Re: [osol-discuss] Garrett D'Amore joins Nexenta. if needed, can nexenta become the fork?
Roger Bisson wrote: James, All OS platforms benefit from the network effects of wide usage. I agree, but 'free as in beer' helps here more than 'free as in source'. To the average developer, Solaris (which, in a closed model, you would have to pay for) I think you are confusing 'free source' with 'free download'. I do think its in Oracle's interest to allow hobbyists and micro-companies to have a very low-cost solution and the admin costs may mean that 'free as in beer' is no worse than a low cost license. If we agree that applications often drive the widespread adoption of a platform, then application developers make the primary decision. Sun's decision to make Solaris generally available at nominal or no cost (and by Solaris, I also refer to OpenSolaris) places SunOS (and complementary technologies) in the hands of developers at little cost; similarly when compilers and development tools are given away freely as in Netbeans, JDeveloper and the like. I agree. I think making it widely available (and 'supported' on non-Oracle hardware) is very important. But the number of developers who need or even want the Solaris source code is rather small and making it (the source) available doesn't necessarily have any impact at all on whether ISVs can or do use it. To some extent ISVs will do whatever they need to target their market anyway and if big companies use Solaris and want software for it then ISVs will supply it. Personally I'd like to see a product priced at OS/X but without the nasty legalese - but the big issue there is really that Sun never managed to leverage the binary interfaces and their industry position to get a better hardware supportstory on modern motherboards than Linux. (I don't want support for basket case old hardware - but good support for what's on the boards now, including embedded graphics, gig-e and RAID chipsets). Compare this to HP-UX, and AIX which require access to more specialised technology platforms. As a developer, I do not have access to these platforms and am unlikely to develop software for them unless there is a specific need to. The growth of Windows in the data centre is a prime example of technical folk recommending technologies they are familar with. Indeed, though in fairness Windows is a rather good server platform and works well, and has particular advantages with Windows clients, not least with the integrated security. Also from memory you can sign for access to remote dev hosts for both HP-UX and AIX so you can beetstrap to the platform without buying your own. Oracle could easily make Solaris available for charges comparable to Windows Professional, and Windows Server (circa £100, and circa £800) and attract sales but it would also require Oracle to make a commitment to support (and provide warranties with respect to) non Sun hardware and offer wide driver support which it may not be prepared to or have the resources to do at this time. Indeed. But i don't think 'open source' will fix that. The sort of hardware that's very attractive as a user is often only supported marginally on Linux and really needs the hardware vendor to provide timely support. I am not sure what you paid for your IPX but it would have cost several thousands of pounds in or about 1992. It cost a small fortune. Worse, the crappy C++ compiler cost as much as MSDN Universal and didn't come with a year's worth of support and upgrades. What are you running OpenSolaris on? VBox and native on a PC. But I have some E420Rs and E280Rs I'd like to run it on too. Shame they're so noisy. Open sourcing allows, to some extent, the company to concentrate on the stuff that matters to it (like the underlying service and infrastructure technologies) to support "big iron" technologies that it can sell to organisations, while reducing its cost of supporting commodity technologies such as graphics cards etc and slick user interfaces on cheap commodity equipment for developers: Xorg, Gnome, KDE etc all of which are open source projects. Here we disagree. I don't want crappy partial support of modern nVidia and ATI graphics - I want the realy full-fat blob that also runs the card on Windows 7, and that needs core engineering and will be easier to organise with NDAs and joint engineering resources. I'd rather pay for a consumer product and have that support than the sort of support that you get with 'freedom first' Linux solutions. My personal view is that Solaris is more viable in the long term with the existence and continued development of OpenSolaris, than it would be if OpenSolaris were simply abandoned; the value of the open source model being that those consumers working with OpenSolaris on commodity equipment may be somewhat less demanding than paying customers would be. I agree, but I remain unconvinced that the availability of source for the kernel is very important. This isn't Linux - the kernel interfac
Re: [osol-discuss] Garrett D'Amore joins Nexenta. if needed, can nexenta become the fork?
Roger Bisson wrote: Personally, I am of the view that Sun's original decision to release Solaris in the form of OpenSolaris was absolutely the right decision to ensure Solaris' maintenance and growth as a platform by making it available to developers and technicians (thereby encouraging its application in commerce). Do you have any evidence of this? SunOS and Solaris had a long history of third party support (and use 'in commerce') before that. I would have said that, by virtue of open sourcing, both individuals and commercial enterprises have been significantly more likely to deploy (Open)Solaris than they have been to deploy such operating systems as HP-UX, AIX and SCO's OpenServer, all other things being equal. While I can't see 'its open source' being a black mark for anyone, and hence those people for who its a positive issue would make it a net win, you say 'significantly more likely'. What evidence do you have of this? There's a big danger we can all talk ourselves into believing that open source is somehow necessary. If we can get to a point where companies like Nexenta are able to fund significant development and its rolled into the core, then that's some evidence in itself. But their resources are still limited compared to Oracle. Personally I've been using SunOS professionally since an IPX was a shiny cool thing to have, and I even bought a twinhead sparc5 clone. I really don't care whether Solaris is open source at all - just so long as somebody (or somebodies, I don't mind) can make enough from it to justify spending money on its development, and it doesn't bitrot to oblivion. I'm not convinced that having something open source actually helps that justification (or at least, has helped) - but I do think we should avoid assuming the answer when there is so little real numerical evidence around, and we should support Oracle in doing anything and everything the need to keep Solaris viable. And if that means we don't get things we used to get from Sun, then so be it. I'd rather have a closed Solaris, and a choice, than a runty abandoned Solaris that leaves me using the usual suspects. Let's see what Oracle do in the enxt 6 months, and lend our support to Nexenta and hope they can make a go of it (ie one that outlasts the venture exit points etc). James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Any news about 2010.3?
Simon Phipps wrote: Exactly right. As it turned out it was also comfortably profitable. Sun's terminal problems lay elsewhere. Which bit of the last financial statement shows evidence of this - that the profits came from open sourcing, not merely from software that was open source before Sun bought it or that was already being licensed before it was open source? (Hmm - and in mysql's case, we could also ask 'how much more profitable than spending the billion on gov't bonds'?) James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Any news about 2010.3?
