OpenIndiana lead Alasdair Lumsden resigns
From: | Alasdair Lumsden <alasdairrr-AT-gmail.com> | |
To: | oi-dev-AT-openindiana.org | |
Subject: | Resignation as OI Lead | |
Date: | Wed, 29 Aug 2012 02:18:11 +0100 | |
Message-ID: | <503D6DD3.6010809@gmail.com> |
Dear OI Developers, It is with much sadness that I hereby resign as project lead. I may, if the situation improves under a new project lead, stick around to offer my opinion or occasional assistance, but my resignation is final; I have no wish to return to the project in a leadership capacity. My resignation is primarily driven by a lack of time; I simply cannot commit the hours necessary to maintain a project of this size. I have my life, my health (primarily mental), and my future to think of. But it is also in part due to frustrations with the difficulty of making any progress on the project. OpenSolaris was maintained by a large corporate entity. We however, are volunteers, contributing our personal time to work on a project we believed in. For many of us this was the first open source project we had ever contributed to, myself included. The task at hand was vast, and we were ill equipped to deal with it. But what really, right from the very beginning, upset me, was the lack of interest from the large commercial players benefiting from Illumos, and from those who had been paid to work on Solaris at Sun. Instead, what we got, was grief regarding the name (Project Indiana seemingly being a sore point for Solaris engineers, something I was completely unaware of when we chose "OpenIndiana"), hostility towards IPS, and a total lack of interest, encouragement or friendship from people many of us looked up to when we were mere end-users of Solaris under Sun. Right from the very beginning, Illumos was on life-support. I have no doubt that Nexenta, Delphix, and Joyent in particular will continue to innovate and that SmartOS will be a success, but support for Solaris from the open-source software community has over the past 2 years gone from bad to worse. Only the other day the MongoDB developers responding to an issue with it segfaulting on OI stated "OI isn't supported, use Linux": https://groups.google.com/forum/?fromgroups=#!topic/mongo... I lay the blame of this squarely on the lack of a successful general purpose distribution of Solaris/Illumos. OpenIndiana was my attempt at competing with the Linux distros, but our lack of progress has torpedoed it. Nobody in their right mind would use OI - it ships severely out of date insecure software, lacks some of the most common 3rd party apps such as LibreOffice, and so much simple shit that should just work, such as "pecl install", "gem install", "pip install" or whatever barfs due to nonsense SunStudio flags, to the point you need a background in computer science and compiler flags to get it to work. Not fit for purpose. So what exactly are 3rd party software developers such as the FFMpeg or MongoDB developers supposed to use to develop and test their software on? Buy a SmartDatacenter? Install a storage product? Run it on a database appliance? All of you, Joyent, Nexenta, Delphix, are complicit in the increasing irrelevance of Illumos. OI, even in it's current current state, is by far the most widely used Illumos distro, so by not supporting it beyond contributing to the Illumos core, you've all shot yourselves in the foot. With a fucking shotgun. What's sad is that you don't even see it. It didn't have to be this way. With some assistance we could have made large strides forward - we had lots of solid ideas of how to get things moving. What we lacked was time, graft, and expertise from those who worked on this professionally - items easily supplied by those with deep pockets and plenty to gain from our success. Instead we got the Illumian farce from Nexenta, along with their senior staff claiming OI is an existential threat to their continued existence. And when I asked for help back in November, we got Bryan Cantrill telling us all "when you want to do something, just do it" - rich coming from someone paid to work on all this whilst the OI devs volunteer their personal time, often at considerable personal sacrifice, to work on this stuff. With the ZFSOnLinux port becoming increasingly popular (so many of the Linux users I know are using it), and brtfs/dtrace-on-linux/upstart/whatever else slowly brewing away, even some of the core features of Illumos are becoming less and less important. Yes, the Linux equivalents suck in one way or another, some are completely and fundamentally broken by design, but it doesn't matter - what matters is perception and the typical Linux user is happy with "good enough". When I encourage my Linux-using friends to try OI they laugh in my face. OI and Illumos to them is a dead platform. Add to that our increasingly out of date and poor hardware support due to the march of never ending new LAN/SATA/SAS/motherboard/GPU chipsets and you start to get the picture. I hope, I really do hope, that Illumos does not become entirely irrelevant. But when less and less software works out of the box, and when heavily used products such as MongoDB, Varnish, etc don't support Illumos (regardless of whether they actually work on it or not, what matters is whether these projects will help end users when they have problems), and when OI disappears and there's nothing left but a handful of fringe distros or niche products, what then? You think Riverbed are going to maintain Stingray (Foremly Zeus) LB on Solaris, or any other commercial software vendor develop for it, when nobody is using it? Well, I've said my piece. This has been weighing on my chest for some time and I am glad to have gotten it off. I am not doing this because I want to start a flame war, I just had to say it or it would have bugged me for the rest of my life. I would like to thank from the bottom of my heart those of you who have volunteered your time to work on OI. For those not mentioned directly, you know who you are and it has been a pleasure working with you. I hope we can continue to keep in touch. I would, in particular, like to thank Richard Lowe for his unwavering support. He is without a shadow of a doubt one of the kindest, selfless, helpful and wise people I've had the pleasure of dealing with throughout this journey. He was always there to help, and to provide a modicum of sanity when all hope seemed lost. Without Rich, OI would likely not exist, and we all owe him a very large debt of gratitude. I would like to also thank Alan Coopersmith for his support and impartial help. His presence on IRC provided much comfort to all of us, and