Havoc replied to my earlier blogs regarding FD.o quite succincntly: we don't have to worry because FD.o isn't a standards body.
but hold on now... i never actually claimed that FD.o was a standards body! i simply claimed that it's where standards get made and that DConf was looking to become a desktop standard. that's the whole point of DConf. the title of this blog starts to make sense, doesn't it?
so let's examine whether or not FD.o really is about standards or not. there's a whole list of standards over on FD.o. and the main FD.o webpage says:
'freedesktop.org is not a formal standards organization, though some see a need for one that covers some of the areas we are working on [..] Unlike a standards organization, freedesktop.org is a "collaboration zone" where ideas and code are tossed around, and de facto specifications are encouraged.'
so Havoc is right, it's not a standards body. but it is where we have created many of our standards and continue to host the documents that define those standards. call it what you will, this is a dead parrot.
Havoc also said, "freedesktop.org right now is the equivalent of Sourceforge, essentially, except that projects have to be desktop-related." ignoring the fact that this is only true on a strictly technical level, the question is begged: is that what we need, another SourceForge? i mean, don't we already have three or four of those around? and if we're honest with ourselves, isn't that what most people are involved at FD.o for, to bridge between the projects? again, this isn't a SourceForge-for-the-desktop, it's a dead parrot.
there are couple of things Havoc nailed, though. first, that the "platform" is not about technology pimping. amen. but here's the real nugget of gold:
"While some people may have other ideas, I will again reiterate that KDE and GNOME can veto anything by simply not going along with it. That's the core reason I don't understand any paranoia here."
he's absolutely correct, we can simply decide to not use something. the problem is that if the FD.o collective puts out major pieces of software with the intent of bridging systems (and that's the whole point of DConf) and yet fails miserably due to flaws in the process, FD.o loses value. this is what i don't want to see happen.
it seems that both Waldo and Havoc think i'm afraid of FD.o pushing bad ideas on KDE or GNOME. but what i'm really worried about is FD.o becoming irrelevant. because then we'll have put ourselves right back to where we were all those years ago before we had a place to discuss and come up with standards, even if they were de facto.
22 comments:
Dude, first you write a dumb article like:
" How To Kill Open Source on the Desktop?"
Dumb the title and dumb the content.
Andnow you are spreding FUD about FD.o?
please, get over your self.
> Andnow you are spreding FUD about
> FD.o?
actually i've been one of the very vocal proponents of FD.o for quite some time, and am known for being so. i'm not out to do something bad to FD.o here. you'll note in the entry here that my concern is that FD.o is in danger of losing some of its relevance. i don't want to see that happen.
and it's not Fear, Uncertainty and Doubt if it's an accurate description of what's going on.
i know it may be uncomfortable to consider these things, but i'm not big into pretending everything's fine only to wake up one day to an mess that we can no longer fix.
> please, get over your self.
look, if i was doing this to stroke my ego i'd praise FD.o endlessly. my inbox, and people like you, say that this is actually the least egotastic route i could probably pick. but i'm fine with the occasional bit of abuse as long as it also opens the necessary conversations with those who are actually in the know.
btw, next time you post to my blog i would appreciate it if you would try and engage in the conversation rather than simply vent your random anger. =)
I offert you my opologises if I said something you didn't like.
I'd like to know that if you don't like where FD.o is heathing what do you propose to keep the interoperability betwen Free Desktops?
Hmm...
My challenge to Havoc would be the following:
"Read discussions of Gnome/KDE usability on sites such as OSNews. They frequently tend to mention FD.o and the Gnome users who mention it will declare that FD.o is a standards organisation.
Even if you might not agree, when presented with FD.o product, these users will consider that product as implicitly a standard and that understanding will propagate. Ultimately, FD.o becomes a standards body for defacto standards through engendered mindshare.
You are going to be seen as a standards body, regardless of your opinion, and need to take on the social responsibilities of that position."
In addition, the idea that no-one forces solutions that are technically or usability-wise inferior to existing solutions on users is erroneous. That is true in a full meritocracy, but fails in heirarchical development structures. Look at the number of users and developers who left the Gnome project with the migration to 2.0, which they felt was highly regressive; that pattern has been the hallmark of Gnome developments for the past several years, not limited to Spatial Browsing and re-built Nautilus in general (loss of emblems, etc.), file selectors without address fields, loss of menu editing, etc. etc. etc. etc. As a KDE user and application developer, I manage to avoid that, and the iterative improvement allows me to work and only have my productivity and experience improved. I'm not discomforted with each new release, but made happier still.
