Dylan McCall: New stuff for ubiquity-slideshow!

Remember that slideshow when you did a fresh install of Ubuntu Karmic?


For Lucid, ubiquity-slideshow is rocking, if I may say so myself :)

First of all, a quick refresher: this is a really simple project with the goal of providing an introductory slideshow that runs when people install Ubuntu. The ubiquity-slideshow packages are all content packages, then Ubiquity displays that content at the right time.

The slideshow is implemented with Webkit, since all the cool people use Webkit. (It also renders things nicely, it's flexible, quick, we can quickly throw it on the web, people can make content really easily, and Javascript allows us a good split between interactive goodies and actual functionality like installing the operating system).

As far as content is concerned, I try to keep this different from other introductory slideshows in the content that I maintain. It doesn't try to sell the product as people install it, instead providing some exciting points of action for getting started. Not as vague as you get in Windows' slideshows, but not an in-depth how-to either.
The idea is really summed up in the first slide: Through this I want to encourage new users to explore Ubuntu and really discover how awesome it is (instead of what buttons to press).


(I admit it: I am obsessed with whitespace)


So: Lucid... Lots of new stuff here.

First of all, the project is now providing separate slideshows for Xubuntu, Kubuntu and Ubuntu installations. It's also really easy to add more, so anyone who wants in on neat slideshowy action can!
I'm hoping to get an Ubuntu Netbook Edition slideshow in there, too. It isn't perfect at the moment, so if you have any suggestions for what needs to be done or said, please let me know or write to the ubiquity-slideshow mailing list on Launchpad!

Translations aren't horrific to work with any more. It's still possible that strings may change before documentation freeze, but the gist of it is we just have one translation template that has all the text for each slideshow. You can go through the list on a rainy afternoon and have your favourite distro's installer slideshow completely translated for your locale.

Speaking of strings being changeable, if you see any errors, things that sound strange, things that are useless, things we need or things that you could say better, please file a bug or let me know. I keep a recent copy of things at http://people.ubuntu.com/~dylanmccall/ubiquity-slideshow-ubuntu

Michael Forrest and Otto Greenslade, from the design team, sent me a mockup and some graphics to fit with the exciting new Ubuntu branding. I worked them into the CSS and played with the text a little. Thus, behold, the proposed new look for the Ubuntu slideshow!
The whole thing can be seen on the web. It is a little bit larger than the one in Karmic.
Imagine it in Ubiquity, without a border, inside the awesome new Ambience theme, gracefully connecting to the title bar. Mmm...

Keep in mind this is still a proposal, just fully implemented (granted a few clunky bits).
Any constructive feedbacks or cries of "stop, you maniac!" are certainly not in vain :)

Thanks for listening!

Categories: Computers | Linux | nix | Planet Ubuntu | Ubuntu

Dylan McCall: Why menu bars are evil, take 54304

Seif Lotfy posted excitedly about Ian Cylkowski's in-depth Nautilus redesign mockup.

I found it very nice and refreshing, too. However, I noticed one strange thing: the menu bar just up and vanished. Ian`s incentive intrigued me:

Hidden Menubar: I'm not going to take sides here on whether we should still be having a menubar in applications or not; it's another minefield of opinions and flaming. I'm personally fine with a menubar inside the application, but I also happily use applications that tuck away the menubar under a single icon (think Google Chrome). But I do think that we should have the option here. In my mockup, all the menubar settings can be brought up with the settings icon (first icon after the pathbar). But if you would like to see the menubar permanently then this, too, should be an option.

Being one of those people who considers the slightest growth of an options dialog a crushing defeat, I just have to debate that. Hopefully this won't cause the flame-war Ian predicts. That would be a bad start for my first aggregated blog post. (Hi, Planet Ubuntu!)

What I would like to do is go over why Chromium's menus are great, and why they absolutely are not just "tucked away under a single icon" but intentionally designed that way and inseparable from that particular approach. It's about simplifying the menu so it relates to the two distinct subjects the user interacts with: the application and the current page.

There are millions of things I dislike about menu bars, but that would take all year to discuss. (And I have different things to fix). I'll just vent about two problems which relate closely to each other:

• They traditionally consume 100% width. A small menu bar looks wimpy. Half of that menu bar will inevitably be wasted space no matter how hard the developers try to cram stuff into there.
• There is a strange urge to duplicate everyone else's menus for consistency and have as many of these as possible. For some reason the uniqueness of an application is expected to vanish at the menu bar, which becomes an abstract world with words like "File" referring to web pages, photos, applications and video clips alike. Menu bars are popular things for accessibility, so I wonder if this abstraction helps in that regard or hinders.

