UTF-8 solution

UTF-8 migration > UTF-8 justification > The solution

The solution

Once explained all the problems with the schema used until now and viewing the benefits (immediate and future) of the migration to UTF-8, in this page we are going to organise a bit the whole process of the conversion. Every required step will be defined briefly (you will be able to know more about each one in next chapters). Here it's:

• Documenting and discussing plans: All the stuff you have read up to now and the rest of this Wiki page, plus active discussions in the Developer forum will be the sources to understand the whole process. Aren't them enough?

• Gathering detailed information from the community: To have a real perception of how the UTF utilities are going to work under different configurations, one simple test has been created. By running it against as many Moodle servers as possible we will be able to detect potential problems before release.

• Stabilising min requirements for Moodle 1.6: With the information gathered in the previous step and based in the LAMP platform evolution current Moodle requirements have to change. Not only the UTF-8 migration but everything will be benefited with the use of new versions of LAMP. Details are available in the next page.

• Creating a common text library to be used everywhere: A lot of PHP functions are not UTF-8 safe and we must provide Moodle with their UTF-8 compliant alternatives. All these specialised functions will grouped in a new library to maintain it easier. More information in the next page.

• Moving language files: Every language pack needs to be converted to UTF-8 plus other adjustments to run smoothly under Moodle 1.6. To maintain everything working and trying to make things easier to translators (isolating them from technical and complex CVS operations) there is a plan. Just read it and, if you want, discuss it too in the Languages forum.

• DB migration: All the contents stored at Database level must be converted, from their original charset to UTF-8. This conversion can be performed following different paths, depending of the type of site (mono or multilang), its original charset and the DB backend (MySQL or PostgreSQL).

• Recoding PHP scripts: Moodle code itself needs to be modified to use new string manipulation functions and to work properly under an UTF-8 environment. A list of changes and their level of implementation. The Developer forum is, once again, a good place to discuss all this stuff.

Now, let's see the initial steps of the process.