GTK fesses up – this ain’t for you; Qt takes over the world
LWN.net’s Nathan Willis, who previously covered this blog’s viral Arch’s Dirty Little Secret article a few years ago with unusual courage and honesty, has an article back from August which covers several talks at GUADEC 2013, wherein lead GNOME developers talk about the limited uses and ill future of GTK.
In my clear view, the Red Hat corporation has declared itself sole owner of the community-developed GTK project, and is driving it into the ground, making it unusable, probably at Google’s bequest. Their greatest vision for it is making a desktop clock. Any apps larger than that are pushing the usability envelope. GIMP, the original creator of GTK, need not apply.
Meanwhile, Linux developers are flocking to Qt. Yet it should be noted that as soon as Digia aquired Nokia’s Qt, they pledged to become Google’s bitch everlasting. Today, they’re very excited about Chromium. They are controlled by large corporations who make all the decisions and decide the directions. Where do you think that will lead? Why do you think Google didn’t buy Qt themselves? Short of cash? Why use a pawn like Digia?
To me, all of this powerful corporate drive to support ‘cross-platform’ development is merely a game to turn Linux into Windows – to make it so it doesn’t matter what you run, you’re still running a Google product. Google is the new Microsoft. It amazes me how many Linux users think Google is their friend. The Linux community has really become nothing short of stupid, absorbing corporate press releases like populations absorb propaganda. They can’t see even the most obvious attacks, and give their full support to their own demise.
I think it’s safe to say that any spirit of freedom and diversity that once drove Linux is dead. The new people entering the realm of development in Linux are just Windows developers looking for a larger base and more money, or simply corporate whores tearing it apart for short-sighted, malicious goals (which they themselves understand very poorly). They care not for any of the principles that made Linux what it is, or was.
So Linux has been lost because the community has failed to protect it and help it grow. And this isn’t just about toolkits – the infection goes deep into the kernel, udev, the init system, and other areas. In the next few years any remaining GNU Linux users who even know what a principle is, will need to find a new home.
Meanwhile, while you still have a non-Google-implanted brain, you might want to try to figure out why corporations want to (and have always wanted to) completely control the software and abilities of your computer. And you might want to consider differences between Windows and Linux beyond how widgets look. They once represented very different visions of the personal computer.
Updated reading:
53 Comments
Sorry, the comment form is closed at this time.
IgnorantGuru's Blog 
To make matters worse, Debian is on the way to adopting systemd as a defaut init system.
The discussion of Technical Council is still ongoing, but it looks like the choice is going to be systemd, ‘because of it’s technical superiority’.
http://www.phoronix.com/scan.php?page=news_item&px=MTU3NDc
http://bugs.debian.org.nyud.net/cgi-bin/bugreport.cgi?bug=727708
Just some discussions:
http://lwn.net/Articles/578208/
http://lwn.net/Articles/575672/
It is funny that on http://bugs.debian.org.nyud.net/cgi-bin/bugreport.cgi?bug=727708 some people were mentioning that GNOME needs logind that needs systemd as an argument for systemd.
Also, IG, I think that you’re being too pessimistic. I think that there are enough of sane people that there would still be thriving community-driven projects such as OpenRC and probably a handful of GTK and Qt alternatives.
Even if when Linux is being de facto ‘windowsized’, I think that in absolute numbers the free software community projects are thriving.
Maybe someday we’ll see something like 95% Linux-based ‘windowsized’ desktops and something like 2% true Free Software desktops. I don’t think it is a lost war to fight for this 2% mark. The fact that we’ll be able to use true Free Software when we really need it, in times of hardship and struggle, is a huge boon in itself.
Yes, Debian is going to systemd, which is not surprising. While Debian has some strong freedom-loving influences, it too is ultimately corporate-controlled, including via universities. The university scene used to provide a degree of resistance, but universities are of course in corporate pockets, especially these days. As would be expected of a corporation the size of Google (not to mention who really controls Google and why – see NSA spying and the myriad of other links between dark agendas and Google), their infiltration and destruction of Linux is very well thought-out. At this point I view it as history – past the point of stopping it.
Aside from the mere stooges who simply support whatever their masters tell them to (eg this jackass from Stanford), the arguments over the technical merits of systemd are ridiculous and beside the point. Of course it works well in some ways – THAT’S THE BAIT, IDIOTS. Without bait, why would they bite? Yet all they can manage to argue about is how the bait tastes (same for Qt fanboys). It’s looking beyond the bait to THE HOOK that requires actual intelligence, and that is very lacking, especially in those making these decisions. IOW the upstream practices of systemd devs (more Red Hat developers), and the way they’re creating a monolithic system stack that virtually no one can escape from is simply beyond the IQ of these people.
I’d say this is a good time to archive GTK2 and other tools made back when tools were community-developed. You can say that all of this can be forked, but the reality seems to be that little effort or organization exists outside of corporate (Google) control. Linux is owned. That said, maybe some interesting kernel and toolkit forks will eventually grow out of this, but it will be on a small scale. Either say goodbye to large-scale polished tools, or become a de facto Windows/Android user. Which many Linux users already are – there is no sense of values at all. Just controlled users (including the developers). Do you know what universities now use to teach beginning programming? Adobe Flash.
Using Debian is certainly convenient, but they’re not going in any direction I care to follow. I would rather abandon a lot of the polish and start moving toward a BSD variant, or some fringe Linux-like kernel and tool set that is actually open. What it comes down to is who is making the development decisions. In mainstream Linux, it’s now all corporate.
Have you used BSD already? I’ve also been pondering such a switch… but I haven’t tried it out and am unsure how well it “just works” for normal desktop user purposes.