his insights were always highly valued. My thanks go to Garrett D'Amore; without his stellar efforts creating Illumos things could have been catastrophically worse for us all. I hold him in high regard and in no way hold him responsible for the current situation with OpenIndiana, even if he did help spawn Illumian. I'd like to thank Jon Tibble for his dedication to OpenIndiana, and for his hard work, especially with the pre-stable releases, which was greatly appreciated. Jon is a first-class citizen of the community and I hope he will continue to work on the project even if I'm not at the helm. I'd like to thank Andrzej Szeszo for his contributions. His deep insight into complex parts of the distribution, along with his persistence and capacity for tinkering, have unstuck the project many times. Again, without his help OI may not have come as far as it did. I'd like to thank Guido Berhoerster for his hard work on JDS and his support in getting the project off the ground - again without his help we would simply not be here. I'd like to thank Albert Lee for his help in the beginning of the project, indeed Albert was responsible for pulling an all-nighter to get our first release out. We once again owe him a debt of gratitude. Lastly, despite their lack of a handle on what's happening with Unix/Linux distros in the real world beyond kernels, I'd like to thank all those who have contributed to Illumos, without which OpenIndiana would not boot. You are the real heroes. I may have complained bitterly about our little distro being ignored by you, but you have my respect and thanks for your unique talents in developing a truly amazing kernel that we all love dearly. I will continue, through EveryCity, to provide hosting for OpenIndiana's infrastructure. I also hope that a new project lead will step forward to look after things, and that they can carry the project forward. If no viable new lead steps forward then I would encourage the OpenIndiana developers to hand responsibility for it over to the Illumos Foundation. Finally, I wish Illumos every success. Ultimately Illumos is what matters, OI was only ever going to be a vessel for delivering it's power to end users. May it go from strength to strength and get the recognition, attention and user-base it so rightly deserves. Regards, Alasdair
(Log in to post comments)
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 14:38 UTC (Wed) by davem (guest, #4154) [Link]
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 15:30 UTC (Wed) by gtirloni (guest, #85631) [Link]
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 17:24 UTC (Wed) by louie (guest, #3285) [Link]
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 16:31 UTC (Wed) by rriggs (guest, #11598) [Link]
I spent over a decade developing big data software, primarily on Solaris. I have seen workloads that bring a 64-core Solaris box to its knees with massively multi-threaded, big-memory footprint applications. Linux handled the same load with aplomb.
I get a sense from the email below that these truths are finally sinking into the author. I still sense a little denial in there, but he's moving on to acceptance.
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 30, 2012 4:34 UTC (Thu) by ringerc (subscriber, #3071) [Link]
When most Solaris users' response to a new box is "Argh, get gcc and the GNU tools on here, change my PATH, aaaaaaaaah" you know you have a problem.
A really good move - about 5 years ago - would've been for Sun-before-Oracle to have licensed icc from Intel and dropped it in as a sane replacement for SunStudio. That alone would've made a surprising difference IMO.
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 30, 2012 4:35 UTC (Thu) by ringerc (subscriber, #3071) [Link]
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 31, 2012 22:39 UTC (Fri) by khim (subscriber, #9252) [Link]
Sadly icc is in the same bucket as SunStudio. Really. Only three compilers matter today: MSVC (it's awful, but people support it because they have no choice), GCC (it works everywhere and it's also compiler for Linux), and Clang (for one but important reason: Apple is popular enough and Apple pushes Clang). Everything else is a no-go (or rather: no-go for general-purpose OS, niche projects have their own pet compilers... even icc is used not just to run SPECs).
Five years ago the only choices were MSVC and GCC. Why Sun tried to kill itself by refusing to accept reality is beyond me. Sun succeeded, BTW.
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 31, 2012 10:40 UTC (Fri) by trochej (guest, #35052) [Link]
What makes me wonder is, how all of comments in this thread I've seen look just like typical trolls. So you're happy with your Linux boxes. Cool, I'm happy with my Linux boxes too. But I'm also very happy with my illumos boxes. While Alasdair choose pretty poor way of putting it, he is right in many aspects.
Also, while I believe Linux is going to have equivalent of DTrace one day, and of ZFS (and maybe even ports of them), I am using them right now, stable, fully productive, with no strings attached to Oracle whimsical view on community. Go figure.
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 31, 2012 16:09 UTC (Fri) by mabshoff (guest, #86444) [Link]
So, would you be able to tell me one specific work load which falls over on a current linux kernel? You have seen 'loads', so it cannot be hard.
> What makes me wonder is, how all of comments in this thread I've seen look just like typical trolls.
Really? I am not sure which websites besides lwn you frequent, but this is not what I would consider trolling, especially 'all comments.'
> So you're happy with your Linux boxes. Cool, I'm happy with my Linux boxes too. But I'm also very happy with my illumos boxes.
Well, nothing wrong with that. If it works for you no one around here will tell you that you cannot use it.
> While Alasdair choose pretty poor way of putting it, he is right in many aspects.
Yep, I saw the discussion between him and various other Illumian people when it came about and I was surprised it took Alasdair this long to quit tbh.
> Also, while I believe Linux is going to have equivalent of DTrace one day, and of ZFS (and maybe even ports of them), I am using them right now, stable, fully productive,
See above: if it works for you: great. But you should be able to notice that despite ZFS and dtrace being open sourced in 2005 it has not really helped Solaris. And whatever technical advantage Solaris might have had it has shrunk by quite some margin since then.
> with no strings attached to Oracle whimsical view on community. Go figure.
