MonthOctober 2010

As novidades do kernel 2.6.36

No dia 20 de outubro de 2010, o finlandês agora cidadão americano Linus Torvalds apresentou ao mundo a versão mais recente de seu kernel livre, o Linux 2.6.36. Com o codinome Flesh-Eating Bats with Fangs em homenagem a um morcego que recentemente invadiu a residência dos Torvalds — alguém comentou no blog de Linus que “o Batman finalmente invadiu a casa do Pinguim” — esta nova versão do Linux foi completada em 80 dias, precisamente a duração média dos últimos lançamentos de versões estáveis do kernel. Porém, com uma versão -rc a mais — recentemente, essas versões só chegavam até -rc7, mas alcançaram -rc8 desta vez — a versão estável “pareceu demorar mais do que de costume”, alegaram alguns desenvolvedores.

O kernel encolheu??? Que bom!

Pela primeira vez nos últimos vários anos, uma nova versão do kernel Linux traz menos linhas de código do que a anterior: 13,49 milhões, contra 13,54 milhões na versão 2.6.35 e 13,32 milhões no Linux 2.6.34. Curiosamente, o número de arquivos aumentou de 33,3 mil na versão anterior para 34,3 mil no Flesh-Eating Bats. Além disso, pela terceira vez consecutiva o número de commits que compõem o kernel ficou abaixo dos 10 mil, mais precisamente em 9501.

A redução no tamanho do kernel se explica pelo admirável trabalho de faxina que vem sendo feito na base de código em constante evolução. Os avanços na remoção da BKL (Big Kernel Lock), por exemplo, foram significativos nesta versão.

Vamos às novidades!

AppArmor, finalmente

AppArmor é um sistema de “controle de acesso obrigatório” (Mandatory Access Control ou simplesmente MAC). Desenvolvido pela empresa Immunix nos idos de 1998, ele foi mantido pela Novell — que adquiriu a Immunix — de 2005 a 2007, quando a fabricante do SUSE Linux Enterprise dispensou seus desenvolvedores.

O AppArmor é ativado por padrão em diversas distribuições, sendo os principais exemplos Ubuntu e SUSE Linux Enterprise (Desktop e Server). Incluído no kernel, ele passa a fazer companhia aos demais residentes SELinuxTOMOYOSMACK.

Desktops mais rápidos, mesmo sob pressão

Na área dos desktops, uma série de patches tornam o escalonador mais competente com relação à redução da latência máxima dos processos — nenhuma relação com os patches de Con Kolivas. Juntamente com as novas workqueues implementadas no kernel, o uso de múltiplas tarefas concorrentes num único sistema poderá ser significativamente mais rápido. Por último, um conjunto de patches finalmente dá alento aos pobres coitados que possuem dispositivos de bloco muito lentos. A partir do kernel 2.6.36, os sistemas conectados a esses dispositivos não mais ficarão praticamente incapacitados de reagir quando seus dispositivos — mesmo os USB — demorarem a responder.

Vídeo

Se você possui uma GPU integrada a seu processador Core i3 ou i5, poderá utilizar todos os benefícios do suporte ao Intel Intelligent Power Sharing, recurso presente nesses hardwares que permite um melhor uso — compartilhado, como sempre deveria ser — entre esses dois componentes do “pacote” presente na pastilha de silício. Os benefícios são uma espécie de “overclock” automático da GPU quando a CPU não está a todo vapor, ou então CPU e GPU mais frias e economia de energia quando não estiverem ambas sendo usadas ao máximo.

Ao trabalhar com GPUs AMD Radeon, o Linux enfim se tornou capaz de ler os sensores de temperatura dessas GPUs, entre outras novidades.

Já no campo da concorrente Nvidia, o driver KMS Nouveau passa a incluir suporte — embora ainda básico — às GPUs Fermi presentes na recente série GeForce 400.

Controle remoto via infravermelho

O uso de dispositivos infravermelhos no Linux é, historicamente, trabalhoso. Porém, isto finalmente está mudando. Com os avanços iniciados no Linux 2.6.35, o kernel agora oferece interfaces utilizáveis pelo utilitário LIRC, que por sua vez faz a interface com transmissores e receptores infravermelhos. Com isso, os drivers, antes mantidos no espaço do usuário junto ao próprio LIRC, estão sendo migrados, pouco a pouco, para dentro do kernel.

