Re: [osol-discuss] The last of the 5000?
Since it seams the website team is effectively gone (see Elaine and Derek's
goodbye messages elsewhere), it may be hard to pull something like this
together. Instead, how about a flickr tag ("#opensolaris"?) so anyone can
post their own pictures and mashup an image stream?
-John
On Fri, Aug 20, 2010 at 3:36 PM, Ian Collins wrote:
> Would it be possible setup a page on the OpenSolaris site where we could
> upload our photos, wearing our faded and possibly too small first 5000
> T-shirts?
>
> Just a daft idea for the weekend,
>
> --
> Ian.
>
> ___
> opensolaris-discuss mailing list
> opensolaris-discuss@opensolaris.org
>
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Re: [osol-discuss] When's source included?
If you build a distro and make the resulting binaries (an ISO image...)
available on a website, all you need to do is provide a compressed tar file
(or an ISO image of one) of the "make clean"d source tree you used - there
is no reason to make the source tree part of your binary ISO image
Since the source tree tar file will be large (and not very interesting as
time marches on), a better way is to provide a pointer to the same IllumOS
or Oracle or ... versioned mercurial repository that *you* started with and
make a gzipped file of just your "diffs" ('hg export' or 'hg merge' as
appropriate). An interested developer could then grab the common/unchanged
sources from the same place you did, apply your changes and build their own
copy.
The goal is to make it so that someone else could build on your changes in
the same way you built on the work of others.
-John
On Sun, Aug 22, 2010 at 3:52 PM, Andrew Greimann wrote:
> Or, could it be include all source and work with/on binaries, leaving
> 90-95% of the system unmodified?
> ...
> Please shed more light on what it takes to make a distro legal. I'd
> appreciate it. :)
>
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Re: [osol-discuss] place yer bets on who leaves next...
2010/8/20 Matthias Pfützner > Let me add some numbers... > ... > that a high percentage might be working on Solaris might provide us with an > approach to a number close to 1000 engineers... > Of your 1000 engineers, maybe 100 were the senior leaders in innovation, vision, drive and ability. At Sun, the makeup of that "club" was exceedingly dynamic, to be sure, but it was a meritocracy - if you were *good* and had job/product/whatever performance to prove it, membership was open; nobody had to leave to make room for you. >From what I have seen (and I have no visibility into the current numbers or membership), a significant number of the distinguished engineers and fellows that were there when Oracle took over have left. 30%? 50%? More? I don't know either, but the Names that are making the headlines all come from that small club... IM(ns)HO, losing that many top performing engineers to the competition will do more to harm Oracle in both the short and long run than anything that might conceivably happen due to "premature product and feature exposure" due to open source community involvement. Nobody really cares if a company lays off a bunch of its low level staff, but losing half of the technical leadership of a technical company is a disaster. Oracle may have bought the trademarks and rights to the code, but the real value of an acquisition is in the minds of those who produced the products in the first place - long term engineering excellence isn't a commodity that can be cheaply purchased or easily duplicated. Don't forget that the easiest way to make the books look better in the short term is to get rid of all those expensive engineers - you will immediately see a 10%-15% rise in profitability because you no longer have to pay the cost of development. Of course, after 24 to 36 months of coasting, you will be dead, but given the Street's myopic focus and short term memory, who cares? Just buy some other company and start everything over again... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Oracle shuts down open source test servers
On Thu, Jul 29, 2010 at 9:31 PM, Edward Ned Harvey wrote: > > > Edward, STOP! be nice! > > I'm not being un-nice or anything. I'm completely serious. If Ken really > believes what he says, "run away from oracle products," then why > participate > in this discussion forum? > Sorry, Edward, but the OpenSolaris Community *IS NOT* an Oracle product. Our community was formed around the idea of an open Solaris development partnership with Sun, and the people here shouldn't have to apologize for being concerned about the future prospects of that partnership now that Oracle has taken over. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] SchilliX-0.7.0 ready for testing
[I *really* doubt that Alan misunderstands SVr4 packaging in any way whatsoever...] You are right, from a purely technological authorship perspective (stuff written 20-25 years ago by AT&T -vs Sun), patching, network aware installation, package repositories, automatic dependency computation, bundling and co-installation and all the other things that transform a simple file archiving system into a usable and useful system installation and feature provisioning system are not really part of the SVr4 packaging system - which is rather the point, isn't it? Without those "additional features", pkgadd and friends are pretty limited; an uncharitable person might even say a packaging system without the ability to upgrade/patch a system is rather pointless. -John On Mon, Jul 26, 2010 at 1:03 AM, Joerg Schilling < joerg.schill...@fokus.fraunhofer.de> wrote: > > You missunderstand the SVr4 packaging system. It does not include the > concept > of patching. "patchadd" is a script on top of the packaging system. It you > like > to get rid of something "evil", just throw away this script > > ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] SchilliX-0.7.0 ready for testing
On Sat, Jul 24, 2010 at 11:50 PM, Stefan Parvu wrote: > IPS tries to be a network pkg management system. It needs to be simple > ... list of good ideas ... These are all things that could easily be addressed, and aren't per-se reasons to abandon IPS. If it was important, these things could be done. With the new on-disk format, I'd assume it would be easier to maintain a mirror using the existing mirroring tools... As Machiavelli said, change is opposed by those with a vested interest in the status quo... > What comparative analysis have you done Last time I looked, probably 5 years ago, and if I recall correctly, the basic work flow in the SVr4 package tools was something like start pkgadd, pkgrm, pkginfo, ... process read /var/sadm/install/contents into memory install/remove/whatever a package read /var/sadm/install/contents and merge/modify the lines into a new /var/sadm/install/contents file if the world was modified exit process for every single package installed. The thrashing over the /var/sadm/install/contents file was more than huge, and is one reason why patch management in zones sucked rocks so badly (tens of hours to days to upgrade a system with many zones...). There were some simple comp-sci style modifications that could/did speed things up by orders of magnitude, but they required replacing the /var/sadm/install/contents text file with some sort of structured object (database, btree, whatever) that was incompatible with way too many things that had come to depend on the project private contents file. There were proposals to cache the file in memory (ala nscd), to make it a special kernel driver (ala /etc/mnttab) and others, none of which (IIRC) made it past the proposal/prototype stage. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] SchilliX-0.7.0 ready for testing
Ken, et al, > IPS was shoved down the community's throat in a heavy handed and decidedly > not FOSS manner. What shoving? It seems to me that those of you who are wailing and moaning that you want a version of OpenSolaris with SVr4 packages are conspicuous by the absence of an OpenSolaris derived distro based on SVr4 packages. Even Joerg has put sweat behind his dislike of IPS and produced Schillix with his own build and packaging set. Hmmm, does that mean you will now claim Joerg is shoving his schilly-tools vision down the community's throat in a heavy handed and decidedly not FOSS manner because he didn't consult with you about his decision? FOSS isn't a democracy, and it is certainly isn't a republic - it is a despotic form of self-selected government where voting rights are earned by the measurable contributions one makes; the one(s) who put in the most effort make most, if not all of the rules. If you don't like how a community is run, fork the code and start your own, with you as king (or queen). Even if Oracle isn't, and Sun wasn't, playing in our sandbox, our community was formed around the idea of governance by a meritocracy: "Code wins". While I'm one of those who wished that the decision to put IPS into OpenSolaris-the-Sun-distro was something the community as a whole had some say in, rather than being a decision made behind closed doors, it really is no different from Joerg's choosing to use his schilly tools to build his distro or Nexenta's to use Debian/RPM - both those decisions were made by the people putting their sweat into making a distro, and those of us who were only backseat driving on the mailing lists got deservedly ignored. Myself, I am coming to understand - and respect - IPS more and more as I use it. It is hard to compare IPS to SVr4 packages because it solves a larger set of problems. For all intents and purposes, SVr4 packages are not much more than file archives with meta data and conventions; worse, most of those conventions are so poorly understood that the features they try to provide (cross package dependencies, upgrade paths, multi-platform/architecture support, ...) are effectively unusable AND unfixable. Speaking from experience with Sun's Release Engineering package audit tools, if you were to create an audit tool for packages that validated ALL the SVr4 and Solaris package requirements and conventions, and ran it against every existing Solaris SVr4 based package out there from Sun and Blastwave and all the other vendors, enthusiasts and suppliers, I would wager that close to 100% of them would fail the test - even if you could somehow agree on what all those requirements and conventions were. In other words, after 20-some years, I believe it is time to admit that the SVr4 package ecosystem can not really be fixed or evolved. In much the same way that SMF was both better than, and a different beast from rc3.d/* and ZFS was better than and more than ufs, IPS is better and different from SVr4 packages. IPS is an ecosystem that SVr4 packages never attempted to be, and (IMHO) never could become. Even so, at least in OS2009/06, SVr4 packages and IPS work together on the same system. You can install Blastwave packages, and you can grab updates to your OS from IPS repos, and both work. Where's the gripe? If you don't want to produce IPS style stuff, don't. If, despite your product's lack of the value-added IPS features I have come to expect, I choose to use it, I can still invoke pkgadd, install it and use it. Of course, I can also choose not to, and shop elsewhere. Of course, if your argument is that nobody listens to your gripes on mailing lists when you don't contribute tangibly towards a solution, then I'd have to agree - those doing the development *shouldn't* have to listen to those who simply chatter; why do you suppose so few developers hang out on OpenSolaris-discuss? We put up with Joerg because he votes with his code; what are you voting with? :-) -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Community distro mailing list
A reminder that we *do* have a distribution community here, with a distribution-discuss alias that is a good place to hold "community distro" conversations... http://mail.opensolaris.org/mailman/listinfo/distribution-discuss -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Community distro
On Tue, Jul 20, 2010 at 1:59 AM, Joerg Schilling wrote: > Bart Smaalders wrote: >> . Joerg (et.al.) This all sounds like crying over spilt milk, as they say. Your arguments seem to be of the form "THEY didn't need to move away from SVr4 packages; instead THEY could have..." rather than "here's what SVr4 packages do that IPS doesn't". That is, your gripe seems to be that you weren't consulted, that your ideas weren't incorporated, the project went in directions you wouldn't have gone, and that you do not know what problems it is solving - none of which are terribly convincing arguments in an open source community... The fact of the matter is that nobody *DID* step up and fix/redo/replace/improve the SVr4 packaging stuff. Instead, a group of somebodies went off and built a better-for-them mousetrap to replace SVr4 packaging. While even I have quibbles about *HOW* it was done, the fact remains that *WHAT* they did is very good; it is a clear and (to me) overwhelmingly better system that SVr4 packaging was - or ever could be, given SVr4 packaging's inherent flaws and malformed, in-the-field package metadata. In *theory*, with effort that never materialized, SVr4 packaging coulda, woulda, shoulda... It didn't. But. as I said before, none of this really matters if we don't even have a community distro that can be built and installed! If you can't even boot the OS, can't even get to a prompt, then who the hell cares about what kind of packaging system is there *THAT YOU CAN'T EVEN USE*? Let's get the basics working first, and be as open as we can towards others. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Community distro
Shawn, Joerg, Richard, et.al, IMO this is not the time or place to argue about this level of detail - like it or not, IPS is part of what we have and isn't going to magically go away anytime soon. Neither are SVr4 packages. What we don't have is an up to date buildable, downloadable, installable *community distro* that uses EITHER of those two technologies. It seems to me that our energies are better spent getting to the point where we have the above before we balkanize and fragment over whether or not it should have some particular feature :-) Crawl before we walk and Walk before we run. It seems rather pointless if we don't even try crawling or walking, but instead sit in front of our computers and bash each other about how things would be better if only we could run in different directions... -John On Mon, Jul 19, 2010 at 10:16 AM, Shawn Walker wrote: > On 07/19/10 02:50 AM, joerg.schill...@fokus.fraunhofer.de wrote: > >> "Richard L. Hamilton" wrote: >> > ... > > ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Top Solaris developer flees Oracle- from the reg
On Thu, Jul 15, 2010 at 9:21 AM, Bart Smaalders wrote: > On 07/15/10 02:27, Sean . wrote: > >> Was just reading this on the Register myself. This is not good news >> for the future of Solaris at all never mind OpenSolaris. >> > > Many of us very much like Greg... he was a VP here, not a developer, > and he has a great opportunity that he wanted to pursue. We wish > him luck in his new role @Cisco, and we'll continue building a better > Solaris under different (but also familiar) management. > > +1 This seems to simply be the usual "every 6 or 12 month reorg" that happens everywhere in the valley, and not an Event That Will Shake The World. Looking behind the scenes, Greg was never a Top Solaris Developer in the same way Bart or JBeck is; it probably wouldn't be far from the mark to assume Greg simply got tagged with the VP slot when the music stopped and everyone else left. Calling the VP of Solaris "the top Solaris developer" is a fallback to Sun's perception that it was a technology company, and thus its development team managers were technical people. Since many of them *did* come up thru the hardware and/or software developer ranks, this wasn't much of a stretch; many of the directors and VPs I worked with were brilliant technical people - though, to be honest, there were some techno-idiots there as well :-) -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Yet another warning about behaviour on this list
On Thu, Jul 15, 2010 at 8:33 AM, Alan Burlison
wrote:
> Actually I think the TOU is pretty clear. The particular messages that
are
> of concern contain personal abuse directed at other list subscribers.
The TOU contains 21 clauses over several pages that include umbrella
statements like "Host reserves the right to ... take any action Host deems
appropriate". You keep referring to continuing personal abuse that, other
than one outburst several days ago, I don't see, and you are threatening to
unilaterally and without warning shut down our community discussion list(s)
because of it. Obviously your skin is thinner in places where mine is
thick, and your tolerance for heated discussion in times of stress is set to
a different level than mine.