Yeah, Oracle might not be a popular company in the ex-Sun community to put it mildly, but unlike Sun they understand that you have to make a profit to keep the lights on. Last I head they had about 500 people in the Solaris group and were hiring and it takes a lot of people to keep an OS relevant. So while many if not all of the ZFS and dtrace engineers quit when Oracle took over Sun that does not mean that they can sustain an OS.
Cheers,
Michael
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 31, 2012 18:17 UTC (Fri) by dlang (guest, #313) [Link]
I've been using Linux, AIX, and solaris at work on servers for the last 15 years, and when I started there were tasks that did not work nearly as well on linux as on the other *nix flavors, there is very little difference nowdays, especially once you consider linux running on the Sparc or Power hardware.
On the Sparc hardware, Linux long ago passed Solaris for anything except possibly the beefiest hardware (I remember showing my Sun Sales rep a Sparc box running Linux and seeing his jaw drop at how much faster Linux was than Solaris, and this was over 10 years ago, but on what was at that time low to medium end Sparc hardware.
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 8, 2012 8:21 UTC (Sat) by Pawlerson (guest, #74136) [Link]
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 8, 2012 10:33 UTC (Sat) by mabshoff (guest, #86444) [Link]
Yep, it is.
> However, the must funny part was about slowlaris beating Linux. Keep dreaming till you wake up and realize slowlaris is dead.
Meh, I hate it when people use derogatory terms for OSes since it weakens their argument. While on average system call time for example seems to be faster on Linux than Solaris 10 or 11 on the same hardware, it is a long way from Solaris 7 and 8 days where it could be rather bad.
Last week AIX come up and some comments called it dead, too. IBM sells over 5 billion power hardware for AIX a year (compared to about 8 billion a year for the Linux hardware market overall) and in the second quarter 2012 it gained 6 points market share in the Unix server sector. Sure, it fell 10% in revenue compared to the second quarter 2011, but every other Unix vendor fell more. This has in part to do with upgrade cycle, i.e. Power 7 -> Power 7+, T4 -> T5 for Oracle and the latest Itanium for HP. The Unix market overall is shrinking but calling it dead is far from the truth. There are also still plenty of main frames around simple because there are workloads where they excel and people called the main frame dead in the 90s, too. And I am sure any hardware vendor would love to have a 5 billion a year hardware product range with juicy margins on top of which it could sell quite a bit of services, too.
Large SMP Unix boxen will just simply not go away. Savings for using Linux vs. Unix seem to matter rather little once you figure in for example the cost of an Oracle DB on that box. On top of that add risk avoidance and potentially a little FUD and who in their right might would want to migrated a number of large DBs for potentially little gain? It is different for green field deployments, but decisions are done very differently in a conservative enterprise setup. So AIX, Solaris as well as HP/UX will be with us for some time to come ;)
Cheers,
Michael
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 8, 2012 12:27 UTC (Sat) by khim (subscriber, #9252) [Link]
The Unix market overall is shrinking but calling it dead is far from the truth.
The proper term is "zombie", I believe. This is lock-in schemes at work. Think War of Currents: AC distribution networks won back in XIX century yet some customers still bought elecricity from DC networks in XXI century!
UNIX may linger for similar amount of time, but it's relevance is the same: existing customers keep the thing alive, but there are few (if any) new users.
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 8, 2012 13:36 UTC (Sat) by mabshoff (guest, #86444) [Link]
Yeah, inertia is a powerful force :)
> UNIX may linger for similar amount of time, but it's relevance is the same: existing customers keep the thing alive, but there are few (if any) new users.
I don't think it will last quite as long as DC current, i.e. even 20 year old technology is woefully out of date, but still lurks in some dark corners of data centers. I have never seen one, but I heard stories that some places keep some PDP hardware alive to run certain jobs. My own experience closest to these stories was some BS 2000 mainframe sitting in the corner of a data center to do certain jobs, no idea what exactly, but I counted myself lucky I had nothing to do with it.
As long as Unix vendors make money, they will sell and support it. There are really only three vendors left that matter (IBM, Oracle, HP) since Irix went into really deep maintenance mode and Tru64 is officially no longer supported. The BSDs do not really matter commercially in that market, but they are important for the custom storage and routing operating systems for companies like Netapp, Isilon (bought by EMC recently) and Juniper to name a few. Illumos is competition to the 'storage BSDs' and I would personally like to see it prosper since it is at least open source while the filesystems developed by Netapp and Isilon are not. Linux is certainly also used more and more in storage applications, so it will be interesting to see how commoditization will affect that market. James Bottomley has been giving talks on the economic forces of open source, i.e. look at LinuxCon Japan 2012 iirc for a recent example, and in those he argues that products that face commoditization tend to have the aspects that aren't value added parts open sourced. And the filesystem and routing function in operating systems seem to be at this point, i.e. five or ten years ago a storage box was big bucks, these days you can build one with open source without much trouble. So there has been price pressure on commercial vendors since you really start to have trouble to charge extra for a dedup feature in your array if you can just get an open source one. There is still plenty of gear to be sold in the high end, but I am certain places like Netapp fell the pressure.
To get back to Unix: I personally think that HPUX will be the first to go of the three left since Itanium's future looks more than a little bleak if you have read the discovery documents Oracle published [1]. HP has attempted to port HPUX to x86-64 twice ([2] and [3]), but killed it again and again, so I think it will be the first to go as Itanium hardware support fades. Solaris has x86-64 and IBM seems to be able to sell enough Aix on Power despite the lack of workstation models which traditionally has indicated that a platform is in trouble.