PS: A while back I was curious to see what’s up with Minix (and other such microkernel projects), but it seems still a research project and not useful for everyday “typical user” purposes, alas. (E.g. still no functional graphical browser.)
That’s the challenge – if you don’t use mainstream polished Linux, it’s like maintaining an early 20th century motorcycle. Good luck getting anything else done; just keeping a running system will take all your time.
And they know this. This is why as soon as Linux became popular on the desktop it came under deep attack. It’s funny how many Linux users wished for its popularity. Now they have lost it via that popularity.
Yet it’s getting so ridiculous that I would rather play with fringe systems and waste some time. At least you keep learning that way instead of being a brain-dead Windows user. Either way it takes a lot of time to maintain these systems. Windows (and these days, Linux) takes time because it’s garbage. So might as well spend time on something interesting.
I did try FreeBSD, and I liked it quite a bit. It reminded me of Arch Linux, and resembles Gentoo. I had some problems with proprietary drivers of course – another catch-all for drawing users into mainstream OSes. But at some point I will be willing to replace hardware, or it will expire. My NVidia card, which was one of the problems, has since died, so I’m down to one driver for my laser printer.
The other problem with FreeBSD is that many people run Linux libraries and software on it to have complete software. So there is a connection there, and I wonder how pure it will remain. Not to mention the large Apple connection to BSD, Apple being at least as ugly as Google.
The Linux hurd kernel sounded promising too when I took a look, but like you say these are largely research projects. Trying to use them on a daily basis is much more time and energy consuming (sort of like Linux used to be, hmmmm). It takes quite a team to develop and maintain something like the Linux kernel or GTK. And such a team has to be impervious to infiltration and division – a common tactic when corporations decide to take over free work. Debian’s fail being a good current example, as well as GTK.
The world of Linux is a microcosm of the larger world and all its ills. Same diseases – centralized control, loss of freedom, corporate ownership of everything, etc. In a word, corruption – so its hard to fix it. I guess we need to build better things for ourselves instead of relying on corporate masters. Easy to say but takes real work, and most people are not willing to invest in that kind of thing. That’s why they lost Linux. Now it’s just another disposable ill-working corporate product, let’s face it.
You have given no arguments in favour of keeping SystemV. Here’s why systemd is right for Debian:
* Every service restricted by cgroups, namespaces
* Prevents memleaks from affecting the system
* Prevents services from sniffing devices
* Restricts capabilities of services to prevent system compromise
* Keeps all services jailed without requiring a chroot or different IP
* Every service prioritised appropriately for oomkiller
* Every service protected by the software equivalent of a UPS against packet dropping
* Replaces (x)inetd
* Efficient spawning of processes
* Dedicated daemons restarted upon crash
* True SELinux support, without a stupid hack
Please tell me how SystemV is superior, regardless of who wrote the code. Also, please show me how systemd is non-free.
This is also what I question.
Sure, systemd is coming from one source and hence susceptable to upstream control over acceptance or not acceptance of patches, but its license is not preventing a fork. Look at Mate, Nemo etc.
I don’t think that in next years the Linux users will have to seek other system. That’s groundless fear.
What I think will happen, is that GTK and Gnome will be forked for the very same reason of having the features that actually are interesting to the users.
It would be a fatal mistake to even accept such thoughts into mind as “giving up” on GTK, Gnome, init system.
I know you wrote a udisks replacement with sole motivation to make it a systemd-independent component. That’s FINE, even if runs via insecure SUID, where systemd is exactly against that. Its FINE, it gives choice, should systemd fsck up we will fork it or switch back to components, that people like you create. Please do not even think of giving up, just stay aware and share your thoughts.
Thank you!
I agree that, unfortunately, GTK+-3.0 goes wrong but, practically speaking and in spite of suspicious speculations, Qt seems quite promising to me. I’ll be happy if “Linux developers flock to Qt” because it’s much better than GTK+-3.0.
That’s called bait. Now find the hook.
Obviously corporations like Google have the resources to produce sophisticated toolkits with lots of bells and whistles. But they make all the decisions (regardless of the nonsense they feed you about it), and will drag you where they want, now and in years to come when you’re fully invested in their tools. They’re thinking way ahead, don’t worry. Just because it works as you want now, doesn’t mean it will tomorrow, or that you’ll have any say in the changes.
As developers or users we all like easy, powerful tools, don’t we? Great bait.
Anyone who imagines that Digia or Google or whoever cares one bit about software freedom is seriously deluded, and seriously abandoning their principles (if they ever had them). But nothing you’ve said indicates you care anything about that either.
Now, this seems to be rather an emotional speculation, which could be good just as a warning if no mistaken for a fact.
And your comments could be mistaken for something besides a fanboy’s cheerleading. So what? Learn some history and maybe my thoughts won’t seem like mere speculation to you. But it’s obvious your only purpose is to tell us how wonderful Qt is, no thinking required.
Google is your best friend. Go back to sleep.
As Debian found, it’s impossible to even have an intelligent conversation that discusses a topic anymore. Just ridicule and nonsense denial. Corporate whores have inundated the Linux development scene so much that it’s the only point of view heard. Well I’m giving another point of view as a free software developer who is NOT a corporate whore. Too bad.
Sorry, I thought you appreciate a rational discussion. The words you chose made me realize I mistook your intention.
>Sorry, I thought you appreciate a rational discussion. The words you chose made me realize I mistook your intention.
Come on, Tsu, I’m sure IG didn’t really mean it.
Being sincere is of more importance than being polite.
If you have left discussion due to being pissed off, please return, as OP critic was not pointed upon you, but upon a greater enemy behind bait.
If you don’t come back, then it would mean you have no more arguments.