Compressão de memória mudou de nome novamente

Compcache? Não, o nome mudou para Ramzswap. E quando finalmente foi incorporado ao kernel 2.6.36, o dispositivos de blocos compactado na memória RAM recebeu um novo nome: Zram.

Para usar o Zram, basta ativá-lo nos drivers staging e carregá-lo (chama-se zram). Com isso, cria-se ao menos um dispositivo de blocos compactado, em /dev/zram0. É possível encomendar mais dispositivos por meio de uma opção de carregamento do módulo zram. Para definir seu tamanho, não é mais necessário um utilitário, pois seus controles mudaram dos ioctls para o sysfs. O resultado: basta usar echo 100000 > /sys/block/zram0/size para definir um tamanho de 100 MB para o dispositivo zram0.

Nota: em testes na minha própria máquina, este método não funcionou, pois o sysfs não permitiu a gravação de novos dados no arquivo /sys/block/zram0/size.

Matador mais competente

Quando seu sistema utiliza toda a memória RAM disponível, ele passa a recorrer ao espaço de swap. Quando acaba o swap… alguém (algum processo, é claro!) precisa morrer. O matador de processos se chama OOM (out-of-memory) Killer, e não é sempre tão inteligente. Porém, seus desenvolvedores perceberam que é sempre bom ter um pouco mais de inteligência, mesmo que seu trabalho seja apenas apertar um gatilho. O OOM Killer agora é capaz de tomar melhores decisões a respeito de quais processos matar a fim de liberar a tão valiosa memória.

Novos processadores, nova arquitetura

Uma nova arquitetura de CPU, que ainda nem existe na prática, chamada Tilera, já conta com suporte oficial do Linux. É mais um caso que demonstra a rapidez de evolução e adoção de novas tecnologias típica do Software Livre e, em particular, do kernel Linux. Se a Tilera entregar o que promete, poderemos ter um salto interessante no poder computacional disponível em máquinas comuns.

Já nas arquiteturas tradicionalmente suportadas, o Linux agora funciona com CPUs Tegra da Nvidia, baseadas em ARM.

Novas opções de compilação

Os novos alvos de compilação oldnoconfiglistnewconfigalldefconfigsavedefconfig já foram usados pelos desenvolvedores para eliminar mais de 200 mil linhas de código referentes a arquivos de configuração padrão para as várias arquiteturas suportadas pelo Linux. De certa forma, é como se eles pudessem agora usar um diff dos arquivos de configuração padrão, em vez de uma cópia completa do arquivo de configuração padrão somada às configurações padrão daquela arquitetura específica. Viu como a redução do código foi boa? 🙂

Virtualização

KVM e Xen são grandes concorrentes, mas cada vez mais semelhantes. Os desenvolvedores do KVM acrescentaram alguns recursos que permitem uma espécie de suporte reduzido à execução do Linux como Dom0. Seria este o começo da união entre os dois principais tipos de hypervisors do kernel Linux? Resultados das atuais discussões na LKML poderão ser vistos em uma versão bem próxima do kernel.

Fanotify

A tão prometida e desejada varredura de arquivos sob demanda — imagine isso num sistema antivírus — está mais próxima: o novo sistema Fanotify entrou, embora “de leve”, no kernel. No entanto, vem desativado por padrão em virtude de discordâncias quanto à sua API. A tendência é que o Fanotify substitua as duas “gambiarras” usadas até o momento (sem grande sucesso) para esse mesmo propósito, fsnotifydnotify. O Fanotify é uma estrutura do subsistema de sistemas de arquivos capaz de enviar mensagens ao espaço de usuário (por exemplo, a um software antivírus) informando sobre a abertura, alteração ou fechamento de arquivos. Assim que ela ficar pronta, podemos esperar que seja muito usada em diversos tipos de programas.

Sistemas de arquivos

Quem acessa compartilhamentos remotos via NFS já testemunhou, a partir do kernel 2.6.30, os benefícios do cache local de sistemas de arquivos remotos. A boa notícia: se você usa CIFS para isso, finalmente sua hora chegou. Sistemas de arquivos CIFS agora já têm suporte a essa estrutura, chamada FS-Cache.

No campo do já antigo sistema de arquivos Ext3, a onda retrô levou os desenvolvedores a tornar padrão, novamente, a opção de montagem data=ordered. Ela oferece maior segurança e menor desempenho do que o “antigo novo padrão”, data=writeback.