And in the end what matters most is applications, which I suspect is the reason that the x86-64 port of HPUX was killed, i.e. why port to HPUX on x86-64 if you can run the software on Linux? If you look at the articles you will see that HP wanted to offer RHEL 6 compatibility so that Linux apps can run. I have no idea how they wanted to do it, i.e. visualization, kernel run time support via personalities or whatever. And this is really the issue with Illumos, to get back on topic, i.e. will software vendor support commercial software on it? I doubt that it will happen for a significant number of software packages and Oracle will certainly be hostile to them since the uname is still identical to Solaris. I am sure plenty of that unsupported software will run, but I do not think many companies will take a gamble to run their Oracle DB on a Illumos derivative :). Even free software project have started to have issues, i.e. there was a patch on the Mesa list to disable some features on OI/Illumos due to missing headers that worked perfectly fine on Solaris 11. It didn't go in. I have seen similar conflicts with patch sets to the user space code to OFED for example, too.
In the very end the weakness of the Illumos/OI community can be IMNSHO demonstrated via two issues: (1) no support for Sparc (2) the Illumos port of KVM does not have AMD support, i.e. it is Intel x86[-64] only, roughly two years after it was released. I wish them luck since I believe they will need it.
Cheers,
Michael
[1] http://www.oracle.com/us/corporate/features/itanium-34670...
[2] http://www.theregister.co.uk/2012/06/08/hp_ux_on_x86_proj...
[3] http://www.theregister.co.uk/2012/05/23/hp_project_blackb...
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 8, 2012 18:26 UTC (Sat) by khim (subscriber, #9252) [Link]
I have never seen one, but I heard stories that some places keep some PDP hardware alive to run certain jobs.
It's even worse then that: PDP hardware is too expensive and rare today thus they run that stuff on emulators. I know someone who helped big customer to speedup their PDP-11 emulator for Itanium. Wonder what they'll do now when Itanium goes the way of dodo itself...
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 9, 2012 7:34 UTC (Sun) by mabshoff (guest, #86444) [Link]
Hehe, I have been around the block a couple times, so those PDP stories are not something I heard about last week and I am sure are still told to this day as yarn to frighten the young sysadmins. Emulation makes sense given the space and cooling requirements and it probably runs much faster in emulation, too.
> Wonder what they'll do now when Itanium goes the way of dodo itself...
I am sure you will be able to buy Itanium hardware from Intel until the end of the decade with the last CPU we know about (Kittson) speculated to appear in 2016 according to [1]. Throw in an additional die shrink like it happened for PA-RISC [but not for Alpha :(] and 2022 or 2024 does not seem unrealistic for a date until you will no longer be able to buy Itaniums. That will give time for the people running OpenVMS or NonStop and the strange main frame OS vendors that ported to it to get onto something else ;).
Cheers,
Michael
1. http://www.theregister.co.uk/2011/04/13/intel_itanium_ser...
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 9, 2012 2:46 UTC (Sun) by dlang (guest, #313) [Link]
I'm sure that many IBM shops have so much invested in AIX software that they don't take much advantage of it, but it's an option.
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 9, 2012 7:22 UTC (Sun) by mabshoff (guest, #86444) [Link]
Yep, but I suspect that the number is relatively small. I have never seen any estimates, but given that RHEL as well as SLES is available on Power it must be worth it. There certainly is Linux on Power, i.e. Watson, the system that beat the reigning Jeopardy champion ran on 10 racks of Linux/PPC [1]. There is also a lot of Linux in HPC, i.e. Blue Gene with the third generation Blue Waters coming [2] (aka Blue Gene/Q in marketing terms). Even though none of the three generations is not really a 100% Linux system since only the IO nodes run Linux while the compute nodes have some minimalistic kernel [3].
IDC gives estimates for operating system market share in the server market (see [4] for the second quarter 2012) and while all those reports should be taken with a pinch of salt (or discarded like the insane Itanium projections from the late 90s :), these do seem to make sense.
I quoted my numbers above from memory and it seems AIX is probably slightly less than five billion and the Linux hardware market clocks in a little higher at around ten billion a year. Ironically systems like the Sparc based Fujitsu K (HPC 500 current second most powerful system) also count there since it does run Linux and not Solaris. All the other deployments of those Fujitsu HPC systems I have read about also run Linux even though it is also offered with Solaris. That should tell us something about the HPC market and the demand for Solaris in that sector ;).
> I'm sure that many IBM shops have so much invested in AIX software that they don't take much advantage of it, but it's an option.
Yep, with virtualization there is the possibility to run mixed workloads and given how much effort IBM spends on KVM/PPC it seems that there is demand. Linux on S390 certainly seemed to have helped IBM in the main frame business.
Cheers,
Michael
1. http://www.novell.com/promo/suse/ibm-watson.html
2. http://www.ncsa.illinois.edu/BlueWaters/
3. http://en.wikipedia.org/wiki/Blue_Gene#Architecture
4. http://www.theregister.co.uk/2012/08/29/idc_q2_2012_serve...
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 16:53 UTC (Wed) by rsidd (subscriber, #2582) [Link]
If I wanted a non-Linux I'd use a BSD (in fact I probably will install one on my work machine, fairly soon). FreeBSD has both ZFS and Dtrace. Dragonfly BSD has a very impressive filesystem of its own, Hammer. Both have active and enthusiastic communities behind them. Meanwhile, Linux, whatever its faults, is rock solid and runs for months without a reboot (it would be years, but I upgrade to the newest distro now and then) -- and supports pretty much all the hardware out there. Why would I want to try OpenIndiana or any other variety of Solaris? The (ex-)OpenSolaris folks should try to answer that question first, before turning on members of their own community.
