Here are my rough notes from DebConf 10. Nothing should be taken as a direct quote of the speaker who is supposedly associated with the talk as my own thoughts are interwoven with their points, the person switches around between they and I and we all the time… I made them for my own reference and publish them in the hope you might find them interesting but they must not be relied on. Happy hacking :P
(Hmm, Advogato doesn’t process MarkDown. That sucks.)
<h1>2010-08-01 Sunday - DebConf Day 1</h1>
<h2>Intro to DebianDay</h2>
Free software means we control out computers. Interaction is part of what it means to be human. We make and use tools, that’s also essential to humans. More and more we interact through our tools. And as computers become more important, they effect democracy.
Free software has the potential for anyone to contribute. But having a libre license isn’t enough; OpenOffice for example is veyr largeplex, C++, and the freedom is not accessible. Free software is just the beginning of the idea that we should own our tools.
We need to change how we think about computer literacy. Debian people work together to build a fully free OS, but we continue to talk about USING a computer, and not IMPROVING a computer.
We want read/write tools.
When we teach anyone about computing, we ought to teach programming and customisation. Programming itself is typically designed by and for engineers. That’s useful and needed. Like reading and writing, we have specialists and even writing systems for writing fast, but we also have vernacular language for everyone.
The big next step is to make programming something everyone can and wants to do.
Other ways to contribute are submitting bug reports, writing documentation, organising things like DebConf.
Deebian is a great example of democracy for developers, an international community, run on a constituion (DFSG).
So.. Should Debian be the OS for computer literacy for everyone? OON, Debian is for technically skilled people who want a level of control over their computer that only free software can provide. OTOH, Debian wants to be accessible to non-experts.
How can that happen? If Debian wants bug reports from newbies, can te ‘technical wall’ be overcome?
Q: phones?
A: Telecomm companies are corrosive to networks. I have no mobile, I have an N810 with Wifi that Iuse for VoIP, its Maemo is Debian based I think…
Promoting Free Software in Government - Andy Oram, Editor at
<h2>OReilly - praxagora.com/andyo/</h2>
The government of Massachusetts 2003-2007 used OpenOffice.
USG spends 13billion on software licenses. Half what is spent on guided missiles.
Network effects! Gov choices effect a lot of people. Proprietary vendors know this, FSB can get a lot of benefit from these network effects.
Whitehouse.gov adopted Drupal, 1/2 of us hear this.
US Government staff associate “free software” with untrusted downloads of dubious origins. “open source” is what they understand and trust.
Ration arguments against: Inertia and fear of change, change costs and time sinks, government managers are risk averse, and proprietary vendors will say anything they can.
More sophisticated arguments: Access for all. Vendor independence. Archiving. Customising for special government needs; USG was banned from setting cookies, which meant it couldnt use YouTube; when USG posts a video, youTube will not post a cookie to the user. Security; libre can be just as secure.
Cloud computing becomes a problem especially with DMCA restricting opening proprietary media formats. Libre formats is key.
Passionate arguments are crucial, even if the rationale is air tight, because unexpected problems will come up
Its great to have backing from the top, ideological backing. Via Libre Foundation in Peru, CISL in Brasil, helped influence their governments. You need a lot of training infrastructure to do this right.
In Peru, 1996, there was a practical problem from inconsistent proprietary software used to do taxation which meant the tax law was applied differently in different areas. In 2001 Villanueva had a bill, passed because of a huge media campaign on citizen rights.
So practical and ideological arguments combined are key.
Managers to implement the vision. understand how to interact with developer community. Get managerial credit, like OpenSourceForAmerica.org is an award that managers in federal USG can work to win.
The new ‘challenge’ approach: USG announced an app content, publishes open data sets in programmable formats. Healthcare datasets in the USA and Brasil have done this.
forge.mil was made with CollabNet, and DoD sees the need for flexibility in the field and thus libre software wins on libre, not on cost.
The procurement process is a “waterfall” model, like when the gov builds a guided missile, but the “challenge” approach can overturn this. Must put good dataon the web in an open format. Must have a bit promotpeople know the oty is there. Sustaining the prject is another challenge, since politicians like the big splash annoucement but incremental improvements are hard to get press on. Community pressure can keep things going, an NGO non-profit custodian.