I’d suggest to push on Sta.li and efl, if efl wasn’t going to eat more c++ for cairo and wayland (which, though, is the only decent software produced by RH nowadays… Yep it’s the xorg you’ll have to use with Intel-NSA graphic cards in future, but it remains good software all in all). My life seems too short to work on every fucking single thing has to be worked to have a free and the least dumb operative system. Depressing.
> My life seems too short to work on every fucking single thing has to be worked to have a free and the least dumb operative system. Depressing.
This is due to the insanity of the process. There was some really smart stuff going, and if it had been allowed to evolve, we wouldn’t have to be wasting time reinventing buttons in 2014. Some of my users still use console mode! We don’t need all these endless cycles of so-called ‘new stuff’. What we need is some intelligence and building to last. Just like in physical manufacturing, a false economy is very expensive. That’s what gives us the feeling that no amount of effort could change it. But it’s more of a social problem than a technological one. That’s why we failed to solve it.
Just like in automobile engines you see ridiculous internal combustion engines still being used simply because of a corporate control structure, this kind of control creates an artificial technological ceiling that continuously depletes resources, in this case human resources. That is depressing (literally, it depresses resources). It inhibits growth in the way of a poison. Who Killed The Electric Linux?
So most developers do feel depressed about developing in such an environment. It’s monotonous instead of genuinely creative. My bags are packed. Not that I’m racing about it – too lazy for that. But clearly this ship be going down. You’ll believe me when you feel the chill of water on your ankles. :)
> “Intel-NSA graphic cards”
Please provide source.
Proprietary source micro-code is closed.
I’ve been thinking about all of this discussed here too. I have used Linux exclusively for about 14 years now and all this nonsense is making me, well, quite sad. As a result, I pretty much have decided to drop Arch in favour of Slackware or if that won’t work out well, go FreeBSD. I’m not sure how Slack and Gentoo are going to do in this mess, I guess we’ll find out in somewhat near future.
All in all, I’ve pretty much lost every bit of interest in what is happening with init systems, sound servers and big corporations, as well as modern computing in general. All of that is ruining both my hobby and my job along with my favourite OS and kernel with the daunting atmosphere surrounding.
So, that said, I have moved my interests back to my trusted old Commodore 64, and projects like AROS. At least I can have fun with those. My PC is pretty much left running XBMC and doing some cross compiling, and it’s fine by me. Anyway, some sort of split will happen in Linux world, for sure. It’ll be very interesting to see what future will bring us regarding this matter. Hopefully something interesting (and sane) will come along. Projects like stali certainly seem interesting.
Ah the golden age of the 80s. People who didn’t live through it don’t get it – I pity them. I too had more fun on my Vic-20 with all of its 3.5K of free RAM, which is a sad statement to me.
To think that back then we actually thought computers had a future. Never saw Idiocracy coming.
It is sad. To me it feels like everything that was fun, inspiring and exciting with computers has since died and disappeared almost completely. It probably went into the same place where common sense and individual though have disappeared too. It’s all about Facebooks and such nowadays, and hardware itself has lost it’s meaning. Beauty of design and engineering is not relevant, as long as these devices give people what they want in a most braindead way and produce as much money as possible and optimally even more.I think this all in a way is a side effect of our whole civilization going down. Corporates are milking the last drops out of us and I don’t think there’s nothing that can be done with it especially since it looks like people are raised to not think from ground up. Not going deeper into that somewhat off topic subject, but I would like to encourage people to get more involved in the machines of the past, and the few interesting projects still respecting certain ideals. The fun is still there and who knows, maybe it inspires one to do something actually productive (with modern hardware), instead of fighting with shitty toolkits and web browsers thousands of times bigger than a fully functional operating system.
Anyway, it was quite the amusement to see the look in the face of my co-worker (A young C# developer) when I talked about tape drives and all the other awesome stuff of the past systems to him. Probably thought I had gone even more wrong in my head. I think it was around year 1998 when I got my first Intel hardware. My Amiga 1200 motherboard had given up and I did not have the funds to either get it fixed or buy a new one. Instead I got an old PC from a friend practically free and settled with that one. It took about 5 minutes to realize Microsoft way was not meant for me, so I ended up installing Red Hat 5 (?) and been Linux user since then. I had previous experience of BSD so I was quite happy and somewhat impressed with Linux back then. It did not take that many years forward when it started getting obvious how things would go with Linux when the big corps started to get interested and devs from Windows side started to come pouring in without any standards or respect (of course not everybody was like it). Still, I never figured that things could go this wrong. Back when one particular init system hit Arch, I remember thinking “Lucky we have Debian around if this thing is forcibly pushed down our throats, surely they would not get into this nonsense”.
@ignorantguru
Your blog and comments stinks of non-sense and speculations and all miss is your tin-hat. If red hat is controlling the world with systemd, then you have missed the fact that they were already controlling the kernel, xorg, gcc , glibc, gnome DE for years. Did you wake up from a deep coma and now you cry foul? If you are half as smart as you claim yourself to be, you should have switched to BSD or whatever a long time ago. Its a free market not a free world, in case you have been deluding yourself with the all the idealogical bull crap comments you have posted.
Rather weak. But at least you’re trying, giving it your all.
So you’re really upset and angry that we’re discussing something you already know. Hmmmm… doesn’t add up. Doesn’t explain your reaction and all the upset, does it? Rather sounds like someone stepped on your toe (otherwise known as truth – pain, denial and all that).
Plus you used the phrase “free market” in a non-sarcastic way in 2014. FAIL Careful, your true colors are showing (what was that I was saying about corporate whores?) I really doubt that you can understand any of the finer points being discussed here, and that is the real source of your anger and frustration. Your imagination simply cannot fathom anything beyond the usual tools of exploitation. You’re afraid. It’s a dark wood for you and you don’t like people running into it or even talking about it. Just obey the corporate model. They control all and always have. Thank God, we’re safe.