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 18:38 UTC (Wed) by alan (guest, #4018) [Link]
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 23:22 UTC (Wed) by vonbrand (guest, #4458) [Link]
I for one would much prefer some concrete examples of "bad design" (to think over and perhaps fix or justify) than some vague handwaving. Oh, well.
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 30, 2012 12:07 UTC (Thu) by mabshoff (guest, #86444) [Link]
Yep, the same applies to a lot of the BSD folks' attitude too, i.e. 'Linux is a mess', etc, and I have yet to see a list of specific issues in the kernel. You see criticism for udev and systemd, but to be honest they fix real issues and while they might be not the most beautiful design in some people's eye they get the job done. And no one forces you to use those two for example :)
Another point the descendants of Solaris seem to miss is that despite the open sourcing of Solaris 10 starting in 2005 it has not reversed the trajectory Solaris had. Many of the project participants will tell you that without the bureaucracy of Sun things would have turned out differently, but I have my doubts it would have made that much of a difference. Illumous and its derivatives are doing well as the basis for storage systems, but that does not magically turn it into a general purpose OS. If you look at their mailing lists you see development for ZFS/dtrace, storage and network drivers, but little else.
And Oracle seems to be most interested in milking the revenue stream from the install base of Solaris on one hand and at the same time move toward an application model with all the Exa* systems. And that certainly will not help more software to be supported on Solaris, i.e. the mongo DB reference made in the original email. And as the Solaris markets shrinks less and less commercial applications will be supported. Oracle seems to do ok with Sparc hardware development, i.e. see today's T5 hot chips talk, but it is pushing it into the massive DB server direction it profits from the most while its use in other sectors of the enterprise market is limited. And I am sure Oracle is making money on Sparc and will do so a long time into the future, i.e. I would not be surprised to see them selling T-somethings in a decade. Ironically most of the CPUs used for Exa* systems are x86-64.
If the rise of Windows as a server operating system to nearly 50% revenue of the server pie and Linux to 25% teaches us a lesson it should be that 'good enough' and 'familiarity' are potent forces. And while the time when Solaris was superior to Linux in every regard has passed a long time ago, any advantage Solaris might have today (i.e. ZFS, dtrace) is temporary as Linux equivalents mature. Btrfs is getting the needed enterprise hardening now while perf and LTTng 2.0 are really not behind dtrace any more imho. So give it another year or two for btrfs, time is on our side.
'Good enough' has killed many better alternatives, otherwise we would all be talking about OpenVMS :).
Cheers,
Michael
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 30, 2012 12:59 UTC (Thu) by drag (guest, #31333) [Link]
People don't give a crap if the file system is better unless they are using their system as a file server.
It's vastly more important for a system to be usable then better. Users using their applications are the #1 most important thing a OS has to well and everything else pales in comparison. Better 'usability' can come from better design, better familiarity, better compatibility, or a accident of history. Linux is more usable then Solaris for most people so most people are going to use Linux if they are required to make a choice. Linux does a lot of things terribly, but the massive work people put into it pays off in the end.
The 'Good Enough is better' mantra comes from people looking at software from a upside down perspective: Concentrating on technical differences first and user using applications second.
:)
Solaris innovations ported to Linux
Posted Aug 30, 2012 8:33 UTC (Thu) by CChittleborough (subscriber, #60775) [Link]
ISTM that Alasdair was criticizing the ports of various Solaris innovations to Linux, not Linux itself:... some of the core features of Illumos are becoming less and less important ... the Linux equivalents suck in one way or another, some are completely and fundamentally broken by designI think he has a point.
Something equivalent to DTrace, for example, would be wonderful to have on Linux — but it would require an enormous amount of work. Instead, the Linux developers are incrementally adding facilities which over time may well approach DTrace's functionality. It's a cathedral/bazaar thing.
(Personally, I'd like to see a Linux implementation of Doors, an IPC mechanism in the old-fashioned Unix style: powerful, efficient, rather minimalist. Full disclosure: I started that Wikipedia article.)
Solaris innovations ported to Linux
Posted Aug 30, 2012 11:31 UTC (Thu) by clump (subscriber, #27801) [Link]
The caveat is that it appears Oracle is playing games with licenses.
Musings and Open Questions
Posted Aug 29, 2012 17:20 UTC (Wed) by tstover (guest, #56283) [Link]
Large scale software engineering requires extraordinary resources. Perhaps it is the way of things that only so many such projects are sustainable at once, or perhaps not. There was a time when many people like myself couldn't wait for total "full spectrum" world wide Linux domination. Today I would argue that such an aim is the wish of a fool. Every time a technology is left behind, a project abandoned, a dream left to wither and die - the grip of monoculture tightness around our neck.
Remember when "all of us" couldn't wait until computers were so cheap that everyone could have as much hardware as they could ever dream of? In such short order we filled landfills with yesterday's treasures. Just as fast in so many ways, we abandoned skills, craft, and respect for our tools. Seemingly forever adrift in a future of unhackable, disposable, inefficient consumerism.
Remember when "all of us" couldn't wait until the Internet was omnipresent? Remember that it seemed that it would take forever? Well that came so very fast, and with so much praise that so few even noticed all that was lost (hint: watch BBS the documentary).