I like my KDE and Linux desktop. I love it, and I get lots and lots done. More than under Windows or OSX (which I use regularly). I don't want to sacrifice that at the alter of better 'usability' or standardisation that means losing the stuff I need because someone decides that it's better for me.
Give me a better config system that makes my life tangibly easier, and I'll be right behind you; 100%. Make my life as a user and developer easier.
> I'd like to know that if you
> don't like where FD.o is heathing
> what do you propose to keep the
> interoperability betwen Free
> Desktops?
my proposal is to fix FreeDesktop.org. i really, truly hope that is what we can manage to do here.
if it isn't fixable (which would only because the people involved decide they don't want to fix it, which may be because they like the way it is), then we do have a hard set of questions ahead of us.
we need some place to collaborate on common specifications and technologies. if FreeDesktop.org becomes irrelevant another initiative will take its place.
Open Source is resilient this way. unbreakable, even. but i don't want to see us waste time and effort or the valuable capital in the FreeDesktop.org brand.
Aaron, what exactly are you suggesting?
At the moment, as far as I can make out, you believe:
1. fd.o is a good thing.
2. People on fd.o lists are not universally clueful.
3. These people, whilst not clueful, are still capable of making software. The software will be bad due to the badness of the authors. People will believe this to be blessed by fd.o
4. a) No one will use it and fd.o will lose credibility
b) People will use it and everyone will have bad software.
I think this line of reasoning is extremely tenuous.... I can't see 4 b) happening. 4 a) is just wrong : there are plenty of dumb and failed things developed on fd.o, gnome.org and kde.org that no one uses. Until something is really blessed (which only happens when something is widely used), no one loses credibility.
So what is the fix? Make fd.o a closed organisation? Only allow blessed hackers to comment in lists or to volunteer? What exactly could allay your ridiculously overextended fears?
So I think this all ends up being bluster to cover for what sadly seems to recur every so often : the KDE victim complex - "Wah, they are going to make me use inferior objects!", "They stole our SuSE!". Sorry if this isn't how you intend to come across, but that is what it ends up sounding like.
> 1. fd.o is a good thing.
yes.
> 2. People on fd.o lists are not
> universally clueful.
this is true everywhere, but is not the center of my argument.
> 3. These people, whilst not
> clueful, are still capable of
> making software. The software
> will be bad due to the badness of
> the authors. People will believe
> this to be blessed by fd.o
and this is where you start to diverge from what i've written. perhaps i'm not being clear enough; if so, i'm sorry.
the point is that the people on FD.o are being insular and not behaving like an organization that is trying to make standards.
it has nothing to do with bad people or bad software. the software can be great but still be useless to the projects that most need or want something like it.
> 4. a) No one will use it and fd.o
> will lose credibility
> b) People will use it and
> everyone will have bad software.
those are two of the negative possibilities, yes. they aren't the only possibilities, however.
> Make fd.o a closed organisation?
did i say that? no. why? because that would be stupid.
> Only allow blessed hackers to
> comment in lists or to volunteer?
you're batting two for two now.
> What exactly could allay your
> ridiculously overextended fears?
"ridiculously overextended fears." *sigh* and your mother wears army boots. look, if you want to have a constructive conversation, do so. if you want to throw tomatoes, go somewhere else.
i've described what would allay my concerns, which are shared by MANY in the community, btw. and that would be to practice outreach, to not develop (de facto) standards-as-software within FD.o and return to being a group primarily concerned with cooperation and collaboration.
> the KDE victim complex ... Sorry
> if this isn't how you intend to
> come across, but that is what it
> ends up sounding like.
you are neither sorry nor are you being straight.
i'm not speaking about this being an endangerment to KDE. it isn't. it's an endangerment to FD.o and that concerns me.
i've said as much in these posting several times and i'm getting very tired of people such as yourself trying to deflect this anywhere they can to avoid looking hard into the mirror that is reality.
Outreach.
To whom? KDE and Gnome developers are on the lists. XFCE and rox and I think Gnustep are on the lists. I believe that Mozilla and oo.org do pay some attention too.
Who exactly is this outreach to be directed at? Samba? Wine (who also have project members on the lists)? And what form should it take? I don't think a lot of these other projects will really *want* people coming along *before* there is anything even there. At the moment, DConf is in the initial requirements stage : I think it would be great if Samba, Wine, and other projects could be involved. Spamming their lists at the "we have nothing" stage does not seem likely to generate results or goodwill.
