GSoC 2012 Improve Usability Chelen

=Improve usability and get rid of little bugs=

The main goal of this project is to improve friendliness. The idea that improving usability very well is possible in curing many very little bugs or sometime in adding very little features. That is why I will focus on this goal and improve undo/redo functionalities which can have incorrect behaviour.

We can notice that story editor lack of undo/redo but I will not work on this one as this widget should certainly completely change soon. To perform undo, we have to add the action on a stack where we can "pop" or "push" actions to undo or redo. Qt have already some classes to preform these kinds of actions.

As there are always something to do to improve usability, there will not have an exhaustive list of all issues. However, I will do this proposal with those which look to be the most important as user may encounter this regularly. A more important list can be found Here

Undo-redo
When we speak about user-friendliness, it also means that Scribus has to do what the user expect it does. We can notice that it remains some bug in undo/redo functionalities.

Currently, undo/redo is absent with some functionalities (changing style character for example) or have an abnormal behaviour (with pattern/graphic or about save functionality). As we can see with those examples, there are some improvement to do in these cases and especially with undo/redo when we resize/rotate a frame or a graphic.

Make easier user handling
To improve usability of scribus, we have to avoid repetitive actions and quicken access to different linked actions. Even if regular users used to do with these actions, new ones can be disturbed by this kind of behaviour as it is a waste of time.

In this case, we can think there are a lot of different thinks to do but in reality, it is pretty much the same in all different cases. Indeed, whether adding some space after a paragraph or enable/disable pre press or remembering rename or replace and many others, what the user want is just that scribus remember a default behaviour.

The second case is when the user has to click 10 times to do what he wants or when he has to go in another panel like with contour definition. In all these cases, we have to reduce the number of manipulations needed by the user to do an action.

Implementation
We can separate this in three parts:
 * Studying how the undo and the user interface are managed nowadays
 * Reading some part of the code where undo is correctly managed
 * Collecting information about Undo/Redo framework in Qt
 * Discuss with Scribus team what parts of the code have to be improved first, either the user interface or the undo/redo management
 * Coding new features
 * Fixing well known bug related to user handling
 * Fixing undo/redo management bug
 * Improving undo/redo functionalities where it is possible
 * Adding a new control panel for undo/redo to enable user to see and control easily his actions.


 * Finalisation
 * Execution of a bench of test to be sure that this code can be included in the trunk
 * Documentation of the code the enable the team to maintain this part of the code
 * Merging the code in the scribus trunk

Project deliverables
Here is the full list of what I expect to do.

Undo/Redo

 * Text
 * Export to 1.5 the text frame undo/redo already done by Cezary on 1.4
 * Character style modification


 * Graphic
 * EPS graphic
 * Combine/Split polygons
 * Rotation, resize of shape


 * Option
 * Overprint/Fusion mode/Arrows in PP
 * Action pdf


 * Layout
 * Creation of master page/import of page
 * Selection of frame after undo
 * Frame modification


 * Miscellaneous
 * Correct management of save flag after undo
 * Replay an action
 * add undo/redo panel

Make user handling easier

 * Remember usual behaviour
 * Multiple-duplicate/locked copy
 * Master page import (http://bugs.scribus.net/view.php?id=10263)
 * position after switching
 * Rename/replace style
 * Pdf options after export
 * Cursor position after insertion
 * Paragraph layout


 * Enable user to do things faster
 * Apply to all layers
 * Add multiple master page
 * Improve click on spinBox
 * Align shape

Contingency
As some actions might take more time than expected, I have selected some parts which can be postponed after GSoC.


 * Add undo/redo panel
 * As it really has a place in undo/redo project, I hope I can implement this during the summer


 * Replay an action
 * This is a part which can take a lot of time and it is on the edge of the undo/redo subject.

Time Line
As the main goal of GSoC is to involve students in open source project. There are no reason to wait until June before beginning this project even if I will have more time during the summer. To make user handling easier and improve undo/redo management, I will have to discuss all the time with scribus team to be informed about user difficulties to be able to fix them as a new part of the project.

April:

-Define a list of main tasks to improve usability and what should be done for the mid term evaluation.

-Have a deeper understanding of the undo/redo management in scribus.

-Fix one or two bugs to improve my Scribus code base knowledge.

May:

-Fix one undo/redo bug to see what is involved when I will work on that.

-As this is the month when I take my exam, I prefer not to overload my timetable but I may do more.

May 28-June 10:

-Fix undo/redo issues with text and options actions.

June 10-June 25:

-Fix undo/redo issues with graphics actions and other miscellaneous actions.

June 25-July 6:

-Be sure that everything is good with undo/redo and improve performance.

July 6-July 10:

-Redaction of the mid term evaluation.

July 10-July 24:

-Enable Scribus to remember user default behaviour.

July 24-August 7:

-Make faster user actions on scribus.

August 7-August 13:

-Final test and merge with scribus trunk.

13-30 August:

-Pencil down: Preparing the final evaluation and polish all the work.

After GSoC
I really want to work with Scribus team even after GSoC and I will continue to fix bug and help you in all these tasks. Also, I have some ideas to improve Scribus but I cannot do it in GSoC as the timeline is to short or I may need a deeper knowledge of the code base.

For example:

-When I used Scribus, I used to work with other persons and it is really inefficient to work on different computers and merge different files at the end. I think it could be great to enable scribus to work through the network. Like googledoc but with scribus. This project has to mature but Qt supplies a lot of classes to do network stuff.

-Add openMP instructions to make scribus work in parallel and improve its capacities. It has to be added with the multithreading subject.

-Add a way to update scribus automatically on the computer without reinstalling everything.

Biography
I am Pierrick, second year at Telecom Bretagne, a renowned engineering telecommunication school in France.

I have been using Scribus since more than one year to do a newspaper and a magazine.

I have already some experience in developing with C/C++ and Qt sometimes for fun, others for school project http://code.google.com/p/qrypt4all/ As I had work with other guys, I have already used cooperative tools like svn/git.

It is the first time I am involved in a big open source project like this one. To show my capacities, I have fixed two bugs in scribus:

http://bugs.scribus.net/view.php?id=10571

http://bugs.scribus.net/view.php?id=7793

The second one is now merged in scribus trunk.

I know that GSoC is a real job and I hope to have at least 35h per week on the project

Contact Information
Name: Pierrick Brunet

irc: Chelen

mail: pierrick.brunet@telecom-bretagne.eu

skype: Pierrick912