Remember when "all we wanted" was to just get to use Linux for everything? Well (despite these nearly daily popular hallucinogenic alternative reality articles to the contrary), that day was long ago. Linux essentially has killed just about everything else. So is that good or bad?
It's simply nature. Things live, they die, and evolve. Here's the thing though - we are humans, we exorcise dominion over nature, not the other way around. (That is we aim to anyway.) We annihilate entire species, then place others into protective breading programs. Why? Because there is inherent value in the diversity itself!
The scope of human endeavours into computing is such a rich history to which we are all heirs. The question is what is it that we choose to do with this legacy. At the very least, know and let it be known that there are those whom - regardless of their motivations and intentions - do try to take the roads less travelled even if we perceive them as adversaries. Our hearts should always long for exploration.
So the next time you start to work yourself into another heart attack about say, Gnome 3, consider this. Sure not all ideas are valid, but if it were not for all those other desktops & window managers, the cage would be closed on many of us.
---
As for what remains on the Solaris branch on the tree of life, I would like to hear from those in the know. What exactly is out there that is worth "supporting" at this point? I remember being very intrigued at the NexentaOS concept (gnu/linux sans linux replaced with solaris kernel), but after that initial release it never made it to another VM for me. I currently have the OI on an older box that I did actually verify some things of mine at least build on, but that was long ago. The debian kernel variants might be worth loading up. Someone actually wrote me the other day using "solaris". I wonder what that even means now. Is there a download on oracle?
Musings and Open Questions
Posted Aug 29, 2012 17:32 UTC (Wed) by louie (guest, #3285) [Link]
Our hearts should always long for exploration. That's completely right. The question for OpenSolaris, of course, is and always has been - what are you exploring? GNOME 3, for better or for worse, is exploring some very different ideas about what a shell looks like and does. OpenSolaris appeared to be exploring fairly well trod areas - open-licensed operating systems - with no good explanation of what new ground they were going to cover (other than dtrace and ZFS). If they'd been able to explain their vision - or, in your terms, explain what new ground they were exploring - they would probably have attracted people who were interested in joining them in that exploration. But they never were able to do that, as far as I could tell.
Musings and Open Questions
Posted Aug 29, 2012 17:38 UTC (Wed) by tstover (guest, #56283) [Link]
Musings and Open Questions
Posted Aug 29, 2012 21:29 UTC (Wed) by Fats (guest, #14882) [Link]
Big smile.
Musings and Open Questions
Posted Aug 30, 2012 6:51 UTC (Thu) by deepfire (guest, #26138) [Link]
> to wither and die - the grip of monoculture tightness around our neck.
This makes for a wonderful quote, IMO.
Amen.
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 17:30 UTC (Wed) by xnox (subscriber, #63320) [Link]
Wait did OpenSolaris / OI have an awesome init system or is this just a misuse of upstart out of place?
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 18:52 UTC (Wed) by mabshoff (guest, #86444) [Link]
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 19:25 UTC (Wed) by drag (guest, #31333) [Link]
It's was a bit disturbing when I realized the similarities, but oh well.
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 19:23 UTC (Wed) by clint (subscriber, #7076) [Link]
http://wiki.openindiana.org/oi/Creating+new+services
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 29, 2012 18:34 UTC (Wed) by landley (guest, #6789) [Link]
Remember when Sun abandoned the workstation market (thus cutting off its developer base), then killed Solaris-x86 (thus cutting off what its developer base had migrated to), then brought it back but killed it off again a few years later because they'd literally _forgotten_ the outcry the previous time? (And then gave their salesdroids a bigger comission on each Solaris X86 seat sold than the customer paid for the license?) I guess it really does take a direct meteor strike to kill off a dinosaur.
OpenSolaris was always a Sun internal political thing. The Sun Civil War a decade back was between the once-unified "Sparc/Solaris" block and the "children of Java" (which included openoffice, looking glass, Sun's windows guys... Yes Sun had a Windows faction, which arose from the "Windows JDK" folks but later included the "OpenOffice has more deployments on Windows than anywhere else" guys). Java was always bigger than Solaris, but Sparc and Solaris were where the political power was in Sun and they killed anything that was a threat to them, keeping Java on a tight leash that seriously neutered its potential. That's why Java and such weren't open sourced for so long, the Sparclaris guys squashed it (Danese Cooper's entire job at Sun, for 6 years, was fighting against this). But when Bill Joy went crazy Sun bought co-founder Andy Bechtolsheim's startup that sold Opteron servers and Andy made 'em port Solaris to the Opterion. That split the Sparclaris faction down the middle.
The real tipping point seems to have been the Windows guys getting massive payments from Microsoft (for shilling for SCO during that FUD mushroom cloud):
http://www.microsoft.com/en-us/news/press/2004/apr04/04-0...
http://news.cnet.com/Sun-gets-second-Microsoft-patent-pay...
Money is power in a company like that, and when the rise of Opteron left Solaris isolated they struck back by forcing Solaris open source (kicking and screaming) to politicaly neuter them (and as payback for the Solaris guys had done to Looking Glass).
If you want to get a feel for Sun internal politics and a bunch of good history, read http://www.blinkenlights.com/classiccmp/javaorigin.html
Rob
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 30, 2012 19:21 UTC (Thu) by sumanah (guest, #59891) [Link]
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 31, 2012 16:18 UTC (Fri) by mabshoff (guest, #86444) [Link]
I knew the 'have you ever kissed a girl' line, but until now was unaware that it was written by Cantrill. Ever since in his LISA 2011 talk (see http://www.youtube.com/watch?v=-zRN7XLCRhc for example) he has rubbed me the wrong way mostly due to his comment on Linux filesystem developers being amateurs, i.e. the clown college remark toward the end of the talk.