Please, I'm not being funny here, please could you state some examples of concrete actions that would satisfy you? If this counts as a deflection, I do not know what would not.
Looks like you not only don't like DCONF, you want others don't like it too.
If DCONF if adopted for GNOME, XCFE and others KDE won't have other choice to adopt it too, so, looks like you are starting the batle already to avoid this adoption.
DCONF is not finished yet, I suggest you to waith to get finished before complaining.
Not another Monty Python fan?! Watch those battery-farmed Llamas at your barbecue!
Anyway, I think this problem can be tracked down to one thing - a lack of communication. What is Freedesktop? Is it a standards body? Does it want to become one? How on Earth do you become one? Does it produce standard implementations? All these questions spring to mind and everyone appears to say very different things. Think of some of the things that maybe led to the KDE APPEAL stuff here (a really good idea I think and a good starting point).
I would advise that you corner Havoc and ask him to describe Freedesktop in one sentence, and one sentence only. I'm looking for something like "...to promote communication, interoperability, standards and common technology between all the software projects within what one would normally think of as a Unix/Linux operating system". Don't let him get out of it, as working from that is key.
Don't put Havoc down too much though, as I think he really wanted Freedesktop to be a collaboration point first (even if he wanted something else later) rather than any recognised official body, so as not to put too much pressure on people and projects with sensibilities. Everyone has them. I think he wanted to turn the noise level right down on what would have inevitably occurred if he'd announced Freedesktop wanted to be a standards body of any kind.
Unfortunately, things have gone too far the other way now and people don't know what Freedesktop, or what anything like DConf, wants to be. Consequently you've got smoke and mirrors about what Freedesktop is or isn't and people using all the uncertainty to bludgeon the use of their own pet projects by the back door (or appearing to). From then on it's a downard spiral of accusation, counter-accusation and mistrust. See above for some nice examples. From there, Freedesktop, and any work that goes into anything there, becomes meaningless as everyone contracts extreme NIH syndrome. This happens everywhere.
Try and get Havoc, and people like Waldo, to stop that runaway train before the momentum is too great. I know open source people aren't too great on nailing things down too tightly, but that's really what needs to be done here. People need to know where they are. Once that is done no one can conceal any hidden agendas and we'll know if Freedesktop really is dead.
I just wanted to say that I agree with aseigo on this... I like the original idea behind fd.o but I'm not so happy with it being the dumping ground for everything the gnomes doesn't want to have in their cvs.
> KDE and Gnome developers are on
> the lists.
some are, yes. but until i mentioned LDB and Ian to Waldo, that was a complete unknown to the goings ons. now they're talking and that's awesome. but saying "we've got 2 Foos and 3 Bars" ignores the fact that there are actually hundreds of each. and as FD.o continues to spread out into more and more territory, it will be increasingly important to be able to tap into those hundreds and pull the appropriate people together.
> Samba? Wine
yes, i mentioned Samba. Wine is a great example too. CUPS wouldn't hurt. MoFo, OOo, etc. the whole desktop needs to at least be aware and have the informed opportunity to add something.
has anyone done a survey of what all these projects are working towards? or are we just working in our little patch of grass blind to the rest of the software pastures?
> Spamming their lists
yes, this would be non productive. but finding out who is working on configuration issues would be a good first step. that wasn't even done in this case.
nor has it been done in past FD.o efforts and it has disinfranchised a number of quality developers who otherwise would have been (and hopefully still can be in the future) terrific allies and contributors.
additionally there are the hard questions of "are we making standards, de facto or not" or "we are just writing some desktop software with not real standards goals". the message is very conflicted at the moment and this causes distrust where it shouldn't exist. i have defended and promoted FD.o within KDE for a good while now (ask people who were at aKademy in 2004 =) and i'm burning out on it because FD.o is making my job harder not easier. =/
> If this counts as a deflection, I
> do not know what would not
no, i don't see your comment as a deflection at all. thank you very much for that =)
@ segendum
very well said and i really resonate with a lot of what you wrote.
> I think this problem can be
> tracked down to one thing - a
> lack of communication
in part because i don't think everyone working within FD.o knows what they want to communicated. some don't know what they are aiming for, some don't really care, some do care and do know but are concerned about communicating it correctly. at least this is my impression from speaking with a number of people involved in various bits and watching FD.o's progress and discussions.
I wish people would give up the "You suck arguments!".
What really is awkward is that the point of FD.o at least to me, is to bring people together. It's a pointing to standards body. There are standards out there, they get improved and we point to them and bring these loosely related things under one roof and try and get things consistent.