MA OpenOffice? There was initial interest but needed a specific reason. Kriss and Quinn ad a patent on the Office 2003 format. They did a thorough evaluation of all depts. A few people could keep MS office for Macros and advanced features. Most people can use OOo and ODF.
Conversion was demonstrated to save money when an auditor was brought in unusually to check and they agreed that it saved money.
MS saw that OOo lacked accessibilty features for the visoin impaired. MS got a coalition of visoin impaired people to protest OpenOffice. They didnt see they could ADD the features because its libre software.
This shows knowing the strengths and weakensses is crucial, K&Q could have preempted this objection.
It went pretty well. Eventually the state de-funded the project. K&Q left, and the IT dept said MS OOXML is an acceptedable format.
OOXML was created in direct response to the standardisation of ODF. The standardisation process was a travesty, but it proved the argument that open formats has in governments.
In 2005, Munich was prodded by the End Of Lifefor Windows NT in 2004. They did a cost analysis that showed switching costs were higher than license upgrade costs over 5 years, but they still chose libre software for strategic reaosns. They really looked DEEP, below software to the organisational behaviour.
They got high up backing, and did a 5 year cycle with 3 years for planning.
An article of this talk will be in the “Journal of Information Technology and Politics”
HOWTO:
- emphais strategic goals
- optain budget holders support
- hire mangaers with deep insight into libre software
- develop PASSION to ride out rough tpathces
Q: In Albany NY, “Open Government” is seen as publishing on the web. they dont think about libre software users. would a law help?
Archving - can the government open its own data in 50 years?
Peru drive was driven by passion for freedom and interestingly MS tried to push on rational arguments for practicalities and failed.
<h2>Freedom in Education Panel</h2>
Matt: “Freedom is taken, not given.” Deschooling! “Most learning requires no teaching.” Not ask what should someone learn, but what kind of things should we be in contact with in order to learn. Open Education Resources are about that.
Michelle: Library perspetive.
?: USG can make 15 year contracts with Microsoft. That is decided at a managerial level. High shcool teachers tying to teach students were using online collaborative tools more and more; rogue teachers would get youtube unblocked (its totally blocked in NYC schools) and use it to teach with. Paul XXX has a book “Pedagogical Freedom” and theres a conference in Palastine later this year that will talk about this a lot.
?: school management software, bedina (?)
Michelle: There were many insittuionally purcahsed iPads at a recent Librarian conference. DRM and eBooks are awful for libraries, so seeing them is sad. being really supportive of tiny movements to free software and ‘spreading the love’ is great, instead of condeming people for uneducaed decisions.
Matt: Educators often share many values of free soffware, and focusing on the educational benefits of software freedom is key. they have other things they are trying to do WITH software, that are more important than the software.
Q: Isnt their reason for using iPads, ‘ease of use’? Matt: Maybe. its a legitimate concern, But a lot of teachers dont choose their software, theit IT dept does. Q: Mkaing changing habits easy is good Matt: Yes. With shcool computers, you cant install anything, access anything, and they dont work.
Q: The read-write loop in the FS community is great; when kids say “it would be great if” or “it sucks that” then teachers can tell them they CAN interact with the people who make it and make it happen, thats really great.
Phil H: We tend to assume poeple have heard of the idea of free software. in the UK a school found kids swapping libre software and tried to stop it, published a county wide leaflet about it, and when FS people contacted them nicely, they emailed mozilla and asked them to stop because it made it hard to tell kids they cant share :)
Michelle: Librarians want to avoid DMCA Cease & Decist letters. Publiciing that libre software is available and legal to copy is needed.
Q: SO much emphaiss is put into makin things like windows that are GUI and ‘just flow’ and LaTeX helps people to participate by being very highly scriptable.
Q: Firefox Wiki has a great page on how to open it up and hack on it. Drumbeat is trying to get people into school to show they can hack on it.
Q: Apple is pushing hard into education, iTunes U for example.
<h2>John Sullivan on the FSF</h2>
FSF does a lot of positive things and some negative when they are really needed.