My uncertainty is how you are going to interpret these terms, and which one
of these is being violated by the "personal abuse" and "badmouthing" you
claim to be seeing?
3) You are prohibited from posting or transmitting to or from this Website
any unlawful, threatening, libelous, defamatory, obscene, pornographic, or
profane material, any software virus, worm
This seems obvious and not open to interpretation, but then we get to vague
city:
or other material of a disruptive or destructive nature,
Hmmm, what do you consider to be disruptive? A post complaining about
Oracle? A post complaining about someone complaining about Oracle? A post
complaining about the complaints? Maybe you don't even care about this
clause... Is my calling a fellow list member a "fanboi" destructive - or
factual? Where in the TOI does it say I can't badmouth Joerg by calling him
a Sumo Fat Boy?*
(c) forge headers or otherwise manipulate identifiers (including URLs) in
order to disguise the origin of any Service or Content transmitted via this
Website; (d) misrepresent your affiliation with a person or entity;
Hmmm - does this mean that if I post under a pseudonym ("Anonymous
Coward"...) I'll get booted? What if I use a screen name that disguises my
true identity (dogbreath, MilkmanDan, sickness, nrubsig, to name a few on
IRC...) ?
(e) disrupt the normal flow of dialogue or otherwise act in a manner that
negatively affects Users' ability to use this Website;
Is this rat-hole about list etiquette and management disrupting the normal
flow of trolling and fanboi-ism on this alias? Will it get me booted?
Rather than opening this can of uncertainty worms, IMHO it would be better
to respond publicly to those who violate the TOU's norms. A simple "R)eply
to all" by you, as moderator, saying "This is the moderator. This message
crossed the line again. You were warned privately on xxx July 2010, so you
have been placed on moderated posting status for XXX period of time". This
would do a few positive things for the community: It would make any
discipline transparent, and thus accountable, and it would provide specific
reminders and expectation setting to the rest of us as we strive to stay
civil with each other.
In my mind, part 3 of the TOU is not where we should be focusing our
attention. Rather we should take a big dose of clause 7: "ALL SERVICES AND
CONTENT ARE PROVIDED BY HOST ON AN "AS IS", "AS AVAILABLE", AND "WITH ALL
FAULTS" BASIS ONLY. YOU ASSUME ALL RISK OF LOSS OR OTHER HARM FOR THE USE
THEREOF", delete any messages we don't wish to see (bye, Ned...), and stop
expecting someone else to shield us from the real world.
Thank you,
-John
* reference to an OSCON beer-inspired party stunt several years ago :-)
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Yet another warning about behaviour on this list
On Thursday, July 15, 2010, Alan Burlison wrote: > Despite repeated warnings, some people are continuing to badmouth each other > on this list. As explained previously, this is not acceptable. We've been > warning people who have overstepped the mark, from this point on we won't be > doing that, we'll just be immediately closing accounts and ... In order to educate others on where the bar is set in your mind, could you please provide a list of example behavior that you feel are over the top? The TOU is long and short on specifics, and now you are going into "shoot first vigilante mode" .., Pointers to messages would be fine so as to not regurgitate the spew... Thanks, John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [distribution-discuss] Community distro
On Wed, Jul 14, 2010 at 7:07 PM, Paul Gress wrote: >> Great idea. Not what the current OGB, as the OGB, is chartered to do. >> The OGB would really have to quit the current system and reform in >> some other guise, or a new body would have to take this on. > > Maybe this is whats planned if "they cut their heads off". I didn't agree > with that decision, but now I'm starting to think of it as a well planned > decision. The best of all worlds, IMHO, would be for us (the community) to build and maintain a website of our own that hosted a minimalist distro of our own, with IPS repos of our own, using only the master mercurial (and/or whatever) source repos mirrored from within Oracle. Explicitly NOT a fork, and explicitly NOT an Oracle product, it would provide desperately needed direction and coherence for our community. In addition to, but separate from it, there should be an "Alan Cox" style branch where community contributions to the core OS could be evaluated, packaged, committed and maintained. Continuing my fantasy, there would be a set of related IPS repos that contain community generated packages for each of these two distro "styles" (ala the source jucr and/or Blastwave and/or SunFreeWare). Imbibing fully of the Kool-Aid gets me to a point where one could use the Distro Constructor to craft a personalized distro, say with KDE and ZFS or Crosbow, ZFS and no graphics/desktop support at all. At least that's my vision. Whether it is motivating and inspiring to others is a big unknown. One significant hurdle is that not all of the current OGB members feel they can devote the significant quantities of their limited energy, resources and time that such an undertaking would entail - a situation that applies, I'm sure, to many others in our community... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] More negative press: Oracle maintains silence on OpenSolaris' future
The comments *there* are pretty balanced - what a surprise :-) -John On Wed, Jul 14, 2010 at 10:34 AM, Peter Tribble wrote: > > And we just made Slashdot. The fact that it took them 2 days to > notice is slightly worrying. > > ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] OGB Threatens to Shoot Itself In The Head
My last post here on the topic. Come on over to OGB-discuss if you want to seriously work towards a better solution. Giovanni Tirloni wrote: > they are saying "sorry, under the current terms we've no purpose so it's useless to pretend we should hold any official position". Bingo. Many Somebody Elses wrote: > I wasn't on the call, but ... As expected, the uninformed nonparticipants jump in with no real understanding of the issues, having *done* nothing in our community but post email. Talk is cheap, code wins. Fanboi and Oracle cheerleader wrote: > Well, i was under the impression the OGB were already working things out with oracle because of the following email from july 3: >"Hi CHOSUG Members The last official contact Dan Roberts had with the OGB was as part of his IRC chat conversation during the March 2010 election. The quote "is already sharing all the information he has" is literally and factually true - as he has absolutely NO information to share, he has indeed shared it all. He has had off the record, explicitly unofficial contact with several board members in private where he also said pretty much nothing. Makes you feel warm and fuzzy all over, doesn't it? BenR (whom I have enormous personal and professional respect for) wrote: > What exactly do we have to gain or Oracle to loose? All Oracle does > is runs out the clock, the entire OGB resigns, and then the one little bit > of control the community has is gone. What motive, other than a benevolent > act to garner press attention, does Oracle have to comply? We've just > made their job easier. I'm sorry, but this smacks of a collaborationist viewpoint - don't annoy Oracle, they might get mad and do something really bad to the community... Bullshit - the damage was already done on Feb 6 here in the US when all the Sun employees became Oracle employees and were told they could no longer freely discuss open source products under development with the community that was helping create them. I suppose we could do what you suggested on IRC and scale back our meetings to once a quarter: "Hi, we're all here for our 4th quarterly OGB meeting, Uhm, except for anyone from Oracle. Status update: We still can't get our website infrastructure updated, we still don't have a community distro, and still haven't heard from the Man. Any business? nope, but it's time to start another election cycle, hope we can find 7 more suckers to continue this charade for another year..." -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Oracle Solaris: General FAQs: seems from july
+-- On Tue, Jul 13, 2010 at 7:58 AM, Paul Gress wrote: | So I think not a Fork is required, but a "Cloning" is required. This new project | should be called maybe "Desktop Clone-aris" Where what we do is try to | compile the source and recreate the missing closed binaries. +-- Please join the distro community and help make this happen. Joerg, Moinak and the rest of us will welcome you with open keyboards! Like you said, there are a few obstacles in the way of doing a completely open distro right now (closed source I18N, some NFS stuff, ...) but, in general we are mostly there. All it needs are a few more people who would rather code than post emails on mailing lists and forums... +-- | I know there are people with the resources to start this. This new project should be really open and show progress openly and allow easier participation. +-- Not drastic at all, but a way for our community to start taking charge of its own destiny. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [opensolaris-discuss] Were to from here?
+-- On Mon, Jul 12, 2010 at 9:49 PM, Hillel Lubman wrote: | there should be no moral dilemma in forking, if Oracle/Sun is not playing their | role in the partnership with the community. +-- As a matter of principle, I see no problem with forking as long as the stewards of Oracle's community don't use their Oracle-chartered role and authority to advance that agenda. In other words, I (John Plocher, private citizen) could easily champion a community fork, but I (John Plocher, OGB Chair) must not. I was elected to govern the OpenSolaris Community, as chartered by Sun/Oracle, under the terms of a constitution that upheld that charter. Using my position on the OGB to seduce the community away from that charter would be wrong - even if I could justify or rationalize my actions by pointing out other people's missteps. The morally right thing to do, were I inclined to support a forked community, would be to resign from the OGB and then devote my energies towards making the forked community successful. If, as you say, the partnership is indeed dissolved, then I simply need to acknowledge the fact by resigning, at which point I am free to go my own way. If a new community is then formed, I'm absolutely sure its success (or my standing in it) won't be based on something as trivial and meaningless as my having the title "OGB Chair" at the time :-) -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] warning about mail on this list
+-- On Mon, Jul 12, 2010 at 8:59 PM, Fredrich Maney wrote: | +-- I wrote: | | In other words, it isn't Jive's fault, nor is it the Jive user's | | fault - from their perspective, they don't see the problem. It is only you, | | who chose to use a deficient user interface (mail) to follow their | | conversation, who has the problem | +-- | As a user, by choice, of a "deficient user interface" to participate | in these discussions, I am offended by your tone and assertion. +-- Fredrich, I'm sorry that the combination of grammar and logic so easily offends you, but so be it. I intended nothing offensive in my statement that "from their perspective ... it is only [the mail user] who has a problem". From *A JIVE USER'S* perspective, you are simply using the wrong tool for the job. For other jobs, and from other perspectives, mail may well be the right tool. The same would hold true if you tried to make a mail reader gateway for slashdot "conversations", blog post comments or flickr photo tags - the originating medium incorporates context into its user interface design in ways that most mail readers do not. Forcing slashdot or blog or flickr users to perform unnatural text quoting acts just so *you* can use a mail reader to follow their tags or comments would be unthinkable, wouldn't it? If you still don't get it, the real question here is "why do we continue using Jive as the web-based UI for our predominantly mail-based conversation style", and not "how do we retrain Jive users". Bluntly, shame on us for building poor tools, not shame on them for using them the way they are supposed to be used. -John "I'm not illiterate - I know who my parent's are!" Plocher ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [opensolaris-discuss] Were to from here?
On Mon, Jul 12, 2010 at 2:16 PM, Giovanni Tirloni wrote: > Sorry for sounding so pessimist, it feels terrible to see the OpenSolaris > project going through all of this. It feels worse to be on the OGB and still being able to do so little about the problem. Even the obvious alternative (go do a fork...) is a practical impossibility for the OGB - even though two members are distro-creators in their own right, the OGB as the *OpenSolaris Community Governing Board* doesn't have the resources, time or (IMHO) moral right to hijack the community that Sun/Oracle chartered and start up what would effectively be an organization that would compete with Sun/Oracle in developing a Solaris derivitive... (Which is not to say that someone else couldn't, shouldn't or wouldn't...) Sigh. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [opensolaris-discuss] Were to from here?