What I don't get is, if one initiates a project at FD.o, shouldn't their first priority be checking out prior art? Find as many examples of others who've been down the same road. Then try and get these people to brain dumb, start examining all the things you get and putting things together.
It seems, at least in DConf, this didn't quiet take place, it's more like, lets quickly put together a bunch of folks and see what we get. I'm being cynical, but that seems like a lot of hubris to me.
>> I'd like to know that if you
>> don't like where FD.o is heathing
>> what do you propose to keep the
>> interoperability betwen Free
> Desktops?
>my proposal is to fix FreeDesktop.
>org. i really, truly hope that is
>what we can manage to do here.
Where is the proposal?
I still don't see a clear, precise statement about:
a) What is wrong.
b) What can be done to fix it.
The only clear thing I've been able to put my finger on is that there are some people who aren't involved in discussions about "dconf" that probably should be. (And even then dconf is at a very early stage).
Surely in that case the fix is to track down such people and point them somewhere relevant.
At the end of the day FD.o is just a loose grouping of people. If you think something needs to be done then surely the best thing to do is to either:
a) Do it.
b) State clearly and precisely what needs to be done so someone else can say "that's a good idea" and do it.
When saying "FD.o needs to do x" it is probably also important to realise that means "Someone at FD.o needs to do x" and that ultimately FD.o is just a bunch of imperfect someones all acting with good intentions and limited resources.
Aaron, I will say that it would be lovely to get as many projects involved as possible. And some extremely stupid statements have been made on the lists. (esp re particular implementation strategies. "over my dead body", "well go away then").
I think we do need a clear picture of what is acceptable to all parties in terms of implementation, orthogonal to this particular discussion.
AFAIK, acceptable strategies would be:
* Plain C with no glib, needed glib bits copied in ( with a definite no to GObject).
And nothing else. I would hope, given that Gnome has a hard dependency on Mozilla, that C++ without Qt would be acceptable too. But I do not know. I think people will hide behind "C++ abi is unstable. Ooh." Hopefully gcc4 can help...
But atm, plain C is the price for cooperation. But every time you can expect some to push for C++ and some for GObject. And then everybody gets emotional.
Leaving aside implementation, what I do not understand is this train of thought:
1. I know some interesting config effort is going on in the Samba project.
2. I support fd.o and want it to succeed because I do not believe standardisation is defacto evil, even if it does reduce the total amount of Qt style code in my system.
3. Oh look, fd.o is having its annual config system discussion. They haven't mentioned that Samba thing I noticed...
4. Rather than be involved in the process, by pointing out the samba config project details I know in an open way, I will create a series of melodramatic blog entries decrying fd.o and saying it is sure to become irrelevant.
If high profile community members like yourself take this tack, doesn't this become a self fulfilling prophecy, or at the very least counter productive?
I don't think the problem is a body such as FD.o but rather some of its participants.
It's not a problem that a bunch of developers think of merges, new ideas and all that. It is a problem however those ideas and changes seem to either just appear in code bases out of nowhere, or get recognition based on aggressive marketing rather than technical merits which address real concerns.
My main example would be when Waldo Bastian committed significant desktop file changes to KDE without prior discussion whatsoever, I believe it was the X-No-Show-In: GNOME nonsense or something similar (the least they could have done was make it less brain-dead and use X-Preferred-Environment: KDE) or something like that).
(from Waldo Bastian)
Re: X-No-Show-In
Hi Rob,
There are two keys. One is called "OnlyShowIn" the other is called
"NotShowIn".
I am sorry for not soliciting your feedback prior to their addition. To make
it up to you, I invite you to provide meaningful feedback to my posting on
kde-core-devel regarding KConfig, titled "Re: An analysis about a generic
desktop application configuration management system" of April 12 and my
porposal for KControl, including patch, titled "KDE 3.4: Control Center
scrollbars" of March 1st.
Cheers,
Waldo
Hi, I was just out blog surfing for detailed info on self help when I ended up on your page. Obviously I ended up a little off base, but I am certainly glad I did. If you wouldn't mind, I would like to post your link on my "favorites" page. Should you ever need it, there is valuable information on my site about self help.
I skim a lot of blogs, and so far yours is in the Top 3 of my list of favorites. I'm going to dive in and try my hand at it, so wish me luck.
It'll be in a totally different area than yours (mine is about mens male enhancement reviews) I know, it sounds strange, but it's like anything, once you learn more about it, it's pretty cool. It's mostly about mens male enhancement reviews related articles and subjects.
Nice Blog, My site is:
software
Post a Comment