Having said that, GNOME applications deserve some credit for usually replacing the title of the File menu with Music or Photos. Yet, the generic top-left-most menu is still there in spirit.

As we can see from the gnome-terminal screenshot, though, the File menu's spirit is weak. Challenge of the day: name one file related thing there, then explain why adding and editing profiles should be in different places.

So, Chromium is interesting because it is one of few applications to finally take a bold step against the '90s fashion in menu bars. First of all, the distinct lack of a menu bar clears up the interface considerably. Then they accept that, to an end user, the browser is not interacting with "files" but with web pages. So, the File menu becomes a Page menu. Completely. They also throw away all the baggage that the File menu used to have.
...and that is where the magic happens. Now the menu has a far more powerful, meaningful context. Actual useful functions can be put there that are of importance to people and relevant to what they are doing. Things like zooming, printing, copying and pasting inside pages.

The remaining stuff happens in the context of the Application menu (the wrench). Quitting, opening windows, changing preferences, dealing with bookmarks; anything that is related to the application as a whole, not the current page. Since Chromium was designed with a logical scope, that is everything else it does except extensions.

I think we can do a lot better then the traditional menu bar. Just yanking it out, though, won't do us much good.

Categories: Computers | Linux | nix | Planet Ubuntu | Ubuntu

Justin Dugger: New Homepage Design

I've decided my old homepage was bad enough to revisit now that I've got a bit more content hosted deep within it. I replaced my crappy hand written HTML with tools written this decade, and threw in some amateur visual design.

The software

Firstly, in order to keep the webpage fresh with little effort, I've chosen RSS aggregation as the method of content generation. Since I know Ubuntu and Debian both use Planet, that's where I first looked. But it seems Planet 2.0 is aging, and the fork Planet Venus brings some neat new options. It expands the selection of templates, adds a configurable RSS filter step, and makes the normalization step configurable.

It's also packaged in Ubuntu as planet-venus, making it fairly simple to set up. Deployment was a little tricky, as the package leaves most of the site configuration to the admin. You'll need a config.ini (I used /etc/planet/planet.ini), a template dir (/usr/local/share/planet-venus/theme), a cache dir (/var/cache/planet) and an output dir (somewhere in /var/www typically). Finally, you'll need to set up a cron job to run the static output generation script regularly. The script reads all the feeds and parameters in config.ini, caches the results to save bandwidth on subsequent runs, passes them to the template engine, and places the final product in the output dir.

When building a lifestream style site, you have to be picky about the kinds of feeds you put in or it gets Facebook / Twitter style spammy. This is where the RSS filter step can help; Planet Venus comes with a few filters like 'notweets', and a few stripAds filters to cleanse ads before republishing. It's the same design pattern I talked about before here with Liferea. In the future I could write one to add in comment feeds and then filter out everything that fails to meet some strong quality criteria.

Output templates

Planet Venus's real selling point to me is using Django templates. I've been meaning to learn Django for a while now, and this is a pretty good way to work with the templates portion of Django. And again, the filter pattern pops up. Here, filters take python variables as input; in Planet Venus's setup you have access to feed and item variables, as well as planetwide settings. One example filter might be to simply pluralize a word based on a variable (yes, you can even handle 'y' pluralization). Another example is the urlize filter that adds HTML anchor tags to likely URLs (not so great when you already have anchor tags in the filter's input).

I also use templates to generate an RSS feed. Nothing difficult about it, since the input to templates is basically an RSS feed to begin with. To reduce the probability of bugs, I translated a provided example htmptmpl RSS template into Django, and it's much smaller and clearer to me. Unfortunately, there's a bug in Planet Venus that prevents the use of multiple Django templates. I've reported it upstream, and I'm sure I can fix it or work around it.

Web Design

I also decided to take a look at CSS layout frameworks, to get up to speed on the subject quickly. 960.gs is popular, but it's 960 pixel width assumption works poorly with quirky resolutions found on massive monitors and smartphones. Luckily, I found found fluid960, which is very similar, but implements fluid layouts. It retains the CSS class names of 960.gs, so tutorials and documentation on one translate fairly well to the other. Which is good, because fluid960 pretty much relies on you already knowing regular 960 (I didn't). This presentation gives a good summary of things you might want a CSS framework for, and this 960 tutorial covers what I needed to know.

Color scheming is probably the hardest part for me. It's simple to pick a color pallate that goes together, but there is a higher level opportunity to communicate something through visual design. I could choose a purple scheme to reflect my collegiate experience, or an Ubuntu pallete, but it seems inappropriate for a personal site. I've got a bit of low level coding experience, so I could go with a green on black terminal theme, but it's been done to death ever since the Matrix, and it's basically impossible to beat jwz's version.