On Mon, Jul 12, 2010 at 10:52 AM, Matthew Nawrocki < matthew.nawro...@gmail.com> wrote: > So I listened in to most of the call and it seems that, more likely than > not, that the OGB is going to dissolve. My festering question is as > follows... if and when we turn the so-called steering wheel over to Oracle > for the community... what exactly will entail? Does this mean that we lose > control or what? > > The OpenSolaris Community was chartered to foster a development partnership between Sun and a community of operating system developers around the core components of Solaris. Along the way, other partnerships and communities came into being around other parts of the ecosystem (user groups, desktop, sourcejuicer, distros, etc), and were included under the OpenSolaris umbrella. Unfortunately, circumstances and desires changed within Sun/Oracle; their protracted and across the board silence about all things OpenSolaris has made it impossible to continue as an external development community associated with Solaris, as expressed in our Charter. There are two perspectives here, a practical one and a legalese one. As Alan said, from a practical POV, nothing changes: Oracle isn't playing with the community *now*. Without a partner for the OGB to dance with, the OGB can't actually do anything. While we have a bully pulpit, we can't actually force Oracle's employees to update our community website to support our constitution, to be active and vocal community members or to even provide schedule and content information concerning the distro that bears our name. Ben and others are very concerned that this action *might* provoke Oracle to shut down the community completely and cut off even the meager flow of code that trickles out today. >From a legalese POV, dissolving the OGB triggers a community reset back to Oracle: *1.3.5 Board Dissolution* If the Board membership falls below three, more than 14 months pass between Annual Elections or if the Electorate passes a motion of no confidence in the Board as a whole, then custody of the OpenSolaris community will temporarily revert to Sun Microsystems, which shall, at its sole discretion, appoint to the OGB additional natural persons sufficient in number to increase the OGB's membership to three. Those appointees shall serve only until a Special Election to elect a new Board can be held; their appointed term of office shall expire after 45 days, by which point a Board elected at a Special Election will have taken over. Dissolution will force Oracle to make some sort of public decision - to either reconstitute the community under the current Charter or to disband it and do something else. If it does something else, it will need a different charter - and a different set of leaders. If Oracle continues its silence and does nothing, then at least the 7 of us on the OGB get part of our lives back. Neither of these options are desirable, though. The OGB's hope is that Oracle will see this as a wake-up call to actively support an open development community tied to Solaris, and start interacting again BEFORE we dissolve. We (as private individuals) are passionately committed to an open Solaris development community, and see this as the only remaining tool we (as an OGB) have to break the logjam of indecision and silence that is killing our community. In the best of all worlds, our dance partner will show up, and we will end the evening in each others arms, having forgotten the missteps and awkward moments that troubled us when the music started.* -John [*] Please don't take this analogy too far :-) ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] warning about mail on this list
Edward Ned Harvey, Please, before you condemn everyone who uses the Jive forums as being l0sers and id10t's, go and use it yourself to post a day's worth of mail replies. Find out *for yourself* how hard or easy it is to actually live life the way you are preaching, because, in the end, unless you walk in their shoes and show thru your actions that you merit people's respect, nobody gives a damn what you have to say. Most of us here have lived with Jive's foibles long enough to realize that there is a fundamental conceptual flaw in trying to gateway between a forum where the entire thread history is presented in the user interface in the order it was generated, and where quoted text is discouraged and bad form and a mail thread where those things are explicitly not true. In other words, it isn't Jive's fault, nor is it the Jive user's fault - from their perspective, they don't see the problem. It is only you, who chose to use a deficient user interface (mail) to follow their conversation, who has the problem -John On Mon, Jul 12, 2010 at 7:25 AM, Edward Ned Harvey wrote: > > From: opensolaris-discuss-boun...@opensolaris.org [mailto:opensolaris- > > discuss-boun...@opensolaris.org] On Behalf Of Svein Skogen > > > > As you have been told before, the users using the jive-forum doesn't > > have this luxury without a lot of manual work. Those who are using it > > as > > a mailinglist do. > > > > Please stick to the topic in the mail, and stop blaming the format if > > you dislike the message. > > I am not expressing a personal opinion, such as "copy and paste is too much > work." I am expressing an objective position: If you don't quote > messages, > it's difficult or impossible for readers to know what you're replying to. > > ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The nearest thing to (forthcoming) news
On Sun, Jul 11, 2010 at 8:35 AM, Scott Rotondo wrote: > Please don't speculate about anyone's departure, including whether it was > voluntary or not. This is a reasonable request in general, as we don't want to cause people distress or harm an individual's reputation. In this case, however, it is also a matter of historical record: as part of their taking control of Sun 5 months ago, Oracle intentionally laid off many of the people who were in Senior Solaris and OpenSolaris Engineering/Management positions; they also reassigned many other active community members to non-OpenSolaris jobs. In my mind, RIFs and job reassignments certainly equate to being "sent away". > suffice it to say that Oracle is not actively "sending away" senior Solaris > engineers. It also may be true that things have changed and Oracle is *now* trying to retain and/or grow their Solaris engineering ranks. Time will tell... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Solaris 10 set
On Fri, Jul 9, 2010 at 5:03 PM, Edward Ned Harvey wrote: >> You do understand that the jive-forum interface he is referring to, is >> the Opensolaris.org one, don't you? > > So? The point was that, even with paid support AND a full time staff of several people dedicated to the website, the OpenSolaris jive/mailman configuration and interaction is less than perfect, making the arguments we've been hearing (and hearing and hearing...) that one particular type is obviously better than the other seem pretty stupid. In this case, BOTH the paid software and the open source packages are lame and poorly integrated... In the one case, you need a vendor who responds with a better product that you can use; in the other, you need staff that is motivated and capable of adding a desired feature; OS's Jive/Mailman example simply shows the failure mode when you don't get what in theory should be easy or obvious, for whatever reason -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] The nearest thing to (forthcoming) news
It might be worth going just to find out who the current "head of Oracle Solaris development" really is; it'll certainly be more than the OGB has been able to find out all year... Might it be Stephen Hahn or Tim Marsland or Bill Franklin or Vincent Murphy or Greg Lavender or someone completely new? I don't suppose any non-Oracle community members were asked to be on the panel, either. -John On Thu, Jul 8, 2010 at 5:42 PM, Graham McArdle wrote: > D'oh! I didn't notice the session listing was in multiple pages, and of > course it's after page 1 where the main event is lurking: > "Oracle Solaris 10 and Beyond: Lead Engineers Panel (S317641)" > Abstract: "This session, hosted by the head of Oracle Solaris development and > featuring a panel of core engineers ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Open Solaris going the way of the Amiga
On Tue, Jun 29, 2010 at 9:59 PM, me wrote: > The OGB should really think about this > > We did, almost 3 years ago Since Oracle owns the trademarks (OpenSolaris *and* Solaris...), they can do whatever they wish with them. See http://hub.opensolaris.org/bin/view/Main/trademark -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Move the network related configuration file on Solaris10 but not go into effect
Hot damn - a thread that is actually related to OpenSolaris code !!! :-) Given the difficulties in enumerating all the potential suffixes for what could go after the "hostname." part of the filename, I'd bet that the code simply looks for a wildcard like "hostname.*". If so, renaming the file to "BACKUP.20100628.hostname.ce0" might be worth a try... -John 2010/6/29 Simon Yuan > Hello, > > Our customer said after he moved the file "/etc/hostname.ce0" to > "/etc/hostname.ce0_bak_20100628" on Solaris 10,the ip address which > originally tied to the ce0 still take effect upon OS reset(that is,it can be > seen in "ifconfig -a" output and still be configured on ce0 interface),why > ? The similar issue wouldn't appear on Solaris 9. > > Thanks. > Regards, > Simon > > > ___ > opensolaris-discuss mailing list > opensolaris-discuss@opensolaris.org > ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Oracle Unveils Next Generation Sun Fire x86 Clustered Systems
In Oracle's mind, one is a development work in progress, the other is a revenue product. Guess which one gets the corporate focus? -John Let's hope this is just an omission , so far, but on the OS > compatibility list of all these new servers, Opensolaris doesn't > appears...Only Oracle VM, Oracle Solaris and Oracle Enterprise Linux, > > ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Enough already: Re: So who is ready to be let down?
While it has been slightly amusing to watch all the trolls baiting each other here, this whole taunting and name calling episode has gotten rather old. Could you (collectively) take it outside and off this alias - the rest of us don't care to watch as you decend into kindergarten again... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] JDK in the installer...
> Realistically, it's just for the installer. Exactly what it's doing you'd > have to ask the Solaris folks, but its not used during network/jumpstart > installs, only for the interactive ones, IIRC. > > > One word: I18N -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] PC Emulator for OpenSolaris?
On Tue, Jun 8, 2010 at 1:42 AM, Nikola M wrote: > And who exactly you think could sue an free software projects and over > what issue? Take a look at the last 6 years of legal drama in the Jacobsen-v-Katzer JMRI open source case which has set the first legal precedent involving the validity of open source licenses. Along the way, the open source project lead had to double mortgage his house and has hundreds of thousand's of dollars in legal bills... http://jmri.sourceforge.net/k/summary.shtml If you weren't following the details, you probably have a wildly naive and incorrect view of how open source licenses, copyright and the law interact - I know I did. Things like the judge saying "damages are 'cost times volume of infractions', and free software costs ZERO, so you don't have a case" - and having to go to the federal appeals court to get that view changed, claims and counter claims of copyright infringement, etc all make the legal system something that is foreign and unnatural for engineers - law doesn't work at all like physics. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Is it possible to switch GNOME's license from GPL?
On Tue, Jun 8, 2010 at 12:47 AM, Nikola M wrote: > Therefore i see your question as mostly open source unfriendly in general. > > You never explained why you don`t like GPLed software , I am only > guessing it might have something to do with your intentions to misuse > free software. Some of us are drawn to the FSF's philosophy - and thus the GPL, while others of us find it difficult to embrace. There are a slew of OSI approved licenses, all of which protect the openness of a project and its source code in slightly different ways. For those of us who are not in the "super-anti-proprietary" camp, these other licenses fit our needs quite well. For example, the CDDL does a good job of making it possible for "free, as in Stallman" code to live side by side with "open, as in BSD" and "closed, as in 'we don't have the legal right to redistribute'" code - something not possible with GPL(v2). Just because some of us don't buy into the FSF religion doesn't make us "open source unfriendly" at all. We're just less fanatical. Having said that, the rest of us aren't 'bad' just because we have different ideals; diversity is a good thing, as is tolerance! -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Sun Software Product Map- opensolaris not included
On Mon, Jun 7, 2010 at 6:26 PM, Hillel Lubman wrote: > Is the reason a secret, or no one simply knows the answer? > > The obvious 3rd choice is that the reasons for the delays are known, but Oracle forbids its employees from publicly commenting about the details of any "yet to be released" products, under pain of discipline and potential employment termination. That this attitude is ultimately fatal to open source communities is an irony felt by all involved in the trenches, I'm sure -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Is it possible to switch GNOME's license from GPL?
Changing a license is amazingly simple in theory, yet complex enough in practice to be nearly impossible: The owners of the current copyright simply need to decide to relicense or dual license their code under a new license. Unfortunately, the tens of thousands of people who have contributed to GNOME over the last decade or so probably aren't of the same mind as to whether or not they wish to change, much less on which license they wish to change to. And, unless they *ALL* agree to the same details, it can't be changed. Note that the *users* of GNOME (you, me, the various distro producers...) are not the ones who have a say in the discussion - unless they/we also happen to be contributors... They that wrote it, own it. -John On Sun, Jun 6, 2010 at 9:12 PM, Andrew Greimann wrote: > Out of curiousity, is it possible to convert the GPL-licensed GNOME on > OpenSolaris to the CDDL or MIT licenses? We are NOT talking of LGPL here. ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] SUN not doing well under Oracle.
On Tue, Jun 1, 2010 at 8:33 AM, Oscar del Rio wrote: > Solaris achieved a world record TPC-H 3 TB non-clustered performance result > of 188,229.9 q...@3000gb with a price of $20.19/q...@3000gb." Translated : for $3,800,361.60 you can buy a lot of whatever you want, including benchmark prizes :-) -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] community driven distro...
On Thu, May 20, 2010 at 8:47 AM, Peter Jones wrote: > I am not getting it?are you suggesting the recreation of opensolaris > and > IP repository separate from Oracle? >From my perspective, no. What I am suggesting is that we should finish the work started in the emancipation community, pull in the groundbreaking build and boot efforts from the various other distros (starting with Schillix, Belinix and the IBM/z port simply because Joerg, Moinak and Neale are involved) and produce the tools, docs and code needed to enable others to easily grab the OpenSolaris source code from OpenSolaris.org's mercurial repos, build the bits and boot a running OpenSolaris kernel without reference to or use of closed-bin bits, proprietary-to-Oracle binary repositories or tools. In my mind, *not* being able to do this simple thing ourselves is a liability that needs to be addressed... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] An effort to rally around - building a real, community driven distro...
Re: http://opensolaris.org/jive/thread.jspa?threadID=129430&tstart=15 As Dennis said, "We are in a situation after some five years of the OpenSolaris community where one can not simply download sources and build something runnable. We can not even bootstrap the kernel without special corporate internal knowledge and resources. There are still large numbers of closed bins, trap doors and secret hallways just to build the current OpenSolaris code base." This is an invitation for y'all to come on over to the Distribution community and join us as we brainstorm and code a real, community driven distro into existence. Joerg and Moinak are already there, plotting how they can start by leveraging each other's build procedures and associated source code to allow anyone to easily clone the ON sources from OpenSolaris.org and build a kernel that boots to a shell prompt. Once we get to that point, we'll take another step towards removing barriers and enabling community participation. And another. A bunch of little steps, maybe, but they all add up. Join us! -John Plocher Less talk, more code! ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [ogb-discuss] This is how Oracle treats open communities and projects. Will OGB intervene?