Matthias Pfützner wrote: You (James Mansion) wrote: if you look at Sun's annual earnings documents, you might notice, that most of Sun's revenue and especially margin was generated by big iron hardware. Indeed I had. And the money coming in from software was mostly licebnses to OEMs (think Java), as far as those numbers have ever been published. So, I'm a bit interpreting stuff here. I agree. And I think that revenue did not improve materially from open sourcing that IP - while the move to open source surely gives up a degree of control. apps. And, yes, that worked! Not in the way many hoped it would, but yes, it generated more mind-share, and with that additional HW-sales. Microsoft have never had a problem with ISV mindshare because they made things cheap and accessible. Free is just one form of cheap from that perspective. I absolutely believe that Microsoft's success is not nearly so much related to monopolistic practices as to their early ability to court ISVs and their ability to use ABIs (like VBXs and OLE controls) to create a marketplace for small ISVs. Microsoft took their eye off that ball a few years back but they seem to have recovered. Maybe I'm just an old fart but I recall my dismay at the Byte headlines that OO had failed and components had won. I was an early C++ adopter and it was galling - because it was true. How many businesses ever got anywhere with aftermarket controls on any of the X toolkits? Maybe Qt will create an ecosystem - I don't know. But I think the lesson was that open standards don't create that sort of ISV-friendly environment on their own and the existence of such a market does wonders. Look at the iPhone app shop. Same thing again. So, Oracle now is the second biggest SW-company of the world, they KNOW how to monetize SW, and I hope and am sure, we will see some big monetizations coming out of the assest that Oracle got with the acquisition of Sun. And that in turn will again allow to let OpenSolaris live as well as Java. Larry stated it cleary: "It's not needed to produce margin or revenue directly, as long as it helps generate revenue and margin over-all!" Well, I hope so. I really want Solaris and Java to survive and thrive - hopefully in a form that I can afford. I don't want or need the source code for that to happen and I don't believe anyone outside of Oracle really needs it either - I think 'free as in beer' is enough, though as I've said to Martin its advantageous if the delivery is in a form that allows custom distributions to be created. Time was we all linked our kernels from objects, what was wrong with that? Modularity is the key - not an ability to run cc. And personally I'm entirely happy to pay (a modest amount) for security fixes and upgrades. I'm not questioning whether its worthwhile courting a hobbyist or SME market or even universities. I *am* questioning whether the full monty open source is necessary - or particularly helpful for long term viability, given what the release costs you. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Any news about 2010.3?
Мартин Бохниг (Martin Bochnig) wrote: No, you are missing the point. The idea was to provide a free and open AllInOne A to Z software platform stack, that would be able to compete with LinUX and to win the OS battle. Then that was a mind-numbingly stupid strategy - because: a) it would take so long that Sun would be out of business first (oh, look ...) b) I'm not sure that's what customers really want c) having the best free stack is pointless without revenue This would have increased hardware sales and sales of service/maintenance/support contracts. So - we've given up trying to be like Microsoft, let's try and be like Red Hat? The difference, however, is massive: look at how much IP Sun created, and how much Red Hat create. Red Hat are canny in employing high profile (and high influence) Linux hackers, but how much do they actually fix or write? They look good on stats next to other Linux collaborators, but next to Sun's Solaris investment? If you want Solaris to be Linux then say so. To me, a good differentiator was that Sun had leadership, authority, and resources. And you need those things to avoid tinkering on the periphery, and to enable doing hard or just nasty stuff. Unfortunately, two of those things seem to have caused what some people think is an epic fail in community engagement. Trouble is, I'd rather Sun had those things and could drive Solaris than that they became great community members and mere peers. To you, maybe Open Solaris is an opportunity to play at being an OS provider. But to an OS user like me it was a way to observe the process and see what's coming - and get an affordable and reasonably current release based on the observation rather than marketing getting feature complete for an 'industrial' release. Whether or not source code was available to anyone outside Sun is largely irrelevant to me *providing* the binary bits are available in a way that would enable repackaging in a way like Belinix or Nexenta, which do provide refreshing alternatives of approach. I do think there's plenty of value to be had from openness that doesn't cede control or even necessarily expose sources - its just a shame that there has been some severe issues with expectation management and/or execution in the case of Open Solaris. There the cash would have come from. See, how IBM makes a living? I don't see DB2 or AIX or the OS/400 or mainframe stuff being free. Sure, IBM sell services too, but they don't give away their IP. Looks like they're happy to milk Sun's tho. I'm not saying you can't make a moderate open source business selling services on Other People's Stuff. I'm concerned about what happens when you're the main author, and carry all the R&D costs, and all the pre-market investment risk. Bear in mind that the time delay between the engineering investment and the development technology risk both conspire to mean that you have to get a big return to pay for that R&D - and if the stuff is free then the other guys you've let into the playing field as peers don't have those costs or risks to recoup. Or you can embrace the emperor's new clothes and develop in the open, and lose control and any USP. Maybe Sun could have gone all out for community and given up control - and fired 90% of their engineers to reduce costs in line with a new business model where they're packagers and Innovation Happens Elsewhere? Is that really what you wanted? Could that build ZFS? Java? Its usually bad enough having bike shed discussions internally. Sometimes authoritative leadership with resources is necessary - compare and contrast the Debian ecosystem before and after Mr Shuttleworth's intervention. ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Any news about 2010.3?
Мартин Бохниг (Martin Bochnig) wrote: Solaris code for marketing purposes, rather than creating an independent, community-led, open source project with the ability to make real decisions. I think you're missing the point. What is the benefit to Sun shareholders to have Solaris so open, really? Simple question - where does the money come from? And yes - they could have engaged and formed more of a community. Quite true. But I don't accept that its necessarily relevant, unless you really think that Sun would fire 90% of its engineers and expect most of the development to come from the community and run lean and mean as a distro creator like Red Hat, with a few high profile engineers to show they have commitment and at least some core skills. Whether or not OpenSolaris has a community or not is all but irrelevant so far as I can see because either way its not a revenue generator, and what Sun needed was revenue. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Any news about 2010.3?
