when i write in my blog, it's coloured by my personal knowledge and sometimes i forget to include things that to me are obvious given my day-in-day-out involvement with these things but which aren't so obvious to others. when it comes to the recent xgl issue, i realized that i had not included enough detail after reading this response.
the writer of that "editorial" bit is wrong on several points. first off, XGL is not a new project started by Novell. it was an existing project that Novell hijacked by hiring some (but only some) of the developers and then taking all their work in house, where it is both duplicating and blocking the efforts of those still working in the open. therefore the "getting something working before opening it" argument is not accurate at all.
next, he (i assume it was a he; it could be a she though) compares XGL to luminosity. they aren't the same thing at all. luminosity is not an xserver, it's a layer that runs on top of a given xserver and intercepts the various calls xfake style.
he also says that KDE has not done anything towards this work. he ought to read up on zack rusin a bit. he's a kde developer who has contributed to X and has worked on XGL. more over, we would like to work with this technology as well as consumers of it.
he also tries to frame it as a kde and gnome issue. it's not. red hat, a gnome shop, is as out in the cold as kde is on xgl. we are all losing out on this one.
finally, i find the dismissal of the utility of open development to be vastly concerning. here's a hint: successful open source projects are developed in the open. many otherwise good projects have (and are) failing specifically because they aren't.
as for what KDE is doing about these things .. well, we would like to be involved but we aren't going to be idiots and try and fork off in yet another direction. just because some people feel that's an acceptable response doesn't make it a good one. more productively, we're working with technologies that will benefit from hardware accelerated x servers that will hit end user desktops with kde4. kde will not fall behind in visual presentation. that's not my concern. i just want the open source desktop not to be slowed down (or worse) by the self-centered actions of a few.
so .. what am i hoping for out of all this? best would be if xgl development opened up. second best is if our partners and user base understand what is going on so expectations are reasonable and so they can vote with their feet should they decide to.
a final note: if you're going to "editorialize" on something by writing an aritcle for publication, have the decency to post non-anonymously. identity is the basis of trust, and if you don't feel confident putting your reputation on the line for what you have to say in an article then maybe you should think twice about saying it. =)
Subscribe to:
Post Comments (Atom)
25 comments:
And yet Novell hasn't said anything about this. It still feels like a big rumor to me :\
Mark Czubin
it's not a big rumor, it's what's actually happening. it's common knowledge amongst those involved. i've even talked with people involved with this in novell.
i have no idea why novell isn't saying anything. perhaps they hope it'll just blow over. perhaps they don't care. perhaps they think it's unimportant and don't wish to dignify it with a comment. perhaps they are busy figuring out what to say.
it doesn't change the fact that what i wrote is actually what is happening.
Pseudonymity and anonymity are problems the Internet have in general. Just look at Wikipedia. People just get a stick up their ass about privacy.
a final note: if you're going to "editorialize" on something by writing an aritcle for publication, have the decency to post non-anonymously. identity is the basis of trust, and if you don't feel confident putting your reputation on the line for what you have to say in an article then maybe you should think twice about saying it.
i just lost what little respect i had for you. this sort of backward thinking has no place in any modern internet community.
Well I posted comments on both Xgl articles on osnews, I got the big picture.
I kind of cheer Novell about Xgl, just not it's openess.
And it just annoys me that Novell isn't going with this to the community. At least let us know they are doing it and ask for input even if the development isn't open.
PS: I post anonymous because I don't want to register to everything ...
Mark Czubin
Aaron: Did you try to contact Novell about it? (Or, anyone else tried?)
Maybe that's the way - public letter with a request to open Xgl development?
P.S. Could you find a second to answer my email (the one I sent to you few days ago? :))
I've always felt a fair amount of disdain for those who use a cloak of anonymity (or a pseudonym, or whatever) when expressing opinions, and couldn't quite put my finger on why I felt it, but you've articulated it nicely. Thanks :-).
