In Scrum we trust

we had a Scrum training in the last 2 days with Boris Gloger. It was clarifying to have this training as most of the participants were already using Scrum actively in their projects. Now we know what we did wrong and what could be improved.

the points that catched most of my attention was the rule of working on one backlog item at a time (of course you need to be reasonable here, as there might be situations where it is necessary or better to parallelize, but as a general statement you should avoid that).

another point that confused me was the sizing of the backlog, we previously sized the items in days, but we got convinced that sizing should not be dependent on time as you want to estimate how big your item is and not how fast you can solve it.

one interesting practice to sizing is the poker play: where you use cards with the values of an almost fibonacci sequence to "size" your backlog items among your team.

a very important point is to define beforehand the acceptance criteria with the Product Owner and also check what went well and what could be improved by the end of each sprint.

according to the curse, for having scrum running efectively you need to have:

1. Timebox
2. Self organization
3. Full principled
4. Potential shippable code
   

and for documentation, it is basically the product backlog (PB) & the sprint backlog. \o/

as an interesting side note, the word backlog refers to the containers that are transported by ship and this backlog is also ordered by importance: the most important containers are placed in the bottom in case there is any storm.


the things we are going to improve in our team for the next sprint: do the daily scrum meetings DAILY, time boxed and on time, work on the same backlog item until it is really done before moving to the next one and doing the sprint planing 1, . planning 2, review & retrospective in order to improve after each sprint. In order words, just apply Scrum more strictly :)

the main benefit of the training was to understand the principles of the Scrum methodology. With that in mind you should be able to adapt it to your day to day work, without loosing the agile principles.

Canola beta9

the new beta 9 of canola is out! in this release the main new additions are the UPnP plugin and the "On the Move" playlist.

I find these two new features extremely useful: now you are able to create and edit a dynamic playlist on the go and access your media from your local network seamlessly.

besides that, there is also the new canola-tuning plugin, which enables you to get the cover arts and video thumbnail from inside canola´s settings (no need to run the canola-tuning app anymore).

one really old bug that was finally solved is the PNG thumbail generation! this was a long time on my todo list, but with low priority. the fix included a patch for epsilon, the EFL thumbnail library, and I took the opportunity to send some patches for epsilon upstream.

hope you enjoy, there is more to come ...
br

Fullwork at Canola 2.0 pre-release

we are now working at full speed in order to release canola 2.0. a lot of new features are being implemented, and unfortunately we are not having much free time to report back to the maemo community our work in progress.

I have been working since the beginning of the project with the download manager, the new canola media player, code-named "atabake" and more recently with the UI of the photos screens.

next week we will start posting some screenshots of the features of canola 2.0: podcasts, media player and photos browsing have been through a major "ramp up"

and despite the new look and feel and the features added, canola 2.0 keeps the original spirit: a simple, yet powerful and sexy portable media aggregator.

hope you enjoy the outcome of this project!

Work in Progress: New Canola

hi,

after a long time without updates on this blog, I will give an overview on my current tasks involving the new ramp-up of Canola.

there is a lot of work going on around Canola, as you can see in various posts from Marcelo, Barbieri and Kenneth! That´s an exciting time for good changes and new features :)
Besides the adoption of evas and edje as the UI toolkit, Canola is being rewritten in python and being reviewed from an architectural and functional point of view.

in the first semester I have worked on a cache and download manager for providing easy access to offline content, such as podcast and images, etc. Now I am working together with Artur (MoRpHeUz) on a media-engine for Canola. The main purpose of this media-engine is to provide simple access to multiple backends, such as MPlayer, GStreamer, the current Osso media server and others.

we all know that the default media-player has a lot of restrictions to supported formats, that´s why we believe that multiple media-backends is the way to go for applications that want to maximize the features for the end user. The new media engine will be able to select the best backend automatically for each mime-type, combining MPlayer, for open formats, with GStreamer (we are still waiting for playbin2 to come out) or the current media-engine for proprietary formats and DRM protected content.

in addition to support a wide range of media formats, this new engine will also manage sessions, allowing applications to save/restore state, such as remembering the last url played along with the position. Moreover, the session architecture will end with today´s mess of various applications using and confusing the osso-media-server.

what all applications want is a simple and reliable method of managing the playback of their media and that is what we will do!

BR