The sudden death and eternal life of Solaris
dtrace.org
dtrace.org
You can use illumos today, right now, and have your ZFS, mdb, DTrace and zones. It really is open source and we're a community using and improving it. For 7 years now already.
As illumos is only the basic building block of the operating system (but unlike linux includes kernel + basic userland) one usually runs one of the distributions:
- SmartOS, developed by Joyent as a cloud hypervisor. Supports zones, KVM and lx-branded linux containers. [1]
- OmniOS CE, a minimal distribution targeted for bare metal server installations [2]
- OpenIndiana, similar to the traditional opensolaris. If you care about GUI this is probably the one. [3]
- Tribblix, modern components with retro style [4]
Sure, the user base is smaller than Linux. But that is also true for FreeBSD which is used by Netflix and Whatsapp. Running a different OS than most other people can give your company an advantage and I know that ZFS and zones have done that for mine.
[1] https://wiki.smartos.org/display/DOC/Download+SmartOS
I will say the wikis are not inspiring; many on Illumos, OpenIndiana and SmartOS have pages last edited multiple years ago. For Illumos, I think your info above is more useful than the alphabetical listing of distros on the wiki. I think the marketing on OI's site could better tell the "why use me over Fedora/Ubuntu/BSD?" story, and SmartOS could spend less time selling its public cloud service, and more time telling me how to create a private cloud. Its docs are not super discoverable.
All just observations; take them for what they're worth. I do think it's neat that Solaris has lived on, and it makes me want to throw up a lab server to poke around a few hours. Thanks for your work and the comment.
A lot more of activity is going on in irc on freenode, so if you start playing with your lab server or are just interested in general you might want to drop by in #illumos or #smartos.
- zones allowed for a much higher deployment density and better utilisation of our hardware than other virtualization technology would have.
- zfs snapshots. We've integrated those into our product and when a customer messes up they now have button to instantly restore to a known good state.
Doing both of these things would have been a lot harder on another os.
Potential areas for improvement could be anything from filesystem to process concurrency to virtualization.
Things like:
* Java
* PHP
* Perl
* Python
* Ruby
* Apache
* Nginx
* Varnish
* MySqL
* Postgres
* Node.js
* etc
Not debating here, BTW; I enjoyed my time with Solaris and then its offshoots.
Edit: The thinking here only applies to systems running in production where official support is pretty much mandated.
First: Really, the requirement for vendor 'support' is something I've always had a bit of a hard time rationalizing.. Getting a vendor in for consultancy/implementation help is a bit of a different story though and is often worth it.
Back to support.. The old "Alright, we pay RedHat or Oracle for support for when something breaks" thing.. The reality is when something does go arse over bollocks you'll get two options from your vendor (at least from sun, oracle, redhat, ibm and sap in my experience): upgrade to the latest version, or downgrade until you get a code fix from the vendor. Your critical oracle 10 db running on some older redhat point release shits itself? Not their problem. Hope you have backups. They will take file a bug internally to prevent it happening again but they can't do much more than you can internally to get things back up.
I've been in ops for almost 15 years and I still don't have a single story of a vendor 'saving the day' outside of hardware support (actually, that's not true -- once Joyent solved a problem for us by patching some servers within about an hour of us reporting a problem which remains the single greatest support experience in my life).
We worked out how to do software upgrades and rollbacks a decade ago, if one shitty linux kernel package is enough to down your business then yeah, you're doing things wrong and hopefully will learn from the experience, but paying a support fee per instance isn't going to help you recover or avoid those things.
I think we agree, all I'm trying to say is at least in my experience: You're almost always better spending your money on good staff than support contracts.
Second: In what world do you live in that your sysadmins can write kernel patches during prod outages? Maybe your reality is very different to mine, but adding/changing code has never been the solution to bad outages (and I've been through some gnarly ones, trust me) for me so far...
The people who is going to suffer the most from Oracle killing of Solaris are largely its own database customers. I honestly doubt many Solaris customers are running an open source stack on top of it. It's the closed source applications that will be trouble some.
1. It is less profitable to develop exploits targeting a minor OS.
2. Minor OS may have some specific features that the major OS don't.
We're in a dark ages with a sliver of light. We might be in it for a while. People who have never used ZFS, SMF, DTrace, or FMA, have no idea that they are stuck in the dark.
There is nothing odd in that: contrary to the Linux party line and all the Linux hype, before illumos and SmartOS, Solaris was the most advanced operating system on the planet. As others have very accurately described and pointed out in the comments on the other reduction in force thread, this was sadly lost on the Linux generation which "grew up on the PC in their parents' basement".
Is your startup / company deploying on Solaris? Nope of course you aren't - who is?! Pretty much nobody nowadays - w3techs.com shows Solaris at 0.0005% for webservers for example. Everyone talks about how great it is(to be sure there are some small bits that are pretty dang amazing) - yet for some reason nearly no one actually DOES use it! Perhaps there is a reason for that. Lord knows I have many battle scars from the many real world rough edges that never seemed to get less fundamentally shitty(for starters: path_to_inst and smf I'm looking right at you!)
I definitely respect the amazing work of some of the engineers on the Solaris team back in the day. ZFS and DTrace being two that have made the tech world by in large a better place either directly or thru inspiration. And Linux and the BSDs are definitely better for those selective code gifts! But nonetheless, the world voted with their feet, and Solaris didn't make the cut.
This is a crop out that you're giving on behalf of Oracle.
Oracle does this all the time. It's their modus operandi.
They buy companies/technologies out, milk them for license fees and let it trudge along until it stops making money anymore and either close shop or abandon it to Apache Foundation.
There are anecdotes out there ranging from technologies they've bought to sale representatives having their bonus taken away from selling licenses.
Your statement also just ignore their cowardly act of re proprietrating solaris. They did the same with Java not giving out TCK for Apache open source Java. There are many other instances of hostile actions toward the open source communities.
That's not just Oracle's fault. Place blame where blame is due(java) - but blaming everything on Oracle is kinda silly here. Sun mis-stepped pretty hard in the years after the original .com bubble and what we are seeing now the the final result. I'm shocked it took this long.
Sun made some strange moves toward the end, but there absolutely could have been a product line left if the developer market felt neutral about the buyer and the buyer tried to focus on upsells and professional services. The way Oracle tried to sneak this EOL in is very much evidence of there still being support licenses and professional services money for a few years more.
But it was worth it if the Sun curse took down Ellison. May your foot never leave your mouth again, cloud boy. Now go play golf with network's owner.
I'm shocked, I say shocked, that it took much longer for System V to kill it than you're shocked, you say shocked, how long it took Java to kill it. ;)
It was difficult to get much experience with Solaris or the other commercial Unixes because you usually had to have particular hardware. Even when Solaris became free and ran on x86, it was picky about hardware. By then it seemed to me that even running BSD was kind of an uphill battle, if you wanted a VPS or something.
I'm not a completely neutral party. Sun totally botched a job interview process with me when I graduated, and I came away feeling that there must be a lot of internal confusion. Their public side certainly gave that impression too. I knew they had cool technology in Solaris, like dtrace and zones, but apart from the names you almost never saw someone who attributed their success to these technologies. And then of course Oracle and the slow death.
https://en.wikipedia.org/wiki/Symbolic_link#Variable_symboli...
>Pyramid Technology's OSx Operating System implemented conditional symbolic links which pointed to different locations depending on which universe a program was running in. The universes supported were AT&T's SysV.3 and the Berkeley Software Distribution (BSD 4.3). For example: if the ps command was run in the att universe, then the symbolic link for the directory /bin would point to /.attbin and the program /.attbin/ps would be executed. Whereas if the ps command was run in the ucb universe, then /bin would point to /.ucbbin and /.ucbbin/ps would be executed. Similar Conditional Symbolic Links were also created for other directories such as /lib, /usr/lib, /usr/include.
https://en.wikipedia.org/wiki/Pyramid_Technology
...And the hardware wasn't all that reliable either!
http://art.net/~hopkins/Don/text/gymble-roulette.html
At least you could run Space Invaders on the system console while it was down waiting for repairs.
Although less extreme, Solaris multiarch was not exactly always a picnic.
From when it started becoming a lot less insane:
https://www.perkin.org.uk/posts/multiarch-package-support-in...
People my age and younger just didn't get much access to commercial Unix environments when we were growing up. Once we got into industry, we found it hard to see the justification for high Unixes when we had experienced good results with low Unixes. So the low Unixes have mostly won the war. This story repeats time and time again in our industry: the freest foo conquers the non-free foo, even if the non-free foo is technically superior.
inst> keep *
inst> remove incomplete overlays
inst> conflicts
no conflicts
inst> go
Nobody is asking for a round of applause for Oracle, but the scorn poured on them for this seems misplaced.
I was very negative on Sun, but looking at Microsoft fill the gap eventually, I have to assume IBM would have made me eat my shoe if they got the deal instead of Oracle.
cop out :)
[edited to add firewall, that first gen that no-one used because it was so complex]
Business is fundamentally uncertain, and decomposing failure into avoidable and unavoidable factors is very very hard. This is even harder in industries dominated by network externalities like computer platforms.
On the other hand: if engineering wasn't serving the stated business purpose (things that customers want, at least as far as business can see) but doing semi-academic research on the company dime... well, it's an indirect misappropriation of funds, however cool the results turned out.
Analogy: Sugarcube Industries is in the business of farm built infrastructure and its flagship product is a barn that can comfortably house N horses cutting down on injuries, promoting weight gain by proper thermal environment. I don't know much about horses, bear with me.
Engineering is asked to design such a horse barn. Maybe it's fundamentally impossible to fix heat flows in such and such weather or idiotic in general to try and improve horse health by those means. Maybe business even understood market research wrong, and what people wanted was barns with hooks where farmstaff can hang horse ponchos that will be manually deployed in times of need.
It's a bad look on engineering if instead of delivering the best horse barn on the market, they come up with a three-storey optimized IVF-incubator-horse feeding operation.
----
Unix (a decades-long project whose capital costs are inestimable when considering all the failed companies and free labor) nearly lost to Windows NT because the latter had hooks where sysadmins could hang those needed ponchos.
For example when was the last marketing campaign that you can remember for Solaris?
Besides, they've started killing Solaris since 2010: http://garrett.damore.org/2010/08/hand-may-be-forced.html
> And Linux and the BSDs are definitely better for those selective code gifts!
AFAIK DTrace and ZFS can't be used with Linux due to their license being incompatible with GPL.
Ubuntu packages it in the default distro, your base install can have ZFS as root with no configuration.
For Debian/Fedora I think you need "non-free" packages added.
for debian based: add non-free repo and apt install ...
while installing, there comes a prompt were you have literally no choice but to accept the incompatibility of the licenses (and move on :)
For precisely the reason of Oracle's direction from day1
We run a number of physical hosts using SmartOS (only 20 or so, but they're quite big (512GB, quads)) and on top of those we use a mix of KVM and zones (both solaris and lx) for various workloads (around 400 zones/vms).
It's easily been the correct choice for us. YMMV.
if a company is using solaris, it likely host a Linux kernel and apache in virtualized containers to serve web content.
https://www.youtube.com/results?search_query=bryan+cantrill
It's sad that we live in a world where a company doing good , innovative and useful tech is gobbled up by a company like Oracle who should not exist and do largely on their ability to con ignorant senior management and even more ignorant boards of directors.
Is there ever a technical reason to use Oracle? If there is what percentage of Oracle sales does it account for? Ie if it really is non-zero I bet it's a single digit percentage. Woeful company from the top down is their reputation and they've earned it.
Today, Postgres has come a long way, and distributed databases are much more developed. There is probably much less technical justification for using Oracle than there was 10+ years ago.
Also many companies would pay for Oracle and just use it for CRUD operations. That was throwing money away, no doubt.
That's never really been true. From day 1 it's been neck-and-neck with DB2.
I understand that sometimes you do the composition of many CRUD operations, but this is still CRUD.
But they also provide stronger typing and checking of the data. They provide a programming language neutral way to access the data quickly and efficiently. They index data to make queries on it faster. They do much of this automatically so you think it's all simple CRUD operations, but they also provide ways to explain what they're doing if it isn't working as expected. I can count the number of programs (not DBs) on my hand that I know of that can explain themselves accurately without needing a debugger.
Edit: I don't mean to be annoying. Most people on HN knew this and I'm not assuming you didn't, but while most people just use RDBMS for simple CRUD operations they're losing out on the power that these databases have been designed with.
However, I've come to dislike building business rules using the proprietary languages of the RDBMS vendors. In particular, I use Oracle and have been making a case with my employer to stop creating ties to Oracle by using PL/SQL as our house language. While it provides some efficiencies and can sometimes make coding easier, it can stifle development as easily.
I'm all for expressing data relationships in the RDBMS and using efficiency mechanisms like indexes and cost-based optimization. However, I now prefer business rules to be expressed outside the database.
[EDIT: Corrected anarbi to arnarbi.]
The question I pose to anyone consider this type of architecture is, how often do you switch database platforms?
I work for a moderate-sized university. The applications I support require up-to-date information in order to test modifications. They are read-intensive with few writes within the database. We can afford only one test database for the entire university. We refresh the database from production only once a quarter, at most. My applications would benefit greatly by being able to switch URL's to the production database in order to test a mod. Since our code, written in PL/SQL, gets stored in the database that it runs on, we can't do such a thing without significant effort: dblinks or modifying package names to store in production for testing or a test schema or whatever else you can think up.
Another issue is that PL/SQL is not a robust language for modern development. Its type-system was designed for strictness, which is great in life-critical systems. But, there is no notion of inheritance and no ability to write generic collections.
To mimic inheritance, you can create objects backed by tables and inherit one from the other. It has the feel of being bolted on and yet another tie to Oracle.
One final issue is that PL/SQL is highly subject to the resource management of the Oracle kernel. Often, the kernel is extremely efficient. But, there are times when you need code running in a separate address space, preferably on a separate server. For instance, one application I wrote is a batch system that has a lot of processing rules built into it. Our database is tuned for OLTP. My application is categorized by the DBA's as data warehouse-oriented. Yet my application has to run hourly throughout the day and compete with resources necessary for the OLTP stuff. Putting it onto a separate server, away from Oracle's kernel management, would more than likely help a performance issue we have.
So, the answer to your question is the one you were looking for. But, my concern isn't necessarily switching database vendors.
Your DB either supports ACID or it doesn't. The term "stronger guarantees" is snake oil. It's either guaranteed or is not. That simple.
If table stakes for a database are more sophisticated than CRUD, what the hell is trivial enough to be CRUD then? Seems like a straw man.
Plenty of apps--perhaps more apps than not--use a database as a vaguely typed (MySQL out of strict mode barely gives a nod to them) multiplayer notepad.
My first job was even writing dynamic web apps entirely in Informix.
Do you think of better ways to handle this nowadays? Is it more effective/maintainable to do this in the app layer?
But I don't know, I don't really do this stuff anymore. At Google we certainly have storage and database systems with a lot of non-CRUD features (look at Spanner for example).
Oracle's only redeeming quality seems to be free golf at nice country clubs paid for by Oracle salespeople.
Same applies to their stored procedures languages and deployment options.
Because pgAdmin isn't it, specially after it became an Electron app.
pgAdmin3 wasn't a terrible application. Nowhere near SSMS, but not terrible (other than stability... it's shit there).
pgAdmin4 is unusable.
Which is why the villain in Charlie Stross's _The Jennifer Morgue_ is "Ellis Billington".
Every single first person or second hand account I've heard about Oracle makes like a terrible place to work... is this just people being hyperbolic, or is is truly that terrible?
I'm on the outside looking in and don't have any special internal insight, but they didn't just pull this robo-RIF idea out of their ass spur-of-the-moment. Oracle is the kind of place that maintains a robo-layoff system, which pretty much says all you need to know.
"What you think of Oracle is even truer than you think it is. There has been no entity in human history with less complexity or nuance to it than Oracle."
Based on the stories I've heard about Oracle and Elison, I've often thought that Bakan might have had an inside informant at the company.
https://www.amazon.com/Corporation-Pathological-Pursuit-Prof...
(Saved you a Google)
You answered yourself
Support.com is another big offender.
Any connection with the Stanislaw Lem novel (or Tarkovsky movie) was always a bit unclear to me... But I guess a book about futile interactions with a planet-sized alien brain that doesn't care about you other than mysteriously experimenting with your memories is a reasonable metaphor for the Unix user experience.
But to appreciate the book have to understand that Lem, in his serious mode (he also wrote comedies and light-hearted adventures, which are quite different), is not very concerned about plot, nor about most things American sci-fi has always been about: There are no heroics, no politics, no space wars, no McGuffins, and there's only minimal world-building.
Lem is an engaging writer, but also a somewhat dry one, and Solaris is a sort of essay on science, camouflaged as a novel: Like his other masterpiece, His Master's Voice, it is a critique of science. Whereas HMV attacks the scientific field pretty broadly, Solaris is more of an investigation into the limitations of the human mind: A scientist visits a far-away research station and comes to realize that all previous research (by decades of so-called "Solarisists") isn't even wrong, it's irrelevant. About a third of the book is spent on a wry summary of all the scholarly research that Solarisists have been performing and collecting in journals. Even after many decades, Solaris has funded many scientists, and yielded endless observations, but also zero insight. (At the time the book is set, most scientists have simply given up, and the station is largely abandoned.) Solaris is an inscrutable alien organism that the human mind can simply not understand. Ironically, even the main character comes to think, towards the end, that he has understood Solaris, when it's obvious to the reader that his conclusions are just more of the same, a reflection of our human-ness.
Of course, both movie adaptations got this wrong, or chose to ignore that aspect of the book. They think it's about people reincarnated from memories, and that it's about love. In the book, the "ghosts" are actually dealt with rather simply and matter-of-factly; the scientists find themselves being tormented by them, and quickly move to destroy them. Lem, however, doesn't bother with the emotional melodrama you find in both movies.
I don't remember what Soderbergh's version aimed for exactly (other than being a sort of Last Tango in Paris in space), but Tarkovsky's is very much inward-looking and concerned with the nature of artifice and truth (the "ghosts" are not real, and neither is Solaris, in a way). Tarkovsky didn't really want to create a sci-fi movie, and so you get something which is not entirely convincing.
In a similar tone, I actually love the Tartovsky film. A very weird piece of work, and a hell of a thing to experience a few weeks before a summer spent in Russia...
Also people, please don't downvote parent. It's very fair to say this movie is boring and confusing, and I think it was the very point. Tarkovsky has this style of filming around supernatural / sf things but never actually showing them. Instead, he prefers to explore the anxiety they cause on humans, anxiety which he often represents through long waiting with deceptive conclusions, so I would say that feeling annoyed is quite what is expected from people watching them. It's art, not entertainment.
As for Solaris, it is worth reading the book as well as watching both versions of the film -- they ask and explore different questions. The book is about the nature of intelligence and life whereas the films are more about the human part of the equation -- what does it mean to be a human as opposed to be a "replicant"? Tarkovsky is slower paced, but IMHO encourages deep reflection. Soderbergh's version is definitely more watchable and is actually quite thoughtful by Hollywood standards.
One of the keys to Tarkovsky's movies is that he is concerned with human relationships primarily. Romantic relationships, parents and children, etc. Even when he directed "sci-Fi" movies, they were never about rockets, aliens, shooting, explosions.
This quote from the movie, illustrates that well I think:
---
We don't want to conquer space at all. We want to expand Earth endlessly. We don't want other worlds; we want a mirror. We seek contact and will never achieve it. We are in the foolish position of a man striving for a goal he fears and doesn't want. Man needs man!
---
I remember watching this movie and being disappointed because I expected more sci-fi stuff. Same thing with Stalker, his other sci-fi film. Only after seeing The Mirror (my favorite of his films) started liking his other movies.
Besides Solaris, his meditations on life and society in The Cyberiad remain some of my favourite science fiction of all time.
The Futurological Congress is also a lot of fun; it's one of his comedies, a satiric look at futurism.
His Master's Voice is advanced level Lem. It's less a novel than a fictional autobiography of a scientist, and while the subject matter is fascinating, it's a bit dry.
The Cyberiad is a classic, but very different and not really sci-fi. It's a collection of absurd, surreal fairytales about a pair of robots. It contains math puzzles and rhymes and lots of weird stuff.
Fiasco, Memoirs Found in a Bathtub, Tales of Pirx the Pilot and Star Diaries are supposed to be good.
I found Eden (one of his early works) to be a little juvenile and poorly written. It, too, is about first contact gone wrong.
Like Blade Runner's relationship to Do Androids Dream of Electric Sheep, it was a lot different than the book, but shares some deep ideas, and stands on its own as a great movie.
The scene in the USC ICT's motion capture studio was riveting, with Robin Wright playing a partly fictionalized version of herself, and Harvey Keitel playing her agent, baring their souls to the giant emotion capturing machine.
https://en.wikipedia.org/wiki/The_Congress_(2013_film)
As far as I've read, plot has absolutely zero to do with The Futurological Congress. The only similarity is that there's something in the movie called the "Futurological Congress" and that there's some kind of attack on it.
If you think it's worth seeing, maybe I'll give it another go.
It may be an easier entrypoint than Solaris, at cost of arguably not being as "deep" in its reflections.
Michael Kandel was Lem's translator for many books including The Cyberiad -- including some brilliant poetry and plays on words!
He's such an excellent translator, it would be interesting to read other stuff he's translated! Any recommendations?
https://en.wikipedia.org/wiki/Michael_Kandel
Kandel is perhaps best known for his translations of the works of Stanisław Lem from Polish to English. Recently he has also been translating works of other Polish science fiction authors, such as Jacek Dukaj, Marek Huberath and Andrzej Sapkowski. The quality of his translations is considered to be excellent; his skill is especially notable in the case of Lem's writing, which makes heavy use of wordplay and other difficult-to-translate devices.
http://www.art.net/Studios/Hackers/Hopkins/Don/lem/HorribleP...
Oft, in that wickless chalet all begorn,
Where whilom soughed the mossy sappertort
And you were wont to bong --
A love poem, lyrical, pastoral, and expressed in the language of pure mathematics. Tensor algebra mainly, with a little topology and higher calculus, if need be. But with feeling, you understand, and in the cybernetic spirit.
http://www.art.net/Studios/Hackers/Hopkins/Don/lem/Femfatala...
Probably the biggest problem is the fact that all existing distributions are undermaintained and unpolished. SmartOS is the only exception, but it's not a replacement for Solaris which was a general-purpose server OS. SmartOS is merely a bare-metal hypervisor.
I really hope that some of the laid-off developers will start contributing to the project. It really needs them.
[1] - https://github.com/illumos/illumos-gate/commits/master
On commit activity, we in illumos don't necessarily have the same values as other projects: we really want to get it correct the first time -- a trait inherited from Solaris that we call "FCS quality all the time."[1] As such, commits have to be tested, thoroughly reviewed (four or more reviewers is not uncommon), and (as a practical matter) are more often than not shipping in a downstream repo for an extended period of time before they integrate upstream into illumos. In short, we value quality much, much more than quantity -- and what you perceive to be "low" activity I perceive to be quality activity (and as someone who very much relies on the ability to stay current and deploy into production, high commit activity would leave me nervous).
In terms of hardware support, you need to be specific. Yes, there is esoteric hardware we don't support -- but we in illumos are really focused on the server-side, where there is less hardware variation than you might realize. At the moment, the biggest hardware limitations aren't in the traditional domain of CPU/NIC/HBA/flash/spindle (where we have broad support), but rather in the emerging domain of the GPGPU. And anyone who is interested in engaging in the substantial kernel work in this domain (both to make NVIDIA-based GPGPUs to work in hardware virtualized guests and to allow AMD-based GPGPUs to work natively) should consider reaching out -- Joyent is very much hiring![2]
[1] http://dtrace.org/blogs/bmc/2015/09/03/software-immaculate-f...
This Solaris news is disappointing. I feel bad for those of you who have worked on this project for a significant portion of your career. Hope those who have been working on it find a good home somewhere.
Bryan, do you expect this will affect the direction of SmartOS in any significant way?
Also how well does intel's userland network and storage libraries dpdk and spdk, work with smartos.
Any time a commit to the code base would be made, it needed to satisfy that criteria, else Jeff Bonwick (the father of ZFS, who was the gatekeeper back then) would unceremoniously rip it out. Any kind of breakage, even during normal development, was unacceptable. This engineering ethos lives on in illumos and SmartOS, which is why commits to the code base are all about quality, not quantity. The commits might not be as frequent as elsewhere, but can pack tremendous amounts of functionality and must pass both peer reviews and all kinds of regression testing, not to mention have a complete and usable documentation, before being committed. In addition to all of this, as far as I am aware, no commits to the illumos code base may be made without a corresponding tracking identifier.
If you learn to use zones (which you should) it can run any workload that Solaris can.
On top of that it also runs native Linux containers and KVM for virtualizing everything else.
It's hands down my favorite operating system.
Dream OS for me would be something ala OpenBSD, but built on top of illumos and with all the goods that would come with it.
If your hardware isn't supported and you don't like it, get your ass into the code and build support for it.
Of course, once you already use the software, the situation inverts itself. I do choose only hardware which is known to work well with Linux, since that's the operating system I use, and I have even ported a device driver when that wasn't the case. But that can only happen because, many years ago when I first tried Linux, it worked perfectly with the hardware I had.
If you want to evaluate it but don't have the hardware, try qemu.
Things are not as simple as you suggest.
Amen to that!
Oh, look at Mr "I don't need to use what's laying around" over here.
Always choose your hardware for your software, never the other way around.
This sounds awful, and very long from what would be the legal requirements demanded in the country I live in.
What would such a robot call say?
