Jump to content

KDE PIM/Meetings/PIM Winter 2014 meeting: Difference between revisions

From KDE Community Wiki
Carlsymons (talk | contribs)
 
(19 intermediate revisions by 7 users not shown)
Line 1: Line 1:
= Topics =
= Topics =
===Usability - User stories and scenarios===
===Usability - User stories and scenarios===
Enable user-centered design through user stories and scenarios based on actual users form the city administration of Munich.
Enable user-centered design through user stories and scenarios based on actual users from the city administration of Munich.


Interested: Michael Bohlender; Thomas Pfeiffer
Interested: Michael Bohlender; Thomas Pfeiffer; Achim Bohnet


===QML API===
===QML API===
Line 22: Line 22:
=== Mail Search Testing ===
=== Mail Search Testing ===
Searching within emails still has problems for quite a few people. We could use the collective brainpower present at the sprint to try to get to the bottom of the problem.
Searching within emails still has problems for quite a few people. We could use the collective brainpower present at the sprint to try to get to the bottom of the problem.
Ideas: https://techbase.kde.org/Projects/Usability/HIG/SearchPattern


=== Frameworks release schedule & Repository divisions ===
=== Frameworks release schedule & Repository divisions ===
Line 31: Line 33:
I think we should take a high level overview of what parts we have in kdepim, how they fit together and how this structure helps us getting done what we want in the future.
I think we should take a high level overview of what parts we have in kdepim, how they fit together and how this structure helps us getting done what we want in the future.


This becomes especially relevant with changes in akonadi, the QML API etc. We need to know why we implment what where, and be able to draw clear lines between the various components in order to avoid them getting to complex. For this we need to have an idea how the architecture should look like that we are moving towards (even if this cannot be achieved as short-term goal).
This becomes especially relevant with changes in akonadi, the QML API etc. We need to know why we implement what where, and be able to draw clear lines between the various components in order to avoid them getting to complex. For this we need to have an idea how the architecture should look like that we are moving towards (even if this cannot be achieved as short-term goal).
 
Interested: Christian Mollekopf; Daniel Vrátil; Aaron Seigo; Achim Bohnet
 
=== Async library ===
Current async code sucks and we need better tools. I'd like to have a brainstorming session on that so we can stop writing async handlers all over the place. Things that I'd like to see:
* Composable jobs/tasks
* higher order functions to be used with the composable jobs (map/filter/...)
* reactive collections
* transformations in reactive collections
 
These building-blocks should result a generic KAsync (?) library. I'd like to see PIM spearheading this (I will anyways), since we are writing loads of async code.
 
Interested: Christian Mollekopf;
 
=== Windows / Mac ===
 
Windows build scripts to be found in:
 
 
The overlay for the stable branches of kdepim from Kolab Systems are here:
 
[email protected]:kdab-portages.git


Interested: Christian Mollekopf; Daniel Vrátil; Aaron Seigo
In order to use the overlay, add this to your etc/kdesettings.bat:
 
set EMERGE_PORTAGE_ROOT=C:\<PATH TO YOUR CHECKOUT>\kdab-portages;%KDEROOT%\emerge\portage
 
These scripts point to:
 
https://github.com/kolab-groupware
 
and of course to upstream KDE 4.13 git.


=Agenda=
=Agenda=
==Saturday==
* Mobile / Active
* Akonadi Next
* Email Search
* Windows / Mac
* User Stories (LiMux)
* Vision
==Sunday==
* Repository Split / Merge
* Frameworks 5 Port
* QML API
* Devel & Release Cycle
* Async Library
=Meeting Notes=
=Meeting Notes=
https://notes.kde.org/p/PIM_Sprint_Winter_2014
== Email Search ==
It turns out that the biggest problem with search is that the full-fledged search form has little to do with how search is currently implemented using Baloo.
So while getting the two in line again, we could take the opportunity to improve the search experience for users.
The conclusion was that we should implement search the following way:
* There is a single search bar (which can be used to look either in the current folder or everywhere).
* The search bar takes natural language queries using the natural language query parser by Denis Steckelmacher ([[http://steckdenis.be/post-2013-07-24-integrating-the-query-builder-into-dolphin.html | blog post]]))
* When the user starts to type into the box, they get presented with suggestions for query expressions like "Who, When, What" in a popup layer, and when choosing one of them gets suggestions for more detailed specifications like "From/To", "Subject/Body/Everything" etc.
* The goal is to teach users how they can enter their queries directly without having to click through a GUI
* Results are shown immediately in the same view, which jumps to a search folder
* There is an option to only search in the current folder, by default it searches everywhere
== Frameworks 5 Port ==
* kpimutils will die
* kmime parts => kcodecs
* gpgme single framework (keep pure c++ library)
KDEPIM builds with Qt5!
== User Stories ==
* Michael can do one month of research how a company uses mail to generate user stories.
== Vision ==
The group agreed that it does make sense to create a vision that guides our work. It was decided that Akonadi/kdepimlibs and Kontact should each have their own visions. Some draft ideas can be found in the sprint notes, but they still have to be refined and then agreed upon by the community.