Joerg Schilling wrote: You are correct: without OpenSourcing Solaris, Sun would have been in trouble earlier. What evidence do you have for this? I know there have been externally sourced code contributions, but how much of it needed source rather than the stable ABIs, and how material is it really? More to the point - what revenue did it generate? So you can answer your question "I'm not sure how the open sourcing was successful for Sun shareholders." with a _yes_, as it helped to raise the Sun stock price. Very briefly, yes, but since then I think it hasn't materially helped shift anything that generated revenue for Sun, and Sun's stock hardly trended up on its recent improving quality. I further believe that a closer collaboration with the cummunity (as intended by Sun in September 2004) would have given the additional momentum for Sun to push it into the win zone for a longer time. Well you can believe that, but belief in that is behind a lot of the open source hype, and one thing that's hard to find is concrete evidence of how it translated into a sound business plan and revenue. This is however a lost chance and we cannot roll back time... Indeed, but we can try to learn from it, and I think questioning how and when open source is helpful to a technology creator is worthwhile. Sun's position was fundamentally different to Red Hat and post-NetWare SUSE, since they owned the things they were giving away, and had done most of the development (or had paid for it). I think its worth considering what they could have done differently and why. I can't help thinking that they would have had a better chance by going in completely the other direction and using their grip on Java and Open Office (and mysql, eventually) to try to cut off Linux's air supply and stunt its datacentre growth until Solaris on X64 could get a decent foothold, but its all conjecture, and it would have accepted handing a potential short-term gain to Microsoft. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Any news about 2010.3?
Simon Phipps wrote: Ths thing I find interesting in the article, and indeed in many of your statements, is that you show absolutely no sign of self-doubt about whether open sourcing everything you could actually destroyed shareholder value and drove Sun down the toilet. That's because it did not. See the penultimate paragraph of http://webmink.com/2010/03/08/sundown/ I don't understand. I'm looking at 'we've achieved some amazing things ... despite the success of Sun's open source business, it still wasn't enough to rescue Sun'. That looks self-congratulatory to me, not doubting. I'm not sure how the open sourcing was successful for Sun shareholders. Definitely successful for Red Hat shareholders though. Where's the bit that says 'maybe embracing open source was a huge mistake and we screwed up'? Its one thing to embrace open source by consuming it, but embracing it by taking a huge IP investment and chucking over the wall? (Well, mostly ...) The whole strategy seems to have been predicated on 'the enemy of my enemy is my friend', presumably based on a massive chip-on-shoulder brought about by NT eating your market in CAD and financial workstations. Hopefully Larry's management team will see that there IS some market for a not-Windows alternative for PC-clone workstations and that the consistency, stability of interfaces and compatibility that defined Solaris are a differentiator that can make it more attractive to OEMs than Linux variants. But I'm not hopeful. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Any news about 2010.3?
Simon Phipps wrote: It's been an interesting "does he take sugar" experience watching the conversation about me; I thought I'd interject with a link to a story that has correct information: http://www.infoworld.com/d/adventures-in-it/former-sun-open-source-officer-joins-osi-board-109 Ths thing I find interesting in the article, and indeed in many of your statements, is that you show absolutely no sign of self-doubt about whether open sourcing everything you could actually destroyed shareholder value and drove Sun down the toilet. I'm certain that there was a decision to be made about being mean and nasty to the open source systems that were eating your lunch (and dinner) or whether to cosy up. But I'm not sure you made the right one, because Solaris did run on Intel and Java was a powerful card, and the companies that are still standing are ones who didn't give away all their IP and show their competitors their cards. You might have done the right thing for society as a whole (maybe: we'll see how your big customers fare and whether ultimately there's a reduction in real choice) but I'm not at all sure you did the right thing for Sun shareholders. I'm not one, but I've been a fairly happy user at major banks. (I was a less happy personal customer - I paid as much for your C++ compiler on Solaris 2.5 as I did for MSDN Universal - and guess which got me support and updates for a year? And which was the better compiler?) Personally I don't give a fig whether Solaris is free (in either sense) - so long as it has a version that's affordable, preferably in the ballpark of Windows or MacOS. After all, Esix and Interactive are long gone. It remains to be seen whether that happens. I'd be quite happy for Oracle to dismantle the freedom in exchange for a hundred dollar UNIX with a stable ABI and an NDA-happy relationship between the driver integrators and nVidia and ATI that can deliver Mac-like stability and video integration and performance on vanilla hardware, but I suspect I'll be disappointed since Oracle have hardly courted that sort of market. Maybe I'll just have to get a Mac - I've been resisting. :-( Let's remember: Sun opened a lot of technologies - but I rather think it was done without seeing how to monetise it. And you can't put the genie back in the box. Sun is now history. Maybe it would have been anyway, its speculation, but the management at Sun failed, badly. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Solaris 10 - no longer free
Giovanni Tirloni wrote: Personally I'm an open source advocate so any software business model that is closed will have to try hard to convince me that it's better than all the community can provide in testing, bug reporting, free marketing, free support, etc. You might be an advocate, but have you tried to make money 'doing open source'? Red Hat makes money essentially by providing a de facto standard and having got there first so that third party support got there - and its self-fulfilling then. As consumers, we basically want a near monopoly (for ubiquity and standardisation) and an also ran to keep the main player honest. There are a lot of people with big opinions advocating free software, so long as someone else pays for it, and takes all the business risk. Oracle is a successuful business, and what makes it successful is that it has lots of customers who pay, not lots of advocates who think the sun shines from you know where. But if Oracle is exclusively focusing on taking market share from AIX and HP-UX, perhaps Larry can look at it as a "I have more developers working on Solaris then you" and disregard the outside contributions completely. He has more than Red Hat too, at least in terms of engineers he can focus and instruct. Lets wait and see what he does with it. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Performance of opensolaris
Bruno Damour wrote: The surprising fact is that the performance is still slightly better on my laptop than on the server. This has been true with postgresql 8.3 and now with postgresql 8.4 as well. ... Is there any explanation you would think of ? If you examine the disk in device manager on XP, what is the caching policy set to? ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Error : symbol __0dLACE_WStringEnpos referenced symbol not found
Rekha wrote: fatal: relocation error: file server: symbol __0dLACE_WStringEnpos: referenced symbol not found Killed can some one please help. not able to solve this error.. Guess: find the ACE library you are using, and make sure that you have also recompiled it on the new system, with the new compiler, and the same compiler options. ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] iSCSI Boot for x86
I saw a PDF with some 'what's new' info in it last week, and one headline was 'iSCSI Boot for x86' (or x64 I guess). But I can't find that PDF in the new web site now that 0906 is out - and there is no mention of iSCSI boot anywhere. Does anyone know anything about this? James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Updating 11/2008 to build 108 with Package Manager
Shawn Walker wrote: There are a set of rules about what characters are allowed in the name. Spaces are one of the characters that are not allowed. I seem to remember being bitten by that too, albeit briefly. I seem to remember cursing the lack of clarity of diagnostic. And to be honest, if I'm going to supply an URL and a name, then I really expect the name to be for my benefit and somewhat free-form. The limitation seems out of place. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Any plans to update the "roadmap"?