Since I'm not really looking to break into web design, I went with a relatively muted color scheme that organizes the content without distracting from it. Truthfully it doesn't matter all that much, as experience shows the majority of hits will come via RSS.

Well, that's basically all there is to my automated homepage system. On to more important things, like setting up a calDAV server or a feed processing tool.

Categories: Computers | Linux | nix | Planet Ubuntu | Ubuntu

The Fridge: Ubuntu as project in Google Summer of Code 2010

I’m very excited to announce that Ubuntu has applied as participating organisation in the Google Summer of Code 2010!

We submitted an organisational application, along with suggested ideas for potential projects for students. We also encourage students to come up with their own ideas.

If you’re a student interested in Open Source (or if you know students who are), now is the time to act to get involved in Google’s wonderful Summer of Code program.

Also, if you are thinking about becoming a student’s mentor please visit:

https://wiki.ubuntu.com/GoogleSoC2010

Make sure you read all the necessary information carefully and join the IRC channel and mailing list for more discussion.

The timeline is as follows:

• The list of accepted Mentoring Organisations will be announced on March 18, 2010 at 12 noon PDT / 19:00 UTC and will be posted on the Google Summer of Code 2010 site.
• The student application period begins March 29, 2010 at 12 noon PDT / 19:00 UTC and ends April 9, 2010 at 12:00 at 12 noon PDT / 19:00 UTC. Please see the Google Summer of Code 2010 timeline and FAQs for further information.

[Discuss Ubuntu and the Google Summer of Code 2010 on the Forum]

Originally sent to the ubuntu-devel-announce mailing list by Daniel Holbach on Fri Mar 12 09:59:16 GMT 2010

Categories: Computers | Linux | nix | Planet Ubuntu | Ubuntu

Mirco MĂĽller: OpenGL 4.0 specification

The Khronos group just announced the OpenGL 4.0 spec. That’s some major news! I still have to get my hands dirty with geometry-shaders, GPU-tesselation and OpenCL *sigh*

Categories: Computers | Linux | nix | Planet Ubuntu | Ubuntu

Benjamin Drung: Ubuntu units policy

We finally have an units policy in Ubuntu. I started to work on this issue over a year ago. The first step was to talk to other people (Ubuntu developers and upstream), but the opinions diverged. Neither a consensus was found, nor any result came out of it (except heated discussions). Upstream was not willing to change anything. It was time to contact the Technical Board to get a decision for Ubuntu.

Now we have the policy and we can start filing bug reports and fixing them without discussions about the reasonability of the patches. Let’s get Ubuntu 10.04 (lucid) in shape!

Categories: Computers | Linux | nix | Planet Ubuntu | Ubuntu

Kees Cook: openssl client does not check commonName

I realize the openssl s_client tool tries to be upper-layer protocol agnostic, but doesn’t everything that uses SSL do commonName checking (HTTP, SMTP, IMAP, FTP, POP, XMPP)? Shouldn’t this be something openssl s_client does by default, maybe with an option to turn it off for less common situations?

Here it doesn’t complain about connecting to “outflux.net” when the cert has a CN for “www.outflux.net”:

echo QUIT | openssl s_client -CApath /etc/ssl/certs \
  -connect outflux.net:443 2>/dev/null | egrep "subject=|Verify"

subject=/CN=www.outflux.net
    Verify return code: 0 (ok)

Categories: Computers | Linux | nix | Planet Ubuntu | Ubuntu

Amber Graner: Hug Day/ Bug Day - Targeting Ubiquity





Categories: Computers | Linux | nix | Planet Ubuntu | Ubuntu

Benjamin Drung: Translations of package descriptions

I saw the Translations of package descriptions video from FOSDEM 2010. Every distribution has two translatable string for each package: a synopsis (summary, short description) and a long description. These descriptions differ from distribution to distribution. The descriptions should be shared between the distributions. This will enable us to share the translations of the descriptions.

My idea: Why not letting upstream provide the package description and the translation for it? They should have the knowledge to provide a good description and to update it if required. To encourage upstream to provide the description, we should create a freedesktop specification for it. Quick draft: The tarball should contain a file named package.info. The package.info file should contain three RFC-2822-like fields for each package: Package, Synopsis, and Description. Translation can be stored in package.info.<language> (for example package.info.de).

