Michael Noda wrote:
On 12/21/06, Neil Harris usenet@tonal.clara.co.uk wrote:
Gerard Meijssen wrote:
Hoi, The notion that the numbers used in the Latin script are universal is also a fallacy. Thanks, GerardM
Yes, but it's quite a good approximation to reality. For example, the front pages of ar:, am:,he:, ru:, el:, th: and zh: are clear evidence that use of ASCII Hindu-Arabic numerals is widespread in common practice, even in cultures with their own distinct historic numeral systems.
And if we really wanted to, we can also represent the same set of decimal digits in other writing systems, for example Devanagari numerals, since we already have the localization code to do this ready to go.
I like Stephanie's suggestion of User:?????? <#2352562> as a display format: we could quite easily arrange that User:#2352562 would be an alias for User:?????? throughout the system, with ?????? remaining that user's true name everywhere.
Thus, for the user "Examplename", User:Examplename would be their user page on every wiki, with User_talk:Examplename being their discussion page, and so on... but, for wikis where the script used in their name is not one of the native scripts for that wiki,
[[User:Examplename|Examplename]] <[[User:Examplename|#425256]]>would appear in every context where the machine-generated text
[[User:Examplename|Examplename]]
would have appeared before, and "#425256" (or its equivalent with the decimal numerals in any other script) would also be usable in the software as a synonym for their real-name throughout the SUL universe. That is to say:
- User:#425256 would be a redirect to User:Examplename on every wiki.
- Blocking User:#425265 would have the effect of blocking User:Examplename
- Special:Contributions/#425265 would have the same result as
Special:Contributions/Examplename (and would display "User contributions for [[User:Examplename]]" at the top on home-script wikis, and "User contributions for [[User:Examplename]] <[[User:#425256|#425256]]>" at the top on others...
Note that the numeric tags would only appear on ''system-generated'' text: no user would ever be required to use their alternative numeric tag in any human context (unless they wanted to, of course).
Net effect:
- for most practical purposes, no visible change on any wiki, except in
the immediate context of a username with "foreign-script" characters
- no user will need to rename any of their accounts
- no user will need to create any nicknames if they don't want to
- most users won't even notice the change, since 99% of all editors tend
to have same-script names
- only formerly-unreadable "foreign script" names will appear with
numeric tags on any given wiki, yet
- either tags or names will work in any computer-input context such as
blocking or user page display, throughout the entire SUL universe, but with the system always giving the true name display priority
Just to gild the lily a bit more, we could also add a bit of latitude about what constitutes a "foreign" script: for example, if the community on zh: agreed that they could distinguish Latin characters (which they can in practice), or the community on en: was to agree that they could generally distinguish and remember Greek and Cyrillic characters, the software could be configured so that numeric tags could disappear on usernames in those scripts within those wikis. Or, contrariwise, if a community so desired, numeric tags could appear on all system-displayed usernames, even their own.
I think this can be a fair and readily implementable system, both technically and socially. It solves the problem en.wiki was addressing (clumsily) with the username policy, and is fully compatible with SUL. It displays no bias for or against any particular character set. In other words, I now expect that the SUL implementation team will account for this proposal, or something very much like it, when SUL goes live. If they're not considering it now, I for one think they should. (Although I like the earlier suggestion of salting the account IDs to avoid silly status games over longevity :-) )
I'd just like to express my sadness at the lack of co-operative spirit and foresight shown by some defenders of the en.wiki username policy, and the lack of assumption of good faith all around. While I personally dislike the policy for its heavyhandedness, it's clear that there is a problem that it addresses, and does so in one of the very few ways available *in the current technical environment.* What my fellow en.wikipedians are failing to deal with properly is that the technical environment is about to change, soon and inevitably. SUL is not going to just go away, because it's too useful to too many people. Plans need to be made now to deal with the major sociotechnical changes that will accompany SUL's introduction. One of those changes must be a revision of the en.wiki username policy. That will come about through discussion and (cross-)community consensus, so flinging what amount to accusations of bad faith around, right off the bat, is less than helpful, to put it mildly. We need to shed more light and less heat on this subject, as soon as we can. I hope we can come together, as a community of communities, and rise to the challenge.
-Michael Noda
Thanks, Michael!
As I understand it, the current en: Latin-script-only policy is a sad necessity caused by the need to fight the continual stream of vandalism directed at en:, combined with the limitations of most en: editors at distinguishing and remembering non-Latin names in general, rather than a manifestation of cultural imperialism imposed for social reasons.
This is also a policy that other wikis may be forced to adopt at some point in the future, for the same reason, when they get large enough that their level of vandal account creation gets sufficiently high (for example, can most readers of Arabic read Chinese, or vice versa?)
If a technically feasible system to reduce the social and administrative difficulties of dealing with editors with different-script names could be put into place, I think the need for this policy would disappear, and the policy itself would simply evaporate.
-- Neil