Jim Grisanzio wrote: > location. Just a suggestion. That conversation will be on website-discuss. > I can see how it comes to be there, and at least it is *somewhere*, but it seems wrong to me. Maintenance of a coherent roadmap should be a function of project leadership (and from my point of view a differentiator between OpenSolaris and Linux systems in particular and *BSD as well to some extent, but that's possibly contentious). I'd hope that website-discuss be a place to determine presentation issues primarily, and that there should be a chain: roadmap maintenance -> communications policy -> website presentation issues (And to me the last of these should be on website-discuss) After all, it might be appropriate that the roadmap should be available from a PDF library or in a tracker etc - while these things might all ultimately be available externally solely through the web site, the same could be said of code etc. It might even be that the current roadmap should be available as a package and/or as part of the current system image core documentation. ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] will opensolaris ever be released under the GPL?
Tim Scanlon wrote: > The GPL is crap, I hope to god it doesn't get used. Corporations violate it > all the time, developers don't bother to read it, and it's an awful license > to use if you want to feed yourself with your efforts. > The thing I find mot odd about GPL is that if you work in a company that is quite big - but not so big as to have different business units for tax purposes - then it gives you a free ride. At least it does if you allow linking to non-externally-distributed code as well as 'mere usage'. But if you're the last of the garage developers, you're screwed, or at least liable to get infected. Sort of 'steal from the poor and give to the rich'. It doesn't do anything to help SMEs with no internal resources who desperately need a lively small ISV environment to buy from. No problem for corporates that can afford to hire contractors like me to get what the need though. It just seems upside down to me, at least in terms of how I think 'power to the people' needs to be delivered to avoid a widening have/have not divide. You can't expect the guy running a cornershop to join a developer community but he still needs accounts, payroll, and possibly ecommerce, and he can't possibly be expected to pay for their development axcept as a share where the total cost is amortized across a lot of users. I think the underlying problem is that it tries to give freedoms that some users don't all need, to all users, and in so doing can make it hard for small providers (that some users do need, because they can't afford to pay for the whole cost of custom development) to turn a dollar by spreading their R&D across a number of clients. I think the problematic right that's transferred is that the receiver of the customised software can effectively compete with the customiser (or at least gut their business case) and copy it on to *anyone*. If you could conceive a license that requires all source to be provided to 'customers' of a perpetual basis (ie cut out all the escrow issues, and give them free reign for further customisation - where they can share maintenance with other 'customers' but not compete with the provider) - then everyone involved can get what they _need_. Its unfortunate that the simple 'non compete' that makes it practical also makes it a non-free license, despite that customers having the freedoms they actually need as non-technology companies. Just not the ones the FSF say they must have. Oh well. Never mind. There's more to life than 'open source'. JAmes ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Triple Boot with OpenSolaris as 3rd OS
Glenn Lagasse wrote: > Whichever disk is the 'master' disk. By that I mean, whatever disk is > listed to boot first in the bios (which is where all the other > bootloaders installed to). > Fine - so disable the other drives and install onto a dedicated drive, and then re-enable the others. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Triple Boot with OpenSolaris as 3rd OS
Glenn Lagasse wrote: > The OpenSolaris installation will overwrite your existing Grub in the > MBR. > Yes - but on which disk? If there are three hard drives and each has just one OS, surely you just go into the BIOS and disable the drives except the target one, then do an install 'as if' the only drive, and then do the chainloader trick described elsewhere? I installed Vista first into one drive, then OpenSolaris and then Ubuntu onto partitions on a second drive. I use OpenSolaris as my master boot manager and chainload to the Vista drive or boot Linux from GRUB. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Support for AMD 780G chipset? OMG!!!
Alan Coopersmith wrote: > As soon as ATI releases an open source driver, or enough information > to the open source developers so they can make a driver, we'll have > Solaris support.I have no idea if or when either of those are > likely to happen though It continues to amaze and confound me that a company ith Sun's resources cannot arrange with ATI to release a first class binary driver that supports Solaris and Open Solaris. AMD is a partner, right? Please remember that there are users who could hardly care less whether source code is open. Sun has a potential advantage over its competitors but it seems unable to exploit it. Bizarre. 'There is no open driver' is one thing. 'There is no driver' is quite another. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Why are there no 16 socket (or more) x64 servers on the market?
