Gustavo Sverzut Barbieri

I’m often introducing new developers to Evas, part of the Enlightenment Foundation Libraries (EFL), and one common problem is to understand how Evas, Smart Objects and Edje integrates. Yesterday, while teaching some guys here at ProFUSION I came to the following picture, that describes it well:

Evas exposes just a handful object types, like solid color rectangle, gradient rectangles, images, text, text block, polygon, line and smart object.

Smart Object is the only special type because it’s extensible and developers should implement its virtual methods to have it to behave like regular objects with calls like evas_object_color_set() or evas_object_resize(). Smart Objects do not have any visual representation, they can contain other objects that may have it, like images or rectangles, that’s why they’re represented like a box in the figure above. Their children are handled as if they were in a special layer: outside objects cannot interleave smart object children. It’s recommended as good practice to not expose children to outside, instead provide methods to manipulate these children.

A regular mistake is to forget that just creating objects from inside smart object methods don’t make them children! You must assign a parent to a child with evas_object_smart_member_add(child, parent). You can “unparent” them with evas_object_smart_member_del(child). As children can not have more than one parent, adding a new parent to the child will first remove its parent before adding the new one.

Edje is a smart object that loads a group of parts from a description file often called “EDJ” (or edje file). These files contains a collection of groups, each with a set of parts (evas objects), each part with various states. When a group is assigned to some Edje object with edje_object_file_set(object, file_name, group_name) the parts are created in order (first will go first, thus will stay at the bottom) using the "default" 0.0 state. If some part have relative positioning to the whole available space (Edje object geometry), when the edje object is changed, all the dependent parts are recalculated and have their geometry set. If some part is relative to other parts, then when one of these changes, the dependent part is recalculated and reset as well. One can manipulate edje by changing the state of the parts, this should be done by means of programs, which will act based on signals. Edje will dispatch some signals like "resize", "mouse,move" and "show", but users are encouraged to create their own application signals that can be emitted with edje_object_signal_emit(object, "signal", "source").

The most common mistake using Edje is to touch its children. Never, ever, change the object that you get with edje_object_part_object_get(). That call is for inspection purposes only. You’re just allowed to change Edje by emitting signals from your application, EDJ should have programs to catch that signal and trigger some state change on some parts. Let’s remember that it’s recommended to never changed smart object children, Edje is strict, making this a rule. What can you expect if you break this rule? Before, I said that Edje would recalculate dependent parts when the dependency changes, but I lied: for simplicity purposes, when something change, Edje recalculates the whole thing, and it does so using the state information from EDJ, not the current value of the object. If you change the color of some rectangle from white to black, recalculation will enforce the rectangle to be white. Same for images, if you set the image file to be something else, it will be restored to the image specified in the EDJ.

One variation of this mistake is dealing with SWALLOW parts. These parts enable user to add an external object to be managed by Edje with the edje_object_part_swallow(object, "part_name", external_object). As said, Edje will manage this external object: it will become the object’s parent, it will set clippers, colors, visibility and geometry. After that call, user don’t own the object anymore, he is only allowed to change other, non managed properties, ie: images can have their file set because SWALLOW parts don’t have any image property (unlike IMAGE parts!)

It’s very important to understand these concepts. Real applications are a mix of smart objects and edje, one inside the other, which might led to great confusion if management ownership is not clear.

In the figure above we can see the management ownership to avoid these problems. User manages both the smart object and the background. In order to change the edje he should call some API defined in the smart object. To change the “star” image, user should call smart object that would call edje.

You can also download the printable version of the figure.

You might know that INdT released Youtube plugin as free software, it’s great feature-wise and touches almost every part, providing new models, views and controllers, those with complex use cases like threaded models so GUI will not block during HTTP requests and even options menu.

That’s cool because one can do lots of things (and some users are already showing us some nice plugins!), but we still need some base text introducing people to the concepts, with smaller code, so here it is: Canola URLBookmark source code and text.

This introduces you to some concepts, explains about “plugins.info” and how plugins are loaded and in the end you have a list of URL to play. Of course this hard coded list of URL is on purpose so you take some time to change it to something more useful. If you ask me, I’d like to see UPnP, MPD, Samba, Avahi, Shoutcast and lots more.. I did my part, everything you need to know is there, now it’s your turn.

So it’s that time of the year again, almost summer in North, winter here in South and Google helping free software projects with its Summer of Code. I’m glad some projects I’m involved were accepted, including: Enlightenment and BlueZ.

I’ll be a mentor for Enlightenment and we have great ideas, if you’re interested in them, mail me or go to #edevelop @ irc.freenode.net so we can discuss your ideas, experiences… It’s a great way to get involved in computer graphics and a platform that is growing everyday on mobile systems, with adopters like Canola2 and now OpenMoko!

As for BlueZ, I’m not mentoring, but some friends are, they also have great ideas, things that will benefit every GNU/Linux bluetooth user, from mobile to desktops to laptops, some are really interesting like better audio support. It’s a way to get into kernel and low-level user space world, and get paid for it

So BossaConference ‘08 is over, what a great conference! Lots of great people, some are still around, doing some hacking with us at INdT office, it’s really great to have some time to discuss new ideas, drink some beers and play jokes

Let’s hope next year we can keep it to the level! Congrats to all the organization members.

My old website http://www.gustavobarbieri.com.br/ is now dead and points to this blog. It was ugly and hard to keep updated, using wordpress blog is much easier than hand editing HTML and provide lots of useful features, like rss feeds.