"i just lost what little respect i had for you. this sort of backward thinking has no place in any modern internet community."
first, i'm going to go out on a limb and guess you had no respect for me to begin with ;) and that's fine; my job on this planet is not to earn everyone's glowing admiration.
but furthermore, i have no problem with privacy and anonymity in the general case. i use gpg with my email to be able to have privacy, for instance. and i think being able to comment on things anonymously is of tremendous value.
but writing a feature article and publishing it anonymously is something else entirely in my mind. we are not a community where you need to protect your identity for fear the police will knock on your door, but we are a community based on participation and reputation.
when someone editorializes anonymously on the front of a website as that individual did, there is no measure of responsibility involved. there is no reason to just wank off and spout any utter crap you'd like because there is no ramification, there is no check-and-ballance.
that's reality, and it shows in the general quality of such anonymous articles in general.
if it's not good enough to put your reputation behind, it's not good enough.
to repeat: anonymity is useful, valuable and has it's place. community facing editorials is not one of them.
> Could you find a second to
> answer my email (the one I sent
> to you few days ago?
since you posted anonymously, i don't know who you are, so it makes it kind of hard to know which email you are referring to.
i get a few emails a day as you can imagine =)
By the way, thanks for starting the discussion about the Xgl for the first time since Mr. Smirl quit the Xegl.
If your blog just makes Novell give us a newer Xgl (in the CVS) for all us eye candy addicts to play with then it was all worth it!
Hi Aaron
I remember reading a mail from David Reveman on the xorg-mailinglist that he really wanted to put his latest code in CVS - it was may be in August or something.
I'm sure you have done more research, but from what I read, it more seemed the kind of thing: didn't get to do it (or translate it to - I have more important things than pushing it to cvs (which I believe to be stupid, but anyway))
That's why I'm asking - did you get in touch with David and he told you he shouldn't push it to cvs or were there other answers?
Benjamin
Still, it seems some people did get your point after all, if you browse the thread on 4 :)
On the other hand, I would be interested in your comment on this:
http://osnews.com/permalink.php?news_id=13061&comment_id=76903
Just curious...
when someone editorializes anonymously on the front of a website as that individual did, there is no measure of responsibility involved.
and editorializing with a pseudonym would somehow involve more responsibility? i don't think so. everything on the internet is anonymous unless it's signed with gpg or something similar.
if it's not good enough to put your reputation behind, it's not good enough.
if it needs the author's reputation behind for it to be worth considering, then it's not good enough either.
anonymity is useful, valuable and has it's place.
and the internet is it's place. "on the internet, no one knows you're a dog."
> and editorializing with a
> pseudonym would somehow involve
> more responsibility
if you use it consistently, then yes. when "mango parfait" writes a column, i know what to expect and where "she" is coming from, even though i don't know who "she" really is. but i have her past track record, and i can judge future installments based on that.
you may say that it's so easy to fake, and you're right ('on the internet nobody knows you are a dog') and yet ... in practice identity through name is immensley powerful and useful.
people read this blog because it has my name on it. and my name represents the sum of all i've done under it. and that's enough for people to have "little respect" for me and to even lose those few shreds of it ;)
so ... it works. and it matters.
another case of practice vs theory i suppose =)
http://osnews.com/permalink.php?news_id=13061&comment_id=7
jon smirl is a mal-informed malcontent. he keeps harping about EXA which has nothing to do with XGL. they are orthogonal efforts and the guy who worked on EXA (zack rusin) is actually in favour of XGL and one of the people who has patches that can't be applied.
i know this because he's sitting about 3 feet from me right now.
no, jon's just got an axe to grind and a unique POV on things to along with it.
> he also says that KDE has not done anything towards this work. he ought to read up on zack rusin a bit. he's a kde developer who has contributed to X and has worked on XGL.
Does this mean that Zack isn't working on this anymore? Has he moved from Trolltech to the in house Novell dev.team?
> Does this mean that Zack isn't
> working on this anymore?
he's working on related things, but not directly on XGL, no.