Mike DeMarco wrote: > I believe that Windoz has a problem dealing with more than 4 CPUs > and the hyper-bus architecture for AMD processors extends to 4 slots. > So I do not believe there is a driving need at this time for such a beast. > > What on earth gave you that idea? Its not hard to find what the limits are, look here: http://www.microsoft.com/technet/windowsserver/evaluate/features/compare.mspx Not sure why they quote 32-way. See here: http://www.tpc.org/tpcc/results/tpcc_result_detail.asp?id=105112801 Admitedly, that is not your average PC, and while its fast its price-performance isn't leading, there are no Solaris systems on the list that beat it in absolute performance. Nor, indeed, any SPARC ones. Hopefully Rock will help out, eventually. I think 64 *is* Windows' limit though, so they can't use the 4-thread-per-cpu hardware used in the #1 performer. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [indiana-discuss] I'm sorry, but I just don't get it
Calum Benson wrote: > GNOME's user-admin preferences window, IMHO. I suspect a sizable > number of users would have insufficient knowledge to make an informed > choice, or just no preference at all, when confronted with such a > choice during installation (I count myself among them!). And that's > usually a sign that it's better just to pick a sensible default, but > allow it to be over-ridden later. > > Indeed - but in that case you have to decide what will 'just work' for most people, most of the time. If 'most' of the development for more-or-less-posix-like systems is being done for Linux, then that makes the default decision quit easy: make it so that the simple configure && make && sudo make install works. And that needs a GNU userland. Sadly. I guess you might get away with root being POSIX if sudo remains in GNU land, but I can't help feeling that's a level of complication that has little upside and a big downside. Surely the answer is that, yes, the default is GNU. And right at the start of the install you can answer one question that will enable such questions to those that are able to understand them. All of the people making a fuss (well, apart from those throwing toys because they want a bigger say in the process) know how to answer that question, and what they want. But they probably didn't need dwaf-caiman anyway, did they? I do believe that having the user admin GUI wizard enable simple switching is also valuable - but the default it will use needs to be chosen somewhere too. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [indiana-discuss] I'm sorry, but I just don't get it
Jim Grisanzio wrote: > itself thrives. We started this project four years ago to build a > developer community. That was the primary goal from which multiple > objectives would grow. In fact, the notion of building a developer > community was part of virtually every meeting I attended even a year > before we launched. Also, we always knew we would eventually grow to > have multiple layers to the community, not simply kernel developers. > > Surely, having a kernel developer community is the least of Sun's actual problems. Sun has developers and having most development done in the context of a funded and managed environment is very valuable. What is needed most of all is a *user* community that extends beyond those of us who work in large corporates who have medium and large Sun servers. From that perspective, having a community that contributes to an involving user-land has to be the primary focus, and I suspect that the decisions to call this 'OpenSolaris' and to make it much more receptive to people familiar with packaging for Linux is right on the money. Please, however, don't ignore the BSD community. In particular, I would encourage anyone interested in (Open)Solaris to download PC-BSD and look at the user experience there with the installation and PBIs. Presumably, it would also be feasible to offer a BSD userland, not just a GNU one. Perhaps the installer can allow a choice of GNU, BSD and SysV (or de-jure UNIX or hawever you want to characterise it). James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Re: Re: Re: Re: recommend the opensource anti-virus for opensolaris?
Alan Coopersmith wrote: And this isn't because Unix viruses/worms are impossible or anything. It's only been a few months since the telnetd worm struck Solaris machines - however, unlike Windows, all known worms for Unix exploit obvious bugs in Unix, so the response from Unix vendors is to fix the bug to stop the worm, instead of trying to scan for and stop the exploit code. It's far simpler and more productive to fix telnetd than to try to recognize signatures of every different bit of code that may try to exploit it. That seems manifestly unfair. Windows (and even IIS) isn't usually what gets attacked - the attacks often work on user level components that expose automatic scripting interfaces. Which are handy for some types of extended user applications -- and OpenOffice, Evolution, Firefox and others have many similar vulnerabilities. What *has* been the case is that Windows users have had too many privileges. You may regard telnetd as 'part of UNIX' and play with boundaries about what is in and out of Windows, but that seems a pointless play on words to me. I've run telnetd on Win32 after all. *Any* listening socket or UDP receiver is a potential attack vector, on any system, and such complacency seems misplaced to me. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Multi-boot recommendations
I just added a second disk to my development workstation, which has Vista all over a SATA disk. I've added a 750gig SATA drive and would like to multi-boot Open Solaris and Linux - and ideally FreeBSD. Is there an order I need to do things in? I have the OpenSolaris 'Starter Kit' and also the Sol10 6/06 DVDs. I last tried the 6/06 setup in a VMWare system, but now I'm looking for something I can use to test SMP performance of some C++ code. I'm tempted to download the latest Bellinix or Nexenta and start from there, but I'm slightly concerned about keeping up to date with kernels and libc if I stray too far from a vanilla install. Is there anything to be gained or lost one way or another? The system has an Asrock board with integrated Intel 845G adapter, so I'd like something that's happy with that at 1920*1200. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Re: Open Solaris Distributions
But I'd like to ask: *why* do you - all - maintain different distributions? It is obvious why I created a distribution: It was done in order to create a distribution as none did exist before. Why did other people create _different_ distributions? Jörg Well, you answered a question I didn't ask. Joerg - we have to respect you for forging ahead as you did. But why do you continue to want to have your own? Why did the others start theirs? Well, they all had different objectives. I don't think you can begrudge a group of guys working together in India - the geographical location, language issues, and proximity of a team is an issue in itself. Martin's focussing on SPARC. Nexenta has a focus that is clearly at odds with your standards focus - would you have welcomed a team who said they wanted to use apr and appe Ubuntu? The question isn't 'why then?' but 'what now?'. The zones facility wasn't there then, nor Xen. The soft switch for personalities for nexenta wasn't there then. It should now be easier to have a common core and support all the extant userland flavours, shouldn't it? Isn't the battle to avoid fragmentation and make sure that the core system is improved in the general sense and these different flavours can be supported easily - and referably all at the same time? ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Open Solaris Distributions
??? How is that different from chosing (say) Red Hat over Ubuntu, or at an even more macroscopic level, UNIX over Windoze Its not different - except that those examples all have a lot more traction than OpenSolaris. It seems to me that the current state of OpenSolaris is such that it has enough 'major' issues that a more concerted effort to pull together rather than seperately would make a very big impact. I think pointless Not Invented Here duplication is a shame - I'd prefer more like the satellite projects around FreeBSD - there is (to my eyes) less duplication of core components. Like PicoBSD, all the way to PC-BSD. And I think: Diversity unavoidable. Is simply ducking the question. I mean - have the ego trip if you want, I guess, but is that really what you want to be remembered for? If Solaris is going to be More Like Linux, then this is to me a very negative aspect of Linux as an ecosystem: wasteful, unfocussed, directionless duplication. And mostly in the name of vanity and egotism. I'm not against the diversity - but why duplicate the bits that are not your USP differentiators? James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Open Solaris Distributions
