Amir, I think this is a good idea. For the sake of consistency, we should choose a single standard to follow rather than a hodge-podge of newer standards, older (although still valid) standards, and ad hoc codes we made up on the spot (als, nrm) and custom codes (bat-smg, roa-tara, roa-rup, fiu-vro, map-bms, be-x-old). It also seems potentially confusing to me that we have codes that overlap, for example na.wp and nap.wp, ro.wp and roa-rup.wp, etc. -m. On Wed, Jun 30, 2010 at 1:30 AM, Amir E. Aharoni <amir.aharoni(a)mail.huji.ac.il> wrote:

On Wed, Jun 30, 2010 at 10:30 AM, Amir E. Aharoni <amir.aharoni(a)mail.huji.ac.il> wrote: I agree with you and I was thinking about this issue a lot. Without exceptional cases which should be solved by BCP47 and ISO 639-5 codes, ISO 639-3 codes should be enough for everything which Wikimedia projects need. If we are starting now with standardizing Wikimedia language codes, it would be so. However, after 10 years we already have tradition. It is not reasonable to break links which exist for 2/3 of the time of Internet existence. The most "egalitarian approach" would be redirecting "en" to "eng" and similar. But then it would look like imposing the rule for the sake of imposing the rule. It doesn't mean that we shouldn't do that, but it also means that we should leave that for the future. Wikimedia should work on cultural equality, but it has to do that in cooperation with other organizations.