FSF Mission is more relevant as software controls more and more of daily live. Software in medicla devices, Pacemaker’s software has bugs. Logically it ought ot be libre for people to imrpove as their life depends on it. on the way the cabbie said that GPS companies are bribed by MacDonalds and so on to route people near their businesses in addition to the ads on their maps.
The freedoms for code are fundemental. If you cant communicate technical information if you learn it, what is left of free speech? FSF focuses on these ethical ramifications, rather than ‘free software works better, crashes less, less malware’ - and those things are a CONSEQUENCE of freedom. They are also secdonary, sometimes libre software costs more or crashses more. We should still use it because its an issue of freedom and ethics.
We promote use, documentaiton and development; especially the GNU project.
We have the Free Software Definition, a good set of tests for any program.
We have a bunch of active campaigns right now:
“Working Together For Free Software” is a new one. Historically FSF was focused on hackers who could directly contribute; this campaign is for non-hackers
Whitehouse has moved to free software, especially the website, and publicsed this with reasons aligned with the FSF, wehave profiles of people involved in things lke that. Instructions for people installing and using free software for the first time. The FSF hasnt done this before because a lot of people do it alrady; mozilla does it for FF, openoffice does it for itself.
We measure success in terms of how many people support and agree with the idea of software freedom. Not how many people run GNU/Linux. Not everyone who supports press freedom run newspapers. If they believe software freedom is important they will go to the trouble to use it, if their values are only on convenience and price they will swing back and forth between free and proprietary software. Also we’re running internship positions and doing travel funding for women to attend libre conferences.
GNU Project is a huge 27 year project, going strong. FSF handles the copyrights for GNU programmers. GNU Advisory Committee is a project to take GNU in new directions, like network services that make desktops obsolete. We also want to expand participation, modelling on the Debian community; people contribute to the overall project, not just their own, whereas GNU has program-maintainer silos.
Defective by Design is a recent high profile campaign, joint with CivicActions since 2006. It shows one kind of tactic, our campaigns have various approaches. This is about DRM. Apple dropped DRM for music but now DRM all their applications on the iPHone and iPad. Amazon’s Kindle page, there is no mention of DRM on the page. Governments should require full disclosure of their products, and so we publish logos to label products as “defective by design.” This is the most negative style campaign.
End Software Patents is a very different campaign; forming alliances with many proprietary software companies who are also harmed by software idea patents and who we otherwise oppose.
PlayOGG is making inroads with YouTube.
Free BIOS: We are moving to all our servers to this.
FSF Jobs page: We support people working on free software full time.
Donations: We run mostly on individual donations rather than corporate sponsors.
Q: Why isn’t Debian on the recommended disros list? A: 2 reasons: It distributes non-free software in main, mainly in the kernel. The linux-libre project removes proprietary parts. The thing we are looking for is a commitment to shipping only free software; its okay if it happens but is treated like a bug. This relates to the second thing, recommending proprietary software. Debian does provide easy access to non-free software, easier than the FSF is comfortable with. The core of Debian is about free software, the DFSG.
gNewSense is moving to Debian from Ubuntu to support MIPS which some libre laptop hardware uses.
Beyond Sharing: Open Source Design, by Mushon Zer-Aviv -
<h2>mushon.com</h2>
I used to develop for the web using Flash. I became disappointed when it didn’t support writing in Hebrew. I’m israeli and in 2003 I initiated a petition “The Right To Languages” against Macromedia for not supporting Right To Left languages, and it was about what the web could be. Israelis and Palestinians could join together with their shared culture. It took 7 years for Flash to support RTL languages. A friend siad, you’re asking the wrong question. Its not about how long will it take for Adobe to support us, its about, why can’t we support ourselves? I didn’t have a good answer for this, and I turned to free software.
I’ve since started the “shiftspace” free sofware project. I’m a lecturer at Parson’s and teach web design there.
I’m working on “Collaborative Futures” there, a FLOSSManuals project.
You can go straight to a solution from a problem. Or, you want wander towards it. When many people are fixing their problems, they find they walk over the same area. They can group together and solve the problem “once and for all”
Why publish free software? Its easy to do. Others can help me to get it working in a final state.