Guys, It really pains me to see the sort of finger pointing going on here about the various existing distributions and 'Indiana', and in particular I'm concerned at the prospect of alienation between Joerg and 'the rest' because - as an outsider - I do perceive that he did a lot, when Open Solaris was very young. I can understand that when one works on something for a long time, then to decide that its been eclipsed by others that came after can be hard. But I'd like to ask: *why* do you - all - maintain different distributions? All of these distributions seem to have started with a different focus and prioritisation. And so far as I can tell, all have been successful. And yet as an end user, I have to choose how to align *my* priorities with just one these systems in practice. And sometimes that's a shame. It strikes me that there is a lot of hassle and expense in maintaining and distributing a full distribution. You all get to work on what interests you - but to deploy it you have to pull in everything else too. Would it not be possible to work together more? Isn't the pie so large that you could not divide it up and still have big pools to swim in? (Whoa! Mixed Metaphore error!) Joerg clearly has great skill in hardware interfacing on X86 - and in the POSIX standard user space. Can't he drive that? Martin, similarly, has done much for SPARC and Xorg - can't he drive that? Moinak's team clearly has expertise in LiveCD and the 'friendly' Xfce environment - and a 'noob friendly' userspace - can't they drive that? Erast and Alex have a clear (and valuable) alternate userland and packaging system. And I haven't even considered Dennis, and the pkgsource people. And while we have GNU-ish stuff, what about BSD? These areas seem to me somewhat orthogonal. I can understand why a commercial entity might want to control a distribution that it sells support for - and for which it needs to control branding. But is that what drives YOU? Can we have peace and love? Well, not love - but trust and cooperation. The history of Linux small distributions has been one of towering egos hammering away resisting all the Not Invented Here until they just burn out and the bigger distros pick and choose until the benefits are gone. Can we put away the egos and learn from any of it? If Sun's proposal was to fund hardware and networking resources - and travel and accomodation for meetings etc - and a project manager (not boss, a facilitator) - and cede control to you all, in a way that made sense, would that be enough? Would you be able to cherry pick between you - and divide up the responsibilities? Why DO you have your own distributions? James (No, I'm not really that naive. Yes, this is a windup in some sense. But not entirely.) ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
RE: [osol-discuss] What is OpenSolaris success?
(Apologies for the delay.) >If the amateur masses with a small number of ineffecitvely >managed full time engineers can produce a product that >has been able to threaten a well engineered product such >as Solaris, how would you propose responding to that? Hmm - what innovation? Actually delivering stuff 'free' on such a scale is innovative, but pretty much all of what's in the Linux kernel at the functional level has been delivered by Solaris, AIX, and major ISVs before. Or BSD. Or Windows. There is continuous re-engineering, but its not clear to me that this is innovation, rather than reaction to scalability barriers inherent in earlier 'design decisions'. How would I respond? By managing resources effectively? Look at the duplication and the number of herculean efforts that never get merged. Its shameful. I wouldn't consider contributing to that for pleasure. Good thing most people contributing to the kernel (in non-trivial ways like USB and PCI signatures) are paid for it. My concern in making Solaris development 'linux-like' will give up a valuable management role. Shuttleworth has illustrated how much value there is in strong mangement and leadership, and my concern is that by trying to join the 'developer lead' ranks of 'just show me the code' cowboys Solaris will lose. If you must, then at least: - 'show me the customer requirement' - 'show me the technical design' - 'show me the code' - 'show me the tests and instrumentation' (Oooh look, the c-word. And not 'code') Strong management and leadership is important. And that needs empowerment, authority, and an ability to wield resources. It doesn't imply that volunteer or external resources aren't used, but they have to be reliable. >Especially if a large portion of the market (ie the level >req'd to support bringing Solaris to you) decides that >near enough is good enough? Is that really what you want Solaris to be? Near enough? Don't give up on Solaris' strengths. >If the amateur masses can drive innovation and pursue >bringing more features to OpenSolaris, isn't that something >to be pursued and capitalised on, especially if it acts >as a gap filler? Gap filler? Is that it - just a catchup game and then something else? I think its clearly necessary to distinguish what the development processes are for: - actual kernel and libc core - packaging into specialised distributions - userland stuff Linux development in the first of these has tended towards professional now, by companies with vested interests. Some of the companies doing this are Sun's competitors and won't contribute to Solaris on any basis because of brand association. So don't expect real help here except from people doing specialist consultancy who can perhaps help with things like the bootstrap on Intel Mac - most of the heavy lifting will have to be done IN Sun, or perhaps sponsored by Sun. Don't kid yourself that any amount of openness is going to change that. The packaging is largely an issue of: - build hosting - documentation (Its not clear, for example, how to make a PXE boot client that then uses iSCSI to access swap and private workspace: Sun *has* a diskless boot system but its not what I need. It should be fixable. I can make Ubuntu do that - but the swap will deadlock and what you get is Linux engineering at the end of the day) The userland stuff should come from userland emulation of Linux over and above POSIX to make the masses of badly engineered stuff work more easily. If you could get Solaris to install easier on ordinary Joe's hardware and give it away on news-stands, its a problem that will (slowly, perhaps) go away anyway. And university researchers? They do their own thing. By and large, research has a goal of proof of concept only. Its NOT 'engineering'. Its 'R', not 'D'. Don't overestimate the value of 'R' if you can't follow up with real 'D'. The 'R' _is_ important - but not to a product. >>So? Unless they manage a datacentre or otherwise buy >>or specify, they're just Joe Public. Do you really *need* >>the ones who have that responsibility and flunk it to choose >>based on years-ago amateur hacking relationships? >> > >Wow. You do realise that this is an incredibly over >generalised statement and that it insults a lot of people >you've likely never met, never mind understood what >they've done or do? What's the Solaris user base? Who buys Sun kit? How do you expect that to change? You can't afford to lose your existing base. Make the omelete, break the eggs. >I don't know any 'freedom' zealots inside Sun but I'm >sure they're out there somewhere. The point of that >is to say that just because someone is an open source >developer doesn't mean they are a 'freedom' zealot. I didn't say that. But Linux development is full of it, and you seem to want to ape that. Which throws away a valuable differentiator. >A lot of people I know would be quite happy with binary >blobs, so long as they could be used on th
RE: [osol-discuss] What is OpenSolaris success?