If you need some of the files hosted there, please use http://www.gustavobarbieri.com.br/old-website, although I did keep the links to most useful folders, like http://www.gustavobarbieri.com.br/eagle, http://www.gustavobarbieri.com.br/jogos and http://www.gustavobarbieri.com.br/python

So I was helping one Canola2 user to uninstall the old version and for some strange reason apt-get remove libeet0 libevas0 libecore0 libembryo0 libdownloadmanager0 was breaking with “Abort” message. Ok, use dpkg instead, I said, but since we now have split packages for all the libs we use, you’ll end with a dependency nightmare.

Solution? Hack a quick script to get dpkg errors, parse them and generate a new command line with proper ordering:

#!/usr/bin/python

import sys

pkgs = {}

infile = open(sys.argv[1])
pkg = None
for line in infile:
    line = line[:-1] # chomp n
    tokens = line.split()
    head = tokens[0]
    if head == “dpkg:”:
        if tokens[1] != “dependency”:
            continue
        pkg = tokens[-1][:-1]
        pkgs.setdefault(pkg, set())
    elif head in (”Package”, “dpkg”, “dependency”):
        continue
    elif head == “Errors”:
        break # follows a list of problematic packages
    else:
        if tokens[1:3] == ['depends', 'on']:
            pkgs[pkg].add(head)

def unique_extend(lst, extent):
    for e in extent:
        if e not in lst:
            lst.append(e)

def rm_pkg(p, pkgs):
    rm_list = []
    try:
        ddeps = pkgs[p]
    except KeyError:
        return [] # no deps!

    for d in ddeps:
        unique_extend(rm_list, rm_pkg(d, pkgs) + [d])
    return rm_list

rm_list = []
for p in pkgs:
    unique_extend(rm_list, rm_pkg(p, pkgs) + [p])

print “dpkg –purge”, ” “.join(rm_list)

Not that efficient, but simple enough.

After another heavy working week, we managed to deliver a new beta of Canola2, this time with lots of bug fixing (thanks for reporting!), features and optimizations. Also features 770* support, a new tool to merge albums (collections or albums with similar name) and lots of new media formats (wma/wmv and real media included!).

UPDATE: Since we replaced our download manager, previous packages conflicts! Please remove Canola2-beta1 before installing the new one. If you got it wrong, apt-get -f install from Xterm should fix it.

Hope installation will be smoother this time!

More information at: http://openbossa.indt.org/canola2/

* 770 support: gregale is regular install, but for Hacker Edition (bora ported to 770) you need a special proceeding (ain’t you hackers!?):

1. disable all repositories
2. add the repositories:
   1. deb http://repository.maemo.org gregale free non-free
   2. deb http://repository.maemo.org/extras gregale free non-free
3. update list
4. install canola2
5. remove gregale repositories

These steps are required because Hacker Edition kernel doesn’t provide inotify and we compile bora EFL packages optimized for n8xx (omap 2420, arm1136jf-s).

Thanks to Levi Bard we now have ruby-lightmediascanner, with even API docs!

On the bug fixes front, after Canola2 was released and it was tested with multitude of media I could fix lots of bugs with parsers, mainly on jpeg and id3lib.

Last but not least, thanks to Renato Chencarek we now have ogg support, and I also added png and audio-dummy to register known media files we still have no specific parsers (real, wma, …)

It’s already stable, when I have time I’ll add a non-dual process version of process and check, so people with “correct” media files can use this and benefit from a much faster version (however this will rely on developers adding support to choose both implementations).

And so the time has come, almost 1 year since the first Bossa Conference 2007. That was an amazing conference, with exuberant place near the beach, great speakers and lots of time to enjoy both. Since the scope is well defined: Linux and other Free Software in mobile devices, the speakers are really well selected, we could manage to have free time to talk to each other, without going from one side to another, running like crazy and if you have attended conferences before you know what I mean: these casual chats are the best way to know about new features and plans (maybe unofficial or secret ;-)), elaborate new ideas, etc.

That conference was so great that INdT is now promoting another, actually much improved:

Come and enjoy a innovative conference on Open Source Software for Mobile Embedded Platforms. Located in Porto de Galinhas, one of the tropical paradises of Brazil, the Bossa Conference will happen during 16^th to 19^th of March at Summerville Beach Resort. The perfect combination of developers, pleasant environment and good ideas.

Let’s go! We’re waiting for you.

Information and registration: www.bossaconference.org.

INdT released today under LGPL the public release of LightMediaScanner (LMS), a lightweight media scanner that will be used by Canola 2.0 and, we hope, many other multimedia software!

Lightweight media scanner meant to be used in not-so-powerful devices,
like embedded systems or old machines.

Provides an optimized way to recursively scan directories, handling
the parser in a child process, avoiding breaks of the main process
when parsers break (quite common with such bad libs and tags).

Parsers are plugins in the form of shared objects, so it’s easy to add
new without having to recompiling the scanner.

The scanner will use SQLite3 to store file-mtime association, avoiding
parsing files that are already up-to-date. This SQLite connection and
the file id within the master table ‘files’ are handled to plugins for
relationship with other tables.

Check out the code, it’s really simple to use and write new parser plugins:

• C API
• Python API
• SVN
• GIT - LightMediaScanner
• GIT - Python-LightMediaScanner