Why open the design process? Its hard. “It works for me.” is a common attitude.
Designers using libre software is a chick and egg problem. I am using Mac OS X on my Apple laptop, not Debian, I am one of the chickens. They don’t use it, so they don’t wan tto make it better, so it doesn’t improve, and they don’t use it. What makes coding and writing in a collaborative fashion is the GRANUALAR ARCHITECTURE - for coding and writing (wikis).
That allows for varied contribution: Accessible ladder of contribution, I can fix a missing comma in a wiki. History and moderation create transparency.
Stuart Hall’s 1980 idea of encoding/decoding of language, from one mind to mouth to ear to another mind. My thoughs, a rich framework of knowledge, being encoding into speech, is an encoding process.
Finding a shared language is the initial step in design; graphic, colour, layout, animation, interaction. Coding without a shared language is impossible. Setting standards for collaboration is essential. But standards is in many ways at odds with innovation, and a good standard allows for innovatoin around it.
For me free culture is more about FORKING than verbatim FILE SHARING.
WordPress’s backing company hired HappyCog to redesign the admin area for 2.5, they are one of the best high end web design firms in the world, and the wordpress community took it forward in 2.7 - screenshots to compare. 2.7 is way way more friendly and advanced than 2.5
There was a questionaire for WordPress icons, we did a student project for the 2.7 icons.
Other examples: Grids are a decades old concept in graphic design, the swiss school, and its come to web design recently. the-golden-grid, 960 gridsystem, blueprint.css - www.gridsystemgenerator.com/gs02.php
Apple is the worst for freedom, but the UI guidelines are great.
how to fix this?
Free as in price is nice but no way near enough.
We can’t expect or force the use of bad tools as an ideological statement. The two things go together, we need the best tools and a free software idea to keep with it once it is good enough to work productively with.
We want to focus on our strengths, networking in a community and collaboration. Instead of cpoying photoshop, which is trying to beat adobe at its best game, include features about how to collaborate. we know more about this.
There is some success in web design, mainly with the more code-heavy aspects.
Granualrity is key. My students USE CODE because its EASIER and SIMPLER to make granualr buildling blcoks. Having designers use VERSION CONTROL SYSTEMS is key towards becoming a hacker - Designers should be hackers, not neccessarily developers. This leads to:
Putting all MASTER files online, and using GIT and things like a libre DropBox.
Finally, language. Conducting a networked user research, and we have collaborative tools to do this. Having module, extensible design languages, that support inheritance so they can be used as building blocks, and DOCUMENTING the designlanguages. That’s to encode the designs.
Collaborative decoding; some designers are taional , UX reearch, tehchnical aspects of design, and design best practices - likein typography, a blue text underlined means a link.
I’ve been interested in Scaling Subjectivity.
sometimes this is framed as “Leadership or Openness” - funny apple mouse vs openoffice mouse slide… - and I think we can do BOTH.
Okay, my slides are up at github.com/mushon/osd- presentation
@mushon
Q: Inkscape community has designers involved in the UI and using Inkscape. A: I have been following Inkscape, its an inspiring example. I’m trying it every now and then, its very inferior to proprietary compeittion. I think they are fightig the wrong fight, adding graphic features in the paradigm of proprietary illustration programs - playing to win a frame they havent set. I think they should focus on collaboration features. Adobe Illustrator has stalled development recently. Strategically we can address it from a different angle.
Q: OpenClipArt.org is a side project of Inksacpe community. Is that where you see the collaboration aspect?
<h2>FLOSSManuals.com - Andy Oram</h2>
FLOSS Manuals has pioneered the “book sprint”
Ive tried promoting this idea at ORelly, and these sprints in 5 days make 75-150 page books. A 2nd sprint pushed the CiviCRM book to 300 pages. Its great to do something creative like writing in a lan party setting.
Q: ISBNs? A: Getting an ISBN is very cheap and easy, and then you can get your book on Amazon. Lulu.com and similar PoD places are great.