De forma genérica, o XFS também recebeu vários patches destinados a “melhorar seu desempenho em vários trechos”.

A redundância de subsistemas para lidar com RAID também está finalmente diminuindo. Enquanto o Btrfs recebeu grandes trechos de código do subsistema MD para acrescentar ao futuro sistema de arquivos padrão recursos de RAID 6, o subsistema DM (device mapper) começa a ceder seu código de RAID 5 para outros trechos, de forma a permitir, no futuro, que o utilitário dmraid gerencie também implementações de RAID em drivers de controladoras etc.

Quem usa discos SSD agora também terá mais uma ajuda do subsistema DM, que ganha suporte ao comando discard, capaz de informar que determinada área de dados está disponível — economizando assim um novo ciclo de leitura em todos os blocos que a compõem e estendendo a vida útil dos dispositivos. Tudo isso depende, todavia, da existência de suporte ao comando discard por parte do disco e de sua respectiva controladora.

KGDB + KDB + KMS

Esta sopa de letrinhas tem um significado especial para os desenvolvedores do kernel. O depurador do kernel, KDB (kernel debugger), agora trabalha em parceria com o driver Intel KMS. Isso significa que, na presença de uma GPU Intel que use esse driver, o KDB pode ser acessado a qualquer momento pressionando as teclas SysRqG(confira o resultado neste vídeo).

Futuro

Está em andamento uma iniciativa para aumentar a escalabilidade do VFS (Virtual Filesystem Switch), que já mostrou frutos nesta atual versão do Linux. Espera-se que haja ainda mais novidades no kernel 2.6.37. Da mesma forma, os grandes e contínuos avanços na remoção da BKL (Big Kernel Lock) sinalizam que o kernel inteiro deve ganhar mais desempenho, tanto na versão 2.6.36 quanto nas próximas.

Pode-se esperar também que a integração do KDB, neste momento restrita ao driver KMS Intel, seja incorporada aos concorrentes Nouveau e Radeon.

Este foi o último lançamento de kernel estável em 2010. A versão 2.6.37 do kernel Linux deve ser lançada em janeiro de 2011.

Post original aqui.

Hole in Linux kernel provides root rights

A flaw in the implementation of the Reliable Datagram Sockets protocol (RDS) in the Linux kernel can be exploited to gain root (also known as superuser) rights or permissions on a victim’s system. Attackers can exploit the hole to get complete control remotely once they have broken into the system. Dan Rosenberg, who discovered the vulnerability, has published an exploit for demonstration purposes; in a test conducted by The H’s associates at heise Security on Ubuntu 10.04 (64-bit), it opened a root shell.

Kernel versions 2.6.30 to 2.6.36-rc8 are said to be affected. Linux developers have already provided a patch, in the Git repository, that solves the problem. Distributors will probably be publishing new kernel versions soon. As a workaround, Rosenberg recommends preventing the kernel module from loading: echo “alias net-pf-21 off” > /etc/modprobe.d/disable-rds (as root). Most systems will not be affected as they do not use the protocol anyway.

Rosenberg says the problem came about because the kernel functions in the RDS protocol do not correctly check the addresses given when data are copied from kernel memory and user memory. As a result, local users can indicate a basic address within the kernel for a socket structure. Code can then be written into kernel memory and launched with kernel rights when certain sockets are called.

Just a few days ago, a hole in the GNU-C library’s loader was made public that also allows attackers to expand their rights on the system.

Google dá resposta nerd a crítica de Steve Jobs ao Android

Uma das principais características alardeadas pelo Google para o Android é sua plataforma aberta e baseada no Linux, facilitando aos programadores a criação de aplicativos. No entanto, Steve Jobs – executivo-chefe da Apple – fez críticas à mesma em uma teleconferência realizada nesta semana.

Andy Rubin, chefe da divisão Android, inaugurou seu perfil no Twitter em 18 de outubro com uma resposta à crítica de Jobs. Sua mensagem foi enviada em forma de uma sequência de comandos que equivalem à cópia do Android para sua máquina Linux e sua compilação na sequência, deixando-o prontinho para rodar.