On Mon, May 10, 2010 at 6:15 AM, Dave Johnson wrote: > This is how Oracle treats open communities and projects. Will OGB intervene? While I have not been following this soap opera in excruciating detail, my reading of the ARC discussions was that the ksh-93 project to replace existing GNU utilities with ksh-wrapper AST based versions that were not completely compatible and that would not track future evolution of the GNU utilities was withdrawn for good architectural reasons. Since the ARC case was not approved, it follows that repo putback access for that part of the project would also be withheld - a standard ON procedural action that applies to everyone: No approved ARC case, no putback. To answer your question: Will the OGB intervene? The constitution says (note the first sentence): 3.1 Disputes It is expected and encouraged that groups will resolve disputes by themselves according to their documented decision-making procedures. If a dispute can not be resolved within a group or it spreads between groups, then the Governing Board may choose to intervene. The Board will consider disputes on a case-by-case basis and may decline to intervene. If the Board chooses to intervene, it will resolve the issue at its absolute discretion with no possibility of appeal. Its resolution will be binding on all parties. Given this understanding (which may be flawed, but your posts do nothing to show that it is), this all smells like an overly emotional early Monday morning troll; I see no reason for the OGB to get involved. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [ogb-discuss] Proposal: OGB and Oracle communications for communities
On Sun, Apr 4, 2010 at 12:38 PM, HeCSa wrote: > We need something clear and direct appearing in the opensolaris.org homepage > pointing to the official information about the future of OpenSolaris, Oracle > and the communities, etc. To date, with the possible exception of Dan's comments on IRC during the election [below], there has been absolutely NO official information from Oracle about the future of OpenSolaris, Oracle and the communities, etc. So, to be pedantically correct, the current home page *does* show exactly everything that has been said officially by Oracle: Nothing. Many of us wish there was a more verbose statement. -John [IRC log from http://www.spcoast.com/irclogs/opensolaris-meeting/index.php?date=2010-02-26 edited here by plocher to remove join/quit noise from the transcript] +[12:06] * DanR (~chatzi...@nat/sun/x-tbazbldvpztcwjjh) has joined #opensolaris-meeting +[12:06] DanR: welcome +[12:07] ptribble: Thanks +[12:07] hi DanR ! any word on any official communications? I know you were hoping "soon" :) +[12:07] So we do know a few things now, as I discussed on the OGB call a couple days ago... +[12:07] (How's about that for putting Dan on the spot?) +[12:08] Yeah, thanks... :) +[12:08] So, here's what we can say: +[12:09] * bubbva listens intently :) +[12:09] Oracle will continue to make OpenSolaris available as open source, and Oracle will continue to actively support and participate in the community +[12:09] Oracle is investing more in Solaris than Sun did prior to the acquisition, and will continue to contribute technologies to OpenSolaris, as Oracle already does for many other open source projects +[12:11] Oracle is committed to supporting our customers +[12:12] DanR: Does that include x86, or will the heavy focus on SPARC override further x86 development/support? +[12:13] And Oracle will ensure customers running OpenSolaris have an option for support on Oracle Sun Systems where it's required, though given the very little sales here this will not be something we expect many customers to deploy going forward. Solaris is our focus, on both SPARC and x86. +[12:13] DanR: The recent patch policy changes to Solaris were a major concern; given all the SPARC talk, we were worried x86 would suffer. +[12:14] Given the lack of communication there, it is driving platform change considerations for a lot of shops. +[12:14] Oracle will also continue to deliver OpenSolaris releases, including the upcoming OpenSolaris 2010.03 release. +[12:14] sun.com used to be a.. quite good looking page.. for the past ten years. it had information for both managers and techs.. however, it now seems to be very tough to find tech information for any of the old Sun hardware.. will this be fixed? +[12:14] So OpenSolaris really is becoming Fedora? ;-) +[12:15] it used to be very easy to navigate, now.. not so much +[12:16] The patch decision is aligned with how Oracle does business in other areas as well, patches are delivered for customers but not for free. So it's consistent, that doesn't define a platform future at all. x86 is the core of our Storage appliances for example, we're not going away from it at all. Though clearly we make more money on SPARC so there is more of an emphasis given the customer base. +[12:17] DanR: Completely understandable. My company (and I imagine most others) value consistancy heavily. :) +[12:17] For the web transition, we're working hard to simply things and the same leader for example that ran BigAdmin previously now is in charge of creating the community inside of OTN which will now include a place for admins, developers and architects for Solaris and Systems +[12:17] The lack of communication beforehand means we couldn't buy support contracts before the patches were pulled, however. Hopefully that was just an oversight. +[12:18] well, closing up security pathes can be considered ill behaviour from vendor. +[12:18] tsoome: Agreed. But it's done. Perhaps OpenSolaris security/support policy will change as money is less of a driver there? +[12:19] DanR: what about open development in the future? +[12:19] and it will cut off people from even considering *solaris, meaning no money for oracle anyhow. +[12:20] tsoome: what will cut it off? +[12:20] reflect: The lack of free security updates. +[12:21] hm +[12:22] ptribble: Oracle will continue to develop technologies in the open, as we do today. There may be some things we choose not to open source going forward, similar to how MySQL manages certain value add at the top of the stack. It's important to understand the plan now is to deliver value again out of our IP investment, while at the same time measuring that with continuing to deliver... +[12:22] well you can argue its open source and fix yourself or buy support, but not many people are able to fix things, nor buy support. +[12:22] ...OpenSolaris in the open. This will be a balancing act, one that we'll get right some times but may not always. +
Re: [osol-discuss] pkg scripts are evil :-) (was Some Why?-Questions)
On Fri, Dec 18, 2009 at 3:36 AM, Volker A. Brandt wrote: >> What's needed is a supported mechanism to make it >> easy to do it right. > > +999 When Stephen and Bart and company started talking about IPS leveraging SMF, I suggested that it might be possible to keep the SVR4 pkg pre/post install/remove script idea, add the "usual" SMF start/stop/restart bits and automatically put the whole shebang into the SMF framework at install time. The response was "we don't do scripting in pkgs - they should write their own SMF stuff themselves" I don't see leveraging SMF as a hack - heck, to me, inventing a 1-off meta language to be used just for pkg scripts seems much more limiting and hackish than leveraging SMF Maybe SMF needs to evolve the idea of "svcadm install/remove service" so it would have a semantic place to put the pre/post install/remove scripts, or maybe the first "enable" would simply notice that the service had not yet run its install script; the result would be what people are asking for without the problems associated with "wrong context". -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] pkg scripts are evil :-) (was Some Why?-Questions)
>> The better question is why someone is doing something >> so broken in the first place. > > There is nothing broken about being able to consistently and repeatably > create databases via packages. If you can do so in a way that reliably does so, then there is no problem. Unfortunately, given the diverse environments that pkgs are expected to be installed into, the constraints on such pkg scripts are pretty severe. Having written (and rewritten and re-rewritten) packages to do stuff like this, I can almost guarantee that your package postinstall/preremove scripts will fail chaotically and irreversibly in some relatively common environments, like zones (branded or otherwise), cross-architecture diskless servers and live upgrade (aka pkgadd -R..) - i.e., environments where the pkg runtime execution context is NOT the same as the target environment. For example, on an X86 server which serves SPARC diskless clients, if you pkgadd your package for use on a client and it runs "sql 'CREATE DATABASE foo;'", it will fail because there is no database software installed on the server; even if it somehow found the binaries on the client OS disk image, it would fail - both because SPARC binaries can't run on X86 CPUs AND because the database it needs to create needs to be created in the CLIENT's disk/memory context, not the SERVER's. The command might even succeed, but its side effects would be a disaster. The list of "Oh, I assumed for my pkg" failures related with pkg scripts is astonishingly high - almost anyone can create a pkg+postinstall that works in a carefully preselected environment, but very few are able to make a bullet proof, works as expected everywhere, cross version and cross architecture solution. After looking at (and writing!) way too many pkg scripts and finding serious flaws in their coding/behavior, some of us have come to consider them to be an attractive nuisance that should be discouraged. In my experience, it is always seems to be better to have some sort of "run me first" script that can be run in the client's execution context than it is to force the packaging system to also be a service management tool. Last I checked, IPS includes a mechanism that makes it easy to install and invoke such a script correctly via SMF, neatly bypassing the failures found in pkg scripts AND providing a reliable, repeatable and co nsistent way to manage your newly installed service to boot. Yes, it is a change, and yes, it makes simple-but-fundamentally-broken things harder to create; some would say that this is progress. -John -- This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Nomination of Meenakshi Kaul-Basu for the 2008-2009 OGB
In an attempt to grow the OGB and the OpenSolaris community to include more diverse perspectives, I would like to nominate Meenakshi Kaul-Basu as a candidate for next year's board. Meenakshi is the manager responsible for the open sourcing of the Solaris Cluster product and is a driving force behind the HA Clusters community here in OpenSolaris. She has worked with the clustering product for more than 8 years and has been extremely active behind-the-scenes to evangelize and build their open source community. She is currently the Director of Engineering for the Availability Products Group in Sun. -John Plocher ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [advocacy-discuss] [indiana-discuss] I'm sorry, but I just don't get it
>>> Putting /usr/gnu at the head of PATH causes incompatibilities >> Failure to put /usr/gnu at the head of PATH will cause I'm not at all sure that this change would actually affect me - or most of y'all either. Since our target growth "market" for OpenSolaris is users of linux systems where gnutools are the defacto standard, then it seems clear that *they* certainly won't care about being "incompatible with old Solaris or POSIX". They are the ones who would find not having tar->gtar to be a bug. The only people who will care are the old Solaris users who haven't yet figured out how to set PATH= in their shell scripts and/or don't have their own .profile/.login/.bashrc/.cshrc scripts. Since I have my own .startup-scripts, and they explicitly set PATH, I won't even notice that this change has happened. [It would be nice if this choice was reflected in the new user account setup dialog instead of being hardcoded by the installer, but that is nit...] -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [ogb-discuss] A proposal for ensuring sustained Community Growth and Success
Shawn Walker wrote: > Time and time again it has been said that the OGB can only act as an > "arbiter" of sorts; it is my belief that they must be empowered to > actually *guide* the community. We are learning a lot about how to govern ourselves. One of the things I have learned is that leading is hard. It takes time, it takes commitment, and it takes a very thick skin. "Hearding a clowder of burning cats away from a river of gasoline" is the phrase I've often heard applied. And, compared to software developers, cats sound easy ;-) Another thing I've learned is that if you are not happy with the way someone else is leading, you need to step up and become a leader yourself. This means being able to articulate your vision, convince others to adopt your vision, and then effectively make your vision into reality. (I really wish I was better at this part...) To me, this means a couple of things for all US vocal people: 1) Try to generate specific and actionable proposals for change instead of gripes and whines and endless meta-dialog, 2) Nominate yourself for the next OGB elections, and 3) Become Core Contributersin your CGs and put your leadership skills in action. Become the grass-roots examples of the kind of leadership you want to see in others. Lead by example. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [advocacy-discuss] [indiana-discuss] I'm sorry, but I just don't get it
Shawn Walker wrote: >> Tell that to whoever violated ARC by putting /usr/gnu at the head of >> $PATH in the indiana preview ;) > > No violation of ARC occurred. ARC is not a required process until > consolidations are integrated, etc. On the other extreme, it would be extremely stupid^H^H^H^H^H^H ineffective to bundle up all the changes made to Indiana and produce them at the end of February in one un-reviewablely large "we're done, here is everything we did" ARC case There seems to be a bunch of stuff that got started, is being prototyped and played with, has gotten integrated into a gate somewhere and was shipped as part of a developer prototype release that still isn't on the ARC radar. That's the point that has a bunch of us concerned - not that the review hasn't been completed, but that they many of them haven't even been started! -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [trademark-policy-dev] [indiana-discuss] [ogb-discuss] I'm sorry, but I just don't get it
Sara Dornsife wrote: >> It's all about the name. Back away from "OpenSolaris Developer Preview" >> and this nightmare will end. >> > > And then what? I hope that doesn't sound facetious, I'm really asking > what you see as next steps. Seriously: Mend your bridges with the community you diss'ed: Engage with the Desktop, HPC, Install/Pkging and ARC communities by asking something like "we would like to make the Indiana Project into something that could be named "OpenSolaris-foo - how can you help us/what do we need to do to get to that point?" You could also try having a public conversation on the topic of "We've got a developer preview ready, and would like to position it as an OpenSolaris Developer Preview. How can we do that in a way that engages (rather than disenfranchises) the larger community?" Usually, the flow is Get an idea, discuss it with all of the stakeholders, come to a mutually agreeable conclusion, then act in concert on the result. The Indiana developer preview naming decision seems to instead have done Get an idea, discuss it with executives at Sun, announce the results to the stakeholders as an already-decided fact and start acting unilaterally on the result. The first builds a community, the second erodes one. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [trademark-policy-dev] [advocacy-discuss] Project Indiana and the OpenSolaris name
[Followups to [EMAIL PROTECTED] Joerg Schilling wrote: > Compatiblitiy is less trivial than you might belive but without conformance > tests, we cannot claim anything about compatibility > ... > > A distro alone cannot be a refernce. It must not even be changed for > the compatibility tests. One of my comments on the wiki definition was along the lines of: We could, as a starting place for defining compatibility, simply assert that there is a baseline (installer and a set of versioned packages; a "recipe", if you will) that must exist in any distro if it wants to claim compatibility. Of course, this type of definition is poor, from many perspectives. It is, however, easy to implement :-) It presumes significant sameness between distros - not a bad thing from a compatibility perspective. Same installer, same kernel, same packaging system, same repositories... But, it is a starting point that we can use today, rather than waiting for someone to develop a full blown, ratified test suite... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Project Indiana and the OpenSolaris name
Bill Sommerfeld wrote: > On Thu, 2007-11-01 at 21:12 -0700, John Plocher wrote: >> In addition, the OpenSolaris >> community itself produces a distro that is compatible. > > I think "a distro" is unnecessarily constraining. +1 I used the term in a "at this current time" context; as I've said before, I expect that at some point in time we will have a series of focused or niche distros: Indiana 10/07, Indiana 4/08, ... SomethingMajor11 Kansas 10/09, Kansas 4/10, ... as well as An Appliance thingie A Storage/ZFS thingie A Web Stack thingie etc -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [ogb-discuss] [advocacy-discuss] Project Indiana and the OpenSolaris name
Keith M Wesolowski wrote: > ... > I'm looking past the immediate firestorm here to the big picture. I > haven't forgotten where we are today and I don't dispute the need for > carefully considered action to address it. But it's helpful to think > about where we need to go. > Extremely well said. +1 -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [trademark-policy-dev] screwdrivers - Re: [advocacy-discuss] [indiana-discuss] Project Indiana and the OpenSolaris name