<h1>2010-08-02 Monday - DebConf Day 2</h1>
<h2>Stefano Zacchiroli - upsilon.cc/zack</h2>
Debian started 16th August 1993, a GNU/Linux that was competitive with proprietary OS, good quality, easy to install, built collaboratively by experts, and free as in freedom. There were not many other distros.
Today it was 30,000 packages. 120 ‘derivative’ distros, 11 releases, 900 packagers and 120 maintainers, and 1000sutors. We have the largest number of ports in mainstream distros, and we have 2 non—linux based ports coming. There are lots of distros today.
Other distros release more often, have more uusers, innovate more, more credit and more press… is Debian BETTER? why are you contributing? and is Debian still relevant 17 years later?
Yes, I think so.
Q: because the derivatives ARE our users. And we release 4 times a day!
Phil H: Debian does what I want. 1000 people think it does what they want enough to work on it. Even if the rest of the world changes, we’ll still be working on it.
Stanford: We wanted someone we wouldn’t consume, but we could directly contribute to.
Q: I can directly contribute to this.
MTZ: What is Debian? An OS people use directly. A platform for developers to use ot make things. A project, an activity, that people do. In each it does very well.
Q: Unlike other distros, its mainability. “Eternally regenerative system adminstraiton.” Thats under rated.
So yes, all of that.
Why is Debian better?
Quality: Package maintainers are EXPERTS. No packages are 2nd class, all packages are equal. This is shown in responses of big gov users: “We chose Debian because your packages don’t Fail To Build From Sources.” “We chose X [Debian bsaed distro] because we trust Debian package.]
Lots of technical qualities.
But a more profound philosophical reason is its firm principles: devs and users bound by the social contract. Promoting “the culture of free softare” in 1993.
I like to think of Debain as free from the bottom up. We free the software we distribute, including the firmware in the kernel, and free in our infrastrcuture. no non-free web srevices for users, and no non-free services for developers.
“Do-ocracy” in S3.3.1.1 of the constitution; a individual devloper can make any techincal ornontechnicaldecisoins with regard to their own work.
“Democracy” - S2, Each decisoin in the project is voted.
This means reputation follows wok, no benevolent dictator, no oligarchy, yes a cabal ;) but its close to a true meritocracy.
So we want Debian to Live Long And Prosper. To do this we must fix problems when they arise. #1 problem is perceived as “man power” - why other do things better. is it really the case? Sure, we could use more people everywhere. But we can also use better the man power we already have.
NMUs are a good way to collaborate, and to release in a more shared way.
the other thing is that, with computers, a single motivated individual can do WONDERS. other volunteering areas dont have that advantage.
Reduce inertia: consensus != unanimity. “talk is cheap, show me the NMUs.” Debian has both, but do-ocracy comes before democracy. “be bold. what can’t you undo? ”
More manpower can be helped by having an attractive community. folk lore is yo uneed a thick skin here. distrowatch.omc says discussion here can be “uncultured”. every time you are grumpy, fail to ack the good, pinpoint bad, fail to give creit, fail to thank people, insult someone… you kill a “lurking” potential contributor. thick skin != strong hackig skills, it means losig manpower.
Some don’t want new users, more bug reports, we “just” want full down expert DDs from day 1. free software is participatory, and involvement grows in small steps. The “universal operating system” has users everywhere, of every kind.
Some derivatives have 10x as many users as us. In an ideal world. every derivative contributor would also contribute to Debian, and every Debian patch goes upstream to original projects. So lets reach out to our derivatives. We have a ‘derivatives front desk’ to do this.
…
Welcomging non-DD project members. The ‘best OS’ is software, translatoins, graphic, music, etc. we can REWARD them more.
Q: DDs come from users. Also derivs with 10x the users of debian. they become devs of their deriv, and then become DDs. A: We have a flow of DDs from Ubuntu users. I dont think we should as a goal offload users to derivs. We have a responsibility to have direct users.
1030 Davis GPLv3
<h2>Bradey Kuhn on GPLv3</h2>
I can’t represent SFC, FSF, GNOME, SFLC although I have and do work for them and stand on their baords.