Thus you imagine it’s a simple problem with a simple solution, and that some establishment will save you. You don’t have to do anything for yourself, and you don’t want me making things complicated by TALKING ABOUT IT.
I know, I know… I shouldn’t play with trolls.
The BSDs are tightly controlled by small groups, usually comprised of arrogant rude people like yourself. From what I’ve seen of the so-called community, it’s not a pleasant or very open/free one – reminds me of Arch Linux in that way too. Nor is BSD free of Red Hat’s influences – they use many GNOME/RH products. In fact it has been their official DE. Yet to their credit they are resisting GNOME 3 somewhat (who isn’t?)
Your reasoning and judgement is appalling. Your sick and beyond help. What you are expressing here is called projection, If only you knew what it is.
See replies below.
What I know is that you whine and complain about what others are saying while contributing nothing to the conversation – which you just did once again. People who enter a conversation to ridicule others and desperately attempt to shut them up are not participants, they are parasites, fear-ridden bullying children, mentally speaking. They try to silence others’ opinions and forestall discussion through ridicule – aka mental desperation. You’re a real proponent of free and open discussion… I can tell by the way you try to prevent me from expressing my own views on my own blog. Yeah, you’re a real winner. I can’t imagine where you stand on this debate or why you’re so upset with people discussing freedom and choices. Bit control freaky, are we?
Yet everyone is welcome to express their opinion here, even children with the manners of pigs. And even me, presuming to share opinions on my own blog. Imagine that.
So go ahead and continue to relieve your bowel pressure right here if it helps – my blog is proud to serve. Maybe someday you’ll have an actual thought or idea of your own – keep trying. Like I said, I can see you’re giving it your all, and I respect the effort.
@hadrons123
>Then you have missed the fact that they were already controlling the kernel, xorg, gcc , glibc, gnome DE for years.
I’d argue that the degree of their contol over the kernel, xorg, gcc and glibc is vastly different from the degree of their control of systemd.
Systemd is practically a RedHat project with a RedHat leader.
>And all miss is your tin-hat.
How about some recent revelations, such as “To protect and infect”? How tinfoil is that?
Is it OK that some of use live outside of the territory of the biggest spy and the most agressive country in the world and even outside of the territory of it’s friends and puppets?
Is it OK that we don’t want to be droned and spied upon?
We don’t want our operating systems components to be under the control of the multi billion dollar corporation and a defense contractor. And we won’t have it that way either.
s/multi billion dollar corporation and a defense contractor/multi billion dollar US corporation and a US “defense” contractor
Also, “defence” is newspeak. RedHat is an offence conractor.
Just in case you were going to say something more about tinfoil hats.
>“When we rolled into Baghdad, we did it using open source,” General Justice continued. “It may come as a surprise to many of you, but the U.S. Army is “the” single largest install base for Red Hat Linux. I’m their largest customer.” (2008)
As of 2008, the DoD claimed to be the biggest customer of Red Hat.
> As of 2008, the DoD claimed to be the biggest customer of Red Hat.
That certainly explains some of the engineering I’ve seen. It’s a war zone.
Aside from the incongruity of a personal computer OS developed primarily for (and I’m sure largely by) the global military machinery, and all the inherent social problems that implies, there is an engineering incongruity too.
While the military (and corporate world) no doubt values uniformity, tight centralized control, obedience to strict protocols, supervision of users, censorship and control of information, having all the keys, etc., the typical home or small business Linux user is looking for flexibility, creativity, privacy, etc. This is why I don’t believe Red Hat should have much of anything to do with my Linux. Yet Red Hat has MUCH to do with everyone’s Linux – they own it.
It also explains the deep incongruity between some of the tools in Linux (eg udisks2) and what users actually need and want in such subsystems. It’s utterly unbelievable the junk that’s being forced on users, clearly because their use of the OS is merely an afterthought.
So it appears we’re not faced with just corporate forces in Red Hat, but it’s the military controlling software development indirectly. Interesting to a point, but all that really matters is to move in some other direction. The way they’re engineering it is simply unsuitable. (And the thought that the military uses such a security-hole-riddled OS for anything is more than a little scary.)
Enjoy that systemd – you’re paying dearly for it.
> This is why I don’t believe Red Hat should have much of anything to do with my Linux. Yet Red Hat has MUCH to do with everyone’s Linux – they own it.
Well said; that phrase helped me form words for my view on it: Red Hat is a for-profit organization working on free, open source software. They pay their developers to work on it, and will pay top dollar for clever developers (especially socially wily people like Lennart Poettering). The “owning” part I agree with, and a lot of people don’t see it. They think “FOSS is a meritocracy or distributed dictatorship. There’s no way any single entity can own the ecosystem!” They couldn’t be more wrong, at least on the second statement.
In a meritocracy, those who do the (best) work get to influence things. The problem with that approach is that it allows a company or an exclusive group of companies (Intel, Red Hat, Oracle, et al) to hire a bunch of developers to do their bidding… and if they all do good work, then they’ve effectively taken over the so-called meritocracy. The only way to keep a meritocracy (or indeed, any merit- or quality-oriented social unit) pure is to outright forbid special groups with influential power. Forbid corporations from injecting developers into non-company projects, forbid cliques, clubs, or factions with over-arching, ambitious goals.
The corporations may well contribute great code, but as long as there are a lot of for-profit entities mucking about with a system, they will control it because money influences many people. If you ever want to know where something came from: follow the money. Socially cooperative groups do not do what Red Hat has done with GNOME, systemd, etc. Socially cooperative distros don’t do what Arch Linux did and Debian runs the risk of doing.