>And they start to find it hard to make time to spend time on >anything that isn't work - ie open source projects. I've seen That *was* my point. >this happen on a number of occasions. And contrary to what >you're suggesting, they don't turn up an opensolaris, sometime >later (or at least none have, yet.) I think you misunderstood. I didn't think I was that unclear! *I* think the Linux 'model' is flawed. Take away the amateurs, and you have a small number of full time engineers, who are in Linux' case diluted by ineffective management and lack of shared purpose. Sun is in a *good* position. Why pander to the amateur masses? They won't make Solaris better through code contribution. It *would* help to make Open Solaris buildable and maleable by ISVs for particular application scenarios, particularly embedded server devices, but once again that's a professional contribution. >You're missing the point of the importance of students and >what happens in the open source world. I suppose the closest >I can come up with as a model is how people pick and choose >their political tendencies and once chosen, they almost never >change. I disagree. I have a different viewpoint now than I did then. >If the community is going to grow then we need to grow it from >the roots by planting seeds in the right places where they will >grow. I don't *care* if it grows. I just care if OS is a good platform for option pricing and trade management. I want Sun to (continue to) do a good job. I don't see community as germane to that. Openness, transparency, early access, and free as in beer - sure. Community? Don't care. I'm a (would be, to some extent) user of Open Solaris. I don't care about developer love-ins. If you want a great big commune then great, but please differentiate by keeping focuss on users. The comments of the election candidates over binary compatibility *are* very reassuring. >have some sense of community? (Chances are they never stopped >reading the email, etc, they just stopped being really active.) So? Unless they manage a datacentre or otherwise buy or specify, they're just Joe Public. Do you really *need* the ones who have that responsibility and flunk it to choose based on years-ago amateur hacking relationships? >people but I can't see that happening. Linux's appeal is more >than just the GPL and failing to understand that will lead to >failed attempts to copy it. I don't want you to copy it. That's the point. The whole developer community focus is just that - focussed on developers. Its obvious that the users' needs are NOT foremost - look at the state of the documentation of the majority of projects. Precious few even have a clear statement of the problem they are trying to solve. >I don't know any open source developers who would consider >the needs of users as being unimportant. The point of my >comment (which you seen to have missed) isn't that developers >are more important than users but that open source projects >dont need managers, directors, VPs, etc, to drive them and I disagree. Someone needs to tell the 'freedom' zealots that users generally want 'Just Works' and don't care if the code was compiled by nVidia and shipped as a blob. Now I appreciate that Theo makes a good point - for a particular scenario of an auditable security device. But I'm quite happy to use XP for my convenience and a POSIX system where that makes sense. >developers than its users, you'd be in for a rough ride. Really? I've been asked to find the bug and send a patch rather than been given any constructive help many times. The point is made as 'we are all volunteers'. Well, I just don't care - if you are part of the 'we make the OS' crowd then saying that to the 'we use the OS' crowd doesn't cut it. The compensation arrangements inside the 'we make the OS' crowd is really not my business, and I don't see why my expectation of an OS product should be clouded by it. >What you're confusing here is what OpenSolaris is with what >Solaris is. That is the extent to which Open Solaris is of any interest, yes. Solaris, but more malleable and open (as opposed to 'opaque'). For playtime I have Ubuntu, FreeBSD, and NetBSD. For work, I need something better. >it by normal means is fraught and that it needs to live for a >lot longer yet before anyone can even attempt to judge it. On this, I agree completely. James -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.9/719 - Release Date: 12/03/2007 08:41 ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
RE: [osol-discuss] What is OpenSolaris success?
>complete with some suggestions for the future and what I think >it will take for it to really succeed - mostly time. Darren, You suggest that it is a 'problem' that contributing to Open Solaris is a contribution to Sun. As a user, that is precisely one of Open Solaris' strengths. And if that means that the sort of person who has a problem with that doesn't contribute, then I'm personally overjoyed. Good riddance. Let them work in their communes. When they grow up and get lives (and significant others, and kids, and all those things that make adults' time precious) and careers that depend on technology they'll start to understand. Its not as if the stuff is hoarded by Sun. Its not as if Red Hat don't benefit from contributions to Linux. You seem keen on students. I doubt anything I wrote as a student or newbie grad would survive any sort of review I'd make now, though. Do you really let interns loose on Solaris internals? >management at Sun needs to step back, to the point where managers, >directors, VPs, etc, should have no involvment with the project >as an employee or agent of Sun If you really want Open Solaris to be 'just another' open source OS, then fine - but what you're saying is that developers matter most, and that sucks. Users should matter most, and developers (particularly on open source projects) are notoriously bad at putting 'mere' users needs first. Management MUST give strong guidence to make sure that Open Solaris remains close to 'Sun Solaris' and that 'Sun Solaris' aligns with users' needs, whether or not that alignment means doing uncool and boring things. Which it will. If Sun loses focus and cannot ratin (and to some extent regain) traction in the datacentre then its lost. Its much more important that it succeed at that than that Open Solaris is a 'successful open source project' - who gives a toss about that? I'm not a Sun shareholder - but I have been using Solaris in trading systems for nearly 20 years and I *do not want* to be at Red Hat's mercy, or to lose the API and system stability that means that the same old ISV products still work - and so do my old applications. Sun are in a stong position - there is good datacentre penetration and the product has a good ISV portfolio. It needs desktop 'wow' factor but Sun can also put clean air between itself and the Linux hordes by embracing the very things they reject as not being 'open' enough (like I care as a user!) and being a third power with Microsoft and Apple. I can't for the life of me see why anyone would want to be 'me too' and more like Linux, geek cool or not. But I can see how *users* wanting an alternative to Windows would choose a system that has first class support of every feature of their nVidia and ATI motherboard and GPUs, a vendor provided office suite, and vendor provided development tools. And PostgreSQL. Mr Gumpy -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.8/716 - Release Date: 09/03/2007 18:53 ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
RE: [Fwd: Re: [osol-discuss] GPLv3?]