Given that he was an instrumental part in porting and merging KVM into the Illumos base I have questioned his judgement since he does not see a problem in merging GPLed code into a CDDL codebase. That decision convinced me that they would have a hard time as a commercial product since the legal risk of an imho obvious GPL violation would be too large.
Cheers,
Michael
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 31, 2012 15:10 UTC (Fri) by cdmiller (guest, #2813) [Link]
"graft"
Posted Aug 30, 2012 13:43 UTC (Thu) by sumanah (guest, #59891) [Link]
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 30, 2012 14:02 UTC (Thu) by ripratm (guest, #86482) [Link]
If you want a general purpose Illumos OS, OmniOS is actually being updated regularly (unlike OI).
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 30, 2012 22:58 UTC (Thu) by peterj (guest, #86487) [Link]
Alasdair is right that GNU/Linux is a monoculture - just because it's OSS doesn't prevent this.
Some of this is active antagonism to anything other than Linux. Witness the attitude of MongoDB in the link Alasdair posted: "you're not using Linux, we don't want to hear from you". And cases where Linux developers have deleted all the code not needed for Linux on previously portable code.
This is fostered by GNU's "extend-and-embrace" approach - lots of seemingly nice but non-standard extensions that lock you into GNU. Whether intentionally or not, lots of OSS developers write "C" code that can only be compiled by gcc and scripts beginning "#!/bin/sh" that fail on POSIX shells. Libtool and autotools are classic double- speak - they claim to promote portability but actually do the opposite - the code might be portable but autotools and libtool still endeavour to make building the code as painful as possible unless you have a GNU toolchain (preferably on Linux).
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 30, 2012 23:27 UTC (Thu) by dlang (guest, #313) [Link]
I disagree with this. The ecosystem requires lots of experimentation, some of those experiments result in things that are _worse_ than what was there to start with.
failure of experiments is the flip side of being able to experiment.
some failures are small (how many people write perfect code the first time? how many people have never written a patch that broke something?)
some failures (distros, the OI project) are much larger. It's too bad that it took so much time and resources, but even in failure there are good things that happened
There were a lot of people who got involved in FOSS software who had not done so before. Yes, we will probably loose some of those people as the one project they were working on goes away, but I'll bet we don't loose them all. And even those who we loose as developers are probably now far more willing to use or recommend FOSS software than they were before getting involved.
> Alasdair is right that GNU/Linux is a monoculture
It's funny to read this at the same time I'm reading that the reason that we haven't had the "year of the Linux desktop" is because Linux is so fragmented :-)
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 31, 2012 2:48 UTC (Fri) by marcH (subscriber, #57642) [Link]
OpenIndiana lead Alasdair Lumsden resigns
Posted Aug 31, 2012 11:04 UTC (Fri) by drag (guest, #31333) [Link]
The solution that has come up so far is to put a HUGE amount of effort in repackaging the same software a dozen different times to make up for those relatively tiny differences.
Meanwhile people that actually are trying to move things forward are always met with derision and are lambasted for not following a imaginary and continuously shifting unix methodology/ideology that is used to justify all sorts of weirdness in the system. The flip side of the coin is full of people whose main approach to deciding system design is simply to not decide anything at all and make users and developers deal with a thousand different possible combinations of this-or-that.
It is not so much a monoculture as a monodisfunction.
But in the end I still prefer it to the alternatives.
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 2, 2012 21:04 UTC (Sun) by nix (subscriber, #2304) [Link]
Libtool and autotools are classic double- speak - they claim to promote portability but actually do the opposite - the code might be portable but autotools and libtool still endeavour to make building the code as painful as possible unless you have a GNU toolchain (preferably on Linux).This is obviously an entirely different autotools and libtool than I have used, perhaps from a parallel dimension or something. The autotools (and gnulib) maintainers are some of the most portability-crazed people I have ever heard of.
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 3, 2012 11:13 UTC (Mon) by schily (guest, #60311) [Link]
libtool handles "portability" on Linux but does not even handle basic ld features like -L and -R options correctly.
All a bit more complex projects have problems compiling on Solaris because usually -R is not used and the linked binaries thus will not find their libraries and this is definitely not a Solaris caused problem.
BTW: even "autotools" is a step from autoconf into the wrong direction. Autoconf (as documented by the FSF is based on self modifying makefiles - a technique from the 1970s). What the FSF calls "automake" is definitely not an automake program but something that makes people believe that it will create portable makefiles but does not. What it really does is to propagate code that depends on bugs and extensions of gmake.
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 5, 2012 23:57 UTC (Wed) by nix (subscriber, #2304) [Link]
Well, this is the kind of attention to detail that gives you (and Solaris, alas, if people should think you have anything to do with Solaris, which thankfully you do not) a bad name. (If people want to see what Solaris hackers are actually like, I recommend talking to Alan Coopersmith. A more charming and helpful man was never born.)The -R (DT_RPATH) handling in Libtool was very carefully thought out specifically so as to work properly on both systems with an /etc/ld.so.conf-equivalent (like Linux) and systems which prefer DT_RPATH (like Solaris). If you have a problem with it, I might recommend you report it: but configure / make / make install with a variety of prefixes and DESTDIRs works perfectly in my testing. (There have been bugs in this area, but not for a long time). It is definitely unacceptable to just use DT_RPATH, since that would torpedo Linux systems. Instead, we provide an implementation which permits you to use -R to get the effect of DT_RPATH, while instead providing whatever implementation is preferred on the system being used (including 'none' for systems like Linux, if the user so requests). That's kind of what Libtool is all about, after all.