People hate change. Most people run stable disros. I switched back to Debian last year, blogged about this, and I originally switched because Debian stable was too old. I prefer to run stable whenever I can. I was at a USENIX conference talking to Miguel de Icaza, and he said “your not running woody, you’re living in the past!” ;)
So, no one should be FORCED to upgrade anything. an upgrade is a trade off between bugs fixed and new features, and stability and familiarity. when you are sick of the bugs.
licenses are like this; we find bugs and fix them. we think of useful new features, we add them. The tough part is that legal codes have so much ‘weight.’ And thanks to disney and others, copyright is a powerful and frightening tool. copyleft turns that weapon around to free people instead of restrict them.
So what bugs do GPLv2 have?
Distributos can REFUSE online distrobution of source. under v2, they can do this, put the binaries on the web and force you to write in. v2 talks about a software “medium” - that doesnt men the network. and they want a reasonable fee for posting you a CD. some distributors do this to just annoy people. they make improvements, they sell the binaries, and you can only get the source if you buy the download and then pay for a source CD.
In v3 S6b and S6d discuss network distribution. 6b takes care of the silly ‘online binaries, written offer’ thing. 6d makes ‘side by side’ more explicit. so that’s 2 bugs closed.
v2 is naieve about patents. most copmanies with patents agggressively enforce them, even if they PRETEND they dont. patent shake downs happen a lot in back rooms and never get any publicity. we have industry wide initiatives but some only want “kernel and c library safty only” or “only the really bad ones”. RMS has a nice essay, “Patent reform is not enough” on that. patents kill the commercial side of libre software. v2 has ‘magical sectoin 7’; the patent license is only IMPLICIT. RMS saw getting a patent judgement
USA-centric. lawyers say “the language does contempplate the internaitoal nature.”
the termination cluase is nasty. the 60 day automatic forgiveness if no one complains, and 30 days after the first complaint from a copyright hoder, is something free software busiensses ought to be running to.
compatibility. permissoin to combine 2 works under different licenses. Apache 2.0 is a great license but not GPLv2 compatible. v3 works hard to approach universal compatiblity with permissive licenses. Artistic 2.0, Apache 2.0, and others are v3 compatible.
exceptions are tricky to do. v2 has no exceptoins for this. we all did them anyway, they were clunky. GCC had a bunch of different ones. in v3, exceptoins are read as a text of the license. it allows “license hacking”. and if you screw up, by making it more restrictive, then v3 allows users to set it aside. Eg, the pacifist licenses. Under GPLv3, if the exception is like that, users can rewrite it to be good.
it ignores embedded software. this was the center of v3 controvery. an EARLY draft ws very highlighted, and the final draft is very watered down. DRM provisoins is onlhy for ‘consumer products’ and it ONLY says the vendor mst tell you how to jailbreak the device. but the vendor can void your warantee immediately. you can be boted off their network; if you modify the softwar to make trouble on their network, they can boot you off. Google’s distribution of HTC Dream and Nexus One HAVE NO Gplv3, but they are v3 compliant. Because Nexus One shows you that Google CAN comply with GPLv3 in the USA and Europe, the regulators don’t have a problem with alowing firmware mods. The Droid X which is totally locked down got jailbroken immediatley anyway.
“Mobile Exception” - its possible but no recommended, and evne if i was forced to choose between GPLv2 and GPLV3+MobileException, I’d chose the later.
LGPLv3; there is no reason not to switch. That its GPLv2 incmpatible is fixable with dual GPLv2+LGPLv3-or-later licensing. This is a ‘soft’ way to begin GPLv3 adoption. Getting away from LGPL2.1 is important.
and there is no web services clause. this was meant to be in the main text, and Google was the main force saying they would fork EVERY GPL project if we did that. so its in the AGPLv3. the basic premise is that if someone can legitimately use it, they must have acess to source. yes, AGPL need a little more care, i have yacto-reader which is a debian package that’s AGPL compliant in a debian package that complies by default.
Thanks debian because you and downstreams have DEFAULTED to GPLv3-or-later for all GPLv2-or-later stuff.
Q: Afferro based on publiv performance? A: No it uses the right to modfy to put this requiremnt to distribute source to network users.