Sorry for the wall of text; I guess I’m just agreeing with you and interjecting some of my own ideas. Great blog post! If you ever start or join an initiative to create a better, freer OS, please include me! I’d love to be part of something beneficial to the FOSS ecosystem.
Hey sporkbox – sorry I haven’t kept in better communication! I was thinking of your angle on this already.
The whole DE concept is poison to meritocracy – it’s exactly that clique you mention, and as BwackNinja described, they are engineered to be exclusive and to own the system. You get software not because it works well, but because it’s part of the club. And you’re forced to have it, and all the little daemons (demons). That is the only purpose of DEs – they are completely unneeded junk in Linux.
Speaking of money, I see gnu.org is begging for $450,000. After observing how Red Hat and big money controls everything in Linux, including GNU Linux, and after exchanging a few words with Stallman, I am completely convinced that GNU is a fraud that holds Linux back. Classic ‘own the resistance’ tactic. It’s just another of those clubs. GNU’s page says “Free software is a cornerstone of any modern free society. We build this foundation.” Yeah, build it with some more of Google’s money, not mine. They don’t need our money, they just act as a sink drawing money from any genuinely progressive, free projects.
Don’t trust anything big anymore, whether they call themselves “free” or not. It has about as much meaning as “green” labels. It is very doable to control and own so-called ‘free’ projects, as you observed.
There is a huge fraud in the way Linux is presented to users. It is a for-profit, for-control OS. And Red Hat is re-engineering it daily to be yet more closed and Windows-like. Virtually no one outside of the big money club can affect it in any substantial way – they won’t let you in. And it is not engineered for its users. It’s a fraud.
I think this is a good time to go retro and SLOW DOWN development. Let some dust settle – developers need a break anyway, most are haggard from the environment Linux has become. I hear it a lot. My new strategy is to hang onto older versions of Linux and software, and stick with those tools as long as possible. Avoid the new stuff! Hack it, disable it. It’s poison across the board. I don’t think we have a good OS to use at this point – Red Hat seems to control the BSDs as well! We are at a place of no good options, and the options that exist are being torn apart daily. I think falling back to older tools and forking them, progressing from there, is much more wise than adopting anything from these corporate/military/mainstream influences, no matter how sweet it tastes at first. But it’s not easy – they’re engineering everything as non-optional, just like your Arch systemd experience. It’s a takeover.
Walls of text welcome – I’m not afraid of reading, and many of these concepts can’t be discussed intelligently without some expounding, unless you’re just regurgitating known talking points ad nauseum. I’d rather hear what you think. I’ve learned a ton from this thread already and it’s really given me a lot to think about. I’m sad to see the truth, but better the devil you know, right?
Many Americans don’t want to be spied upon either – it’s a global problem. And it’s erroneous to believe the American people are in control of their civilian or military government agencies and contractors (as much as egos like to believe otherwise), or that the agencies are in control of themselves. Shadows and mirrors. Same for other nations. We’re all facing global tyranny from unseen parties operating through governments, corporations, media, etc. – it’s a very ancient story throughout history. These days, who controls the most powerful computers (which are far beyond civilian awareness) control most of the resources and dictate most of the activity on the planet. Anyone who knows the power of computers should understand this. And don’t think that the people who control such hardware just use it to spy, they engage in sophisticated social engineering, drug testing on populations, control of political courses, manufacture of major events, you name it. You didn’t think they were letting all those quantum CPU cycles go to waste, did you? Why do you think they need that kind of hardware? Just to read your email?
Spies will always try to spy. Today’s civilian hardware and software is simply no match – it’s hobbled, to put it mildly. Broken toys for children. They only advertise the spying and get people worked up as a means of social control – the NSA is watching YOU. You can try to create secrets but I view privacy as mostly a social convention to be honored – a kind of respect. It can’t really be enforced, information is free by nature, as the hackers used to say. Make your secrets somewhat open. If someone invades your privacy, they’re the ones who have the explaining to do. As usual, mutual respect goes a lot further toward creating what you want than use of force. You don’t have to compete against the NSA or defeat their quantum computers – they are their own worst enemy. Just stay aware of them and don’t be so naive about the world you’re in.
But I draw a line between that kind of inevitable spying and these entities controlling every creative choice I make on my PC, determining all of my software and my complete view of the world, and basically enslaving me in a false economy. That’s where we can create our own solutions rather than sucking on a government/corporate tit. That’s where resistance is not futile. Let them spy on and watch that – it’s all open. Eventually even they might drop their war games and join us in a finer solution to resource distribution than thuggery.
You can’t control what they do, but you can make choices for yourself.
Scientists rarely understand the political or social implications of their work, anymore than spies do. They are often emotionally naive, easily misled and used, focusing on myopic details. Developers and the whole intelligensia surrounding computers and Linux is much the same – sadly naive. And that is the challenge we’re facing. If you think something based on the principles that Linux is based upon won’t come under coordinated attack from interested parties, you’re naive and likely part of the problem. If you do understand this, you employ certain principles to protect it: openness, diversity of solutions, freedom of participation, self-governance, etc. Basically, all the things that are now lacking in Linux development, which has become closed and centrally controlled.
>It’s erroneous to believe the American people are in control of their civilian or military government agencies and contractors.
Absolutely spot-on. I have a deep respect to Americans and the values on which the country was founded. I also deeply people and culture of European nations and consider myself part of it.
And my deepest respect goes to the people in every country that are opposing the shit that their own governments are involved in. These are the people that are changing the world to the better. It is much easier to criticize other countries and much harder to actually try to influence the bad things that your own country is doing in your own name.
>That’s where we can create our own solutions rather than sucking on a government/corporate tit. That’s where resistance is not futile.
Actually, all these years I’ve been an opensource/free software user rather than developer. Yes, I’ve made occasional bug reports, feature requests and a tiny few bugfixes, but that’s all.