Para quem ficou curioso, veja a sequência citada por Rubin:

mkdir android ; cd android ; repo init -u git://android.git.kernel.org/platform/manifest.git ; repo sync ; make

Ubuntu – Ask Mark classroom

This week Ubuntu guys is supporting the Ubuntu Open Week, and below, follow the chatting “Ask Mark Shuttleworth” today.

Mark: hi all! very glad to be here looking fwd to your questions

jcastro says: people are already wondering about http://blog.canonical.com/?p=446

Mark: congrats and thanks to everyone who helped build 10.10, it was a tight but very well-run cycle

jcastro says: if you want to kick it off talking about unity and utouch I think that would be a nice way to start things

Mark: okdokey. 2 years ago we had a flood of PC manufacturers wanting us to help them build their own OS’s everyone wanted a Linux and they wanted them all to be different and they wanted them all to be built on the core of Ubuntu we did engage with some of them but we saw that world graudally fragmenting, and that path wasn’t going to make linux a world class, strong competitor to the established proprietary platforms so we decided to put all of our own effort into a focused designed and engineered UI for netbooks that started with UNR which evolved (with a clean sheet at one point) into Unity it’s come together quite well for 10.10, we didn’t get it all done as we hoped and there are issues on certain hardware but feedback is generally that people love the design and direction want bugs fixed and want it to work on all the hardware possible so, that will be our focus in 11.04 also we’re starting to see a new generation of mouse, essentially, that brings touch to netbooks too there wasn’t any great open source touch framework anywhere and we have a view on touch beyond basic touch, towards “gesture languages” so uTouch the beginnings of that are in 10.10 it will evolve for 11.04, and it will get easier to integrate with normal apps so, you should be zooming and scrolling with touch in 11.04 all over the place we’ll also integrate window management and touch which is pretty slick to see in action the video gives you a taste how’s that for an intro?

jcastro says: Great, on with the questions

autif1 asked: What is the future of my favorite OS with embedded devices? Specifically – the guru plug and the sheeva plug – these are ARM based computers.

Mark: ARM is now a fully supported architecture in Ubuntu the ARM ecosystem is coming together in something called Linaro, and Canonical is very much part of that Linaro is a forum to get stuff done, not a consortium or a new distro it’s where we can set a roadmap for a unified ARM kernel, and set the pace for the ARM toolchain in 10.10, for example, the whole of Ubuntu is built with GCC that includes patches from ARM that makes everyone’s life a little better, but a little more complicated it helps get those patches upstream faster, because they’ve been exercised at Ubuntu-scale which is good so, you can count on ARM support in 11.04 and the foreseeable future.
next!

Pendulum asked: Today is the 41st annual World Standards Day with this year’s being focused on Accessibility standards. Are there plans to improve Ubuntu’s accessibility and to bring things like the Ubuntu website in line with web accessibility standards?

Mark: yes, accessibility is important, please file bugs where we let you down on that front for 11.04, a11y is one concentrated push for the Unity team, for example    we need all the help we can get, though    there’s no commercial case for it, we do it because we think it’s important commercial engagements related to it would help, and folks on the team who are interested can make a big, big difference
next!

IdleOne asked: Hey Mark! I wanted to ask about the Code of Conduct and when we will be seeing translated versions for the rest of the world to be able to read and sign in theyre own language?

Mark: good question. i don’t know, but i’d like to hear flacoste’s view, he leads the LP team i suspect the main issue is prioritisation of that in LP but LP is open source, so… also we should generalise that feature, so teams can have things like service level agreements and can ask people to commit to them
next!

nisshh asked: is anything further happening with ‘Windicators’?

Mark: it’s in the queue just not a top priority, with everything else moving on i’d like to see it, but i’m not going to force it when i know we have other things to juggle we already have the AppIndicators protocol all we need is a variant of that to associate the indicator with a window and a plugin for (your favourite window manager) to agree to render the indicator.
next!

ssj6akshat asked: Is Canonical profitable yet, or How much time more until it is?

Mark: no, and some 🙂 it’s important that Ubuntu have a strong commercial footing that gives people confidence in the future of the platform it helps build the base of investment in the distro and Canonical is a good partner to our community, I believe, so Canonical’s health is good for the community too we chose to take on multiple things: servers, desktops, ARM which creates contention and slows down the march to profitability but it also makes Ubuntu more valuable as a cohesive platform and i’m still confident we will break through on each of those fronts
next!

mhall119 asked: What is Project Harmony?

