debian's APT tool is an awesome technical solution to an important problem that few others have attempted to solve, much less successfully solve as well as Debian has. apt-get scales up to distributing and installing 5-digit numbers of packages by 3-digit numbers of developers to 8-digit numbers of users, and Debian does an admirable job of ensuring continuous service from devices running their software. I've used it while wearing developer, administrator, and end-user hats for twenty years, and for what it does, it's far superior to anything else.

The problem is that APT isn't a solution to a problem anyone not participating actively in the Debian/GNU project expects to be solved. It's a workaround at best, and it's a failure (i.e. the opposite of success) when measured against several important performance standards.

APT needs minutes of CPU time on a PC-class machine to navigate through a carefully curated (i.e. expensive and error-prone) database of packages labelled according to their dependency relationships. A single-package install on a smartphone-class machine could easily drain a fully charged battery just trying to solve the dependency graph, let alone doing all the IO to perform a non-trivial dependency upgrade.

Assuming APT doesn't finish a package install request by saying "sucks to be you, I can't figure out how to install that," the result can still be disappointing. The configuration APT generates can be specific to an individual user or server instance, which has severe QA and support implications even when it doesn't fail partway through in some disastrous way (which is a thing that happens sometimes, and must be planned for).

In order to get any sort of reproducible behavior and mitigate the technical risks, one has to create chroots and use filesystem snapshots to replicate the various possible development and runtime configurations, back out of bad install/upgrade disasters, and sometimes even to run the applications as well. This would all be worthwhile to build a unique custom setup or development prototype, but in most cases the intended result is to just run a configuration similar to some well-known template shared by thousands of users--and a monolithic binary download is a much better way to get that in practice. Once you've crossed the line and implemented containers or VMs to resolve dependency issues, APT is only useful as a dumb package downloader.

APT edits the runtime in ways that carry elevated risks of catastrophic failures (e.g. bricked device) compared to its competitors based on mostly read-only (i.e. monolithic "system" image) or logically isolated components (i.e. chrooted containers). Often the last successful interaction an end-user has with the device is answering "yes" to the question "Do you want to install this upgrade?" and that's not a story anyone wants their end-users to tell from a first-person perspective.

APT relies on its maintainers to provide a detailed catalog of all the non-trivial dependencies between packages as they are found "in the wild." Everyone else considers the wildness to be the root cause of their problems, and tries to solve that instead. They integrate dependency packages into a single domesticated whole, simplifying the dependency graph down to a handful of (possibly very large) nodes. Ruthless decisions are made to control the size of the dependency graph: we will have feature X, feature Y in the next release, feature Z is deprecated and users will be forced to upgrade or opt out of accepting future releases, no inter-package dependencies are permitted without the permission of the vendor, etc. Vendors didn't invent these strategic restrictions because they're greedy or evil--they *discovered* them, because they're trying to build products that work well enough to sell.

Using apt-get requires specialist skill, and that skill isn't getting any cheaper over the decades. Today, machines range in price from $5 to $1500, and most users' data is worth not more than $100 (thanks to ransomware vendors for researching this for us!). APT easily absorbs $1600 in support costs when something goes wrong, making it more expensive to fix than to throw away. There's no guarantee that spending $1600 will fix APT either, so throwing it away is a more predictable way to spend money.

The one advantage APT used to have was the ability to do in-place live upgrades on a server with a high probability of success. Debian used to be good at that sort of thing, but they started failing in 2008 and haven't gotten better since. If every upgrade carries the risk of a service interruption, it's both cheaper and safer for a server operator to just start over with a new machine, validate all the old applications on a completely clean install, and cut services over from the old machines at some network layer above the individual machines. This is now SOP in the industry, and it means that even if Debian and APT did still support live upgrades well, nobody needs it any more.

APT could have a place among the set of users who will spend thousands of dollars worth of their own time to upgrade a $20 router into a $200 one. There are vendors who serve those users (Raspberry Pi is especially successful here), but even these vendors tend to build monolithic device-specific images on SD cards instead of integrating their changes into the base distribution.

this is sad

Posted Jun 20, 2016 2:02 UTC (Mon) by pizza (subscriber, #46) [Link] (6 responses)

> APT needs minutes of CPU time on a PC-class machine to navigate through a carefully curated (i.e. expensive and error-prone) database of packages labelled according to their dependency relationships. A single-package install on a smartphone-class machine could easily drain a fully charged battery just trying to solve the dependency graph, let alone doing all the IO to perform a non-trivial dependency upgrade.