I was more or less concentrated on empowering non-techies for them to be able to try to change society for the better.
Being a programmer, I’ve been drifting towards towards doing social sciences in my spare time. But, in the light of the ongoing changes in the GNU/Linux landscape, I’m starting to think about reconsidering this path.
Social changes to the better would be impossible without people’s unfettered access to open and independent computing platforms.
What a lot of people are missing in this discussion is what drove me away from every linux distro. Systemd can be considered ‘dangerous’ because of how little being a distro means. Systemd is a wonderful technical project, I’m not going to hesitate to say that and I have never hesitated to say that on this blog, and that’s why I use it. But, if things went south, there would not be enough effort put into a fork of the project to maintain something that we could still continue to call a technically positive project.
Ubuntu is interesting because on the other side of the NIH syndrome they seem to have and the CLA, you can see that they’ve decided to be different and take ownership of their distro rather than being tied to upstream and the status quo. It’s unfortunate that I don’t like many of their decisions.
Going even further downstream, we’ve got the users. People, or rather, Linux users, no longer take ownership of their computers. The “Linux for Human Beings” concept breaks exactly why I use Linux at all. People for the most part are only leveraging “free as in beer” and not “free as in freedom”. I’d pay my way for freedom, but if they got the same system, most people using Linux wouldn’t care that it wasn’t open-source and they didn’t get that freedom.
Hey, IgnorantGuru, how much time do you have to work on something with me? My free time has gone down, but how would you feel about participating in a new distro project? I’ve got most of the process hammered down and made interesting decisions regarding its construction, but I’d want someone to talk about it with. I’ve got no intention of it being some “one distro to rule them all” lofty goal, but rather something to bring together like minds and still be useful even if I was the only one using it. How about it?
Most Linux users don’t know much about an init system – they simply aren’t part of this conversation at all. So you have developers making these decisions which will have grand effects on the entire direction of Linux. Or had effects – to me this is history.
Used to be Linux users were under-the-hood and interested. If the developers went seriously astray they heard about it from sysadmins and users who knew what was going on, and development also moved more slowly with more communication. Even documentation – imagine that. Yet now developers hear from large segments of Windows+Linux users and developers about app features and widget appearances, and have bosses telling them what to destroy next. The Linux users who actually know some of what is still going on and remember what working software looks like are drowned out in the noise. We no longer have a palpable influence over the developers of the systems we’re using. We’ve lost Linux. We’re mere passengers. “Everybody knows the boat is sinking; everybody knows the captain lied.”
Obviously if enough people feel this way they will create solutions eventually, and they already are. But I still think a lot of ground has been lost – it’s like getting kicked back to living in the stone age again, kicked out of the very house you built. But some price has to be paid for the amount of control (total?) that Red Hat was given.
The myth that we’re using a non-corporate, freely and openly developed OS is just that – myth. It’s a corporate product, and if you, or even a large group of you, think you can contribute to it in any way you like and influence its direction, try. Then you’ll find out who makes the decisions in Linux, including GNU Linux, and where you stand.
If you want a freely and openly developed OS, then start making one, but it ain’t Linux.
As for distros, the main value of a distro to me is the packaging. Packaging is deeply linked with the init system, and many distros are derivative of a few of the main distros via packaging. It’s a good place of attack that affects all distros, giving them hard choices. The hard choices will continue and will increasingly affect users who thought they were immune in GNU land. That’s when the old soldiers will rise and start coding again. ;) Or finally collapse, either way it works out. Most likely there will be a falling back to older ground, regrouping, despair and the gnashing of teeth.
I’d like to hear more about your distro ideas and maybe participate in some conversation at least. Maybe start a github wiki, or you’re welcome to go a bit off-topic on the spacefm forum – it’s related in a larger sense. I think a lot of people are thinking “time to make a distro!”, but you probably have the experience to do it well, and spacefm users often have good ideas to share. You might also look over some of the micro-distros that others have going and see what you could add.
The main challenge I see in such an undertaking is that most micro-distros are derivative – they don’t distribute all their own packages, etc. So if you’re derivative you’ll inherit much of the init system, and all that comes with it. In your case, you’re more of a systemd fan so that doesn’t scare you.
Like a kernel or init system development team, creating a distro with its own packaging and servers, support, etc requires large-scale resources. Many people working together in an organized way. Non-corporate-controlled entities have a real problem developing the needed momentum these days, and the environment helps assure that in many ways.
As for Ubuntu, I agree that despite being Debian-derived, they went their own way, flawed as it was. I am definitely not a fan of upstart. I’m more of the mind that if you don’t have something genuinely better yet that doesn’t threaten the whole fabric of Linux, leave it alone! This is the one option they can never use, and why they always destroy everything good with their frantic activity. They don’t know when to leave it alone. And an init system that depends on dbus is shit – jfyi. It ain’t better. :p You just like new complicated toys to play with, BwackNinja, so the bait gets you as it was designed to. But it’s poor design from a standpoint of simplicity, security, etc. It’s the beginning of how you develop something like MS Windows – a closed system, closed by its complexity and breakability and monolithism. But that’s all developers are taught anymore – it’s all they know how to make. (As you lauded systemd, that is my general reply, not to go way OT.)
I’d be going the scarier non-derivative route. Pacman is one of the best package managers for quick development and my personal favorite, so things would end up a bit Arch-ish (with a lot of things based from their PKGBUILDs), like my current system is and it is well documented how to make your own mirror. Serving a small userbase wouldn’t be too much for me to start with, particularly because I don’t intend for a great amount of traction. What would also make updates easier (particularly for me maintaining my own packages) would be a statically compiled system. There’s overhead with updating packages when libraries change, but nothing would break in the meantime which is an issue I battle any time something like libpng updates (running something like “for file in /usr/lib/*; do [ $(ldd ${file} | grep “not found” | wc -l) -gt 0 ] && pacman -Qo $file; done | sort | uniq” has really been my only recourse in times like those). I’d also go for a fully LTO compiled system and at least try to use musl as the libc implementation.