[I'm close to 200 messages behind in this discussion at this point...] Brandorr wrote: >> The Linux world asserts that this is a feature. Do you see it as a >> bug or a feature for OpenSolaris? >> >> If it is a bug (as I believe), how do we fix it? > > John this reads as if you believe the existence of Nexenta, Shillix, > MartUX and Belenix are bugs? If so, I don't think it's the best way to > get your point across to Joerg. ;) At this point, all of "our" distros are relatively compatible with each other, unlike the general Linux world. If we can define (and effectively sustain) a compatibility culture, we won't have this bug. > John, in our trademark discussions you stated that OpenSolaris should > not be used as a standalone noun. As in Joerg's screwdriver analogy. > Am I missing something. If we define a compatibility story, then, as Joerg says, it makes sense to call the "thing that inplements and deploys exactly and only those compatibility bits" OpenSolaris. I don't believe for a minute that Indiana 10/07 is such a beast, nor that Indiana 3/08 will limit itself to being only that subset. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [indiana-discuss] [advocacy-discuss] [trademark-policy-dev] Project Indiana and the OpenSolaris name
Brandorr wrote: > Now you are telling another $foo distro maker, that all the other $foo > distro makers that they are marginal $foo players. And that only Sun > can make a real $foo distro? > > Am I misinterpreting your statement? Yes. I'm trying to say: I am telling another $foo distro maker that all the $foo distros are either compatible, derivative or incompatible $foo distross, as defined by the OpenSolaris community. In addition, the OpenSolaris community itself produces a distro that is compatible. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [ogb-discuss] [trademark-policy-dev] [advocacy-discuss] Project Indiana and the OpenSolaris name
Alan Burlison wrote: > Alan Coopersmith wrote: > >> I agree two weeks seems arbitrarily short, but given the timelines, >> think that December/January-ish is about as late as we can push it >> unless Indiana agrees to choose another name for it's March release. > > My post was a proposal, a stake in the ground - no more no less. If > there's a better suggestion, I'm all for it. As the instigator of the branding project, I'd love to get to that point in 2 weeks. But, I'm not sure the discussion there has managed to progress beyond the "massive email about past, present and future slights, omissions and bad behavior" stage and into a serious discussion about specifics. At this point, only three or four people (myself included) have made specific suggestions and/or proposals; they have been incorporated into the wiki. Unfortunately, judging from the majority of the discussion, nobody else has read any of it. If this doesn't change, I'm unwilling to take my half-baked, decidedly one-sided perspective and rush it out for a vote as if it was something that reflected the intent of the community. It needs to honestly reflect that intent, not try to dictate it. And I need y'all to help it become such a proposal. I'm also unwilling to push this off until Indiana is at the point where it couldn't be modified to reflect the outcome of a vote; I hate playing chicken with projects that are trying to game the system. So, +1, except that your 2 week deadline may need to be something like in a month; most certainly before the end of the calendar year. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Democracy and open source - nit - Re: [advocacy-discuss] Project Indiana and the OpenSolaris name
Brian Gupta wrote: > In other words, Ian seems to have decided that democracy is a bad way > to run an open source project, and wants to install himself as > "benevolent dictator". OpenSource efforts are invariably meritocracies. Those that /do/, lead. Ian and the OpenSolaris Project are out there /doing/. They chartered a Project to do this, found several CGs to endorse their vision, and have just delivered the first distro built by the community out of the community's source code. If you want to be involved in the democracy, go make yourself heard as a core contributer in the install and packaging CG so you can be directly involved in those decisions. Since you are a core contributer there, I assume you have done so; since you are complaining here, I presume such efforts didn't generate the results you desired... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] screwdrivers - Re: [advocacy-discuss] [trademark-policy-dev] [indiana-discuss] Project Indiana and the OpenSolaris name
Joerg Schilling wrote: > This is also a good example as "gimme a screwdriver" has the same level > of incompleteness as "gimme OpenSolaris". The Linux world asserts that this is a feature. Do you see it as a bug or a feature for OpenSolaris? If it is a bug (as I believe), how do we fix it? If you don't think it is a bug, then we will have to agree to disagree. I don't believe this disagreement is necessarily fatal to our ability to work together. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [trademark-policy-dev] [advocacy-discuss] Project Indiana and the OpenSolaris name
Joerg Schilling wrote: > So you like to call SchilliX OpenSolaris? Absolutely yes - one of my objectives in this branding effort is to find a way to allow this (or something similar). -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [trademark-policy-dev] [advocacy-discuss] Project Indiana and the OpenSolaris name
Joerg Schilling wrote: > You make an important mistake here: Indiana is not a community initated > distro but a _Sun_ initiated one. You keep asserting this, as if anything done with or by people working for Sun has no validity. Bullshit. If 95% of the people working on opensolaris things are Sun Employees, then (in your perspective) 95% of the things done here aren't community efforts. Hey, we are OpenSolaris Community members also! > SchilliX was the first community initiaded OpenSolaris distro but Sun > did not like to help with this distro. I'd say, rather, that Sun had its hands full with launching the whole opensolaris effort, and didn't have the time, resources, connections to the right people and/or legal ability to do the things needed to make Schillix happen. Transforming that complexity into a disparaging "Sun did not like to help" is a bit much. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [advocacy-discuss] Project Indiana and the OpenSolaris name
[EMAIL PROTECTED] wrote: > we don't seem to be doing enough to facilitate the other distro's > existence at opensolaris.org. +1 ! Its not the decision that matters, it is *how* the decision was made. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [advocacy-discuss] [trademark-policy-dev] Project Indiana and the OpenSolaris name
Martin Bochnig wrote: >> This is like calling the automobiles built in the BMW Factory "BMWs". > > Only when considering the 2nd layer (the distro-creating projects and > associated ones, rather than OS/Net src base maintainers). That is exactly what I am doing. I'm NOT trying or intending to associate the name OpenSolaris with simply the OS/Net src base... > Then the OS/Net sources should not have the identical name "OpenSolaris" > any longer. Where is the OS/Net consolidation currently being named "opensolaris"? I agree with you - if it is, then it should be changed. The community name is Nevada. Sometimes spelled "nv" The common name is either ON or OS/Net. The BFU image is snv_76 or nv_76 Inside Sun, it was called ON; sometimes "the WOS". The name "Solaris" was always used for the complete system that was sold to users; Solaris always included a GUI desktop, utilities, nameservers, routing and email daemons, etc etc etc. > If you want to be FullyOpen, then you may have to accept this. Being a FullyOpen community member, I accept that others in the FullyOpen community may do things to which I disagree. FullyOpen does not mean "everyone has a veto". It also does not mean "every decision is made by consensus". Sometimes people see a need, and go on and fill it, whether or not others agree with them... > And honestly: Do you understand / consider the existing distro projects > as hostile competition, rather than as (positive, supporting) part (and > partially also parts-supplier) of the whole ONE ? I see them as being positive, value added and supporting alternatives to a community that is focused on providing a best-of-breed generalized reference distro; however, they are NOT the primary focus or design pattern for the majority of the community. > That's almost pretty much a joke, given that I thought about using Ford instead of BMW in my analogy, but I didn't because I don't like Fords. I should have, though, because then I could have concluded by saying Of course, custom car makers like Ferrari, Lotus and Rolls Royce don't consider Ford to be a competitor because their customer base simply doesn't buy Fords. I consider MartUX, SchilliX and Nexenta to be similar to those boutique automobile builders - they are what you install and use if you have requirements that are not being met by the Fords of the world. And, you know what? Many people use Ford or Chevy engines in their customized vehicles - Ford is also a pretty large parts retailer too. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [advocacy-discuss] [trademark-policy-dev] Project Indiana and the OpenSolaris name
Martin Bochnig wrote: >>> Renaming "Indiana" to "OpenSolaris": Wouldn't that be like renaming the >>> brand "Crysler" to "Automibile"? >>> IMO a "Crysler" is not equal to "Automobile". It would be a subclass of >>> it (with more nested subclasses and then n-millions of instances/leaf >>> nodes). I think you have it - exactly. This is like calling the automobiles built in the BMW Factory "BMWs". This is the OpenSolaris Community. We construct a distro out of the various source code parts that are manufactured in our community. It is natural that the thing we produce somehow carries our name. The alternative is that we are simply a parts supplier for other automobile companies. The Linux Distro Hell is flawed, IMO, because it sees itself as a community of parts suppliers rather than as a producer of cars. This may be good for those who love building their own vehicles, but sucks rather large rocks if you just want something reliable to get you to work every day. I can understand why boutique hand-assembled auto manufacturers might not like to see BMW enter the market... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [advocacy-discuss] Project Indiana and the OpenSolaris name
Even better, contribute and make the project something that reflects your values - make it something you can vote *for*. -John > On 31/10/2007, Alan Burlison <[EMAIL PROTECTED]> wrote: >> What will be the point of having a vote on something that is a fait >> accompli? ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [advocacy-discuss] Project Indiana and the OpenSolaris name
Isaac R. wrote: > I think the question of "getting access to" OpenSolaris could be > addressed by allowing (anyone interested in doing so) to > make that decision by looking at a matrix with requirements > (horizontally), and how various distro's satisfy those requirements > (vertically). I tend to agree, but the devil is in the details... Could you take a stab at producing this matrix - or at least the column labels for the features/requirements that you might expect to see? A concrete example would be extremely useful about now :-) > Assuming each of the distros were using Nevada as the kernel (which is > available through the OpenSolaris project, and which they do), then all > distro's deserve to be referenced as OpenSolaris-based, even if such Sounds like your definition of "compatibility" is closely related to "has the same kernel"... I'm looking forward to seeing what your important requirements might be. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [trademark-policy-dev] [advocacy-discuss] Project Indiana and the OpenSolaris name
>> Joerg Schilling wrote: >> I have problems if this was not labelled with "Sun" as this would cause >> harm to other existing OpenSolaris based distributions. > Shawn Walker wrote: > I have yet to see any qualifying statements that indicate exactly > *how* other distributions would be harmed. I *think* Joerg is referring to the classic channel partner -vs- direct sales problem - if the OpenSolaris Community has its own distro, where is there room for other distros to compete? The answer, of course, isn't simple. The status quo changes, and we all have to change or be left behind. As an awesome first non-Sun distro, Schillix broke ground that made it possible for there to /be/ non-Sun distros. But, that was 2 years ago, and finally the community is getting itself up to speed. Rather than being a private effort run outside of the OpenSolaris Community, Indiana is producing a distro within the community itself. (It is interesting to note that of these 6 initial distros, only the SX and Belinix teams seem to have put in the effort to transform their outsider distros into something done within the community) In the end, though, this is a loosely structured community, driven by those who "do" rather than those who "talk". See a need, fill a need. Sometimes there are competing efforts and one succeeds while the other doesn't. Othertimes, both succeed wildly. It is all about choices. If Joerg or any of the other initial-distro leads had so desired, they *could have* created an OpenSolaris Community/Project to host and develop their distros; chances are that if they had, their efforts would now be the ones we would want to call OpenSolaris. Ironic, no? -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [advocacy-discuss] Project Indiana and the OpenSolaris name
[Followups to trademark-policy-dev, please. To post you will need to subscribe by first sending an email to [EMAIL PROTECTED] -John ] Ian Murdock wrote: > ... The first step to a branding program is to define > the OpenSolaris binary core, and I invite the community to help define > it, using the Indiana bits as a first approximation, with the > understanding that it is OK to make mistakes, leaps of faith and > simplifying assumptions as we figure this all out. > Followups set to [EMAIL PROTECTED] . (At 8:04pm this evening, just as Ian was typing up his email, we experienced a ~5.6 earthquake here in San Jose. The USGS says it was effectively right under our house (9km down and 4km east, but who's counting? Coincidence? I don't think so! Thanks, Ian! :-) Ian makes a compelling point that a distro made up of everything on opensolaris.org should be called opensolaris. The question still seems to be if this view can be reconciled with Joerg's and Brian's (placeholders for many, I'm sure) minimalist perspective (i.e., OpenSolaris - the operating system - is only the kernel, libc and a shell). Maybe we don't have to reconcile them, because they are /different/ things. Which of the following are OpenSolaris? Duh, they all are. They simply have different audiences: The OpenSolaris Operating System: At the minimalist end, we have a "miniroot" consisting of just the stuff needed to boot and get to a shell prompt on a specific device. The audience for such a "distro" seems limited to those developers actually working on a particular device. Think PowerPC and CellPhones. Think small number of dozens of people. The OpenSolaris Operating System: Moving up in the world, this miniroot gains enough drivers and userland bits to become the basis for a dedicated appliance. Since the needed bits differ based entirely on what the appliance is supposed to do, and there presumably isn't any need for the user to add new functionality to a given one, the audience for such a distro is also limited to the small set of developers actually working on the appliance. Think routers, web servers, mail servers, model railroad empires; think small number of hundreds of people. The OpenSolaris Operating Environment: At some point we have a miniroot, drivers and enough userland to produce general purpose computing devices. Although one size could fit all (XXXL?), it seems reasonable to postulate laptop, desktop, blade, cluster and enterprise variations. Each of them will be characterized by their own recipe, optimized for the task at hand: Laptops care about X and GNOME, web hosting servers care about Apache, Glassfish and python. Unlike the device and appliance distros, these general purpose distros are targeted at the volume market with the expectation that their users will want to add 3rd party features to their systems. Think volume distros. Think millions of people. From a compatibility perspective, it is probably OK to ignore the embedded device and appliance distros - there really isn't any expectation that a user could take an arbitrary precompiled binary package and install it on them. This leaves the general purpose systems. If you take all their "recipes" and compare them, you will find a large set of common features/packages. This is what I an thinking of when I say "compatibility Core" for the OpenSolaris Operating Environment. Today we have SX, SXDE, Schillix, Belinix, MartUX and Nexenta as examples of various targeted distros. If I have a binary program (say, oracle or my company's accounting package...), and I want to pick a distro, Should I /expect/ my application to just work on it? /Will/ it just work? Does the distro owner have any expectations in this regard? and most importantly, How would I tell? This implies that the branding needs to communicate something about compatibility, and it should also be sensitive to the distinction between Operating System and Operating Environment. I'm going to sleep on it and see what the morning brings before I go edit the wiki... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [trademark-policy-dev] I guess "the community" decided to go with my original suggestion?