As for 'self modifying makefiles', you really do have no clue what you're talking about here. Autoconf is based on M4-driven generation of shell scripts. Automake is based on Perl-driven generation of makefiles (which do in fact warn about non-POSIX-make features, but most people sensibly don't care because GNU make is portable and most non-GNU makes are a horror show: here I explicitly include the coredumping halfwitted feature-free pile that is Solaris make: I have fairly frequent nightmares about the ten years I spent forced to keep things working with it).
Nothing whatsoever in GNU autotools is based on 'self-modifying Makefiles'. Perhaps you are talking about the dependency generation in Automake-driven makefiles, which does rely on generation of makefile fragments which are then included, but does not require self-modifying a makefile, unlike the 'make depend' system it generally replaced.
The makefile-fragment scheme may be 'a technique from the 1970s' but I'll note that this is both a non sequitur and untrue, since makes didn't have the ability to automatically reload their makefiles if their include dependencies had been modified until long long after that. Furthermore, it doesn't matter how old the scheme is: it works, and it is completely transparent to the user, unlike the horrible old 'make depend' system, which worked fine until you deleted or moved a file with generated dependencies, whereupon everything fell down in a screaming heap until you hand-hacked the dependencies or did a complete 'make clean' and rebuilt. Clearly you haven't read Paul D. Smith's excellent paper on this algorithm (due to Tom Tromey), which clearly describes the reasons why the scheme was implemented as it was.
(Actually, sorry, perhaps I was wrong: quite possibly you have read that paper. I'd forgotten how immune to reasoned argument you are. My apologies. For a moment I thought I was talking to a rational being.)
(corbet: sorry about the tone here, but I consider it important to refute schily's untruths, and further that, to be blunt, driving off schily before he converts the entire site into a single giant flamewar is more important than civility in this instance, not least since schily has long since proved himself incapable of civilized discourse. We are talking about someone who got himself banned from the Debian bugtracker(!) for repeated kibozing and trolling, after all.)
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 6, 2012 10:45 UTC (Thu) by schily (guest, #60311) [Link]
Attacking people for a bug description is more than a fauxpas....
Doing this with a non-personalized nick is even worse.
BTW: this was different before aprox. 2004. Before that time Linux people generally have been very polite and helpful. Since then there are so many unfriendly people that these people unfortunately dominate the rest in the public.
Looking at your reply shows that you don't know what you are talking of.
- If you like to associate -R with an environment, it is LD_RUN_PATH.
DT_RPATH does not exist.
- Linux uses the old runtime linker that was written for a.out
This is why linker-caching and path setup exist. ELF has a RUNPATH
property that is better than the method in the old a.out system.
- It may be that recent libtool versions are better, but there is too much
software out that uses non-working versions
- It seems that you just know too few things about GNU autoconf.
Knowing that it is based on m4 is not sufficient to understand that
the documentation from the FSF in how to use it is based on outdated
technologies from the 1970s.
- Using "configure" to create something like config.h (as done in the
schily makfile system) is OK
- It seems that Paul Smith copied the ideas for a more modern
autodependency handling from Sun's ideas and implementations in
SunPro make from 1986 and from the Schily makefilesystem that
implemented this method in a portable way in the 1990s. He however
implemented the support in a gmake in a buggy way as gmake spits out
warnings for things that are expected due to his strange algorithm.
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 6, 2012 13:04 UTC (Thu) by mpr22 (subscriber, #60784) [Link]
It may be that recent libtool versions are better, but there is too much software out that uses non-working versions
That's not an argument for claiming that libtool is bad. That's an argument for claiming that the software that uses bad versions of libtool is bad (and I will cheerfully admit that there's a lot of badly maintained software out there that, if it runs on anything except Linux, does so only by serendipity).
ELF has a RUNPATH property that is better than the method in the old a.out system.
And the GNU dynamic linker, used on Linux, supports it.
OpenIndiana lead Alasdair Lumsden resigns
Posted Sep 8, 2012 16:08 UTC (Sat) by nix (subscriber, #2304) [Link]
Let's go down your fun little list one by one, as far as it is worth it to go.
- DT_RPATH does exist. It is the name of the ELF dynamic tag that is used to implement various linkers' RPATH flags. (It sucks because it wires paths into binaries in a way that is impossible to override without binary editing: DT_RUNPATH sucks slightly less, but still wires paths into binaries, which is less than good).
- The glibc runtime linker was rewritten from scratch for ELF support. It doesn't support a.out, and never has. Linker caching exists because it provides a huge speedup over searching at runtime; its search path exists to allow that to be customized; and RUNPATH is a mild abomination, as you would immediately realize if you had ever had a binary with a runpath that had a path baked into it (often a build path) which was pointing to an NFS server that had become unavailable. I've used Solaris for many, many years, and believe me the number of bugs we had reported that were attributable to mis-set LD_LIBRARY_PATHs (which were *all supposed to have the same value* and which on Linux would just have been set in /etc/ld.so.conf) was simply astonishing.
The rest of your comment is as far as I can tell pure ad hominems, arguments from authority, and straw man attacks, not worthy of a response. ("based on outdated technologies from the 1970s", I'll bet if I ask the actual authors, the only people who might know, they would disagree.)