I’d also go with a pure wayland system and a custom shell, which should be rather manageable and not take so much time to compile like a full xserver stack does. Now is the time when that’s actually starting to be manageable to do.
The biggest thing about systemd for me is that upstart is absolutely unintelligible if you actually try to leverage its features, and sysv stuff is horrifically redundant to deal with. Things like tmpfiles are problems that people looked at as “that’s just how it is” rather than dealing with them in a complete fashion. Ubuntu has even taken components of systemd despite refusing to switch. Socket activation isn’t particularly appealing to me, though if it doesn’t take too much I won’t complain that it’s there. I also made sysv-like scripts for some desktop processes that I wrote and kept needing to restart, so systemd as a session manager is rather attractive to me. I’m contemplating writing a simple, mostly systemd-compatible init system. Partially that’s just because nobody really wants to do something like that and bash the whole system when the interface and file structure is more sane than anyone else’s. The biggest PR issue with systemd is that it is one repository and one build system – if it was modularized properly, then people would have less reason to hate it.
I do like new complicated toys to play with, but I don’t like it when people start taking those toys away or making them less fun to play with. Though it’s an old topic by now, the recent changes in Gnome (the gEdit changes in particular affected me a lot more than they should’ve, considering how little I use it) have only pushed me to want to do this more. We also have wayland porting dominated by big distros and I want to send the message that those with more of a personally tailored system shouldn’t be neglected and stuck on X.
OT: I like that you’ve taken an interest in mpv. It is the best media player at the moment and actually doing interesting things, but has nowhere near enough people interested in it and its progress.
What we really need is a protocol that separates packaging from any particular init system, or even distro. It is time to objectify the ‘distro’ and packaging concepts? But this may be where Linux reveals more basic design limitations. Red Hat is basically using the dependence on packaging to own the whole system.
I find changing libraries and the way older software is continuously being broken in Linux to be in need of review. I envision some kind of sandbox that each program runs in, which contains all the components it uses. Duplication would be handled on another level (eg only needing one copy of a library, or only loading one instance when possible, or multiple versions when needed.) The system is breaking too much software too easily – it’s (deliberately) becoming a high-maintenance development environment where one solution precludes another (lack of diversity). Apps need virtual systems that actually serve their requirements instead of blacklisting them.
Yet of course playing well together is not the agenda du jour.
One of the most interesting questions I asked myself while contemplating this system is “where does the operating system end and the applications start?” Part of my plan is to have a separate /app (and a ~/app) which would house folders that would be a whole app and its dependencies. Statically compiled, their dependence on the rest of the os would be near nonexistent and dropped into place without even worrying about package management.
Once you get into the realm of graphical applications, things are almost entirely self-contained, other than at runtime. Mac OSX and its application install system shows us that. However, it fails in that it cripples the terminal in the process. I’d go for creating a couple FUSE directories – one for /usr/bin and one for /usr/share/applications. To handle manpages in different directories, mandb could probably be used effectively. Beyond that, there aren’t really any directories that are used globally for things other than system daemons. Things do get confusing when you talk about applications using centralized configuration systems like dconf. The deeper you delve into the system, the more you see how all the biggest desktop environments are constructed in such a way that they can’t be separated into individually working component parts and weren’t engineered to scale well as anything a human was ever meant to manage on the filesystem.
More than an distro project, I find all of this to be an interesting challlenge. Failure is disappointing, but having tried is well worth it regardless.
https://en.wikipedia.org/wiki/GoboLinux ?
I’ve seen GoboLinux, but I find no benefit in hiding the normal filesystem hierarchy, and I don’t think symlinking lots of files is the right way to go. I also don’t like their structure – something like /Programs isn’t so terminal friendly like /app would be. This results in a system more meant to be looked at as clean rather than idealizing real cleanliness by itself. The symlinks also point to that because it isn’t functionally any different than a normal linux system – it just looks different when you look at what files aren’t hidden.
tl;dr – the set of ideals behind the way they are implementing the GoboLinux filesystem are incompatible with mine.
Truth in labeling?
>We also have wayland porting dominated by big distros and I want to send the message that those with more of a personally tailored system shouldn’t be neglected and stuck on X.
There was a very informative talk about poor X security on 30c3: X Security – It’s worse than it looks where the speaker was talking about tens of X security bugs he has personally found, reported and helped fix. As a bright side, he has said that every bug report was taken very seriously and fixed by devs very quickly.
On the other hand, in the same talk the speaker has said that there are as many serious security flaws in GUI frameworks. For example, Qt that, in contrast to X developers, Qt devs fail to even recognize, let alone fix.
I’ve lost all hope for GTK+ when they’ve added dbus as obligatory dependency.
The grand goal is to make masses into mindless consumers of informations and for that there is no need for fully featured desktop, just dumb boxes with browser on them. It’s already happening – chromebooks or whatever they call it. Nowadays there is also a crazy trend of reinventing stuff in javascript which is horrifying to me (for example writing emulators in javascript – just because one can it doesn’t mean it is necessarily a good idea.) Instead of getting nice standalone applications, we get another layer of abstraction with browsers and we can’t even get the thing in binary form, we would need to keep webpage source, which may be obfuscated to save on download size or to prevent easy edits and don’t work without server side code.
On linux side the biggest problem is “Year of THE desktop” thing. It’s insane ideology dumbing once working tools to common denominator, which for some developers is a person who sees computer for first time. It’s insanely stupid – making an advanced tools into a toys for 3 years olds, because God forbid they can change default settings or delete their own data because they are stupid and can’t read.