Mark: Harmony is an effort to simplify the forest of contribution agreements into a few, well thought through trees at the moment, there are literally hundreds of contribution agreements (also called copyright assignment agreements, because amongst other things, that’s usually what they involve) i believe contribution agreements are really important to stimulating a healthy ecosystem of corporate involvement in the long tail of open source they are not important for the linux kernel, which will always be cool and sexy and in many cases mission critical for so many companies and individuals you will always have a flood of contribution but they are important for many of the things we want to be there, in quality and to “just work” i worry that this is badly understood by the broader community there are some myths about open source most of the work is done by folks who have a genuine commercial interest in seeing it done in many cases, that interest is tangential to the ownership of the code but in many cases, it’s not for example, compare Qt and Gtk Qt has a contribution agreement, Gtk doesn’t for a while, back in the bubble, Sun, Red Hat, Ximian and many other companies threw money at Gtk it grew and improved very quickly then they lost interest, and it has stagnated Qt was owned by Trolltech it was open source (GPL) but because of the contribution agreement they had many options including proprietary licensing, which is just fine with me alongside the GPL and later, because they owned Qt completely, they were an attractive acquisition for Nokia all in all, the Qt ecosystem benefitted and the Gtk ecosystem hasn’t so one of the problems with contribution agreements is that they never had a strong lead GPL, CC both had clear leadership and become widely adopted we’ve gathered the legal counsel of lots of the top open source companies we’ve looked at hundreds of contribution agreements most, the vast majority, of them look very similar they talk about copyright, patents, and code but because they were all written by different lawyers who “just wanted something that works for them”, they aren’t general Harmony should produce one, or two, general contribution agreements perhaps with options, like some of the main open content / code licenses that way, when you get to a project, if they have a “standard” agreement, you know quickly whether it’s OK for you or not i don’t actually think anybody who has found a bug in X and made a patch has said “oh, I’m not going to contribute it because I believe in the GPL and they are under the MIT license” and similarly, i think contribution is the right thing to do when you participate in a project that requests it there are some exceptions, in the case of things like plugins which could be whole works in their own right but if you’re making a patch to someone else’s codebase, and they own the whole right to that codebase the generous, and imo right thing to do is to contribute the patch in a way which does not change their rights, or yours which is under a contribution agreement we’ve signed many of them, we have a policy that we always do only exception ever was a weird, nasty agreement by some company i’d never heard of that said something impossible which we declined, and i think they fixed so that’s Harmony.
next!

ssj6akshat asked: What do you think about OMG! Ubuntu!?

Mark: rocks
next!

ean5533 asked: Hardware issues aside, has the response to Unity been mostly positive? Would you choose another direction if you could go back in time?

Mark: it’s been flattered, critiqued and emulated, in equal measure 😉 all are important, i think the flattery is nice – people like that it’s clean, the pieces fit well together, layout and space are considered the critique is a very good guide to where we need to direct effort performance on GL fallbaks where the hardware or drivers are not sufficient the design decisions we made around file access need careful testing and iteration and the emulation, well, that’s the sincerest form of flattery and perhaps it’s the only way we could realistically have helped those projects which embrace our ideas, after they work because sometimes you just can’t convince folsk any other way than to Just Do It.
next!

sebsebseb asked: Will 11.04 be using Gnome 3’s Gnome Shell by default in the desktop version?

Mark: we deferred Gnome3 adoption from Maverick, and with retrospect and hindsight i’m very glad we did we’re reviewing the status now in prep for UDS testing with users and chatting with developers we need to settle the question before the end of UDS.
next!

sebsebseb asked: Favourite Ubuntu release before 10.04 and why?

Mark: Dapper Drake. It set our course for LTS’s and being a whole platform.
next!

mobster asked: Why Kubuntu is getting less love?? For example no software center and ubuntu one?

Mark: because it would cost farmore than i can justify i do love the kubuntu community, and spend what some would consider an unreasonable amount on doing certain things twice there is no philanthropic benefit to having TWO free desktops out there that won’t help more folks embrace free software neither is there much commercial benefit in having two free desktops so, ask yourself, on what basis do you feel that we’re letting you down? on what basis do you feel you have a right to expect something else? i admire KDE and Kubuntu, I enjoy using KDE occasionally and hanging out on #kubuntu-devel and i like the people, except occasionally the odd super-self-interested muppet who expects me to singlehandedly make his wet dreams of technology kfuturism come true and that’s that.
next!

danyR asked: About the new ubuntu iconset, any updates? I’ve read somewhere that some inicial sketeches were supposed to be launched, is it still planned to 11.04?