=Organization=
=Organization=
Line 43: Line 116:


LiMux Office
LiMux Office
Sonnenstrasse 25
Sonnenstrasse 25
80331 Munich Altstadt-Lehel
80331 Munich Altstadt-Lehel


From Hauptbahnhof (main station) take the U1/U2 to Sendlinger Tor or the U4/U5 to Stachus (Karlsplatz). Follow the Sonnenstraße until you see the green sign of the Goethe Institut. Enter the passage. The entry is on the right-hand side. The LiMux office is in the 6th and 7th floor of the building; the Sprint is on the 7th floor. Doors should be open, but there will be several Sprint info sheets with phone contacts and directions.
From Hauptbahnhof (main station) take the U1/U2 to Sendlinger Tor or the U4/U5 to Stachus (Karlsplatz). Follow the Sonnenstraße until you see the green sign of the Goethe Institut. Enter the passage. The entry is on the right-hand side. The LiMux office is in the 6th and 7th floor of the building; the Sprint is on the 7th floor. Doors should be open, but there will be several Sprint info sheets with phone contacts and directions.
Line 70: Line 146:
'''Hauptbahnhof -> Hostel:''' take any any S-Bahn except S7 towards "Pasing or Laim" exit at "Laim".
'''Hauptbahnhof -> Hostel:''' take any any S-Bahn except S7 towards "Pasing or Laim" exit at "Laim".


'''Hostel -> Venue:''' take any S-Bahn towards "Hauptbahnhof" (main station). Ride 5 stations till you reach "Karlsplatz/Stachus".
'''Hostel -> Venue:''' take any S-Bahn towards "Hauptbahnhof". Ride 5 stations till you reach "Karlsplatz/Stachus".


===Catering===
===Catering===
Line 79: Line 155:


There'll be a continental oriented breakfast for Sat + Sun available; some rolls, bread, hazelnut spread, jams, a bit of cheese and cold meat.
There'll be a continental oriented breakfast for Sat + Sun available; some rolls, bread, hazelnut spread, jams, a bit of cheese and cold meat.
===Pictures===
You can Download and Upload pictures from the event here: https://apollo.geekaliens.com:8001/public.php?service=files&t=b91dfb35065c1c3aaa96841efb38cd9a
You can use the WiFi password from the event.

Latest revision as of 00:56, 9 February 2015

Topics

Usability - User stories and scenarios

Enable user-centered design through user stories and scenarios based on actual users from the city administration of Munich.

Interested: Michael Bohlender; Thomas Pfeiffer; Achim Bohnet

QML API

Discuss the future (architecture) of QML APIs for PIM stuff.

Interested: Michael Bohlender; Daniel Vrátil, Christian Mollekopf

Active Mail/ KMail Mobile

Discuss the future of Active Mail.

Interested: Michael Bohlender; Daniel Vrátil; Thomas Pfeiffer

Akonadi Framework

We already discussed this a lot during Akademy, this time I'd like to look into solving specific problems that Kolab guys have run into while preparing Akonadi and KDE PIM for Munich. I also want to actually sit down and design some of the major features/changes we are planning.

Interested: Daniel Vrátil; Christian Mollekopf

Mail Search Testing

Searching within emails still has problems for quite a few people. We could use the collective brainpower present at the sprint to try to get to the bottom of the problem.

Ideas: https://techbase.kde.org/Projects/Usability/HIG/SearchPattern

Frameworks release schedule & Repository divisions

Another discussion on how we exactly split up the repositories for frameworks (we know already many parts, but some are still left). Additionally we should discuss options for the release schedule.

Interested: Christian Mollekopf; Daniel Vrátil; Aaron Seigo

PIM Architecture and Design

I think we should take a high level overview of what parts we have in kdepim, how they fit together and how this structure helps us getting done what we want in the future.

This becomes especially relevant with changes in akonadi, the QML API etc. We need to know why we implement what where, and be able to draw clear lines between the various components in order to avoid them getting to complex. For this we need to have an idea how the architecture should look like that we are moving towards (even if this cannot be achieved as short-term goal).