Dave: If the exception DOESNT be too restrictive, can it also be removed? A: Yes you can always pare back
Dave: How do we write a new exception - for fonts document embedding? A: Speak to me :)
Q: if you have 2 programs, one GPL+exceptoin and one GPL, they are not compatible. BDale: adding an exception makes 2 licenses, is the view of our legal team. Q: Me and fontana identica streams talk about requirement
Dave: GPLV4?
A: I have 4 bugs for GPLv3, so GPLv4 will start when its like 12.
Dave: why wasn’t there a GPL2.5 in 2002?
A: RMS wanted to do GPL2.2 in 2002,and it would have affero in the clause. we urged him not to do this and do the GPLv3 process instead.
—
<h2>Michael Bank mbank@debian.org on GNU/HURD</h2>
It works, on some hwardware and in qemu.
<marcus> Jeroen: you are a Hurd developer. Being insane is part of the public image.
Good recent summary on the H blog.
Thomas Bushnell <bsg> and others including RMS did original design. HURD isn’t akernel, its a userspce system that needs a microkernel. Mach was chosen, although Thomsa didn’t like it thatmuch. Ronald McGrath worked on it with him. RMS didn’t code on the core.
…
Samuel Thibault showed up and fixed just about everything; imeplemtned TLS, ported GNU Mach to Xen and implemtned PAE, fixed many crashes, ported Xorg, run GSoC projects, and pushed the design concepts.
Thomas Schwinge pushed the GNU infrastrcutre side, getting new committers approved and migrated to git.
1700 Davis Ubuntu
<h2>Mark Shuttleworth - Unity</h2>
Unity is a culmination of several year’s work. There has been an explosion of interest in GNU/Linux on netbooks. Hardware vendors want want simplied UIs. I’m now not the CEO and I work on parts I like, including Unity’s design. London design team push this; aytana is the umbrella project for this.
Unity originates from considering the UI of game consoles; computers as devices you use for a limited range of tasks. Bigger icons (because touch UI was seen as a future path). The categories are from the typical GNOME applicatoins menu. Places also mapped from GNOME. But we just present the pieces they are interested in. Its been picked up in a few places, JoliCloud and others. For the next revision, we want to relax it a bit; Places became an item in the categories list.
Starting quickly was a real priority. It would be for dual boot preinstalls, so people can become familiar with a free alternative. But no Add/Remove, nothing for interacting with files in the traditoinal way. So we did more work on a “dash”, a full screen place for dealing with lots of files. This was done and went into Maverick a month ago.
In the design process we embraced our constraints; maximising vertical picxel space, using maximised windows, and building for touch as future devices will be have that input.
Our sources of inspiration: console games have no constraints, nothing is “mission critical” unlike GNOME and Linux kernel which must not break. we have a light theme for style and feeling, the idea of light coming through the screen, and use alot of OpenGL and accelerated hwardware. We want search for everything with GNOME Do and Google.
We’ve decided to adopt a “global menu”, for fsater access on small screens and verticla space savings. We build a DBUS protocol and plugins for GTK and QT to export a program’s menu, so a program feeds it to be rendered by Unity. XUL and OpenOffice don’t use GTK or QT so that work is yet to come. This has been published as a freedesktop.org standard. We’ve done a lot of work for 10.10 Ubuntu release to standardise menus across applications. We also have a GNOME Panel applet that makes testing this feature very easy on regular GNOME desktops.
We also have menu indicators. GNOME is murky, KDE is cleaner. left or right click, respond to move movements, anything could be done in a custom way. Now everything in the top right is one consistent menu. this benefits accessibilty.
Ayatana has a goal for ‘persisntent awreness’ with fewer and classier indciators.
…
“Bamf” is a new API for knowing what applications and windows are there.
Summary: Design discussion is on ayatana list, hosted on launchpad.net, and shipping in Ubuntu 10.10 Netbook Edition.
Q: Chrome cross over? A: Google started with Ubuntu and moved to Gentoo Portage. There hasn’t been any cross over in terms of UI either.
…
A: Apple create desire, we don’t Q: I think we’re in a positoin to be BETTER THAN APPLE within 2 years.