Brandorr wrote: > A compatibility test, that says, "verify that this is the same exact > set of bits" is an audit, not a compatibility test. You are defining > compatibility in such a way that only the thing being tested against As the one who wrote those words, along with an email stating that: > What if the test suite was simply an audit that this distro did, in fact, > get built out of the required packages from the required repository? > > Not quite as hard as a full functional test (and maybe not quite as > useful in the long term...), but potentially good enuf to get to the > next stage... isn't in violent, 100% agreement with what you are saying? We don't have any tests now, and I didn't have time to whip any up out of thin air, so I stuck the issue somewhere where it wouldn't get forgotten. It sounds like you have energy to run with this - please do so; flesh out some structure behind this topic on the wiki... > Currently I can take Blastwave packages compiled for Solaris 8, and > install and run them on Solaris 8, Solaris 9, Solaris 10, SX[CD]E > b1-75, Nexenta, Martux, etc. Are these distros incompatible or > compatible? I'd say that we *want* them to be, therefore the question to ask is what the "branding quality control" tests need to look like to make that happen? (Though blastwave isn't necessarily a good example - it carries along its own compatibility core set of packages that isolate it from the host OS...) [I added some of your comments to the wiki - please elaborate on them there...] > But then again this whole compatibility discussion, predicates that as > a community we have consensus on a need for an OpenSolaris reference > distro. I don't recall any community vote saying that need exists. How > can we write guidelines, presuming the existence of such a beast? By doing what we think is best, by trying, failing, trying, succeeding, trying... Where was the community consensus that said we wanted to do ksh93, DTrace, the bugfix for [XXX, YYY and ZZZ], etc? - you won't find it. Development efforts - especially open source ones - don't work that way, they work by people seeing a need, and going off and addressing that need. If they did a good job at it, the community adopts their work... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [trademark-policy-dev] I guess "the community" decided to go with my original suggestion?
Brandorr wrote: >> so again I ask you - please make a counter proposal. Be part of >> the solution! > > Proposal: No distro will be named OpenSolaris. I updated the wiki to include this > John's proposal, as I understand, is that only people/teams that can > use the OpenSolaris trademark as a standalone brand, are people/teams > that make a product that only consists of Sun's binaries. My bad - As I noted in previous email (but not on the wiki), the "binary" stuff was trying (poorly) to capture the idea of binary compatibility (if you use the recipe /and/ the packages in the repository referenced by that recipe...) Since the whole recipe/repository thing is causing confusion, I deleted it, and instead added more to the note in parkinglot#1 that we need to define and work on a good definition of 'core'. > (Please note the use of the commas. They can be replaced by the word > "an" if that makes it clearer for you). I did - and updated the wiki to match. > It also seems that many feel that the current definition of > OpenSolaris shouldn't change. That's the problem - there /is/ no current definition of OpenSolaris. There are a dozen or more, and they are all different. This project is trying to fix *that* problem. Please recheck the wiki - I hope I've captured your main concerns. if not, please add inline comments/questions... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: OpenSolaris as a seperate entity to Sun (Re: [osol-discuss] Re: [ogb-discuss] Re: [advocacy-discuss] Re: Logo/Mascot for OpenSolaris)
[EMAIL PROTECTED] wrote: Isn't anyone in the least bit interested seeing OpenSolaris actually be able to employ people or pay for things itself rather than depend on the good will of Sun to do it all? At this point, with all the backlog of things that need to get done to make things work logistically, I'd have to say no, I'm not the least bit interested in having the OGB, the ON community, the gatekeepers, the Hg repository team, the web infrastructure team or the ARC community (to name a few) spend any significant effort on it at this time. These people are already spread way too thin to take on the significant additional effort needed to create and manage a 501c. Having said that, if /you/ are interested and motivated, feel free to do the legwork yourself. Be a leader, form a consensus, do the research, etc. Remember, actions speak louder than words. Even mine :-) -John 501(c): Been there, done that, don't want to go there again... ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Project Proposal: Open Sound System
Dev Mazumdar wrote: The rumors are true, we're planning on open sourcing Open Sound +1! Adding icing to this cake is that Dev has been actively working with several of us getting things ready for an OpenSolaris ARC review of OSS as part of their initial code contribution! Of course, this means that we in the ARC community are going to have to figure out the logistics of making this happen - the OGB discussion about ARC reviews seems timely indeed! Welcome aboard Dev!!! -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Re: [approach-discuss] Re: [sfwnv-discuss] Like minded individuals let's start working on an "OpenSolaris.org" distro right now.
Joseph Kowalski wrote: However, if the purpose of this discussion is to detail the costs, while the benefits are being pursued in some other discussion, then I'm all for it. Thats why I'm here at least... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: [ogb-discuss] Project Proposal - (what is/was Indiana)
James Carlson wrote: I do wonder a bit what the distributors of far-afield variants ought to think of this, particularly because of the naming issue and the clear differences among the existing variants. If I were working on Nexenta, would I be pleased, uninterested, or disgusted? Should I be I note that none of them are raising objections in this thread :-) jumping at the chance to have my own One True Way stamped as the 'official' OpenSolaris way to build a distribution? You are obviously concerned about something "bad" happening here. I'd suggest that you either need to actively lead this project away from that "bad" path, or step back and see what happens. I'll bet that, if the project is indeed going down that "bad" path, it will fail rather quickly (and good riddance to it, if so). But if it instead does "good" things and succeeds, we all will benefit. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: [ogb-discuss] Project Proposal - (what is/was Indiana)
James Carlson wrote: The part that you snipped away was where John Plocher was asserting that this case would somehow prevent future distributions (such as the ones we now have) from even starting. I don't believe that's the case. I don't think I was asserting anything like that. It was 3am, so I may have been more unclear than normal, but where in the following text that Ian "snipped" do you find that assertion: > Why does Indiana have to meet these requirements when Nexenta, Belinix, > MartUX, Schillix, SX, and all the other OS.o distros don't? ? I just don't see it that way. Is having a community endorse this project a "barrier?" And one that necessarily causes undue hardship? This project is at least as complete as several of the others that have been approved with no discussion in the last couple of weeks. So far, I haven't seen anyone trying to obstruct. Other than Keith and you telling Glynn to "go fetch another rock", because you don't like the rock he currently has. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Do we even need a reference OpenSolaris binary distro
Darren J Moffat wrote: For something to be *the* reference distribution for OpenSolaris is quite a significant status. Exactly why do we even need that status ? Because many people have asked for something like it and some other group has decided to go scratch that itch. Do you really need any other reason? What is so special about the reference distribution that it can't be one of the existing distros ? Someone wishes to do their own distro and do it their own way. See the recent thread on why the existing distros were started... What special status should a reference distribution actually have ? What is the implication to other distros if they do things differently to other distros ? None and None. My personal opinion is that I don't want to see a reference distro, at this time I just don't see any value in it. Then don't work on it. Put your energies elsewhere. Vote with your feet. But please don't shut down the people who /want/ to work on it. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: [ogb-discuss] Project Proposal - (what is/was Indiana)
James Carlson wrote: What I understand Keith to be saying (and what I agree with here) is that if you're going to do that in the name of OpenSolaris itself -- not just "PlocherX" but "OpenSolaris Reference Release" -- then that's logically something that ought to be a deliberate decision of the community, and not something that "just happens" or (worse) "happens because some executive at Sun says so." That "executive at Sun" (Ian M) is also a member of the OpenSolaris Community, and modulo his contributer grants, just as much a player as you or I. Where in the community's charter or constitution does it say that creating /an/ (not /the/) OpenSolaris distro that can be reused by others as a reference must obtain "special" approval by the OGB or OS.o community at large? I'm not sure it's necessarily an unmitigated good thing to have a single privileged reference release (what happens to distributions that decide to innovate in a different direction?), So it boils down to the fact that the OGB [or at least some of its members] either don't like the name of a project or object to some aspect of its goals. Neither is within the purview of the OGB - which exists to resolve non-technical disputes, and not to create technical mandates. but, as a community member, I'd like to see something more concrete about what the reference will contain before deciding whether to endorse it. The OGB exists to make sure the OpenSolaris Community grows and does cool things. Get out of the way and let Indiana grow and do cool things, or let it die and fail - but let it do those things by and to itself. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Project Proposal - (what is/was Indiana)
Within Sun, for the last 17+ years, anyone could initiate a "project" at any time by simply sending a proposal email to a single internal alias. It required absolutely no prior approvals, no sponsorship, and no endorsement. (Though, while it is /always/ a good idea to have your manager be aware of what you are doing, that is more of an HR issue than a project approval one :-) Even so, many considered even that amount of effort to be "too difficult". The only "downside" is that we have had to develop a set of "is it dead yet?" heuristics to determine if a project is effectively dead and not simply dormant or focused on internal prototyping or otherwise still flying under the radar. No big deal. Even with between 500 and 900 projects being proposed every year, the system isn't overloaded or inconvenienced in any way. -John Alan Burlison wrote: [EMAIL PROTECTED] wrote: Yes, that seems to be to cumbersome; I would prefer OGB not to be involved in project creations as long as projects are started under the wings of a community and the community is not disfunctional In that case (if my reading of the OGB minutes is correct) Roland's ksh project would never have got off the ground. I don't understand why a community is necessary for a project to exist, sorry. ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Solaris as NetBSD?
Girts Zeltins wrote: I want to say that Solaris must be as NetBSD and must be available for all platforms. I don't think you will find too many who would disagree with you. I would note, however, that it took several painful years to evolve a multi-platform NetBSD from BSD's initial Net/2 dump. Years where many committed and talented people provided the blood, sweat and tears to develop and maintain the variety of ports that you have come to love. We have the beginnings of a similar culture here - SPARC, X64, PowerPC, System/390 ports are all in various stages of functionality and usability; the OpenSolaris code base is already structured to support multi- and cross-architectural needs. There have been several discussions about porting OS to appliances, game systems and other CPUs - don't take my list above as the final word on what all is happening! What do you see that we are missing here? Other than more experienced and committed developers to work on the ports themselves, I can't think of any barriers or cultural disconnects... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: [ogb-discuss] Project Proposal - (what is/was Indiana)
Keith M Wesolowski wrote: If you want to create an OpenSolaris reference distribution, or any distribution that advertises itself as having that status, Why does Indiana have to meet these requirements when Nexenta, Belinix, MartUX, Schillix, SX, and all the other OS.o distros don't? Why are you throwing up logistical barriers to this effort instead of facilitating them? Or is it that we simply /like/ eating our young? -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: RE: backspace key not working on Java environment
If Sun continues with the [DEL] key being used to backspace-delete stuff, that's just being unreasonable. +1 15 to 20 years ago, we were our customers, and it was easy to map their needs onto our own personal preferences - after all, we were selling to enginerds just like ourselves. Those days are mostly long gone. Say goodbye to them and get over it. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [SVOSUG] Project Indiana, get the skinny direct from the source!
But att audio is good for that, no? What's att audio? AT&T is the conference call vendor. They used to be called SBC, PacBell and other names in the past. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: BASH as root shell
Gerard Nualla wrote: I see... Let me try that TCSH too.. Gets more interesting... :D Wanting to change root's shell implies that you are logging into Solaris as root instead of using su or RBAC. In general, on Solaris, logging in as root is seen as undesirable simply because it can lead to sloppy and/or dangerous behavior and actions. Better to login as yourself, with any shell you wish, and use su on those infrequent times that you need to admin the system. With that admonition aside (as with all free advice, it is worth exactly what you paid for it), there are still situations where you need to log in as root. It used to be that root's shell had to be /bin/sh so that, if your /usr partition got corrupted and you had to boot without it being mounted, you could still log in single user and fix things (/usr/bin/bash wouldn't be there in that case, and root logins would fail with a "no shell" error message - quite a bummer when you ran into it...) In S10/OpenSolaris, the system is smart enough to detect that failure case and fall back to /bin/sh if root's shell is not found, so it is perfectly safe to change root's shell to something else. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Unix wizard pissing contest...
Joerg Schilling wrote: There is no bashing, GNU tar is non-portable Pleeeze - GNU Tar is portable to everywhere that has GNU tar. For the purposes of this discussion, this includes ALL of the OS's we are talking about: Solaris, OpenSolaris, the OpenSolaris Distros, RHEL, Fedora and all the other Linux derivatives. Leaving aside for the moment its bugs and the mental state of the users who have - for better or for worse - decided to use it, the fact remains that the transition between a Linux environment and a Solaris environment is more difficult than it needs to be. We need to address that disconnect. There seems to be two migration paths that are possible: Have developers migrate from Linux to Solaris, or Have developers migrate from Solaris to Linux. Which do you see happening - both short and long term? If what is happening differs from what you would prefer, what needs to be done to bring the two into alignment? -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Unix wizard pissing contest...
UNIX admin wrote: if I can learn once, and use anywhere and if I can write once, and use anywhere ergo I am more productive But this community is /not/ the Solaris admin community, it is a development community focused on improving OpenSolaris. That means changing things - hopefully for the better. If all you want is the status quo, and for it to never change, you may be in the wrong community :-) This whole conversation is so last century; hell, I lived thru an identical one ~20 years ago, although at that time the players were stodgy old BSD -vs- the upstart ATT-USG; the "opensolaris" at the time was this new distro called SystemVr4 (aka Solaris 2.0) Go figure... Instead of spending so much TALK on why this or that status quo is better than the other, how about instead spending the EFFORT to improve things? -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Unix wizard pissing contest...
If I understand correctly, you are saying that a Solaris user can become a Linux user with ease, but not vice versa. Do you consider this to be a strength or a weakness of Linux? Neither. It's a strength of Solaris, in that Solaris breeds a mindset that is portable to HPUX, *BSD, Linux and many other platforms. You learn to work with a set of tools that are present on everything, as opposed to being dependant on particular features of a particular toolchain. A new Linux user would probably learn to use the more modern "ip" tool to manage interfaces, whereas as a Solaris admin would use ifconfig which will work on all of the above. And the awful stereotypes come alive! This sounds like a bad B-movie (is there any other kind :-) with a plot that combines the following cliche's o It was hard for me to learn, therefore it should be hard for everyone else to use, o The only tool I have is a hammer, therefore all my problems must be nails, and o Job security through obscurity If the more modern "ip" tool makes me more productive in solving my business critical problems, what value is there in being retro-macho and restricting myself to less efficient tools? Instead of "least common denominator" sysadmin, maybe it would be more productive to invest in improving the baseline. And isn't that what this discussion is really about? -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [SVOSUG] Project Indiana, get the skinny direct from the source!