> I’ve lost all hope for GTK+ when they’ve added dbus as obligatory dependency.
Almost, and they tried to make it so, but last I’ve seen dbus is not required by GTK3, though it will spit out complaints on stderr. For example, SpaceFM is a GTK3 app that does not require dbus at build or run time. GNOME otoh probably requires dbus. And some GTK+ functionality no doubt does.
In the recent Debian init system discussion I’ve found a link to a hilarity:
If you’ve got some free time on your hands, I encourage you to watch this talk from 27C3 (2010) given by a big university system administrator, where Lennart Poettering constantly interrupts criticism of Pulseaudio, GNOME etc.
“Lecturer, after talking about major Desktop breakages: GDM >= 2.21 starts a full-blown GNOME session, it does it for a simple log-in. I can see some of the benefits behind it, but I really wonder why do we need it. Some people don’t really want GNOME [Applouds in the room]. We canot say that this acts like some sort of library pre-loader.
Poettering: May I respond now?
Lecturer: Yes.
Poettering: OK, so, do you hate handicapped people? Do you hate people who do not speak English? Do you only care about your own use case for GDM? Because there is this thing: GDM needs to be a full session for a couple of reasons. First of all, we need to start screen readers. We need to start all kind of other accessibility stuff […]. A screen reader pulls in an audio stack, because you actually need to play audio, that basically pulls in pulseaudio. Pulseaudio might even pull in bluetooth and the bluetooth stuff. If you log in on the network, then you actually need to sometimes download the LDAP and user list and stuff like that: how are you going to access the network if you’re not running something like the network manager applet.
Lecturer: Why do we need network manager applet?”
Poettering is always referring to “our customers have asked us to be able to do this and that” as an explanation to bizzare design decisions that are forced down our throat. You can clearly hear that he is interested only in his customers, not Free/OSS software users and developers.
Etc. etc. etc.
I cannot believe how Debian can seriously discuss being subsumed by this.
I watched that a while ago. The issue comes down to asking “do we want an application for a very specific use case, or do we want something that will work in all use cases?” It’s not right to say that just because you don’t use a specific feature it’s bad that it’s there. However, the issue comes down to one of modularity and one of consistency. No one complains about the situation with something like the Linux kernel because it is either too scary for them to try to handle or they know it well enough to understand its modularity when they compile it.
The only issue I can find in Poettering’s argument is that those issues are almost entirely non-issues in a /desktop/ Linux system. They are issues that really only come up when you’re talking about the system being set up and administered by a third party rather than by the person using it. Frankly, the kind of people opposing him are those tightly coupled to the idea of a computer they are using being one that they control. When you buy or build a desktop or a laptop, you’re going to have a hard drive or drives that you boot from, with an operating system you installed that you already localized to whatever language you feel like speaking this week. Network booting is more of something associated with corporations and eccentrics.
It’s hard to be a real Linux geek on a system where you aren’t also the administrator. It shows that there is a strong focus on the users that /aren’t/ the developers, that aren’t the traditional audience of Linux. The geeks are marginalized, and that’s not so surprising because they’re the only group that’s capable of fending for themselves anyway. Everyone else is at the mercy of whatever hierarchy that they’ve subscribed to.
The loud and proud geeks are marginalized, while corporations and individuals who complain about the default wallpaper are the ones catered to. Most people are okay with these changes, but you hear a lot of anger because the people who feel like they represent Linux *ahem* GNU/Linux as a way of life and not just an operating system and a means to an end are the ones who are complaining.
>However, the issue comes down to one of modularity and one of consistency.
To me the issue comes down to the fact that I look at this as if it was an Embrace, Extend and Extinguish in the second phase, i.e., the technical merits of the decision to go with consistency are clearly overshadowed by political risks of doing so.
“Do you hate handicapped people? Do you hate people who do not speak English?”
Sheesh, what a transparently bogus level of rhetoric Poettering stoops to. That’s like a parody of nationalist attacks against protesters “If you disagree with the government, you must hate freedom.”
I’m amazed he can actually talk like that for real and apparently some people take him seriously.
Something to count against Qt here – it looks like Qt Multimedia functionality (https://qt-project.org/wiki/Qt_Multimedia_Backends) has been built with a hard dependency on pulse audio in Debian (see my bug report here: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=740451) – from the Qt side the configuration files are set up so that if pulse is manually requested via a platform config, or no config is defined and pulse headers are present, pulse completely takes over audio and ALSA is not enabled even if its headers are in place.
Found this when ppsspp (a new emulator for the PSP) would hang in the Qt interface when sound was used on this pure ALSA system (https://github.com/hrydgard/ppsspp/issues/5132#issuecomment-35830879).
Looks like gthumb, a competent image viewer, is being bricked by GNOME dumbness now – no more normal menu and toolbar in v3.31 (https://mail.gnome.org/archives/gthumb-list/2014-March/msg00006.html) – can’t control/configure the program in a normal desktop environment like XFCE4??
Fullscreen images are no longer fullscreen – a frame with significant padding surrounds the image?? See changelog: https://github.com/GNOME/gthumb/commit/f79c0f10d39e87cafd76d9c1ef6cd74e3d07bf9f
This has made me pretty angry, not looking forward to finding other shit. I think of forking, but its just an image viewer… no way can I justify the time to do that… very frustrating.
I’ve never tried gthumb but geeqie is a capable viewer with some extensibility.
In other news, another udisks priviledge escalation.
Thankyou – a bit of fiddling around and it does proper fullscreen now. Needed to hack the desktop file to stop it opening separate images in one instance, but SpaceFM makes that easy ;) First time I’ve used that in fact.