If you have been dowloanding code from the mercurial repos, you would have notice some spooky coincidences… most package MCLs update all at the same time!
Well you can relax, our repos are not posses. This is the result of Nokia’s package owners delivering contributions to our repos in a centralised manner.As you can imagine delivering updates to over 30 million lines of code every other week is a rather complexed operation.
We have been Working with Nokia to improve the contribution channel to make it reliable (we now consistently recieve contributions every 2 weeks). We are now moving towards the next big step, an automated package-base publishing system. You can now see the first live pilot on the access security package repo. Continue reading “Improving Code Contributions”→
I have previously blogged about the quantity and quality of features that the community is planning for Symbian^3, but how about Symbian^4?. Let’s have a recap on where we are today!
Symbian^3 is almost there
Symbian^3 is nearing Functionally Complete (FC is in Q1 2010, likely to be February) and the contributions have come fast and thick over the last month. Of the 43 package feature tracked in the integration plan for Symbian^3, 30 have already been contributed to the foundation code line. These are:
Multipage Homescreen – Provide multiple home screen supports
3PC – Bearer mobility support in 3PC and adoption
WDP – Proven Writeable Demand Paging platform support
HD video – Support for files of over 2GBs that will enable HD video
One-click connectivity – Simpler connection dialogue that requires one click only
Song recognition and music store integration with radio application
Remote contact look-up – plug-in framework to allow easy integration of remote contact look-up
You should expect all of them to be included in the S^3 PDK3.0.f (planned for week03). In addition, QT for Symbian is now available from the foundation repositories under LGPL and it will be included in future PDKs (Product Development Kits).
I have been using Mylyn now for a while and it has been great to help keeping on top of large amounts of bugzilla entries. I’ve used it to manage the integration plan for the Symbian platform, which describes the expected contributions into the platform.
One of the benefits that Mylyn offers developers is the reduction on Context Switching. Context switching is not only costly for software programs, but also for humans working on concurrent tasks. Mylyn provides a great integration with the IDE that allows developers using Eclipse to substantially reduce the wasted time on switching between application and work tasks.
However, I am not a developer… so Mylyn did not really provide me with any improvements in this area. Hence, I decided to download the Tasktop 30-day-trial standalone version and see if I could have get some time-saving by exploring the additional features for “Task Context”. Here are the results:
The first obvious advantage is that it doesn’t really load all the rest of Eclipse/Carbide functionality that I don’t use and that eats a substantial amount of my manager’s-spec laptop memory. Hence, the first improvement is better working speed!
With the aim to share and highlight as much relevant information as possible, I would like to bring to your attention the week44 Release Plan! We have regularly published updates to release plan since May(ish) in the Release Council. In an effort to bring it up to the surface, we have created a landing page for it.
I have refered on the past to the capability of reporting metrics against the Symbian Platform. Finally we are there!
This week the team produce the first snapshot of Bug , Code size and Code Churn with real data. The snapshot is to be presented on Friday to the Release Council for feedback. Let’s have a look at it in more detail:
These are graphics auto-generated from Bugtracker Metrics, it allows you to drill down to as much detail as you want. The data in the snapshot is frozen from an specific date, but you can re-run the queries live from the bugtracker metrics interface.
Now that we can measure it, we can start introducing some standard defect management practices.
One thing I keep getting asked is “where do you get the data for the integration plan?”. In the past, I have mentioned that how features are added to integration plan, but from today I am publishing together with the pan: the raw data used to create it, the scripts that provide the data and filters that we used to manage the information available.