Christopher Mahan wrote: What's the chance of your manager eating the cost of a recording/transcript then? At a starting point of $50/hour for audio to written transcript production, and going up beyond $150/hour depending on accuracy, timeliness and other attributes, probably not. I don't want to attend; I want to know what was said. Paraphrasing, "I don't want to expend the effort to participate in the community, but I want someone else to, so I can benefit from the things that the community does. Oh, and I don't want to be bored by the result, so make it easy and interesting for me." Putting it that way, it all starts to sound a bit selfish, doesn't it? Taking a step back, though, I see that the Local User Group concept is showing its age in the tiny digital flat world we live in. Or, maybe, it is an aspect of making this tiny digital flat world we live in a bit less impersonal and more approachable. In my opinion, an interactive user group session is very much more than simply a sterile recording/transcript of one part of a presentation - it is a mix of who you sat with, the side comments/conversations, the emotional recharge of being with other like-minded enthusiasts, the bidirectional feedback loop where your comments influence other people's future behavior, etc. You simply can't capture all that in an online read-only text transcript. Some people thrive on high volume aliases like this one; others live just to spread flame bait, while others couldn't give a damn about anyone who isn't actively committing code. Different strokes for different folks, y'know... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
initial draft 1-pager for Indiana (was Re: [osol-discuss] Sun to make Solaris more Linux like)
Brian Gupta wrote: So, I'd expect to see some documents from you that describe this product. Have you a 1-pager that describes the basics? Ok, I think I see where the vision is going to hit a wall. There is definitely a process disconnect between agile development and Solaris development. In agile development, you begin with a list of simple requirements without a detailed description of how you are going to get there. I don't see the disconnect. In fact, I think we are in complete agreement. A 1-pager is simply a template for recording a "list of simple requirements without a detailed description of how you are going to get there". How else can a group of developers even start down a path unless there is some common understanding of where they intend to go? Nothing in this "list" is intended or expected to be detailed or in depth; rather, it shouldn't take someone more than an hour to pull one together. In fact, here's my first pass on it. Anyone else - please feel free to take it from here and modify things to match your understanding. (yes, this cries out for a shared wiki - anyone?) -John Project Description Indiana is intended to be a product - an OpenSolaris distribution with a regular release schedule. Project Team: Needed... Risks and Assumptions The relationship between Indiana and the other OpenSolaris Distros is unclear, as its relationship with Sun's various Solaris releases. There is probably a dependency on the Emancipation project. Unresolved issues include release binding, compatability and stability constraints, etc. We assume that we can start with an existing distro and work on improving it rather than starting from scratch. Business Summary Problem Area Think of this as the next natural step in the open sourcing of Solaris that began in 2005. In other words, the source has been in the community for a while, and now we're moving the binary version and related machinery into the community too. Why? Because even in open source, it's the binaries that people want. Furthermore, we're not presenting OpenSolaris as crisply as we could be.. In particular, people familiar with how Linux works (and that's A LOT of people) hear the name "OpenSolaris", assume it's the community version of Solaris, and are confused to find out that isn't actually true. This is actually just one part of what I've been referring to as "the familiarity problem" (formerly known as "the usability gap" until I realized that "usability" was relative). I.e., that while Solaris has compelling technology, that technology remains somewhat inaccessible to users that are familiar with the Linux environment (and, again, that's a BIG market). Addressing the familiarity problem is another big goal of the Indiana project. In other words, the goal is to make what Solaris has to offer available to the larger market that by and large is more familiar with Linux as things stand today. Market/Requester Ian Murdock/Sun Marketing? How will you know when you are done? There will be a new "Indiana distro" available every 6 months. Technical Description So, what will be the big features in Indiana? You tell me--and, indeed, a discussion of features could be a great way to actually get off on the right foot here given the somewhat rocky start so far.. My list: packaging, installation, GNU userland alongside "Solaris classic userland", and laptop support (see what I mean that there are already people working on these things?). The big feature from my point of view though is the 6 mo. timed release cycle. Timed release cycles have done wonders to introduce predictability into other open source projects (e.g., Gnome, Ubuntu). And 6 mos. is the clear winner in terms of frequency among Linux community/developer distros--it's just enough time to do interesting work AND have a reasonably long hardening period so the thing is stable. Details In Scope Needed... Out of Scope Needed... Packaging & Delivery Needed... Dependencies Needed... Resources & Schedule Prototype Availability Using agile development methods, we would like to start a sequence of develop/feedback cycles to refine this proposal, validate our initial assumptions and clarify constraints. We expect to form an initial project team, select a base distro and do some initial experimentation within the next couple of weeks. ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
And that would break... what, exactly? (Re: [osol-discuss] Sun to make Solaris more Linux like)
Ian Murdock wrote: > And that would break... what, exactly? We don't know. We know it has the potential of breaking scripts that, for better or worse, parse the output of "/bin/df". It can (and has been) argued that those scripts are already not portable to Linux, since the output is different there. This implies that the portability argument is rather aimed at the existing Solaris user base as they upgrade their systems to newer versions of Solaris, and this change would make it harder for those customers to upgrade their systems. In some sense, this analysis doesn't matter because we already promised our existing customers that we would /NOT/ change these interfaces out from under them in future Minor releases of the ON5.x component of Solaris. Note that we (the community and the ARC) could easily approve a case NOW to change df's default. The ARC opinion for such a case would say "This project ... is approved for integration into a Major release of the ON Consolidation". The question then becomes one of what this "approval with an integration constraint" really means? Since the Nevada development gate was chartered as a Minor release, by our own rules, we can't simply change interfaces that we promised would be Stable. This means that we can't allow this kind of change to go into Nevada. However, it could go into a Major release chartered gate, if we had one. We don't have one today, but we could get one by either A) Re-chartering Nevada to be a Major release gate, or B) Leaveing Nevada as it is and chartering a new Major release train (Indiana, anyone? :-). With a Major release gate, the project team would now have a place to integrate their incompatible changes. Of course, the ramifications of putting a Major release train into motion are complex and non-trivial: One big danger of a Major release is that the result might be inappropriate or unusable by some segment of our existing customer base. This is balanced against the danger that restricting ourselves to only a Minor release might be inappropriate or unusable by some segment of our potential new customer base. Do you want the Red pill or the Blue one? -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Sun to make Solaris more Linux like
Andre van Eyssen wrote: It's a dose of hell for anyone learning to write scripts. I don't believe we need an envariable to enable this for several reasons: A) envariables don't scale - they are per-user, per-system, per-problem band-aids. Over time and over systems, this path leads to chaos. and, more importantly, B) we already have precedent in the system that this is an OK behavior: % uname -a SunOS sac 5.11 snv_56 sun4u sparc SUNW,Sun-Fire-V440 % /bin/ls Old sr.20070223.txt sr.20070406.txt sr.20070518.txt sr.20070119.txt sr.20070302.txt sr.20070413.txt sr.20070525.txt sr.20070126.txt sr.20070309.txt sr.20070420.txt sr.template.txt sr.20070202.txt sr.20070316.txt sr.20070427.txt sr.20070209.txt sr.20070323.txt sr.20070504.txt sr.20070219.txt sr.20070330.txt sr.20070511.txt % /bin/ls | cat - Old sr.20070119.txt sr.20070126.txt sr.20070202.txt sr.20070209.txt sr.20070219.txt sr.20070223.txt sr.20070302.txt sr.20070309.txt sr.20070316.txt sr.20070323.txt sr.20070330.txt sr.20070406.txt sr.20070413.txt sr.20070420.txt sr.20070427.txt sr.20070504.txt sr.20070511.txt sr.20070518.txt sr.20070525.txt sr.template.txt % -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: Multiple development trees... Re: [osol-discuss] Re: About Project Indiana
Josh Berkus wrote: Frankly, I don't see the opensolaris community doing backport work at all. Why would we? Backporting is for supported versions, where (presumably) support customers will pay for someone to do the boring backporting work. I guess I'm not seeing why this is a problem? Both Brian's and your messages seem to imply that the developers doing the backport work are somehow not full members of the opensolaris community - I'm obviously not too comfortable with such an exclusionary perspective. I'm worried about us developing into an OpenSolaris community where Sun's engineers are relegated into a small corner where they only work on Sun's "Solaris" additions in an OpenSolaris fork, while the rest of the community sees itself as being predominately of and for non-Sun-employed developers. Witness the "it was said/done by [EMAIL PROTECTED], therefore it must be a conspiracy to undermine OpenSolaris" type comments that have cropped up several times in Ian's Indiana thread... I'm also worried that we at Sun haven't yet figured out how to make a product/distro based on OSCE. IMO, we have all the enterprise stability OSEE mindset we will ever need - so much so that it is blinding us to all those new OSCE-based opportunities. Maybe that is what Indiana is intended to be. Its late; my spring allergies are raging and the kids have all gone to bed - I should follow their example myself... Maybe things will look better in the morning :-) G'night, -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: Multiple development trees... Re: [osol-discuss] Re: About Project Indiana
Brian Gupta wrote: OSEE - OpenSolaris Enterprise Edition (classic Solaris) OSCE - OpenSolaris Community Edition (swiss army knife distro) Please don't take my comments as throwing cold water on your strawman; rather, try to use them to help drive a deeper common understanding. I agree with you - we really need to have this kind of discussion. I would say you have one main branch (let's call in OSM: OpenSolaris Main). the OSEE and OSCE distros would take what is in that main branch and make modifications to that base. (Preferably just adding to it.) In poor "word pictures", I think we have this today with either (a true source code management system view) OSM = Solaris10RR baseline as of June 2005 | +- OSEE = Solaris10 update releases +- OSCE = OpenSolaris-Nevada (Minor Release) or (a cherry picking model requiring manual back porting of desired features from Nevada) OSM = Solaris10RR baseline as of June 2005 | +- OSCE = OpenSolaris-Nevada (Minor Release) | +- OSEE = Solaris10 update releases (pseudo-child of Nevada) +- Solaris Express +- The various existing OS.o distros One could evolve either of these into a more radical scenario where we charter a new Major release: OSM = Solaris10RR baseline as of June 2005 | +- OSEE = OpenSolaris-Nevada (Minor Release of ON5.10) | | | +- Solaris10 update releases (pseudo-child of Nevada) | +- Belinix | +- Schillix | +- Martux | +- OSCE = OpenSolaris-1.0 (Major Release) | +- Solaris 3.0 (aka SunOS 6.0...) +- Nexenta (Don't ask where Indiana fits here - I haven't a clue. As it is, I'm probably maligning Moinak, Erast, Joerg and Martin :-) When any addition is to be made to a OSEE or OSCE, it must be evaluated for integration into OSM. (With input from the community) In the ARC world view, part of this evaluation is to see if the "scope of change proposed" matches the target's "scope of change allowed", based on the expectations set by the project that first introduced the things being changed, the interface and release taxonomies, and which (if any) things are going to be changed incompatibly. Think of this as: You promised us that XXX would exhibit , and now you wish to break it. The "magic decoder ring" says you can do so only in a release tree. This means that, depending on their release taxonomy bindings, changes that are allowed in OSCE might not be allowed in OSM or OSEE. (duh! :-) I would expect ISVs to port their apps to OSes that Sun distributes and supports. Today Sun distributes and supports Solaris 8 Solaris 9 Solaris 10 Solaris 10 update 1,2,3,4... Solaris Express Solaris Developer Express Historically, ISVs (and Blastwave, too :-) tended to support only Solaris 8, counting on binary compatibility to let it "just run" on S9 and S10. I'd guess that the number supporting S10 is still ramping up. I wouldn't expect anyone to be offering SX or SDX support at this time, though I assume that many are playing with it "in house". The $64K question is whether any of them would support a non-Sun distro; just as interesting is whether or not any of the various distro-producers would care and/or whether there was any expectation of compatibility between distros. You wouldn't need an entire team, as a lot of the work would be in OSM. The thought would be that Sun developers would focus on OSM, and OSEE, with a few bodies dedicated to OSCE. Most of the community work would be done in OSM and OSCE, with the goal to backport any universally useful OSCE changes into OSM, so that OSEE can leverage that work. When I read your comments above, I get the impression that there will be two rather disjoint communities - the Sun- dominated OSM/OSEE one and the non-Sun dominated OSCE. If this is how it works out, what would motivate the OSCE crowd to do the backport work, especially since it would undoubtedly involve more work? -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Sun to make Solaris more Linux like
[It seems that all I have today are questions...] Ian Murdock wrote: It's more the developers (largely in open source projects) deciding what "the standard" is as a side effect of writing their code.. How do we adapt to the new reality? Some thought experiments, building on Mike Kupfer's mail: Given that Linux is incompatible with Solaris in many ways, yet ISVs and customers happily use both, is portability across both platforms really a key issue? In the same light, Posix/SUS/XPG/... all seem to talk to compatibility across alternate branded platforms, yet I don't believe that Linux claims to be one of them. Is this whole standards branding thing more of a Solaris compatibility issue, distracting us from the assertion that some customers no longer care and would rather have "just like Linux", whatever that means :-) ? Both Linux and Solaris have features to die for; conversely, both have features I'd rather die before I'd use them :-) Is there a best-of-both world that we could/should strive for? Or is "being different" (from either camp) a kiss of death? What would be the practical downside of doing a Major release of ON/Solaris/OpenSolaris? Especially if, with Zones and Xen and ... customers could have both the existing Solaris2.10 and the new Solaris 3.0? -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Multiple development trees... Re: [osol-discuss] Re: About Project Indiana