> Has he moved from Trolltech to
> the in house Novell dev.team?
no
When you start talking about identities with regard to editorials, it becomes clear that what you want is to either prevent discussion through shaming them, or engage in character assassination. The strength of an argument does not rely upon the identity of the people making it. The flaws within that editorial can be presented entirely without the need to rely upon the identity of the author, as if even knowing his name would transfer any sort of knowledge to you without either previously knowing his identity or attempting to collect information about him. Do you really care who that person is? Do you care who I am? Do I care who you are?
Whenever people start to talk about the accountability of online identity, I always think of people that buy tabloids at the checkout line to find out what Brad Pitt thinks about sweaters, or people that write comments dismissing ar essays from Eric Raymond simply because he's otherwise crazy.
> it becomes clear that what you
> want is to either prevent
> discussion through shaming them,
> or engage in character
> assassination
wrong again. thanks for playing "accuse aaron of random shit", but your time is up!
> as if even knowing his name
> would transfer any sort of
> knowledge to you without either
> previously knowing his identity
> or attempting to collect
> information about him.
*sigh* could you miss the concept with any greater accuracy than this?
see, here's how "reputation" works: when someone does or says something, it usually makes even more sense taken within the context of where they have traveled.
that's why someone who gets arrested for a crime usually gets less of a punishment if its their first time than if it's their third time.
that's why someone who repeatedly does great things earns respect and fame versus a one-hit-wonder.
that's why when jon smirl starts bitching about EXA and XGL in these blogs i know to read in the context of him grinding his axe.
that's why when you read what i write you know what sort of attitude and bias to expect.
so yes, knowing someone does impart knowledge due to context and pattern of behaviour. that's the whole point of identity. incidently, that's why we sign GPG keys so carefully and build webs of trust with them.
and when one knows they are building a pattern of behaviour which will colour their future encounters, it makes one think a bit harder about what they say and do. it's easy to shoot off at the mouth when there's no long-term affect of that. but when you put something valuable, such as your reputation, behind your words you might just check those facts twice and think about your conclusions instead of flatulating in the direction of a microphone.
identity creates responsibility and increases the signal ratio. that's why identity creates community; random anonymous spurtings don't.
btw, i'm all for personal privacy. i'm also for personal responsibility. and those two things can go quite nicely hand in hand.
Considering how Wine direct-X development went,
I would not be surprised at all if in 2 years we (the community) gives up on Novel and writes our own.
Frankly, I think those who were working on this before Novel should start hacking based on what we have, and then put the burdon on Novel to merge our changes into their code. Force the issue, if we start hacking in incompatible ways they will have big problems.
Ideally though this will get attention on slashdot, and the negative publicity will force Novel to do the right thing. Apple opened up their fork of kthml after we did it, so this might happen. (Though apple was never this closed either)
It really sucks to not have great 3d drivers. It is really bad if you don't run x86, or linux. I've been holding off upgrading my desktop for years because of this issue.
-- Hank Miller
http://osnews.com/permalink.php?news_id=13061&comment_id=7
jon smirl is a mal-informed malcontent.
Thanks for the info :) it is a pity he sounds credible for the uninformed - like most of us (users). Which reminds me: it would be wonderful if someone wrote up a summary of what's what (xgl, xorg, exa, etc...) - perhaps on kdemyhths (in conjunction with KDE does not contribute FUD)?
I'm a daily reader of planetkde, that's why I asked you, but most people won't, and will just spread these unfounded rumours...
Novell is doing the right thing. Sometimes the open source model fails and this is one of these times. Too much politics. Novell decides to do everything in-house by themselves to avoid the BS and egos that come with open source. Especially in the desktop community. I applaude Novell because they will actually deliver a 3d accelerated desktop within a year. While an open source effort results in flame wars on mailing lists and cock size competitions.
Well, it seems Dave Reveman and Novell have commited some (if not all) of the code. It has already been merged to kdrive.
--
japc
Post a Comment