Stephen Harpster wrote: > There are a lot of GPL bigots out there. And you *want* to appeal to them? Seriously - why? Are these bigots running datacentres? Are they running startups that have a hope in hell of actually making money - as opposed to generating PR and then just chewing their VC funds? > a dual-license would make it easier to have OpenSolaris use > that larger body of work. How? The larger body of work won't be dual-licensed, and won't have the necessary extra clauses that would allow combination. And you can run GPL apps on a non-GPL OS anyway. Please, before we start getting to much in-love with the idea of community as an end in itself, can we discuss what - and who - Solaris is *for*. I'm personally tired of 'open source communities' telling me to help fix their OS. It happens with BSD as well as Linux, perhaps more so. And its not condusive to wanting to be an OS user and develop my own apps. Don't join them - please! Be the open source OS that has a clear (and clearly explained) focus on *users* - and if that means that the would-be community members who want to own it, and want to have the same-old 'fix it yourself' attitude, are effectively excluded, then that looks like a Very Good Thing to me. Please let's start with: who are the most important members of the community. Is it: a) free-as-in-freedom campaigners b) coders c) users I put it to you that to deliver to users needs strong leadership (NOT some kind of community democracy) and strongly empowered and directed resource application. This is much more important than whether or not amateurs get the hump over the controls in the process. To my mind, its extremely important that Sun maintain strong control, because Sun *is* experienced with servicing a user base - and I doubt I'm the only user that wants it to continue (and improve). James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
RE: [osol-discuss] GPLv3?
>I think that we ("we" being all of you) should be asking >ourselves what we think about GPLv3. What would it >mean to the community if we dual-licensed? It's now a >possibility that we could attach an "assembly exception" >to the GPLv3 which would let us mix GPL and CDDL code. >This could open up a world of possibilities. What benefit would it bring? Code can't be shared FROM linux with a dual license or an exception - or indeed GPLv3. There might be a benefit from making the base libc dual licensed so that the fun and games associated with the Nexenta distribution are not repeated, but I don't see what extra body of code will become available to OpenSolaris in terms of device drivers that it needs or anything else like that. I'd really rather Solaris DIDN'T become controlled by the 'community'. We have Linux, and *BSD with community driven development and they both have problem fixing the final wrinkles that are not sexy or interesting - but which customers (rather than developers) ARE interested in. I'm fairly sure we're capable of differentiating between cases where ALL the source is open for scrutiny and security audit (eg edge servers) and cases where we want the maximum out of the hardware - particularly with workstations. Solaris can be different primarily by being controlled by a company which can have customer focus and engineering clout and project management process to back up delivery. Sure, contributions that scratch itches can be accepted, but at the end of teh day we need a system with strong binary interfaces that is friendly to binary vendors, be it network or graphic or RAID drivers - and by doing so Solaris can be more customer focussed and distance itself from all the usual bullshit. I really do want an alternative to Windows and MacOS/X. I don't need A.N.Other free-as-in-freedom OS. And ideally I'd like Sun to tune and sell hardware that flies with Solaris, and a second vendor (Novell would have been ideal, too bad!) to provide a software only angle on white box and Dell/HP etc. James ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Re: The FSF Website says "not to use the CDDL"
> I am wondering what value the CDDL has anymore. Why? Why should anyone care whether CDDL code is compatible with GPL, if there is enough CDDL code to stand on its own? What I think is unfortunate is: > module covered by the GPL and a module covered by the CDDL cannot legally be > linked together This is surely wrong, because I can link any GPL and proprietary code I like so long as I don't 'distribute', which since 'distribute' does not seem to apply to intranet use, is not a problem for most users. And where the CDDL code is part of the OS distribution, its even less of a problem. Surely the answer is to say 'Yup, its not compatible, its better' and get on with life. Personally I hope that Sun does not go too far: I [b]want[/b] a Solaris that is [b]controlled[/b] (with absolute veto of what's in and what's out, ultimately) by a [b]commercial[/b] supplier with a strong customer focus that prioritised customers over the great unwashed masses. With it, we - out here - have some hope of avoiding bikesheds and politics, and getting what we need. Free (as in cheap) to use and working on my hardware is much more important than the freedoms the FSF focusses on. There's a space for people who want to think like that, but please do provide a robust alternative, and don't toady to them. This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Re: Re: No google code hosting for CDDL/OpenSolaris
I don't think closed forks are necessarily something to lose too much sleep over. Most code that's written is never used outside of the organisation its written in and GPL doesn't enforce it - contribution of changes back is a self-interest thing that's motivated by laziness with respect to maintaining the fork. Even when there is a commercial justification to maintaining a fork in a product that's sold, the product lifecycle means that eventually the justification will likely be diminished, and then a release will take place - and to me that seems fine, because the enhancement is released, and the investor got a return, and everyone is happy. Its hard to justify major R&D into GPL software - look at the comparative R&D budgets of the Linux vendors compared to Sun. I'd rather have an environment where commercial investment can be justified and made, and trade that for a period of jealousy while investors make their returns. > boost the BSDs and MacOS Now - just what is the point of the MacOS core now? I can't imagine why Apple ported DTrace to their crufty multi-layered kernel. I would have thought it cleaner all round to pop the userspace onto Solaris - they'd find their stuff an easier sell to longtime UNIX shops if they did, I'm sure, and their existing media and publications userbase wouldn't care at all. This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Query re: ZFS
I'm excited by ZFS software raid, particularly with the dual parity facility. Now, I use Solaris 8 during my day job (I'm a developer), and I have a TwinHead SParc5 clone in the cupboard, and I routinely install Linux (and sometimes *BSD), but I'm by no means a Solaris admin. So some questions: 1) Is Solaris getting better at handling controller and disk quirks so that SATA2 disks will generally write-through as they are told? 2) If I configure disks to write through, but normally have write-back in the OS disk cache, and I create a loopback device on a file, and I use the device in direct IO mode and force flushes to it, do those flushes write through to the platter? (Say I'm running Sybase in a Linux container and I'm very old-fashioned in the way that I install Sybase) I'm looking at a picture of a new AMD nForce motherboard, and it has 6-off 300meg SATA ports. And I'm thinking this looks like a fine home server platform, especially if the iSCSI target is easy to administer and reasonably performant on plain gig-e. This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org