Brian Gupta wrote: I proposed in an earlier thread that there I felt there should be two products/distros developed within OpenSolaris.org. - OpenSolaris Enterprise Edition (classic Solaris) - OpenSolaris Community Edition (swiss army knife distro) This is a simple idea that gets hard rather quickly. Coming up with a roadmap/plan that describes what is intended is where we have always run into problems. Some of these imponderables are: Given that we want to have multiple distros, how do we expect that they will be developed? What are the relationships between them? Do we intend to have independent branches, parent-child feature waterfalls, cross-dependent peers, no relationshoip at all, or something else all together? Given a relationship, what is the desired endgame? Does the classic source base go the way of SunOS4.x, becoming a frozen, bugfix-only, no new features platform? Is this simply a request that we do a Major release of Solaris instead of the sequence of compatible Minor releases that we have been doing for the last 15 years? Which of the development trees do we expect ISVs to port their applications to? Both? - if so, are we sure we all understand what drives ISV adoption and customer deployments? Who are the target audiences for each of these releases? If we do this, and it is successful, what does the world look like in 2 years? 5? 10? It would be Sun's digression to support and/or productize either version. (Also, they of course retain the right to make changes when productized, but my hope would be that Sun would propose their changes within the community). Once we have committed to having multiple development branches, we need to figure out how we can avoid doubling the development efforts needed to sustain the code (for bugfixes and patches, feature backports, etc). Of course, it is in Sun's best interest to try and ensure that the work-product of the community aligns with its product plans; IM(NS)HO, it would be a disaster if Sun was forced staff an entire porting team just to maintain its own divergent fork of OS.o ... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] BeleniX meets Indiana (Was: About Project Indiana)
Joerg Schilling wrote: Well, Belenix is a project run by a Sun employee. Do you believe it is not a Sun project? Yes, I believe it is not a Sun project, if, by saying "Sun project", you imply that Sun management made a decision, allocated resources, or otherwise caused it to happen. Of course, this all may change if the idea of BeleniX meets Indiana takes root... As Moinak said, Belenix is a labor of love, driven by a small handful of engineers who take significant chunks of their own personal time to make the project successful. It is not their day job to do so; their job descriptions do not include "develop and work on an OpenSolaris Distro"* Even so, after spending three weeks working in Bangalore, interacting with their management teams, I can say that everyone I met was exceedingly proud of what they have done! -John [*] That's what I love about working at Sun - a job description is simply something to get you started; implicit in every development job is the assumption that you are in charge of updating and rewriting your own job description to take advantage of innovation and change. ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Sun to make Solaris more Linux like
Dennis Clarke wrote: You forgot to add that the Bearded Man has a 72 processor super computer class machine running while the young man has a Dell PC. Yeah, but that young guy actually went out on a date instead of spending the might with a computer... Ha! All you Unix eunuchs take /that/! :-) ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Change charter of OpenSolaris.org to more closely align with reality
Manish Chakravarty wrote: Does this mean that if some developer comes up with a kernel patch (just an example) or some driver, he can commit it to OpenSolaris? After they articulate the architectural impact of their changes, and that impact is determined to be acceptable, certainly. After all, we don't want crap dumped into the source base just to prove that anyone and everyone can commit anything they want at any time. And if the community (not SUN) finds it good, it will be taken in and will be put up in the next OpenSolaris build ? It is rather difficult to separate "Sun" from "the community". If you mean "Sun = corporate VPs setting strategic directions for OpenSolaris", that has already happened. On the other hand, if you mean "Sun = the various developers paid by Sun to work on OpenSolaris", there is no chance in hell of that happening :-) DTrace without Bryan or Mike? X11 without Alan? Docs without Michelle and Susan and ...? Install/Packaging without Bart, Sarah and crew? Not very likely. On the third hand, if you mean "Sun = the various review processes that stand between the source code and total anarchy(TM), then you may have a good point. These processes have evolved over the last 15 years to meet Sun's needs. Now those needs have shifted, changed and grown. What worked well in an environment where people could meet face to face, have regularly scheduled meetings, etc, no longer fits so well in a massively distributed open source environment. This is not to say that we should abandon those processes, but we should work to reinvent them so they work well for us today. What about the bundled software that comes with Solaris Express Community/Developer releases ? Is there any way the community can influence that ? What about the bundled software that comes with Schillix, Nextenta and Belinix? Is there any way the community can influence that? Are these OpenSolaris.org questions, or are they really aimed at the various distros that are built on top of OpenSolaris core technology? I would suspect the answer is not an absolute - it is probably a "yes" if you are talking about "will changes I make to the core ON, X, SFW and JDS components show up in the distros?", and more like a "I don't know, or maybe no" if you wish to influence the addition or removal of other specific components. What if someone has his/her own OpenSolaris based distro ? Can he/she make modifications to it as he/she pleases ? What if such modificiations break API or ABI compatibility with Solaris 10 / OpenSolaris ? Certainly you can do so. Nobody will stop you. And, if you do break ABI compatibility with Solaris, that is your choice. Just please don't imply that you are still compatible and/or complain when various things no longer work with your distro. If you choose to diverge, you are also choosing to deal with the consequences. Who knows - maybe you will be wildly successful and Solaris will follow your lead :-) These are some questions the community must ponder on. Absolutely. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] OpenSolaris Filesystem path standards.
Alan Coopersmith wrote: Brian Gupta wrote: I can't seem to find the document that defines where various things are supposed to go in the filesystem. http://www.opensolaris.org/os/community/arc/policies/install-locations/ Or, on Solaris, % man filesystem -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Re: Re: GPL in BusinessWeek article
W. Wayne Liauh wrote: Oftentimes because an Express release has worked so stably that we forget that it was only intended to be a beta. It is probably time to define some terms, because we may be talking at cross purposes. In the "non-Solaris" world, Beta means something like "The engineers are done adding major features, the product is close to what is supposed to be, but there is a quality, testing and/or customer feedback gap that prevents it from being tagged as a release candidate" Many years ago, Solaris engineers and product managers decided that this mode of development was sub-optimal for Sun's massively distributed and co-dependent development teams. Specifically, it meant that any time that the gate was not in a coherent state, some significant part of Sun's distributed development team would be inconvenienced; further more, every time the gate was allowed to regress from the release quality requirements, it guaranteed that more preventable bugs would escape out of development and grow into customer-discovered problems. This lead to the current Solaris' goal of having the consolidation gates be at release quality levels all the time, and is why the gatekeepers are so quick to back out integrations that break that policy. Of course, if /every/ build is a release candidate from a quality perspective*, then there must be something other than quality that determines whether or not to make a formal release. In Sun's case, that determination is made based on a large number of criteria, including customer demand, ISV and VAR readiness to re-certify, support load costs, migration/transition issues, brand management and, of course, the old "what are the competitors doing" benchmark. The Solaris Express and Developer Express releases are just as much products as Solaris 10 and its Update progeny. The various differences lie in market positioning and customer segment targeting much more than any quality concerns. -John [*] - of course, just because we try to keep things perfect all the time doesn't mean we are always successful at it :-) ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Is there a process document describing a project lifecycle from proposal to final completion?
Brian Gupta wrote: I hope this is the right forum for this question. There is, but you will quickly run into a difference of opinion about the definition of the word "project" - it tends to be severely overloaded. In the OpenSolaris.org constitution Community/Project sense, a project stretches from an idea endorsed by someone all the way to some amorphous affinity-based community of interested people. When I asked the OGB to think about formalizing this lifecycle, they demurred, preferring to work on simpler things first. In the development cycle sense (which many of us, but certainly not all, think that the above should map closely to), a project is simply the term used to identify the collection of planning, discussion, development and integration that goes into making some change into a source tree. /This/ development project lifecycle is extremely well understood, and can be found documented in both the Nevads/ON- and in the ARC- community pages: http://www.opensolaris.org/os/community/on/os_dev_process/ http://www.opensolaris.org/os/community/arc/ http://www.opensolaris.org/os/community/arc/handbook/arc-dev-process/ http://www.opensolaris.org/os/community/arc/handbook/sun-reviewprocess/ -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Micro-optimizing for the wrong thing - coding efficiencies -vs- processor speeds
UNIX admin wrote: And we've progressed... how exactly? Look at Google Sketchup for a great example of the virtues of being able to spend developer energies on 2nd and 3rd order features like intuitive ease of use and great tutorials rather than on counting bytes in rendering subroutines. It is all about costs. Back in your C64 days, hardware was expensive and coding time cheap - or at least cheaper than non-existent hardware. Now, with a couple of 2Ghz processor cores and a few GB of RAM, not to mention 3d hardware graphics engines, outboard IO processors and ubiquitous network bandwidth on a $2k laptop, look at what Apple has been able to produce. Those 200 byte BASIC interpreters and 48 byte renderers etc are now relegated to custom silicon - BASIC-Stamps and GPUs - that are good enuf for the rest of us. Case in point: My first programming job was to support an engineering lab system - running a 4Mhz Z80 with 48KB of memory. It cost about $10k back in 1980. We spent significant time and effort optimizing code so that students could use the system to solve 5x5 and 6x6 arrays of simultaneous equations, instructors could manage grades and the rest of us could simply hack and have fun. Nowdays I can buy a Microchip PIC processor development kit with more processing power than that system for about $100; the all-in-one processor in it runs about $6 (I use several dozen to run my model train layout...). Our daughter has a $60 TI graphing calculator of her own that blows the socks off of that old Northstar Horizon system. Do I care that she isn't learning to microoptimize assembly code on a Z80? Hell no - she is off exploring trig, calculus, graphical analysis of complex systems, robotics and the like. So what if she "burns" all the resources of her Macbook doing inefficient things like Sketchup, iTunes, Java and Robotics? Or if I "burn" a whole PIC doing nothing but driving a few turnouts on my layout? Thats what they are for - tools to learn and build greater things. Besides, next year things will be faster and cheaper still. Is the glass half empty or half full? -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Re: CD burning in Solaris
UNIX admin wrote: My point is, quite simply, if we dumb everything down, I'm glad we've managed to dumb things way down in the last 20 years: No longer do we need to wire up plugboards to program the mainframes. No longer do we need to toggle front panel switches to bootstrap a system. No longer do we need to write in assembly language (or machine code or ...) No longer do we need to ... add your own list ... My point is that a good engineer can identify and remove gratuitous or unneeded complexity (or abstract it safely away behind an API) such that the rest of us can stand on their shoulders and reach much higher. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] How "stable" is b62 compared to SXDE?
Manish Chakravarty wrote: How stable is b62 compared to SXDE? SXDE is "simply" b55-b with additional condiments, so I'm assuming your question is really along the lines of "how do b55-b and b62 compare"? And "Which one will crash less often?" (If you were instead just trying to identify the differences between the two, SXDE bundles more stuff on its install media than you get with the bare-bones download of B62.) OpenSolaris continues the Solaris "tradition" of trying to be release quality all the time. Sun's QA and PIT teams test each build (B55, B56,...,B60, B61...) to ensure that it meets their quality criteria - which include "not crashing" on a large and varied set of test systems. If a build doesn't pass those tests, bugs are filed and (if deemed appropriate) the build is respun, either with the bugfix or without the new feature(s) that caused the failures. This is where b55 turned into b55-b; it was a respin of b55. From that perspective, b62 is a reasonable replacement for b55-b - it passed the same QA and PIT testing, it has many bugs fixed that were open in b55-b, and it has new features. I'd say that it is just as crash-free as b55-b, if not more so. On the other hand, you should look at the list of open bugs in b62 to determine whether or not one of them would be a show-stopper for you. The changelog and flag day pages are also good indications of what is happening under the covers; they can be used to find out what b55-b bugs were fixed in b62... -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Virtual Console new release available NOW!
LingBo Tang wrote:
- What about adding an option ("on" by default) to use the first line of
the virtual console to hold the "window title" ?
This idea is quite interesting. But what will be displayed in that
position? I'm afraid there is no sense to display "console #" for end
users.
Since you already manage backing store copies, how about taking this concept
and making a navigational aid out of it:
When you switch to a different virtual console, display the text
"console #xx" in row 1 column 65-75 for 1 to 2 seconds, then
replace it with the characters from the backing store.
Extra credit awarded for smooth fade-out :-)
-John
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Packaging issues - was Re: [osol-discuss] xpg/bin/tr unexpect output on Sparc?
Peter Tribble wrote: No, we have far too many packages. As background for anyone wishing to address this perceived problem, the theory goes like this: Developers develop features by writing source code. Features evolve at different rates, have different resources available and are parts of different dependency graphs. A component is made up of many features Gatekeepers compile all the source code for a component and produce one or more packages Multiple components are combined to produce products Release Engineering builds a product by selecting a set of versioned components. These components are in actuality simply lists of versioned packages. You are asking good questions about the granularity of these packages. Lets play with the granularity control: What would happen if each component (like ON) were to be delivered as a single large package? I believe chaos would be the result: A) patching and updates would be cumbersome, as any change to a component would require the redelivery and re- installation of the entire component. B) Coordination and testing tends to stratify into a bottlenecked linear process - everything in a component must work before the component itself can be used. C) Developers tend to be drawn into a centralized, waterfall style development model, with a schedule that drags out based on the slowest/poorest performing development team. D) Users/Admins are faced with an all or nothing choice. The other end of the spectrum is just as bad - millions of little packages, all alike, each one a dependency-garnering minefield to trap the unwary. The middle ground we try to walk is to provide packaging boundaries for subsystems and major features that enable distributed and asynchronous development teams. Or, looking at it the other way, packages map to the output of an independent, distributed development team. They are the ultimate "consolidation" boundary - everything within a package is delivered together as a unit and the package is expected to always be self-consistent. Sometimes we screw up, breaking a subsystem up into way to many pieces - or not enough. Othertimes we have no real choice because other constraints force our hand - if the subsystem needs to work with zones or diskless clients, it needs multiple packages. Whether it is targeted at developers or at end users also plays a role. Just my $0.02 -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