You both vastly overstate apt's CPU requirements and understate modern smartphone processing and I/O capabilities.

And a random apt package's I/O requirements are most likely better than a random android app's requirements, especially when you consider that modern Android actually *compiles* the app upon installation, and nearly everything interesting going on under the hood (on Android) happens via bog-stock sqlite databases.

this is sad

Posted Jun 20, 2016 4:42 UTC (Mon) by zblaxell (subscriber, #26385) [Link] (5 responses)

> You both vastly overstate apt's CPU requirements

You're right, I fudged apt's CPU requirement, but not in the direction you're thinking. I'm also aware of current smartphone processing capabilities even though I didn't state them.

I've seen apt keep an 8-core 3.2GHz machine with 24GB of RAM busy overnight trying to solve a dependency problem. That's with just 22,000 packages on a machine many times more powerful than a current smartphone. Feeding a million apps to apt (assuming we ignore the insanity of downloading a million-app package database) would crush its tiny brain.

sqlite trickle-writes a few pages at a time, usually in response to external stimuli (i.e. it's rate-effectively rate-limited by the user). An apt-get upgrade is gigabytes of I/O all at once, and nothing works right until it's done.

this is sad

Posted Jun 20, 2016 12:28 UTC (Mon) by pizza (subscriber, #46) [Link] (4 responses)

> I've seen apt keep an 8-core 3.2GHz machine with 24GB of RAM busy overnight trying to solve a dependency problem.

That sounds more like there's something wrong with the dependency tree due to broken packages.

Granted, some dep solving algorithms are better than others (witness the significant improvement that dnf brought over yum) but even back in the yum days it still took on the order of a couple of minutes to do the depsolving of a major distro upgrade. (And I might add that apt was noticably faster than yum when doing the same sort of thing..)

Still, this isn't exactly an apples-to-apples comparison; your typical GNU/Linux installation has an order of magnitute more software components installed than a smartphone.

And, I might add, after even a minor point update of Android, the system has to "optimize apps" for the new OS (aka "recompile the bytecode") which takes my tablet about 30 minutes for under 100 "apps", leaving the tablet considerably warmer in the process. There's quite a bit of crap going on there.

this is sad

Posted Jun 20, 2016 17:18 UTC (Mon) by zblaxell (subscriber, #26385) [Link] (3 responses)

> Still, this isn't exactly an apples-to-apples comparison;

If you want to understand why users want apples instead of apt, you have to compare apt to apples.

> your typical GNU/Linux installation has an order of magnitute more software components installed than a smartphone.

The "order of magnitude more software components" is the problem statement. Apt is a symptom of this problem, not a solution. Complexity is not competitive.

If I pick a random desktop machine, it has 128 leaf-dependency packages installed, not even double the number found on a modern smartphone (certainly not an order of magnitude more). Each of these is effectively stand-alone with one or two big dependencies (e.g. ten of them are bits of KDE, one is all of LibreOffice, one is most of xorg, one is Chromium, ssh, rsync, etc). Like a modern smartphone, almost half of these packages turn out to be games.

Apt tracks 2484 separate package objects to support the top 128. If I pick a random package and try to upgrade it, apt informs me it breaks 56 other packages immediately, and those have dependencies too. I've done this dance many times before. There will be well over a thousand packages upgraded before it's done, and I'll have to intervene multiple times during the process as apt will pick a solution from 525600 candidates for upgrading this desktop that isn't the one that Debian-QA tested, so it doesn't work.

Most users (myself included) won't do such an upgrade, because the cost and risk just aren't competitive. It's easier, faster, and more reliable to just debootstrap in a chroot or a new root LV and rebuild the local site configuration manually (that's what tools like etckeeper and puppet are for, after all). This will create a new tree of 2484 moving parts that will, unless there is a radical change of direction in the way Debian manages dependencies, also be too expensive to maintain.

This mess is what failure to say "no" to complexity looks like.

this is sad

Posted Jun 20, 2016 17:48 UTC (Mon) by pizza (subscriber, #46) [Link] (2 responses)

Okay, so you're advocating for the "just replace the system" definition of an "upgrade".