Mark: i don’t think we will achieve it for 11.04, no it’s a big program and we haven’t yet started but i know, if we don’t start for 11.10 we won’t finish by 12.04
and i really want it done by 12.04 LTS.
next!

ssj6akshat asked: What motivated you to invest in making free software and debian user friendly?

Mark: hmm… “because the possibility was out there” you know the answer to the question, “why do you climb dangerously high mountains”? “because they are there” life is something we get to use up, once and once only we should do the boldest, scariest, most important thing with our lives that we can dream i felt free software could be all the things we want ubuntu to be: easy to use, free of charge, sustainable, beautiful but nobody else seemed to be interested in getting it there and it wasn’t going to happen by itself it needed a community that was single-minded about THOSE specific goals not the things that people seemed to care about nothing wrong with the kernel community, or the X community, or the other distro communities i just didn’t see anybody who was caring about usability, people, beauty, quality on the desktop if you think something is possible and good and you have the time and resources and nothing more important to do then you should do it and thousands of people seem to agree, becuase they help build it.
next!

bilalakhtar asked: Gnome Shell uses a system of notification that is somewhat similar to notify-osd. When Ubuntu would begin using gnome-shell, would you like notify-osd to be used or the notification system of gnome-shell?

Mark: notify-osd. we designed and built it in good faith it’s compatible with the freedesktop.org standards we did it long before anybody else seemed to care about reinventing notifications we expressed a willingness to collaborate around API’s when suddenly they did now we have good code that works, with lots of apps that use it we’ll stick to it.
next!

highvoltage asked: Why did Canonical ditch the LPI, and will there be any discounts for the new training for long-time Ubuntu contributors (or ubuntu members)? Currently the server training is more than £1000, which is a bit steep for an individual for an online course. there was little demand for individuals getting their own certification and more for something specific to ubuntu that companies could be confident would help their sysadmin teams be productive in an environment where ubuntu was being deployed i’d like to change the forces of gravity and economics

Mark: occasionally, we tweak their noses but in due course they reassert themselves 😉
next!

SergioMeneses asked: Is there an strategy from Canonical to increase local commercial presence in emerging economies (and not via partners)?

Mark: yes! we have an office in Shanghai, we have employees in India and Brazil and South Africa is starting to embrace Ubuntu for education (highvoltage must have had something to do with that ;-)) and i believe in that mission but we can’t be everywhere, doing everything partners are very important to us and where we have the right partner, we are often more effective than we could reasonably expect to be doing everything ourselves.
next!

helger asked: What do you think about services like Flattr? Have you considered integrating something like that into the Ubuntu Software Center?

Mark: they are very cool, and yes.
next!

alecu asked: is there a plan for process isolation for apps installed from untrusted sources (ie, universe, proprietary stuff from the software center)? iOS and sugar from the olpc already have something like this using, say, something like AppArmor?

Mark: i like the idea! you should chat with the right folks at UDS about that, if you can come or raise it on #ubuntu-devel cc mdz 😉
next!

cazo(me): some tablet on the horizon (partner)?

Mark: can’t be

mistrynitesh asked: what are the plans for India?

Mark: have we… ok. We got the new Rupee symbol in the 10.10 ttf-ubuntu-font-family package. first OS in the world to support it natively high-five to sladen and DM for that I think India has the potential to harness FLOSS in a very potent way there is little legacy dependency there is a substantial talent base the only thing that is required is very directed government policy that, however, is challenging in India countries like Brazil might well do better: they too have been experimenting with FLOSS and can more likely translate that thinking into concrete policy that encourages business, universities, schools and government organisations to use FLOSS so, it’s a race to see who is smarter and more organised about this.
next!

danyR asked: What’s the plan with indicator-network and indicator-datetime? Is natty going to be the first linux distro ever to ship without a notification area?