Interested: Christian Mollekopf; Daniel Vrátil; Aaron Seigo; Achim Bohnet

Async library

Current async code sucks and we need better tools. I'd like to have a brainstorming session on that so we can stop writing async handlers all over the place. Things that I'd like to see:

  • Composable jobs/tasks
  • higher order functions to be used with the composable jobs (map/filter/...)
  • reactive collections
  • transformations in reactive collections

These building-blocks should result a generic KAsync (?) library. I'd like to see PIM spearheading this (I will anyways), since we are writing loads of async code.

Interested: Christian Mollekopf;

Windows / Mac

Windows build scripts to be found in:

[email protected]:emerge.git

The overlay for the stable branches of kdepim from Kolab Systems are here:

[email protected]:kdab-portages.git

In order to use the overlay, add this to your etc/kdesettings.bat:

set EMERGE_PORTAGE_ROOT=C:\<PATH TO YOUR CHECKOUT>\kdab-portages;%KDEROOT%\emerge\portage

These scripts point to:

https://github.com/kolab-groupware

and of course to upstream KDE 4.13 git.

Agenda

Saturday

  • Mobile / Active
  • Akonadi Next
  • Email Search
  • Windows / Mac
  • User Stories (LiMux)
  • Vision

Sunday

  • Repository Split / Merge
  • Frameworks 5 Port
  • QML API
  • Devel & Release Cycle
  • Async Library

Meeting Notes

https://notes.kde.org/p/PIM_Sprint_Winter_2014

Email Search

It turns out that the biggest problem with search is that the full-fledged search form has little to do with how search is currently implemented using Baloo. So while getting the two in line again, we could take the opportunity to improve the search experience for users.

The conclusion was that we should implement search the following way:

  • There is a single search bar (which can be used to look either in the current folder or everywhere).
  • The search bar takes natural language queries using the natural language query parser by Denis Steckelmacher ([| blog post]))
  • When the user starts to type into the box, they get presented with suggestions for query expressions like "Who, When, What" in a popup layer, and when choosing one of them gets suggestions for more detailed specifications like "From/To", "Subject/Body/Everything" etc.
  • The goal is to teach users how they can enter their queries directly without having to click through a GUI
  • Results are shown immediately in the same view, which jumps to a search folder
  • There is an option to only search in the current folder, by default it searches everywhere

Frameworks 5 Port

  • kpimutils will die
  • kmime parts => kcodecs
  • gpgme single framework (keep pure c++ library)

KDEPIM builds with Qt5!

User Stories

  • Michael can do one month of research how a company uses mail to generate user stories.

Vision

The group agreed that it does make sense to create a vision that guides our work. It was decided that Akonadi/kdepimlibs and Kontact should each have their own visions. Some draft ideas can be found in the sprint notes, but they still have to be refined and then agreed upon by the community.

Organization

Venue

LiMux Office

Sonnenstrasse 25

80331 Munich Altstadt-Lehel


From Hauptbahnhof (main station) take the U1/U2 to Sendlinger Tor or the U4/U5 to Stachus (Karlsplatz). Follow the Sonnenstraße until you see the green sign of the Goethe Institut. Enter the passage. The entry is on the right-hand side. The LiMux office is in the 6th and 7th floor of the building; the Sprint is on the 7th floor. Doors should be open, but there will be several Sprint info sheets with phone contacts and directions.

We can enter the Limux office at Fr 12:00 but the official start of the event is not until 18:00. There will be some noise as the Limux guys are transforming their regular working place into a meeting space.

Accomodation

A&O München Laim

Landsberger Straße 338

80687 Munich Laim


5 minute walking Distance form "Laim" S-Bahn station.

Checkin from 15:00. Checkout until 11:00.

Transit

Airport -> Hostel: take the S1 S-Bahn and exit at "Laim".

Hauptbahnhof -> Hostel: take any any S-Bahn except S7 towards "Pasing or Laim" exit at "Laim".

Hostel -> Venue: take any S-Bahn towards "Hauptbahnhof". Ride 5 stations till you reach "Karlsplatz/Stachus".

Catering

We're stacked with a bunch of different non-alcoholic beverages and snacks (sweets, fruits, ...). Coffee and tea will be prepared on demand.

There will be 3-4 meals (Fri + Sat evening, Sat + Sun lunch), which will be cooked at the venues (needs 4-5 people). If we choose to go out for a dish - in contrast to cooking - this won't be sponsored by the project.

There'll be a continental oriented breakfast for Sat + Sun available; some rolls, bread, hazelnut spread, jams, a bit of cheese and cold meat.

Pictures

You can Download and Upload pictures from the event here: https://apollo.geekaliens.com:8001/public.php?service=files&t=b91dfb35065c1c3aaa96841efb38cd9a You can use the WiFi password from the event.