That's all fine and dandy. Except that, in order to set up the system to begin with, someone is still going to have to deal with that complexity. And end-users are going to have to live with whatever choices were made for them.

this is sad

Posted Jun 20, 2016 18:05 UTC (Mon) by rahulsundaram (subscriber, #21946) [Link]

FWIW, there are several projects that already do transactions upgrades. For cloud instances, CoreOS is popular. There is also

https://fedoraproject.org/wiki/Workstation/AtomicWorkstation

Pushing binary deltas of images is easier. Layering packages into a underlying image based update systems is a bit more tricky but it is feasible to do.

this is sad

Posted Jun 20, 2016 19:25 UTC (Mon) by zblaxell (subscriber, #26385) [Link]

Not really advocating, just clarifying the motivation behind the simplified dependency model used in projects like flatpak, Limba, etc instead of apt, libsolv, etc. The optimal model is probably somewhere between those extremes: a mixture of aggregation to simplify the dependency graph for common cases, and isolation to allow individual applications and the host system to resolve dependencies without disrupting each other.

It's a safe bet that most users do want to enable someone else to make an apparently monolithic supportable package (or least an aggregation of components that behave as a single unit) out of thousands of mundane technical choices on their behalf. Those that want to make choices themselves self-select themselves away from the rest of us--they just build what they need from source without bothering anyone.

this is sad

Posted Jun 20, 2016 7:39 UTC (Mon) by niner (guest, #26151) [Link] (5 responses)

You start with the assumption that APT is the pinnacle of dependency managers and conclude that if APT has performance issues, the concept as a whole cannot work. Your conclusion is based on a flawed assumption and therefore wrong as well. I'd guess that openSUSE's libsolv is the current top of the line dependency solver which is the reason for why it's not only used by openSUSE's zypper but also by Fedora's dnf and probably others. Until you demonstrate that even libsolv fails in your scenario, your argument stands on a very weak base.

this is sad

Posted Jun 20, 2016 11:07 UTC (Mon) by nye (guest, #51576) [Link] (3 responses)

>Until you demonstrate that even libsolv fails in your scenario, your argument stands on a very weak base.

The problem has been proven NP-complete, so it doesn't matter how many times you cry "but that's only one example!".

Until you demonstrate that P=NP, your argument stands on a very weak base.

this is sad

Posted Jun 21, 2016 11:00 UTC (Tue) by niner (guest, #26151) [Link] (2 responses)

Theoretical limits are irrelevant when it does work well enough in practice. If the dependency solver performs well with millions of packages and comparatively small dependency trees, it doesn't matter if it would take years with a hundred times more candidates. IOW it doesn't need to solve the general problem, only the specific one you encounter in practice.

If it were different, people would not even have to bother trying to write something as widespread available as a syntax highlighter for highly dynamic programming languages like Perl or maybe even C++. Because to fully understand all parts of the source code, they'd have to solve the halting problem. So they don't. They do not pretend to handle completely arbitrary input programs but only those that matter in practice.

this is sad

Posted Jun 21, 2016 12:12 UTC (Tue) by nye (guest, #51576) [Link] (1 responses)

>If the dependency solver performs well with millions of packages and comparatively small dependency trees

That's a fairly major 'if'. It's already not too great with tens of thousands of packages[0]; I don't believe it can ever scale another three orders of magnitude (which is what would be required to be truly mainstream) without a change in approach.

If I understand correctly, package managers have already given up on finding 'optimal' solutions[1] and now use heuristic SAT solvers, but regardless, at some point distros have to let go of the idea that the installability of a given package is a function of *every other possible package*. Eventually to continue scaling there simply has to be some demarcation line that lets the solver prune the tree to something small enough to keep dependency resolution tractable. One way to do this is by separation of a core platform from leaf packages that don't (or rarely) have complex interdependencies, which coincidentally happens to describe how almost every end user actually wants a system to work.

[0]in the sense that both apt and aptitude can take a while to find a solution, especially on a less powerful machine, and aptitude will happily suggest a seemingly endless selection of terrible 'solutions'

[1] in the sense of calculating the minimal set of changes required to satisfy the request

this is sad

Posted Jun 21, 2016 12:22 UTC (Tue) by nye (guest, #51576) [Link]

>If I understand correctly, package managers have already given up on finding 'optimal' solutions[1] and now use heuristic SAT solvers

Just to expand briefly on this: I can't find a quick definitive answer in five minutes to whether libsolv, as used by zypper, is a heuristic solver or a pure solver, but there are unavoidable problems either way: if it's pure then it can't scale efficiently; if it's heuristic then it can't guarantee optimal results, and could potentially fail to find possible solutions.

Reducing the problem space by dividing it into smaller sets with fewer interdependencies is a solution to both.

this is sad

Posted Jun 20, 2016 12:13 UTC (Mon) by zblaxell (subscriber, #26385) [Link]

Even if you made the cost of solving dependency graphs zero, it wouldn't solve the other problems, only work around them.


Copyright © 2016, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds