Re: [osol-discuss] Possible IBM aquisition of Sun
casper@sun.com wrote: DTrace ZFS SMF FMA BootAr IPS IA-Install Incompatible? N N N N N Y Y I think the many people screaming about quota support when it debuted a few years ago, among many other decisions would beg to differ about the incompatible nature of ZFS :) Not that I disagree with the decisions that have been made, but I think your high level summary has smoothed out the comparison a bit too much. -Shawn ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
Martin Bochnig wrote: many other benefits. And you don't need to employ 50 engineers for 2 years to get a written-from-scratch monster like IPS going. I'm not sure where you got this number from, but I think it's about 10x greater than the reality :-) I'm sure the pkg(5) team would love to have had those 50 engineers for two years... :) Your fellow wheelwright, -Shawn ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
Moinak Ghosh wrote: On Sun, Apr 5, 2009 at 2:55 AM, Lurie wrote: ... The user side experience of IPS is no doubt very good but is no different from a good Linux package manager like Smart/Yum (with the exception of ZFS features). From a developer point of view these qualities could have been got by far less effort and far less code/complexity. Actually, it is very different (especially as of build 110) with respect to search capabilities. None of the other, existing packaging systems feature the rich interface that pkg(5) now has for search. In addition, none of the systems feature remote search capability. There are differences in other areas as well, but those won't be as apparent until later functionality is implemented. Cheers, -Shawn ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Packaging Systems was Re: Possible IBM aquisition of Sun
C. wrote: 5) "a secure package manager without any arbitrary post/pre-install scripts" Ok.. so wait a second.. Let's first of all define "secure" because last I checked the IPS authorities aren't signed.. Are they? You're only They are not currently, but this is definitely planned functionality. looking at it from one angle.. If a package is signed and trusted by the authority then the script isn't arbitrary. It was designed and created with a sole purpose. If it *is* arbitrary I don't think I'd There are a significant number of packages from third-party vendors and from Sun itself that were "created for a sole purpose" that often don't work right because of incorrect assumptions made in the pre/post install scripts. So, arbitrary in the sense that it made "arbitrary assumptions" would be correct. backing it. When the IPS repo is handling 10k+ unique pieces of software lets talk. Those scripts are there to add *robustness* to the The other day, you threw out the number of 2,000 unique pieces of software. Well, I have good news for you, if you combine the primary opensolaris.org repository (about 1800 unique packages), and the contrib repository (about 11,505 unique packages), you'll easily be over that 10k+ unique number you're talking about. And you'll find that the pkg(5) system can manage it. think some features are *good*.. Does removing pre/post scripts warrant a new package manager.. Bluntly put.. No.. it sure hell doesn't.. This is why people who have a clue are so pissed off.. Because good people have been fired before, but there's still this rogue team wasting cycles on stuff which could ultimately be spent elsewhere.. So forgive me and other when we are a bit rude and aggressive... I'm sorry if you think this is a "rogue team"; but it isn't. The pkg(5) system was not designed solely to eliminate scripting, Rather, it was designed to create a packaging system that fully integrated with OpenSolaris and took advantage of its significant, unique technologies (such as ZFS) and satisfied the needs of a target audience. As you point out, many packaging systems have been around for a significant number of years. However, what you didn't point out is how stagnant the design of these systems has been, and how they've all stayed rather generic instead of trying to uniquely innovate for a given platform. 7) is very easy to use... ok. you got me there... they were able to make a cli interface that is at least comparable to things which have been around for 15 years.. :) In some areas yes, the pkg(5) system is merely comparable to other packaging systems. However, in other areas, such as the pkg(5) search interface, (especially as of build 110) it is significantly advanced compared to that provided by deb, rpm, or other systems. It also offers a remote search interface that none of those other systems have, allowing users to search for packages without downloading large metadata blobs from package repositories. 8) upgrades the whole system at once, versus just the packages, which ensures you won't have any conflicts ... b) because of the way manifests are created from entire packages just like any normal or sane packaging system you're still just as likely to have unresolved dependencies or blockers. (I don't know for certain a missing file can't pull for *any* manifest even one which provides the same package at a different version) The way dependencies are declared in IPS manifests is not different (as far as I remember) from rpm, deb or other popular packaging formats. So, I fail to see the issue with that. The existing dependency resolution mechanism in pkg(5) is due to be replaced soon with a SAT-solver based mechanism. c) if you argue that is saves bandwidth.. I'll argue that it has delayed being able to easily establish a mirroring system, it forces a It is very easy to create a mirror, and this functionality has been there for several months now. custom daemon, there's no on-disk format (maybe this was resolve To provide the sort of rich, remote functionality that our packaging server provides, you have to have custom software of some sort running on the remote package depot server. The on-disk format was seen as less important than other, significant pieces of functionality and is planned to be implemented at a later date. 10) I've been using IPS since its inception, for how long have you been using it ? This is you just trying to discredit what I've said. As you can see above I think you're just an end user that isn't really clear on what's going on or slightly mistaken in some points. *If* you have used IPS for as long as you say you have I think you'd be a bit more empathetic to certain facts. Did it allow you to cleanly migrate from the original 05/08 release until today? I'm sure at some point you've had to do a fresh install or skipped many updates. There's been a
Re: [osol-discuss] The IBM deal is dead, so...
Pls. ignore the typos. It is very late (early) here ... > %martin ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Acer 1GB Netbook for $99.99 (& Catch)
> > One of the loudest noises among all the netbook
> makers in Taiwan is
> > about Android--everyone is talking about building
> Android-based
> > netbooks on the ARM platform. (This will also
> allow the netbook makers
> > not to be bound by the "must be less than 12-inch"
> constraints.)
> >
> > Android is not Java, but since it is so closely
> related to Java,
> > perhaps someone inside Sun can think of some way
> for Sun to get
> > involved in this potentially disruptive movement?
> >
> > Or at least post something, so that someone else
> close to the action
> > can think of something else? Thanks.
>
> I'm having a really hard time figuring out what
> you're talking about.
>
> What is the "must be less than 12-inch" constraint
> that you mentioned? I've
> never heard of it.
>
> In what way is android closely related to java?
> Android is a tiny linux
> istro for tiny devices. Java is ... well, java.
>
> What "potentially disruptive movement" are you
> talking about, and what would
> be the disruption?
>
> ___
> opensolaris-discuss mailing list
> opensolaris-discuss@opensolaris.org
I must admit that I am having a hack of difficult time understanding your
questions, too. But I believe most in this forum are in agreement with you
than with me.
You really have to read the attached article (if you read Chinese) and a bunch
of other background info to have a full grasp. But I will give it a try anyway:
1. The "12" limitation" is that, if you ("you" means a netbook maker) use
Intel's platform, as a matter of licensing agreement, your screen size cannot
exceed this limit. This limit is imposed by Intel. If you use the ARM
architecture, of course, you are not bound by this limitation. I think the VIA
platform, which also is x86-based, also falls into the same constraint.
2. Wrt Java vs. Android (more important the possible connections there
between), I will leave this to those who are more knowledgeable. Any
additional info will be appreciated.
3. "Disruptive" means that there may be tens of millions of new users in the
first year alone, if things work out the way it was described in that article.
Even Nokia is getting into the Android/netbook business. I mentioned in a
separate thread that China Mobile has over 457 million subscribers. They are
talking about making (via OEM) their own netbooks.
--
This message posted from opensolaris.org
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The IBM deal is dead, so...
Leave a message on his blog :) If it's one thing I do wish he would do, it's use the massive traffic at blogs.sun.com at times like these to plug the lesser known bits of Sun's software portfolio, eg VDI 3, SJS Communications Suite etc. Just get someone to ghost write it, even. 16,000 hits a day to a personal blog is enviable exposure in anyone's terms. -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
C. wrote: Shawn Walker wrote: C. wrote: Do you know how many patches Sun maintains for packages that are never accepted by upstream because they don't agree with the design or implementation? I can tell you with certainty that Sun works hard to ensure contributions go upstream, it's ultimately less work for them and less code to maintain. Yes.. writing an entire package manager from the ground up is *less* work than maintaining the patches. Did someone tell you this or did you "The right tool for the right job" is a common saying; none of the existing packaging tools were right for the job. Hence, a new tool was needed. come up with this on your own? Knowing you're a contributor to the pkg5 team makes you a bit biased, but try to separate your job from reality and stay on the point here.. If you follow the lists and read various comments about upstream history, you can find this out for yourself. Ask the desktop team about the various patches they've had to fight to get upstream or had to continue to maintain themselves for a while. My project affiliation has nothing to do with the reality of the situation. Cheers, -Shawn ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] osol-0906-106a available
> In <1153511638.20901238987785476.javamail.tweb...@sf-app1> > Charles Shannon Hendrix wrote: > What about OpenSolaris? Could the same method (booting from U4) be used > to do an OpenSolaris install, or do you have to boot to the Live image > for that? You can get a usb image of OpenSolaris from Genunix. http://genunix.org/distributions/indiana/osol-0906-110-x86.usb But I do not know how to burn and boot it. -- NAKAJI Hiroyuki ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The IBM deal is dead, so...
On Mon, Apr 6, 2009 at 4:10 AM, Tim Scanlon wrote: > > Also, upon reflection i wish I'd not have cast blame & personal animosity > towards Johnathan Schwartz in the other thread, that was regrettable given > that I do know better. If these forums supported better editing, I'd redact > more of it. > > Tim Tim, didn't you state yourself, that you "know" him from having talked to him a single time? Maybe you aren't aware because it was before your time (implying from your history, as stated in your earlier message), but in 2008 Sun posted a huge loss (related to the MySQL aquisition) Sun Posts $1.68 Billion Loss On Hefty Impairment Charge http://online.wsj.com/article/SB122540145956385597.html?mod=googlenews_wsj How in the world does it fit together with this, then: Why Jonathan Schwartz Sun Microsystems CEO gets $11M 2008 pay package? http://ceoworld.biz/ceo/2008/09/25/why-jonathan-schwartz-sun-microsystems-ceo-gets-11m-2008-pay-package/ Frankly speaking, IF he would just be 50% as good, as you claim, IF he would indeed care about the Sun, IF he would care about Sun's employee's families' lifes, wouldn't he have refused to accept at least a part of his annual bonuses ??? His leads by poor example. He is known to love expensive sports cars, yet his X11 group could not even get a functioning coffee machine in 2007 (the staff had to buy one from their own money). He always says, Sun needs to save and layoffs are inevitable. Then he wastes so much money by accepting it as HIS PERSONAL BONUSES, like Mugabe in Zimbabwe?? That's hypocracy, and THERE ARE LOTS OF OTHERS LIKE HIM. If he thinks layoffs are needed, maybe he should lead by example and should remove the most expensive employe(es) first. This would definitely help Sun, in many ways and on several layers, no doubt about it. Then thousands of engineers and other average-payment employees can keep their jobs, can feed their families, which in turn helps the overall surrounding economy. %martin ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Acer 1GB Netbook for $99.99 (& Catch)
> One of the loudest noises among all the netbook makers in Taiwan is > about Android--everyone is talking about building Android-based > netbooks on the ARM platform. (This will also allow the netbook makers > not to be bound by the "must be less than 12-inch" constraints.) > > Android is not Java, but since it is so closely related to Java, > perhaps someone inside Sun can think of some way for Sun to get > involved in this potentially disruptive movement? > > Or at least post something, so that someone else close to the action > can think of something else? Thanks. I'm having a really hard time figuring out what you're talking about. What is the "must be less than 12-inch" constraint that you mentioned? I've never heard of it. In what way is android closely related to java? Android is a tiny linux distro for tiny devices. Java is ... well, java. What "potentially disruptive movement" are you talking about, and what would be the disruption? ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
Hello Krzabr, The deal is dead. Check this link for further details: http://www.opensolaris.org/jive/thread.jspa?threadID=99304&tstart=0 My recommendation is go and use OpenSolaris as a Major OS. The only tricky part is the same difficulty that exists with Linux and *BSD which is deciding which version of OpenSolaris is the right one for you. A lot of people who want a System V style UNIX operating system with maximum stability in servers (like Joyent) use Solaris Express which is more like classic "Solaris 10" with extra features. People coming from a GNU/Linux background like me tend to prefer either Nexenta (which is like Ubuntu but with a Solaris kernel) or OpenSolaris 2008.11 which is Sun's officially supported distro (you can buy support from Sun for $300 a year and have their programmers help to fix your server when it's broken). Milax is also a very good OpenSolaris distro because it's so minimal (has only the features that you want and no bloat) and uses less resources. The most rapidly developing and technically advanced OpenSolaris distro is probably Belenix as I've noticed that Belenix tends to get certain new features (like NTFS read / write support) before the other distros get them. -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] osol-0906-106a available
I don't have a USB DVD drive. Hate to spend $70 just for an install. Does anyone know if you can put a Solaris DVD on a USB hard drive and boot from it, and do the install like that? I have several of those portable USB drives. Can you copy the Solaris 10 DVD to one of them, and then make it bootable? Alternately, if I were able to install Solaris U4 to one of those USB drives, would it be possible to run the Solaris U6 installer from that, either from a DVD drive or a mounted ISO image? What about OpenSolaris? Could the same method (booting from U4) be used to do an OpenSolaris install, or do you have to boot to the Live image for that? -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Acer 1GB Netbook for $99.99 (& Catch)
> One of the loudest noises among all the netbook > makers in Taiwan is about Android--everyone is > talking about building Android-based netbooks on the > ARM platform. (This will also allow the netbook > makers not to be bound by the "must be less than > 12-inch" constraints.) > > Android is not Java, but since it is so closely > related to Java, perhaps someone inside Sun can think > of some way for Sun to get involved in this > potentially disruptive movement? > > Or at least post something, so that someone else > close to the action can think of something else? > Thanks. Oops, forgot to cite the source: http://tech.chinatimes.com/2007Cti/2007Cti-News/Inc/2007cti-news-Tech-inc/Tech-Content/0,4703,12050901 122009040600216,00.html or http://tinyurl.com/netbook-arm -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Acer 1GB Netbook for $99.99 (& Catch)
> One of the loudest noises among all the netbook > makers in Taiwan is about Android--everyone is > talking about building Android-based netbooks on the > ARM platform. (This will also allow the netbook > makers not to be bound by the "must be less than > 12-inch" constraints.) > > Android is not Java, but since it is so closely > related to Java, perhaps someone inside Sun can think > of some way for Sun to get involved in this > potentially disruptive movement? > > Or at least post something, so that someone else > close to the action can think of something else? > Thanks. Oops, forgot to cite the source: http://tech.chinatimes.com/2007Cti/2007Cti-News/Inc/2007cti-news-Tech-inc/Tech-Content/0,4703,12050901%20122009040600216,00.html or http://tinyurl.com/netbook-arm -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Acer 1GB Netbook for $99.99 (& Catch)
One of the loudest noises among all the netbook makers in Taiwan is about Android--everyone is talking about building Android-based netbooks on the ARM platform. (This will also allow the netbook makers not to be bound by the "must be less than 12-inch" constraints.) Android is not Java, but since it is so closely related to Java, perhaps someone inside Sun can think of some way for Sun to get involved in this potentially disruptive movement? Or at least post something, so that someone else close to the action can think of something else? Thanks. -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The IBM deal is dead, so...
> Also, upon reflection i wish I'd not have cast blame > & personal animosity towards Johnathan Schwartz in > the other thread, that was regrettable given that I > do know better. If these forums supported better > editing, I'd redact more of it. > > Tim As much as I am allergic to this deal, I think the board and Sun's top execs made a responsible move to show that they indeed place shareholders' interest above their own. As to JS, nah, he is much bigger than you and I described him to be. -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The IBM deal is dead, so...
Well, the more I think about it, the more obvious it is to me that the old rules for this sort of deal don't apply so much anymore. Sun converted to open source, and that's made the company a lot harder to blow apart like say DEC was in the 90's... I've done enough research in the past few weeks to where I've become convinced that the structural flaws at IBM would have created a disaster. They also are fundamentally different culturally to an extent that it profoundly affects their basic business models, and I think this mix could have ended up hurting both companies. With the kind of open source Sun's done, with a less encumbered license than a GNU one, some of the failures associated with vendor warlord behavior become a lot less of a risk to customers. I'm a lot less worried than I was a few days ago because I gave the implications of a giant blue screen of death much closer consideration than I previously ever had. Also, upon reflection i wish I'd not have cast blame & personal animosity towards Johnathan Schwartz in the other thread, that was regrettable given that I do know better. If these forums supported better editing, I'd redact more of it. Tim -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] osol-0906-106a available
> In <1813134311.15891238967682081.javamail.tweb...@sf-app1> > Charles Shannon Hendrix wrote: > You got Solaris U6 to install on the Dell T105? Yes. Installation is difficult but it works fine. http://www.sun.com/bigadmin/hcl/data/systems/details/26018.html > When I boot the latest official Solaris 10 U6 on my Dell T105, it boots > just fine but the install fails saying it cannot read the DVD. Yes, I saw the same issue. And I tried an external USB DVD drive to install Solaris 10 U6, because internal SATA DVD drive was not detected after the kernel booted. Informations in Japanese of my USB drive are available at http://www.logitec.co.jp/products/dvd/ldrpmd8u2.html > Unfortunately, SATA DVD is the only method I have for installing. Developpment version of OpenSolaris 2009.06 after b108 can boot with my Dell T105, but Xorg does not work. -- NAKAJI Hiroyuki ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The IBM deal is dead, so...
Thank goodness! I'm glad the deal broke down and I hope the folks at Sun behind it are sacked in the coming weeks. This was a huge embarrassment and shows how the management was failing to do the right thing and get Sun into gear. Selling Sun off is just the wrong direction. I hope this damages any future offers and Sun if forced to just "man up" and get back to making kick ass products and make some money! *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* Octave J. Orgeron Solaris Virtualization Architect and Consultant Web: http://unixconsole.blogspot.com E-Mail: unixcons...@yahoo.com *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* - Original Message From: Tim Scanlon To: opensolaris-discuss@opensolaris.org Sent: Sunday, April 5, 2009 7:50:28 PM Subject: [osol-discuss] The IBM deal is dead, so... http://www.nytimes.com/2009/04/06/technology/business-computing/06blue.html I'm glad about that, I think Sun can do a lot better. Also, IBM's behavior with stimulus funds on one hand, and buying Sun on the other had an overt appearance of fiscal impropriety. There are better matches to be had, but I think that most any deal is going to suffer some problems because a whole hell of a lot of Sun technology is very advanced. IBM's bad offer reflects how the concurrent incomprehension can complicate potential deals. It's hard for a lot of companies to make innovative leaps through a merger, even incrementally. The best example of it occurring recently though that I can think of has been the adoption of ZFS outside of Sun. On one hand, companies like Microsoft are still falling back to closed source patents on DOS fs mechanics, on the other their competitor took a huge leap forward and adopted ZFS. That was risky for Apple, but it worked well because the technology is open source, and that gave them a margin of safety. I can't say I'm a fan of Apple's development practices from a business perspective, but they do have some undeniable clue there too. Aside from waving an Iphone at people, that probably serves as a better example for my point. Another option that should be on the table is the application of stimulus funds, in terms of targeting money, taxpayers would be way better off than they would with say, IBM's half-baked proposals. I'm concerned about mainframey matches, because frankly Sun does a better job with legacy support than most companies, and I don't want to see a system that works thrown out for some of the poor waters other companies suffer from. There's engineering that goes into the process of aging out technology that other companies just don't understand, much less possess. If you're a customer in an era of virtual machines, that sort of thing matters more, not less. The Themis contortions that were accompanying the IBM deal are an example of floundering on the part of IBM over that subject alone. Tim -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The Register: IBM focuses on Sun software trio
There are now early "reports" (in reality more bullshitty secret squirrel stuff) that "talks" are either on the verge of collapse, or have broken down entirely. http://talkback.zdnet.com/5206-10532-0.html?forumID=1&threadID=62908 http://www.basexblog.com/2009/04/05/ibm-sun-deal-collapses-but-should-it-have-ever-gotten-this-far/ ZDnet can be okay at times, don't know about the veracity of the second source. -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The IBM deal is dead, so...
> http://www.nytimes.com/2009/04/06/technology/business-computing/06blue.html Darn it, just made an appointment with a shrink. :-( -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The IBM deal is dead, so...
On Mon, Apr 6, 2009 at 2:50 AM, Tim Scanlon wrote: > http://www.nytimes.com/2009/04/06/technology/business-computing/06blue.html I LOVE YA ALL:))) Thanks God. ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] The IBM deal is dead, so...
http://www.nytimes.com/2009/04/06/technology/business-computing/06blue.html I'm glad about that, I think Sun can do a lot better. Also, IBM's behavior with stimulus funds on one hand, and buying Sun on the other had an overt appearance of fiscal impropriety. There are better matches to be had, but I think that most any deal is going to suffer some problems because a whole hell of a lot of Sun technology is very advanced. IBM's bad offer reflects how the concurrent incomprehension can complicate potential deals. It's hard for a lot of companies to make innovative leaps through a merger, even incrementally. The best example of it occurring recently though that I can think of has been the adoption of ZFS outside of Sun. On one hand, companies like Microsoft are still falling back to closed source patents on DOS fs mechanics, on the other their competitor took a huge leap forward and adopted ZFS. That was risky for Apple, but it worked well because the technology is open source, and that gave them a margin of safety. I can't say I'm a fan of Apple's development practices from a business perspective, but they do have some undeniable clue there too. Aside from waving an Iphone at people, that probably serves as a better example for my point. Another option that should be on the table is the application of stimulus funds, in terms of targeting money, taxpayers would be way better off than they would with say, IBM's half-baked proposals. I'm concerned about mainframey matches, because frankly Sun does a better job with legacy support than most companies, and I don't want to see a system that works thrown out for some of the poor waters other companies suffer from. There's engineering that goes into the process of aging out technology that other companies just don't understand, much less possess. If you're a customer in an era of virtual machines, that sort of thing matters more, not less. The Themis contortions that were accompanying the IBM deal are an example of floundering on the part of IBM over that subject alone. Tim -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
Hello it's my first answer on this forum . I'm not a developer or system administrator , but I'm looking on SUN-IBM fusion and using OpenSolaris for a Major OS . Unfortunatelly we don't know real IBM's Plan of future of Sun , Mysql , Netbeans , ZFS , Dtrace , Sparc , OO and Solaris . Many people should recieve the most unoptimistic predicts , unexlusion that IBM will use Sun's Projects Sources into them software . F.e Netbeans code would upgrade eclipse , and Solaris code would be in future x86 and SPARC port for AIX . More and More IBM might close all Open Source Sun's Projects for Commercial Use . What we Can Do ? Waitng for IBM Decisions about fusion and future of Sun's Software . And Helping On OpenSolaris Developing Process :) -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] osol-0906-106a available
I don't have a Jumpstart server or another Solaris machine to install one on. The Solaris machine is my server system, the only one I have. Everything else is either desktop system (Mac and Windows), or embedded machines for work. I haven't been able to make the machine boot a USB CD, so I don't think that is an option. That's why I was asking if there was some way to boot the install CD and make it use the working driver so I could get installed. What I want to do is put two new 320GB drives in the machine and install Solaris 10 on a mirrored ZFS root. I have all of the files on the system backed up, so I don't need the old install. I thought this was going to be easy... :) -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] osol-0906-106a available
On Apr 5, 2009, at 17:51 , Steven Stallion wrote: Charles Shannon Hendrix wrote: You got Solaris U6 to install on the Dell T105? When I boot the latest official Solaris 10 U6 on my Dell T105, it boots just fine but the install fails saying it cannot read the DVD. It seems to be an issue with the new SATA drivers. I don't understand what broke, because Solaris U4 and U5 booted and installed fine. Unfortunately, SATA DVD is the only method I have for installing. I believe ATAPI SATA is only supported on AHCI controllers at the moment. What controller does the Dell use? Let me modify that reply: The Dell uses an mcp55 pro, and my current install does report it is using AHCI. However, some people say that it doesn't really do AHCI. I have no idea, I was just going by what Solaris dmesg printed. -- Charles Shannon Hendrix c...@widomaker.com ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] osol-0906-106a available
On Apr 5, 2009, at 17:51 , Steven Stallion wrote: Charles Shannon Hendrix wrote: You got Solaris U6 to install on the Dell T105? When I boot the latest official Solaris 10 U6 on my Dell T105, it boots just fine but the install fails saying it cannot read the DVD. It seems to be an issue with the new SATA drivers. I don't understand what broke, because Solaris U4 and U5 booted and installed fine. Unfortunately, SATA DVD is the only method I have for installing. I believe ATAPI SATA is only supported on AHCI controllers at the moment. What controller does the Dell use? Dell uses an nVidia MCP55 Pro, which is an AHCI controller. Solaris 10/07 has worked fine for the last year and there were no problems installing it. From what I have read, the nVidia AHCI drivers are broken in the latest release for SATA DVD drives. It seems like there should be some way to work around this, but I've not found one. Like maybe boot from U4, and then somehow install from the U6 DVD. I have no idea how to make that happen though. I do have a pair of DVD drives, so maybe there is some way I could arrange it? -- Charles Shannon Hendrix c...@widomaker.com ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] osol-0906-106a available
Charles Shannon Hendrix wrote: > The Dell uses the MCP55 Pro. > > I thought it supported AHCI due to some messages from the kernel, but > evidently it doesn't, or does not do so completely. > > That's probably the issue. > > However, it seems like there should be some way to work around this, in both > OpenSolaris and Solaris 10. > > Is there some way to make Solaris use the working driver during install? > > Is there some way I could install the newer version of Solaris from the older > one while it is running, and then copy the working driver into the new system? > > I really need to find a way around this because I need Solaris 10/08 > installed ASAP. > > I would use Open Solaris but I had some issues with it that made me decide to > stick with the Sun releases. > > Are there any kernel boot parameters that might help? Have you considered a jumpstart install? -- Yet magic and hierarchy arise from the same source, and this source has a null pointer. Reference the NULL within NULL, it is the gateway to all wizardry. ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] osol-0906-106a available
The Dell uses the MCP55 Pro. I thought it supported AHCI due to some messages from the kernel, but evidently it doesn't, or does not do so completely. That's probably the issue. However, it seems like there should be some way to work around this, in both OpenSolaris and Solaris 10. Is there some way to make Solaris use the working driver during install? Is there some way I could install the newer version of Solaris from the older one while it is running, and then copy the working driver into the new system? I really need to find a way around this because I need Solaris 10/08 installed ASAP. I would use Open Solaris but I had some issues with it that made me decide to stick with the Sun releases. Are there any kernel boot parameters that might help? -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] osol-0906-106a available
Charles Shannon Hendrix wrote: > You got Solaris U6 to install on the Dell T105? > > When I boot the latest official Solaris 10 U6 on my Dell T105, it boots just > fine but the install fails saying it cannot read the DVD. > > It seems to be an issue with the new SATA drivers. > > I don't understand what broke, because Solaris U4 and U5 booted and installed > fine. > > Unfortunately, SATA DVD is the only method I have for installing. I believe ATAPI SATA is only supported on AHCI controllers at the moment. What controller does the Dell use? ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] NFS Problems after lu to snv_110
Martin Bochnig wrote: > On Sun, Apr 5, 2009 at 10:48 PM, Steven Stallion wrote: >> Thanks Martin. >> >> I added 'subtree_check' to rw exports in /etc/exports on the Linux NAS >> and everything is back to normal (using NFSv3 TCP). >> >> Its a pain to manually hack the options since these are typically >> managed by the device, however its better than nothing (or NFSv2) at the >> moment. >> >> Hopefully someone has filed a CR. > > > Could you do this please? > Currently I don't have a running LinUX installation. > Hence I didn't reproduce the scenario, especially not in a detailed > enough manner. > For you the effort is not big to do it, you have everything powered on > already. > False alarm; I had forgotten my last activate was back to snv_99. snv_110 works fine with NFSv2 mounts. I'll file a CR. Steve ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] osol-0906-106a available
You got Solaris U6 to install on the Dell T105? When I boot the latest official Solaris 10 U6 on my Dell T105, it boots just fine but the install fails saying it cannot read the DVD. It seems to be an issue with the new SATA drivers. I don't understand what broke, because Solaris U4 and U5 booted and installed fine. Unfortunately, SATA DVD is the only method I have for installing. -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] NFS Problems after lu to snv_110
On Sun, Apr 5, 2009 at 10:48 PM, Steven Stallion wrote: > Thanks Martin. > > I added 'subtree_check' to rw exports in /etc/exports on the Linux NAS > and everything is back to normal (using NFSv3 TCP). > > Its a pain to manually hack the options since these are typically > managed by the device, however its better than nothing (or NFSv2) at the > moment. > > Hopefully someone has filed a CR. Could you do this please? Currently I don't have a running LinUX installation. Hence I didn't reproduce the scenario, especially not in a detailed enough manner. For you the effort is not big to do it, you have everything powered on already. Thanks, %martin ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] NFS Problems after lu to snv_110
Martin Bochnig wrote: > On Sun, Apr 5, 2009 at 10:11 PM, Martin Bochnig wrote: >> On Sun, Apr 5, 2009 at 10:03 PM, Steven Stallion wrote: As a workaround it works if you force server and client to use NFS version 2. >>> Ouch. Do you happen to have a bugster ID? When was this issue introduced? >> >> Mhh, see this thread: >> http://www.opensolaris.org/jive/thread.jspa?threadID=99124&tstart=0 >> >> I don't know the bugid. >> Look at defect.opensolaris.org . >> Or may it indeed be that nobody filed this bug as bug yet??? >> In that case it would be something like "unbelievable". > > > Be aware that this topic has been discussed about as early as on March 29th: > http://www.opensolaris.org/jive/thread.jspa?threadID=98644&tstart=75 > > >> Regards, >> %martin >> Thanks Martin. I added 'subtree_check' to rw exports in /etc/exports on the Linux NAS and everything is back to normal (using NFSv3 TCP). Its a pain to manually hack the options since these are typically managed by the device, however its better than nothing (or NFSv2) at the moment. Hopefully someone has filed a CR. Cheers, Steve ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] NFS Problems after lu to snv_110
On Sun, Apr 5, 2009 at 10:11 PM, Martin Bochnig wrote: > On Sun, Apr 5, 2009 at 10:03 PM, Steven Stallion wrote: >>> As a workaround it works if you force server and client to use NFS version >>> 2. >> >> Ouch. Do you happen to have a bugster ID? When was this issue introduced? > > > Mhh, see this thread: > http://www.opensolaris.org/jive/thread.jspa?threadID=99124&tstart=0 > > I don't know the bugid. > Look at defect.opensolaris.org . > Or may it indeed be that nobody filed this bug as bug yet??? > In that case it would be something like "unbelievable". Be aware that this topic has been discussed about as early as on March 29th: http://www.opensolaris.org/jive/thread.jspa?threadID=98644&tstart=75 > Regards, > %martin > ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] NFS Problems after lu to snv_110
On Sun, Apr 5, 2009 at 10:03 PM, Steven Stallion wrote: >> As a workaround it works if you force server and client to use NFS version 2. > > Ouch. Do you happen to have a bugster ID? When was this issue introduced? Mhh, see this thread: http://www.opensolaris.org/jive/thread.jspa?threadID=99124&tstart=0 I don't know the bugid. Look at defect.opensolaris.org . Or may it indeed be that nobody filed this bug as bug yet??? In that case it would be something like "unbelievable". Regards, %martin ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] NFS Problems after lu to snv_110
Martin Bochnig wrote: > On Sun, Apr 5, 2009 at 9:50 PM, Steven Stallion wrote: >> All, >> >> I have been having fits with NFS mounts after lu'ing to snv_110 (and >> snv_109) from snv_99. The NFS server is a Linux based NAS using >> vers=3,proto=tcp. >> >> Once luactivate is used and an init 6 is issued, NFS exports will mount, >> however I am unable to write using a valid UID. If I luactivate the >> older BE (snv_99) (once again issuing init 6), everything goes back to >> normal. >> >> There have been no UID or NFS server changes between each lu activate. I >> have done this twice so far with the same exact result. >> >> Any ideas? >> >> Steve > > > Hello, this is a known and reported bug. > As a workaround it works if you force server and client to use NFS version 2. Ouch. Do you happen to have a bugster ID? When was this issue introduced? Regards, Steve ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] NFS Problems after lu to snv_110
On Sun, Apr 5, 2009 at 9:50 PM, Steven Stallion wrote: > All, > > I have been having fits with NFS mounts after lu'ing to snv_110 (and > snv_109) from snv_99. The NFS server is a Linux based NAS using > vers=3,proto=tcp. > > Once luactivate is used and an init 6 is issued, NFS exports will mount, > however I am unable to write using a valid UID. If I luactivate the > older BE (snv_99) (once again issuing init 6), everything goes back to > normal. > > There have been no UID or NFS server changes between each lu activate. I > have done this twice so far with the same exact result. > > Any ideas? > > Steve Hello, this is a known and reported bug. As a workaround it works if you force server and client to use NFS version 2. -- %martin ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Google for rPath conary "next generation"
Ouch, msg. was sent twice without intention. Gmail's www-interface is not very reliable currently (since yesterday). ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Google for rPath conary "next generation"
> Conary? can someone point to an explanation... its baffling to me, > maybe many others... Google for rPath conary "next generation" http://www.google.com/search?hl=en&client=firefox-a&rls=org.mozilla%3Aen-US%3Aofficial&hs=bbY&q=rPath+conary+%22next+generation%22&btnG=Search Ported test packages for OpenSolaris can be downloaded from here: http://opensolaris.org/os/project/conary-eval/files/ The repo was here: http://www.martux.org:8001/ ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Google for rPath conary "next generation"
> Conary? can someone point to an explanation... its baffling to me, > maybe many others... Google for rPath conary "next generation" http://www.google.com/search?hl=en&client=firefox-a&rls=org.mozilla%3Aen-US%3Aofficial&hs=bbY&q=rPath+conary+%22next+generation%22&btnG=Search Ported test packages for OpenSolaris can be downloaded from here: http://opensolaris.org/os/project/conary-eval/files/ The repo was here: http://www.martux.org:8001/ ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] NFS Problems after lu to snv_110
All, I have been having fits with NFS mounts after lu'ing to snv_110 (and snv_109) from snv_99. The NFS server is a Linux based NAS using vers=3,proto=tcp. Once luactivate is used and an init 6 is issued, NFS exports will mount, however I am unable to write using a valid UID. If I luactivate the older BE (snv_99) (once again issuing init 6), everything goes back to normal. There have been no UID or NFS server changes between each lu activate. I have done this twice so far with the same exact result. Any ideas? Steve ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [ug-bjosug] Buying Solaris computer in BJ
> 你们好！ > > 我是法国人，我叫Laurent（罗杭）。 > 现在我学习汉语在第二外国语大学，对不起，我写英语。 > > Hopefully soon I'll be able to write Chinese only... > I need to get myself a desktop computer, it doesn't > need to be powerful, > but I want to run OpenSolaris properly on it (and if > possible, Solaris > 10 as well). Any advice on where and what to buy, > some brand that would > be sure to work well? And how much it would cost? > Bonus would be a French layout keyboard, is that > possible to find here? > (Ideally - I'm dreaming here :-) - a Sun Type 6 > French keyboard). > > I live in Chaoyang District, close to the crossing of > Jintailu and > Chaoyanglu. > > 谢谢！ > > Laurent 在我的印象中，BJ好像已经没有人在用desktop了。每一个人都是（即使在家里）用所谓的desktop replacement 的notebook。 :-) 你可以到 www.amazon.cn 看一下 HP 540： http://www.amazon.cn/mn/detailApp?ref=BR&uid=168-0575736-1137025&prodid=rlit003154 （约￥4,100，但目前没货）。 这个notebook 只有1GB的DDR2，你可能需要再花个￥150 扩充到2GB。 它run OpenSolaris/Solaris 10 没有问题。 硬碟160GB，应该够用。 用notebook的好处之一是可以带到学校去，跟同学介绍OpenSolaris顺便推广一下OpenSolaris的好处及特点。 As best as I know, no one in Beijing uses desktops any more. Everyone, even at home, uses the so-called desktop replacement notebooks. :-) You can go to www.amazon.cn to check out HP 540: http://www.amazon.cn/mn/detailApp?ref=BR&uid=168-0575736-1137025&prodid=rlit003154 （about 4,100 RMB--or about $600 USD, but is currently out of stock.) This notebook has only 1GB DDR2, you may want to spend another 150 RMB to expand it to 2GB. It has no problem running OpenSolaris and Solaris 10. It has a 160 GB hard disk, which should be more than plenty. One of the advantages of buying a notebook, vis-a-vis desktop, is that you can bring it to school, so as to introduce OpenSolaris to your schoolmates, and do a little bit of (the desperately needed) marketing about many of the super dupers of OpenSolaris. -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
Lurie writes: >> If they wanted (because they wanted) some revolutionary python based >> system like IPS, they could have used rPath's conary, which is in >> development since 2004. > > You've talked enough about Conary with Stephen Hahn before, so I'm not > going into this again. Conary? can someone point to an explanation... its baffling to me, maybe many others... ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] disconnecting hdd from zfs mirror hangs whole
UNIX admin writes: >> I'm using what would pass as junk hardware to many here. Its older >> hardware and cheapo cards... but I've had to use PCI sata cards >> because OSOL does not recognize my onboard sata controller. > > Believe me when I write that I completely understand your frustration, > simply because it is my frustration also. > > This has been a weakness of Sun's engineering for a very, very long > time, as long as I've known Solaris (1994). [Alert1 Alert1, longwinded ramble ahead] I was kind of taken by your comments on this. I'd had it in my mind that some things about the free Solaris haven't changed since my first episodes way back about the time you noted. Mid 90's My usage was as an absolute newbie all around, not just to Solaris but to computing at all. On my 50th birthday (1995) was my first encounter with a computer I owned. Up till then what I knew about computers was that the lady at the unemployment office used one to look at my work records. (I'm a life long and now retired construction worker [Welder and High rigger as field construction boilermaker], High school drop out and general illiterate rolling stone, so had frequent experience in unemployment offices around the mid west and in California) My new wife who also worked in offices of UCSB wanted to get `online' so I bought her a computer, and myself one soon after. Back then, talk of the `information highway' was very appealing. Not many lay people had computers at home yet. Those first machines were windows 95 which was pretty new then. I still remember that it took me about a week to learn to control the mouse. A friend of my new wife, was the `network system admin' guy for UCSB (University of California at Santa Barbara). I had landed on a construction job 20 miles north on a Refinery project for Exxon in 1992 and ended up living there until 2002 or so. That friends son (a young teenager and I struck up a friendship which lasted into his adulthood) was a linux fan (a slack rat). Who soon wooed me into the everlasting tinkering a linux person did back then. So I never really did learn much about windows until much later. That tinkering lead me to tryout a free OS being offered by Solaris. As I recall it cost something like $40 to try it out. The OS was free but you paid for the processed CDs. I still remember when I finally got it to boot up that I thought it was really cool to see the Sun logo pop up on my home machine. The console seemed really retarded even then, compared to the console interface on linux. No mouse, no copy paste from terminal to vi or the like. I tinkered with that OS for several months before kind of giving up on it and trying to learn more about linux. I see today in build 110 that the console is still about the same. Absolutely retarded now, compared to what linux offers. All kinds of unexpected behaviour with backspace delete and such. Still no mouse or copy paste to an editor. Seemed to me that the console would have been vastly improved in some 14 yrs. Especially since it seems there are large numbers of Solaris eggheads that are command line oriented people. I don't mean that as a serious gripe since the gui is easily attainable and Xterms offer a sophisticated command line environment. Its just surprising somehow that the console has gone basically ignored. When, unlike in 1995, lots of linux (gnu) tools are common place on the Osol OS now. Back then it was a big chore for an illiterate to even get gnu tools working on osol. To tie up this ramble: To me the big attraction back to Osol came with a fellow on a linux (gentoo) list responding to a post of mine about building up a home NAS, suggested I look into zfs, and osol as a solution and I've been checking it out for a month or two now. Zfs is truly an amazing tool, but the whole Osol package is a bit daunting for a self taught (with massive wholes in knowledge) tinkerer. I'm still a good ways from having created a low maintenance, solid and dependable NAS with osol.11. Something I can just setup and use for backups of the other 6 machines on my home lan. The Cifs server has yet to produce a visible host on the windows `Network Places' dialog. That's not a show stopper but some of the windows based tools I've experimented with using to backup those OSs, want to be able to navigate to the remote share in limited dialog boxes where typing in an UNC address is not an option. The shares are accessible and writable from windows but only if I type in the UNC address into the location box on `windows explorer' file management tool. (or similar tools) Whereas all other Windows shares, and linux shares, simply appear in the `windows explorer' file manager under the network icon. But compared to the two consumer grade NASs available for purchase, that I've tried, osol/zfs seems vastly superior in too many ways to ignore... even with the vast amount of troubles (AND help) I've encountered. ( well documented in my
[osol-discuss] Installation and packaging (was Re: Possible IBM aquisition of Sun)
On Sun, Apr 5, 2009 at 12:22 AM, C. wrote: > I think you're getting something a bit confused here.. Firstly, being able > to selectively update parts of your system is a design choice and feature.. > The benefit I *think* you're referring to is the file level dependency > resolution. With this it upgrades only the *files* you need to vs the whole > package. The motivation for this afaik is to make the dependency resolution > smarter. I like this feature. But IPS didn't invent it. Conary can and does do this since 2004. In conary it allows you - among many other things - to create so called "virtual appliances". Cut-down precisely tailored minimized/shrinked versions of an OS, which only contains exactly what is required to run a desired application foo, with all dependency-resolution done for you, on a FILE BASED level, not on a package based level. But IPS doesn't support that, and lacks most other great features which the original, namely CONARY, offers. Think of conary as something like SVN for the entire installed binary system. Conary also permits you to generate packages by so called "cooking of sources", similar to what pkgbuild does. And all this under one hood. It was there, it was under some opensource license, Sun didn't use it. This would be ok if they either had used another existing technology or if they would have listened to developers like Moinak. Instead all his other BeleniX ideas have been re-wrapped as "Indiana", and he has been urged away. Somewhere Dr. Hahn had written (I must look it up, it was in 2008): "Moinak, it didn't go the way you preferred, that's bad luck for you, now let us work and be quiet." And this after they stole much of BeleniX's key innovations for their own non-mutually NonCommunity created Indiana "OpenSolaris Community Distro 200n.nn"... Well, I didn't want to start that topic again. Because those in charge do not listen. Have a nice day. %martin ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
On Sun, Apr 5, 2009 at 12:22 AM, C. wrote: > Lurie wrote: >>> >>> Yes.. writing an entire package manager from the >>> ground up is *less* work than maintaining the patches. >>> >> >> This is called "moving forward", IPS is based on new novel ideas, a secure >> package manager without any arbitrary post/pre-install scripts, which is >> fast, doesn't hog the system upon installation and is very easy to use, and >> upgrades the whole system at once, versus just the packages, which ensures >> you won't have any conflicts. >> >> If everyone would do as you suggest and just "copy everything" because >> it's easier, there would be no innovations in OpenSolaris at all. And IPS >> *is* an innovation in my book. I've been using IPS since its inception, for >> how long have you been using it ? >> > > ok.. lets keep this technical and anecdotal free please.. > > 1) I evaluated 15 packages managers and packaging formates. This included > not only the package deliver, but also the build process. > 2) I received a lot of feedback from industry experts like Jeff Johnson and > others to ultimately come up with something RPM5-like > 3) I have a proof of concept integration of this new format integrated with > smart package manager > 4) Smart package manager is *very* friendly to upstream and more than happy > to work with accepting patches for new formats > > Now your points... one by one.. > --- > 5) "a secure package manager without any arbitrary post/pre-install scripts" > > Ok.. so wait a second.. Let's first of all define "secure" because last I > checked the IPS authorities aren't signed.. Are they? You're only looking > at it from one angle.. If a package is signed and trusted by the authority > then the script isn't arbitrary. It was designed and created with a sole > purpose. If it *is* arbitrary I don't think I'd blame the package delivery > system, but the policy for those who are backing it. When the IPS repo is > handling 10k+ unique pieces of software lets talk. Those scripts are there > to add *robustness* to the many facets which in a pragmatic world exist with > open source software. Not every package maintainer is being paid to create > their own makefiles and integrate it into SFW. So.. am I the choir on > this.. actually.. I think some features are *good*.. Does removing pre/post > scripts warrant a new package manager.. Bluntly put.. No.. it sure hell > doesn't.. This is why people who have a clue are so pissed off.. Because > good people have been fired before, but there's still this rogue team > wasting cycles on stuff which could ultimately be spent elsewhere.. So > forgive me and other when we are a bit rude and aggressive... > > 6) doesn't hog the system upon installation > > Is this a technical statement? What is this "hog" you're referring to.. > please try to state facts and not opinions.. or at least give some accurate > measure and comparison.. > > 7) is very easy to use... > > ok. you got me there... they were able to make a cli interface that is at > least comparable to things which have been around for 15 years.. :) > However, to have your argument actually be a valid point you'd have to > compare it to something which is common and not easy to use.. > > 8) upgrades the whole system at once, versus just the packages, which > ensures you won't have any conflicts > > I think you're getting something a bit confused here.. Firstly, being able > to selectively update parts of your system is a design choice and feature.. > The benefit I *think* you're referring to is the file level dependency > resolution. With this it upgrades only the *files* you need to vs the whole > package. The motivation for this afaik is to make the dependency resolution > smarter. What you fail to see is a few points though. > > a) That in itself doesn't ensure conflicts won't happen > b) because of the way manifests are created from entire packages just like > any normal or sane packaging system you're still just as likely to have > unresolved dependencies or blockers. (I don't know for certain a missing > file can't pull for *any* manifest even one which provides the same package > at a different version) > c) if you argue that is saves bandwidth.. I'll argue that it has delayed > being able to easily establish a mirroring system, it forces a custom > daemon, there's no on-disk format (maybe this was resolve recently) and you > *still* can't do an offline install. I'll add to that that ever 2 weeks > updating it from europe is *painful*. It crashes, times out and overall > take a long time.. (and I'm not even talking about the dependency resolution > here) For something which with a straight http pull could be achieved in 10 > minutes takes 3 hours! I'm not kidding on this at all. (I've timed it) > > > 9) If everyone would do as you suggest and just "copy everything" because > it's easier.. > > LOL... read above.. who on earth said I copy stuff because it's easier.. > It's called ev
Re: [osol-discuss] Possible IBM aquisition of Sun
Lurie wrote: Yes.. writing an entire package manager from the ground up is *less* work than maintaining the patches. This is called "moving forward", IPS is based on new novel ideas, a secure package manager without any arbitrary post/pre-install scripts, which is fast, doesn't hog the system upon installation and is very easy to use, and upgrades the whole system at once, versus just the packages, which ensures you won't have any conflicts. If everyone would do as you suggest and just "copy everything" because it's easier, there would be no innovations in OpenSolaris at all. And IPS *is* an innovation in my book. I've been using IPS since its inception, for how long have you been using it ? ok.. lets keep this technical and anecdotal free please.. 1) I evaluated 15 packages managers and packaging formates. This included not only the package deliver, but also the build process. 2) I received a lot of feedback from industry experts like Jeff Johnson and others to ultimately come up with something RPM5-like 3) I have a proof of concept integration of this new format integrated with smart package manager 4) Smart package manager is *very* friendly to upstream and more than happy to work with accepting patches for new formats Now your points... one by one.. --- 5) "a secure package manager without any arbitrary post/pre-install scripts" Ok.. so wait a second.. Let's first of all define "secure" because last I checked the IPS authorities aren't signed.. Are they? You're only looking at it from one angle.. If a package is signed and trusted by the authority then the script isn't arbitrary. It was designed and created with a sole purpose. If it *is* arbitrary I don't think I'd blame the package delivery system, but the policy for those who are backing it. When the IPS repo is handling 10k+ unique pieces of software lets talk. Those scripts are there to add *robustness* to the many facets which in a pragmatic world exist with open source software. Not every package maintainer is being paid to create their own makefiles and integrate it into SFW. So.. am I the choir on this.. actually.. I think some features are *good*.. Does removing pre/post scripts warrant a new package manager.. Bluntly put.. No.. it sure hell doesn't.. This is why people who have a clue are so pissed off.. Because good people have been fired before, but there's still this rogue team wasting cycles on stuff which could ultimately be spent elsewhere.. So forgive me and other when we are a bit rude and aggressive... 6) doesn't hog the system upon installation Is this a technical statement? What is this "hog" you're referring to.. please try to state facts and not opinions.. or at least give some accurate measure and comparison.. 7) is very easy to use... ok. you got me there... they were able to make a cli interface that is at least comparable to things which have been around for 15 years.. :) However, to have your argument actually be a valid point you'd have to compare it to something which is common and not easy to use.. 8) upgrades the whole system at once, versus just the packages, which ensures you won't have any conflicts I think you're getting something a bit confused here.. Firstly, being able to selectively update parts of your system is a design choice and feature.. The benefit I *think* you're referring to is the file level dependency resolution. With this it upgrades only the *files* you need to vs the whole package. The motivation for this afaik is to make the dependency resolution smarter. What you fail to see is a few points though. a) That in itself doesn't ensure conflicts won't happen b) because of the way manifests are created from entire packages just like any normal or sane packaging system you're still just as likely to have unresolved dependencies or blockers. (I don't know for certain a missing file can't pull for *any* manifest even one which provides the same package at a different version) c) if you argue that is saves bandwidth.. I'll argue that it has delayed being able to easily establish a mirroring system, it forces a custom daemon, there's no on-disk format (maybe this was resolve recently) and you *still* can't do an offline install. I'll add to that that ever 2 weeks updating it from europe is *painful*. It crashes, times out and overall take a long time.. (and I'm not even talking about the dependency resolution here) For something which with a straight http pull could be achieved in 10 minutes takes 3 hours! I'm not kidding on this at all. (I've timed it) 9) If everyone would do as you suggest and just "copy everything" because it's easier.. LOL... read above.. who on earth said I copy stuff because it's easier.. It's called evaluating and making a good choice. I can show my evaluation and results.. I have a table comparing various aspects and the benefits of each. IPS has some scatte
Re: [osol-discuss] Installation and packaging (was Re: Possible IBM aquisition of Sun)
On Sun, Apr 5, 2009 at 3:46 PM, Joerg Schilling wrote: > I still prefer C ;-) Count me in. +1 (or a dedicated silent minute) for the "Wait and C" programming language. What would the world be without it. -- %martin ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
On Sun, Apr 5, 2009 at 4:56 PM, Lurie wrote: >> *** before you argue with moinak, be sure that you >> have taken the time to read his blog. *** > > As you have probably noticed from my messages, I never try to argue with > anyone, I'm merely carrying a discussion in which I express my opinion and I > don't make anyone "clueless", "you must be joking", "ridiculous", and so on, > as apparently some people here like to do, there's a big difference. I don't make anybody clueless. But everybody who hasn't read Moinak's blog and website and gates _is_ (or makes himself) clueless. That's all I wanted to express. It was a pointer, not an attack. That's why I said "hi" and "regards". Maybe I should have sent this in a private msg. I'm sorry, it wasn't meant as personal attack. SORRY. Also: In my native language the translation for "arguing" is not all that negative. It means "pro-active discussion full of all the _real_ arguments". And since your messages to Moinak contained a number of weak arguments (here we have it again: "argument", not something negative but simply raw facts / and "weak argument" stands for "argument which in reality isn't any, because things are not explained in a logic and verbose manner" The first time I heard the term "weak argument" was when I studied for the GMAT, just for fun, in my spare time, to improve my ability to process data, it was lots of fun, recommended). Also, I cannot know when you joined our community and subscribed to this and other lists. In such a case my default action is, to be "verbose for redundancy". We all don't know us well enough. I hope there will be another summit where we can drink beer and Vodka. :-)) Regards, %martin ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
> The user side experience of IPS is no doubt very good but is no different > from a good Linux package manager like Smart/Yum (with the exception > of ZFS features). I trust you on that one, and Conary, which was pointed out by Martin seems to be nice as well. > From a developer point of view these qualities could have been got by far > less effort and far less code/complexity. It's not up to me to make a conclusion on that, as people have been payed to do research on whether an existing or a new package manager is the way to go. See for example this mail in particular: http://markmail.org/message/wkbbqp5uo6yfunpe Notably, "There are a few features that seem particularly difficult to accommodate in the existing systems: sparse root zones is the foremost (and diskless in general), binaries of multiple bitness ends up being another... ... If you start dropping those, then other packaging systems do indeed become suitable. " -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
> 5. Usually the design is not cross-platformant, IPS can even run on Windows. > > > Weak argument. > conary is also implemented in python with C backend. > Also runs under everything including Windows. > Also most other pkg systems run on every UNIX. Note the use of "usually", and by cross-platformant I don't mean only *NIX systems, I mean cross-platformant. -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
> *** before you argue with moinak, be sure that you > have taken the time to read his blog. *** As you have probably noticed from my messages, I never try to argue with anyone, I'm merely carrying a discussion in which I express my opinion and I don't make anyone "clueless", "you must be joking", "ridiculous", and so on, as apparently some people here like to do, there's a big difference. > Were you talking about with having "Smart" in mind, or without "Smart"? I've never tried "Smart". > When combining an existing (or new) packaging system > with Smart (no IPS, no conary), you can gain similar speed > improvements, along with many other benefits. That sounds good, although I have a question to ask, why are the main distros, i.e. Fedora, Ubuntu, OpenSuse still on those slow package managers ? > And you don't need to employ 50 engineers for 2 years to get a > written-from-scratch > monster like IPS going. The codebase is certainly not that big, and I'm quite sure there were about 5 *main* devs involved. > If they wanted (because they wanted) some revolutionary python based > system like IPS, they could have used rPath's conary, which is in development > since 2004. You've talked enough about Conary with Stephen Hahn before, so I'm not going into this again. > But as Sun always seemed to have too much time and money, > couldn't ever accept help (many examples!) The last part I completely agree with, and if OpenSolaris will fail then this ultimately will be its undoing, the failure to accept devs help without a complicated process of finding sponsors, going through ARCs and so on, which are fine if you do this as your day job, and not if it's your hobby and you just want to contribute. Although there are a few notable exceptions, Jurgen Keil for example is a great contributor, if there would have been 100 people like that, OpenSolaris would be a completely different beast now and would attract even more developers ;) -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The Register: IBM focuses on Sun software trio
[i][On the bright side, we're getting a lot more press lately./i] Yes. Press that brings up Scotty's shares and SUN down. Win one lose one. Actually, the discussion before this was a good one, for a change. Yes, 'invention is faster at SUN', as someone answered to an earlier post of mine. IPS might even become the greatest package manager of all time, who knows? When I met the chief OpenSource evangelist here, in late 2007, I asked him about the need for a new package manager. "Because it is dated". Not a great argument when you work with Unix. Pressed a tad more, and the answer was, that apt would not know about zones. True. And now, what kind of business do we see from SUN? A perfect re-write of a package manager from scratch. A file system, the last one mankind needs (citation), a break-through new debugger, the final solution for network configuration, a service handling utility that finally says 'good-bye' to last millennium shell scripts in some /etc/init. All hunky-dory. If SUN had the cash that flows around in Redmond, e.g. But SUN is aware that it is not doing well financially altogether. Common sense is enough to surmise that one cannot afford too much luxuries with semi-empty pockets. No chance to commence a project worthy of a century (at least a decade, though). What can you actually sell when you're living on a construction site? Of course, all those new glitzy items are fabulous. But where does the money come from? Especially since - and this is no surprise - none of those glitzy new items is - actually - production-ready. In the sense that I would want to trust my atomic submarine to it, or the accounting of my 5 billion $ bank? If one intends to stay in business, one also has to make compromises. UFS might have done, for some time being. Or, apt might have done. As long as a fully ready new item can be sold, the markets will buy. Yes, this is the responsibility of the management. Or, it would have been the responsibility. Instead, world and sundry was treated to new heights, the all-encompassing cloud. Instead, what we are treated to is 'ongoing work', i.e. half broken. Nevermind, but where is the market sense of some people?? (Actually, about the only thing that was miraculously not re-written was the boot loader. And that's the only item that - at least here - has actually worked flawlessly. This is a great argument, by the way, since it didn't suit ZFS-boot, and still was 'only' adapted, and might not even see the modifications accepted upstream.) Uwe -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Installation and packaging (was Re: Possible IBM aquisition of Sun)
Martin Bochnig wrote: > conary is also implemented in python with C backend. Also runs under > everything including Windows. > Also most other pkg systems run on every UNIX. rpm is not the only > example. Think of Nexenta aka gnusolaris ... > Do you suggest, IPS is more cross-platform, simply because it's > written in python? Isn't C a portable language, too, if written > properly e.g. in standards compliance? Portability is not the result of using a specific programming language but a result of writing things in a portable way inside a portable framework. The build system used by ON and the SVr4 package system is not supporting portability. If you like to run a binary on MS-WIN that was compiled from a C source, you just need this binary. If you like to run a python program on the same platform, you need python. If we like to discuss installation and packaging, we should talk about installation and packaging - the programming language is less important although I still prefer C ;-) Jörg -- EMail:jo...@schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin j...@cs.tu-berlin.de(uni) joerg.schill...@fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The Register: IBM focuses on Sun software trio
Ah well, as long as we are speculating and reading into rumours (and I can't sleep anyway) I may as well contribute :-P Words from Scott a week or so after the initial 18th March rumour: http://searchdatacenter.techtarget.com/news/article/0,289142,sid80_gci1351994,00.html "Sun's McNealy touts open source, bashes Oracle and IBM McNealy acknowledged that Sun's reduced status in the marketplace meant that it didn't have the funds to launch big awareness campaigns with the Gartners of the world anymore, and that most press only reported stories about Sun's poor financial standing. To get the word out about its technology, "we have to rely on word of mouth." "It's a struggle," he said, "but we'll get through it. On the bright side, we're getting a lot more press lately." -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
> Not really, you missed the point. There are places where there is scope > for innovation and people know they have ideas that go a lot beyond > the current stuff that is deserves a clean slate implementation. Like > ZFS. The ideas expressed in ZFS are revolutionary to say the least > and could not have been done by re-using existing stuff. However there > should exist a balance between redo everything and re-use otherwise > one would start re-writing every piece in the name of innovation. > OpenSource is also about a balance between the two. Quite; and how ZFS is introduced doesn't require you to reinstall and retrain your operators. How ZFS is introduced is a wonderful example how disruptive technology can be deployed without require reinstall or much required training. first ZFS was introduced then ZFS becomes bootable (install) and also ZFS can be used by liveupgrade But the system works as before. > I am forced to work with IPS day in and day out at work. I have > submitted bugs with fixes and working on add-in modules. I am very > familiar with the codebase and inner workings of the complex beast, > so I know what I am talking about! > > The user side experience of IPS is no doubt very good but is no different > from a good Linux package manager like Smart/Yum (with the exception > of ZFS features). From a developer point of view these qualities could have > been got by far less effort and far less code/complexity. > I agree; if you look at the new features in OpenSolaris which are disruptive, I come with the follow list: DTrace ZFS SMF FMA BootAr IPS IA-Install Clean slate?Y Y Y Y Y Y Y SA notices? N N Y N Y Y Y Incompatible? N N N N N Y Y Training req? N N Y N Y Y Y Reinstall? N N N N N Y Y Cannot avoidN N Y Y Y Y Y Risk0 0 2 1 2 5 5 Clean Slate: this was new technology developed with a clean slate SA notices: unless you read the (marketing release/release notices), you will not see this new technology. Incompatible: old software and practices work Training req: training is REQUIRED when administrating Solaris Cannot avoid: this feature cannot be avoided Reinstall required: when using this new software, install is needed Risk: is my estimate for the risk to making this change to (Open)Solaris. I haven't seen a lot of problems with FMA, except when you want to repair a system; I have seen problems with SMF and the Boot archive (corrupted repository, out-of-date or unbootable boot-archives). Problems with DTrace and ZFS can, of course, all be avoided because you don't actually need to use them. In the end it comes to compute the benefit vs the risk. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