"uadmin 1 6" was more like it (Immediate poweroff, do not even sync disks)
lom> poweroff
As that is definitely something the os has no cability to block.
Solaris and derivatives are great OS's and far before their time in so many ways.
I remember having a related problem with an ubuntu cloud server a while back. I'd called 'halt' instead of 'shutdown' - they're not the same, and on this machine 'halt' didn't actually send the 'okay, now power off' bit.
I remember being disgusted by the various halt, shutdown, reboot commands and so I learned what the various init run levels could do. That worked until I realized that Redhat and Debian set them up differently. At least init 0 is generally the same...
HCFhttps://www.theregister.co.uk/2010/12/07/mcnealy_sun_and_ope...
"One of the issues I had with one of the other suitors was there was a complete overlap in what they did and what we did, and I could see 100 per cent of the Sun employees getting fired," McNealy told us. He didn't name names, but he was referring to IBM's bid for Sun. IBM had competed heavily with Sun for decades on processors, servers, Unix, Java tools, middleware, and open source. There weren't too many areas where they didn't overlap.
"At least with Oracle, they weren't in the hardware business, the operating systems business — the places and spaces where I saw chance for some Sun employees to keep their jobs, and that for me was an important consideration," McNealy said.
I was very curious about this, and stumbled on further details in https://www.usenix.org/legacy/events/lisa11/tech/slides/cant... on page 15. But read the entire PDF anyway, it's interesting.
Q: On what hardware does Oracle run best?
A: A 35mm slide projector.
Cantrill blames "executive whims, shareholder demands, and a fickle public," but not even a little bit the product designers and engineers who are so attuned to the needs and desires of the market that they have somehow missed the collapse of a Fortune 500 company due to their products not selling and years later are still surprised that the product is being shut down?
Maybe sell to someone else?
Sun had been in discussions with IBM for months about an acquisition, but I believe the offer was lower and there were antitrust concerns (they did compete directly in a lot of markets) [1]. I suspect they would have been a much better steward of the open-source efforts though
[1]: http://www.nytimes.com/2009/04/06/technology/business-comput...
Oracle seems like bad guys, but it's really the board of Sun that had priorities other than the health of Solaris and the long term employment of the engineers. Oracle was a known snake [1].
Also, sale of the entire business of Sun was probably only one of many options on the table.
[1]: http://www.gballard.net/lifes_lessons/girl_rattlesnake/
I remember rumours that MS was in the game, but the fit there was a lot harder to find.
Essentially, Sun's executive team was betting that Oracle had a more stable business future than IBM did.
Mind you, I still use the 'sun' font when booting my linux kernel on framebuffers, I always loved that font.
https://nohats.ca/wordpress/blog/2016/01/24/why-the-sun-spar...
That leaves AIX, which at least has decent hardware to run on. But its got the OS/2 problem too, you can write your app for linux, and use the AIX toolbox for linux apps, and port your app. Although with power8 the move to LE linux, makes moving applications/data between linux on POWER and AIX on POWER more difficult.
Now to get myself one of those POWER9 workstations for home...
Winning Enterprise may be a pyrrhic victory, but we'll see how many server architectures are left in a few years.
Of course Oracle is a horrible company and all but Solaris has been dead for a long time.
For others, like me, who learned Solaris fairly late (2006), Solaris was a testbed for one of the best operating system technologies out there. Modern filesystem (ZFS), dynamic tracing (dtrace), containerization (zones), dependency management during boot (svc) and virtual network stack (crossbow), have been in solaris way before Linux. Most of modern Linux tools, such as eBPF, systap, btrfs, etc are a direct answer to the research done by Sun.
So some are nostalgic as it's their first experience, others, because it felt like a really good operating system a few years ago (despite it's quirks).
This was my case. I was playing around with Linux, both on my own machine at home as well as with a server at my high school (shell accounts being a perk of joining the computer club). While learning about Linux I also picked up some of the history of UNIX. At some point I learned that you could get a copy of x86 Solaris 7 for for < $50 (I can't remember if this was an educational offer or a developer network offer). But I ordered one and was really excited to get to use a real UNIX. I played with it for a while, but ended up going back to Linux for my own computers. Later, in the last year and a half of undergrad and my first year of grad school, I used a Sun Ultra 5 to do the work for and write my first published paper and do much of the work that would end up becoming my MS thesis and my second paper (with some of the heavy lifting outsourced to an Ultra 10 that a postdoc in the research group was using).