Mark: without a legacy systray, i hope so we are building a new GNOME UI for connection-manager, the Intel-Nokia replacement for NetworkManager we’ll have to see, in the final analysis, how it pans out but connman has many advantages in design and testability NM has more road behind it i use the connman bits, and they work well for me with some exceptions ad-hoc networks and i haven’t had much success with 3G though i believe it works for some Google is using a derivative in ChromeOS so, i think it will be solid and i really like the design work MPT did on the indicator and settings, though it’s taking time to implement.
next!

popey asked: We often see figures for how many Ubuntu installs there are, 8 million here, 12 million there. Can you give us definitive (near enough) figures and tell us how you arrive at them? This would help dispell some naysayers who claim we’re making these numbers up.

Mark: no, i have no definitive answer there are stats but we can make those say whatever we want we just don’t do any meaningful tracking or registration anyway, what matters to me is that our users are delighted, whoever they are and however many there are i do believe we have more than either of those numbers but i don’t think anybody knows for sure, except maybe google, and they haven’t said.
next!

TLE asked: Any plans on changing the one-cd strategy, to get room for more standard tools, like say a demon administration tool and a firewall?

Mark: no. it’s a good discipline we need to get better at helping people find things like those tools of yours, after they install and forcing less on them up front
next!

BigWhale asked: So, we started brainstorming for Ubuntu event on Balkans in couple of years. What would it take to get you for a keynote speaker? 🙂

Mark: i’m a pushover, except when i’m impossible so just ask! if it works, i’ll come along but i bet you can get plenty of entertaining speakers without me there are much better speakers in the ubuntu community and commercial ecosystem i did enjoy a week in Croatia once be nice to be back in the area.
nexy!
next, even 😉

jcastro says: time for one more. waiting for a good one.

Mark: hard one.

danyR asked: I’ve been following Ubuntu for years. I’ve also been following the blogosphere for years. I can say, without doubts, that Maverick is the most successful release ever. What’s next?

Mark: Natty!
next!

jcastro says: (Protip: See what’s cooking here: https://blueprints.edge.launchpad.net/sprints/uds-n)    15:57

doctormo asked: Do you consider Ubuntu to be a product of the UK?

Mark: no. Earth barely.
next!

nigelb asked: What do we do at the end of exhausting a – z for naming? 😉

jcastro says: (this will be the last one)

Mark: cyrillic, anybody? thanks all.

jcastro says: thanks, Mark!

Mark: great questions, appreciate the support of the team organising you all do a wonderful job with Open Week akgraner and many others, hugs. cheerio.

charlie-tca says: Thank you very much for that session, Mark. It is always great to be able to listen to you this way.

Ubuntu – Unity and uTouch

One of the most exciting things about the Ubuntu 10.10 release has been the delivery of the Unity ‘shell’ in Ubuntu Netbook Edition. For the uninitiated,  the delivers a very different user experience to that in the main desktop edition. For a start the icons of the most popular applications are permanently featured on the left had side of the screen. This borrows more from the smartphone interfaces but is adapted for use on, in this case, netbooks. So there remains a workspace where users still have sufficient room to watch video, edit photos, create documents, play games, read the web, write emails – all of the usual tasks we use a computer for, day to day.

Everything is optimised however for the more limited screen space. It is sub-optimal for instance to simple port an interface from the full-screen world, shrink it and expect it to be a great experience. Unity does away with the bottom bar for example, that Windows, Ubuntu and Mac users will be used to. This is actually a radical step, but in my experience at least, it take no time at all to forget that there ever was a bottom bar. The result is considerably more ‘vertical space’ for to use  – again maximising the useful area on limited screen sizes.

One of the coolest things though is one that will be experienced by the fewest people at this point – touch. Unity is fully touch-enabled – those big icons are screaming out to have a digit poked at them. But as ever, the boys in the lab, or in this case Duncan McGregor‘s  multi-touch team have gone a step further and created a multi-touch ‘gesture’ library. This allows finger combinations to do groovy things like expand and reduce windows, pull up multiple windows in one workspace, and call up the ‘dash’ automatically. These are in 10.10. In 11.04 we will see a lot more.

Because there are a very limited number of touch-enabled devices out there at present, we thought we would create a video to show some of the features. You can see if below. Duncan goes into more of the detail on the work done and ongoing in his team in his blog. I will let the video talk for itself – it has turned out rather nicely even with the clumsy paws.

Gerry Carr, Platform Marketing, Canonical

© 2017 Ziben IT Solutions

Theme by Anders NorénUp ↑