On Sun, Apr 5, 2009 at 10:24 AM, Lurie wrote: > > While I agree with you about that, I still stand by my point that IPS is a > good thing, and given Sun's need for paid-support repositories, integration > with zones, SMF support, ZFS support, they would've ended with a lot of > patches on their hands. You didn't read Moinak's blog. Plus (my personal amendment): Instead of either writing IPS from scratch, or going the BeleniX way, they could have integrated that functionality into conary, instead of doing that expensive slow new-implementation. > As for the time it has taken, it's not just the development time, once you've > written some complex piece of code in say N months, you could easily > implement the whole thing from scratch in N/M months, it's the slow > development of ideas that should be counted too. > > Moreover, now when most of the code is in place and working, the team can > quickly introduce new features as they are well familiar with the codebase > and know the design well to easily extend IPS. You mean now that the code is in place"? Like ... "Now, after facts have been created without ever having asked the community (made a referendum) for what it wants? "Now after 2 years?" "Now after all those engineers involved had to work on - and get money for - this, rather than on/for something more innovative (IPS is not an innovation, due to the long-existing conary) and more useful?" >> The user side experience of IPS is no doubt very >> good but is no different from a good Linux package manager >> like Smart/Yum (with the exception of ZFS features). > > Unfortunately, I don't find yum to be a good package manager for some of the > reasons I listed above. Bingo. That's why Moinak said "Smart/Yum", rather than "Yum". That's quite a difference. >> From a developer point of view these qualities could have been got by far >> less effort and far less code/complexity. > > In the short run ? Maybe. In the long run ? I don't think so. Please read this: http://moinakg.wordpress.com/2008/09/18/belenix_packaging http://moinakg.wordpress.com/2008/09/22/belenix-packaging-and-repository-directions-part-2/ http://moinakg.wordpress.com/2008/11/22/the-belenix-package-manager/ Also make some research about conary. I wonder if you might not consider changing your mind during the process. Regards %martin http://opensolaris.org/os/project/conary-eval/ ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] When can we expect nv111?
On Sun, Apr 5, 2009 at 12:20 PM, Uwe Dippel wrote: > If memory serves right, it is > 1 fortnight since the nv110. > Usually, the update came before/early in the weekend (as of our time zone). > Or will it be skipped? I've heard it was re-spun (i.e 111a) and hence will be later. -- Regards, Cyril ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
On Sun, Apr 5, 2009 at 10:24 AM, Lurie wrote: > > 5. Usually the design is not cross-platformant, IPS can even run on Windows. Weak argument. conary is also implemented in python with C backend. Also runs under everything including Windows. Also most other pkg systems run on every UNIX. rpm is not the only example. Think of Nexenta aka gnusolaris ... Do you suggest, IPS is more cross-platform, simply because it's written in python? Isn't C a portable language, too, if written properly e.g. in standards compliance? Forgive me, but this argument (5.) was a void joke. Regards, %martin ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] When can we expect nv111?
If memory serves right, it is > 1 fortnight since the nv110. Usually, the update came before/early in the weekend (as of our time zone). Or will it be skipped? -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] disconnecting hdd from zfs mirror hangs whole
> I'm using what would pass as junk hardware to many > here. Its older > hardware and cheapo cards... but I've had to use PCI > sata cards > because OSOL does not recognize my onboard sata > controller. Believe me when I write that I completely understand your frustration, simply because it is my frustration also. This has been a weakness of Sun's engineering for a very, very long time, as long as I've known Solaris (1994). Excellent engineers, excellent engineering practices and processes, but NEVER a product that works 100%, with all kinks worked out. Always phenomenal ideas, but never a 100% working product. About 75%, give or take, is what gets released. Is new functionality needed? Yes, of course. But for once, I'd like to have most of the bugs fixed and 100% working software, than 75% working software and tons of new functionality... because in the end, none of it works correctly when all put together. Last week I wrote that it remains to be seen whether Sun's new religion of "release early, release often" turns out to be the correct path. Now, several kernel panics (b109, b110) and countless hours of hacking later, I'm really beginning to believe "release early, release often" is an approach Google should be whacked on the head for at least 10 times a day, and whoever copies Google... well, fill in the rest yourself. -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
*** before you argue with moinak, be sure that you have taken the time to read his blog. *** On Sun, Apr 5, 2009 at 10:24 AM, Lurie wrote: >> So modern Linux package managers do not have any of these qualities ? > > Some of them ? Of course. All of them ? No. Examples? Details? Were you talking about with having "Smart" in mind, or without "Smart"? > Just a few points: > 1. They usually upgrade a live system (and while it's possible for some to do > a non-live upgrade, the live upgrade is exactly how pretty much all of them > operate by default). > 2. The upgrade process is usually very slow, I've been upgrading Fedora, > Ubuntu, OpenSuse, what they have in common ? A very slow and IO bound > upgrade, during the "update/installation" phase which took most of the time > the whole system slows down, to put into perspective, last time I did a > nightly upgrade of fedora it took around 1.5 hours, while the download phase > took only 10-15 minutes. Then you still didn't need a new implementation. IPS is some type of limited feature-restricted conary-clone. However, much depends on the directed graph dependency algorithm (not [or]) in use. When combining an existing (or new) packaging system with Smart (no IPS, no conary), you can gain similar speed improvements, along with many other benefits. And you don't need to employ 50 engineers for 2 years to get a written-from-scratch monster like IPS going. If they wanted (because they wanted) some revolutionary python based system like IPS, they could have used rPath's conary, which is in development since 2004. But as Sun always seemed to have too much time and money, couldn't ever accept help (many examples!), plus because re-inventing the wheel seems to be fun, as long as they can call it MyNewSunWheel, many resources and time have been spent for something which was everything but necessary, i.e. WASTED. While other groups coul not even get a repaired coffee machine from their management ... (when I was in Menlo Park 2 years ago, as non-Sun visitor). ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Possible IBM aquisition of Sun
> So modern Linux package managers do not have any of these qualities ? Some of them ? Of course. All of them ? No. Just a few points: 1. They usually upgrade a live system (and while it's possible for some to do a non-live upgrade, the live upgrade is exactly how pretty much all of them operate by default). 2. The upgrade process is usually very slow, I've been upgrading Fedora, Ubuntu, OpenSuse, what they have in common ? A very slow and IO bound upgrade, during the "update/installation" phase which took most of the time the whole system slows down, to put into perspective, last time I did a nightly upgrade of fedora it took around 1.5 hours, while the download phase took only 10-15 minutes. 3. The whole package content is downloaded (recently there has been a trend towards moving to downloading deltas only, but still the default in most is to get the whole content). 5. Usually the design is not cross-platformant, IPS can even run on Windows. > Strange!! And how difficult it is to implement no-scripting limitation in an > existing packaging system ... patching something for that would be just as hard as implementing that particular functionality from scratch, and you would have to maintain patches with the upstream (unless you have an ideally de-coupled backend API, which I doubt would be feasible to have once you start adding more and more features), forced to use whatever libraries/languages the upstream uses, and so on. > Like ZFS. The ideas expressed in ZFS are revolutionary > to say the least and could not have been done by re-using existing > stuff. However there should exist a balance between redo everything and > re-use otherwise one would start re-writing every piece in the name > of innovation. While I agree with you about that, I still stand by my point that IPS is a good thing, and given Sun's need for paid-support repositories, integration with zones, SMF support, ZFS support, they would've ended with a lot of patches on their hands. As for the time it has taken, it's not just the development time, once you've written some complex piece of code in say N months, you could easily implement the whole thing from scratch in N/M months, it's the slow development of ideas that should be counted too. Moreover, now when most of the code is in place and working, the team can quickly introduce new features as they are well familiar with the codebase and know the design well to easily extend IPS. > The user side experience of IPS is no doubt very > good but is no different from a good Linux package manager > like Smart/Yum (with the exception of ZFS features). Unfortunately, I don't find yum to be a good package manager for some of the reasons I listed above. > From a developer point of view these qualities could have been got by far > less effort and far less code/complexity. In the short run ? Maybe. In the long run ? I don't think so. -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