Every few months I browse ebay and look at SPARC machines. Though it's not really practically useful, I think it would be cool to have an Ultra 5 again.
I have an ultra 5 in a closet that hasn't been powered on in the current decade. Was thinking of giving it a spin in light of recent news.
I thought about that machine and I wondered which would be more interesting... Trying to get a recent OS on there or maintaining a more typical 90s setup? Somehow I think the latter is more fun.
"Get 'em whilst they're young" - I doubt it was an intentional strategy for Sun, but it definitely helped their sales in the bubble.
It was only when Linux started to seriously kick its ass that Sun engineers took it upon themselves to do some great work and make Solaris interesting again - I remember they had a policy - if something is slower than Linux it was a bug. I guess it all came in too late (including the Open Source part ) and people did not like being tied to SPARC hardware and CDDL was a buzz kill for whatever reasons.
But as we were getting the OS to work as advertised, we had bigger ideas brewing. In Bonwick's case, it was ZFS, for me it was DTrace, for Tucker it was zones, for Casper it was least-priv, etc. I know that it looks like it was in response to Linux and/or as a result of the dot-com bust (and it's impossible to say to what degree these were contributing factors), but I can tell you that from the epicenter it felt much more like a natural evolution than a competitive response. That said, I know that the competition didn't hurt, and you're certainly right about Linux finally causing us to take small-systems performance and system call latency seriously!
Yep - Only ever had one time when it hosed on us (100 people team, many running JVM based applications and some doing load tests on it - IIRC it was 2.6 days.) And there wasn't a challenger in sight as far as SMP goes(h/w and s/w both).
Now, really that product (SunOS) is not the "Solaris" that exists today. Solaris was a somewhat kludgey merger of the early BSD code with the System V tree from AT&T, done as much to settle legal issues as for market reasons. But nonetheless when we all look back to the golden days of unix, we see Sun's logo.
* https://github.com/ServiceManager/ServiceManager/ (https://news.ycombinator.com/item?id=10212770)
systemd was inspired by upstart, according to Lennart Poettering.
I guess that is not the case any more.
[1]: https://arstechnica.com/information-technology/2016/06/hp-aw...
I guess if oracle wants those support contracts, they can't have an operating system that doesn't need it!
https://youtu.be/-zRN7XLCRhc (entitled: Fork yeah! The rise and development of Illumos)
I can't, because I'm not somebody who's never used Solaris.
> As someone who's never used Solaris [...] I'm curious: can someone [...]
Notice that you left out the subject (I), left out the verb (am), and replaced a colon with a comma!
No, I didn't left out anything important to the structure nor did I replace a colon with anything. The thing is that you and I were looking at two different sentences, because OP silently edited his comment (I only noticed that now). The original sentence was "As someone who's never used Solaris or looked into its merits, can someone comment on why all the nostalgia for Solaris?"
"Waiting for the Sun" https://www.youtube.com/watch?v=A0kypyGSKsE
"The End" https://www.youtube.com/watch?v=6FMGYycBAMU
And to quote the also brilliant Jim Thompson (whose Levenshtein distance from Jim Morrison is 5):
We had Joy, we had fun
We ran Unix on a Sun,
But the source and the song
Of Solaris have all gone.(no idea about whether mainframes will continue using SPARC)
LEON pretty neat, you can run it on an FPGA and boot linux on it if you're brave enough.
I think what stood out most was the strongly 90s GUI and focus-follows-cursor, which seemed very strange at first. It seemed like an interesting OS with its own quirks
I still have an old Ultra 10 downstairs. It hasn't been turned on since 2004 or so. I think it had Solaris 8?
To this day I still think Solaris was really well thought out. It had a certain rock solidness to it that I can't really describe. Compare to even openbsd or freebsd I always thought or felt it was more stable (regardless of that was actually the case).
I'm sad that so much work went into such a great OS and now it is basically gone to die. These days if you're thinking about creating a piece of software and want it to last open source is probably the only option.
They sell on the golf course, not the meetup or con presentation.
Corporations are like ant colonies. They are bodies of people who are bound together in order that the whole outlasts the individual, in order to serve the needs of the group. They are made up of individuals with feelings, but they can't "feel" things on their own. They can reflect the shared feelings of the group, but this is just a reflection, like seeing the image of trees reflected in a body of water.
The author is obviously very upset about the whole thing, but personifying a corporation is illogical (even if the Supreme Court disagrees with me)
Reusing names tends to be a terrible idea.
Edit: grammar.