Example package.info:

Package: audacity
Synopsis: A fast, cross-platform audio editor
Description: Audacity is a multi-track audio editor for Linux/Unix,
 MacOS and Windows. It is designed for easy recording, playing
 and editing of digital audio. Audacity features digital effects and
 spectrum analysis tools. Editing is very fast and provides unlimited
 undo/redo.
 .
 Supported file formats include Ogg Vorbis, MP2, MP3, WAV, AIFF, and AU.

What do you think about this idea?

Edit: Alexandre Franke found an existing markup language designed for our use case: Description of a Project (DOAP). Package is name there, synopsis is shortdesc, and description is description. Here is my example in DOAP:

<Project xmlns="http://usefulinc.com/ns/doap#"
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:foaf="http://xmlns.com/foaf/0.1/">
  <name>Audacity</name>
  <shortdesc xml:lang="en">
    A fast, cross-platform audio editor
  </shortdesc>
  <description xml:lang="en">
    Audacity is a multi-track audio editor for Linux/Unix,
    MacOS and Windows. It is designed for easy recording, playing
    and editing of digital audio. Audacity features digital effects and
    spectrum analysis tools. Editing is very fast and provides unlimited
    undo/redo.
    Supported file formats include Ogg Vorbis, MP2, MP3, WAV, AIFF, and AU.
  </description>
</Project>

Categories: Computers | Linux | nix | Planet Ubuntu | Ubuntu

Og Maciel: GNOME Developer Kit, follow up

Just wanted to update everyone who showed interest in the new release of GNOME Developer Kit I announced yesterday. Based on some preliminary statistics I collected in the (less than) last 24 hours, it seems that the VMware image type got the most download, followed closely by the installable ISO format. I guess that was due to VirtualBox being able to use *.vmdk files and some people opting for the free virtualization tool.

Here are the preliminary results so far:

1. VMware image: 42 downloads
2. Installable ISO: 26 downloads
3. RAW filesystem image: 17 downloads

Due to the number of downloads and and comments I received, I felt that I should provide with some background on how to install/remove packages and update your system using the conary package management system. So here you go:

The package management system behind the GNOME Developer Kit is called conary and is considered by many as the next generation package management system when compared to some of the popular options out there. One of the reasons behind this claim is the fact that your entire system is actually completely maintained in a versioned state, and conary is always “aware” of what is installed on your system and what files and dependencies make up the entire “set”. This allows for some pretty nifty operations such as rolling back to a specific state of your system.

In order to check for new updates for your system, open a terminal and run the command sudo conary updateall. conary will then check for updates and prompt you to accept the update or not. Please keep in mind that the first time you run conary for the first time, you will experience a delay as your entire system gets analyzed in preparation for the changes that are to take place. All subsequent actions performed will be much faster, I promise. If after a while you don’t feel like waiting for the prompt, add –no-interactive to the update command to have your system updated automatically.

Now, let’s just say that you decided to install something new, such as Banshee. Easy, just run sudo conary update banshee (remember to add –no-interactive for no-hands updates) and voilá!

Want to know what was actually installed on your system? conary q banshee will tell you what version of banshee was installed. How about what files were installed? conary q –ls banshee will give you a list of all the files that were installed and conary q –lsl banshee will give you the long list with file permissions and modes.

Changed your mind and want to remove banshee from your system? sudo conary erase banshee will take care of that. Want to actually roll your system back to the state it was before you installed banshee instead? sudo conary rollback 1 will rollback your system exactly one transaction. Want to go further back? Just increase that number to represent how many transactions to roll back. Want to rollback but don’t remember what point in time you want to go? sudo conary rblist will display a list of all transactions and what was changed. Note that each transaction is preceded by the letter “r“, so if you want to rollback to the point r.15, then use sudo conary rollback r.15 (and don’t forget that “r” or you’ll rollback exactly 15 transactions instead).

How about searching for a package? If it is something that it is already installed on your system, then conary q [package name] will give you the information you want. If the package is not installed on your system yet, then conary rq [package name] is what you need, though since conary does not yet make use of metadata, you’ll need to know the exact name of what you’re looking for. Now, let’s say you want to find out what package provides the command /sbin/service? Use conary q –path /sbin/service to find out that initscripts:runtime=8.81.2-0.11-1 is responsible for providing it (use rq if you want to search the remote repository).

Well, I think this is enough to get you going. You’ll probably want to install Flash and media codecs to enjoy browsing some sites and listening to your media, so let’s apply what we’ve learned so far and run: sudo conary update flashplayer group-codecs

If you’ve stayed with me until now, you may want to read up on what else conary can do or even consider packaging for GNOME Developer Kit. Your help will be greatly appreciated!

Categories: Computers | Linux | nix | Planet Ubuntu | Ubuntu