Hi all,
last time we discussed this, I think there was consensus to switch the
English Wiktionary to make it stop capitalising article titles, right?
Tim and I discussed this, and I think we should do it soon, because the
longer the wait, the more we will have to fix manually after the switch.
The idea is to
(1) flip the switch (so [[blah]] and [[Blah]] become separate articles)
(2) run a script that renames every page to be lower-case, except for
those pages that contain their own title in capitalised spelling
(3) create a list of those pages that weren't moved, so we can fix them
manually (for example, [[Kind]] will need to be split into two).
Ideas? Discussion?
Greetings,
Timwi
[One more time, posted from the correct address this time]
(Bug report posted on en:Wiktionary:bug reports also, though it probably
doesnt belong there)
Categories don't appear to work on la.wiktionary. Tried to create
[[Category:Γυνή]] ("category of words derived from the Greek root Γυνή"),
and it draws the category box all right but the link for Γυνή appears as a
link to [[:Γυνή]] (with an initial colon; clicking the link takes one to
the entry for Γυνή). One can't go to Category:Γυνή either without being
sent to Γυνή. How do I fix this?
(Furthermore:)
Γυνή is listed in Specialis:Categories, but the link is again to the entry
for Γυνή, not to a category. The same thing seems to happen on the Latin
Vicipædia.
*Muke!
--
http://frath.net/ (my website)
http://kohath.livejournal.com/http://kohath.deviantart.com/http://wiki.frath.net/ (conlangs and concultures)
--
http://frath.net/ (my website)
http://kohath.livejournal.com/http://kohath.deviantart.com/http://wiki.frath.net/ (conlangs and concultures)
For the method that is being develloped on nl:wiktionary, the method of identifying if existing words are capitalised or not, can be done by the defenition of the word; as a word can exist in a language both capitalised and not, we write the word in bold to start with, then the gender, the translation, then the explanation.
This was done to reflect the impossibility to create words in lower case. As all words start off identifying the language, we can have many defenitions on the same page.
One problem is when a word has _many_ translations. 40 is not rare. I think it would be good for a user when he can state what languages to include and what languages to exclude in the presentation of translations.
Thanks,
GerardM
Timwi wrote:
> The idea is to
> (1) flip the switch (so [[blah]] and [[Blah]] become separate articles)
> (2) run a script that renames every page to be lower-case, except for
> those pages that contain their own title in capitalised spelling
> (3) create a list of those pages that weren't moved, so we can fix them
> manually (for example, [[Kind]] will need to be split into two).
(4) create a list of those pages that link to a capitalised word, so
we can check them manually to see if the link should be piped to
the non-capital version.
-Mark-
LS,
I have read many messages about changing en:wikipedia to the next
characterset. I would really love to see the nl:wiktionary changed. I
have some friends who speak Russian, Farsi and Armenian who are willing
to check the words I have at the moment. As I cannot create these words,
I am waitiing for the conversion..
As there are not that many people who will notice nl:wiktionary to be
off line for some time and as we have only 607 words and a lot of other
stuff, it should not take too long to change nl:wiktionary
Thanks.
GerardM
Ray,
You say that all the arguments have been put forward. They cannot be as this system is not finished; it needs refinements. It is not ready. The time for arguments is not there yet.
You talk about consensus; but I am not putting the system forward asking for consensus as it is not finished. It works after a fashion. It is not finished and I do not ask for anything. The only thing I ask for is to leave some articles as a "proof of the pudding"; to show current thinking.
One of the refinements needed, can be found in better code for headers. It has been said that the code will be revisited. My hope is that the TOC and the edit boxes will work better.
You are factually wrong where you say that a word for one language is shorter than a message: Nederlands <-> {{nl}}. The one thing is you know it as Dutch :) . I do not need to know what it is in Japanese; {{nl}} suffices. I have recently added _many_ translations to [[French]], [[English]] in the en:wiktionary. I used the messages and they are as good in English as they are in the Dutch wiktionary. Therefore there are immediate benefits. I added more than twenty new translations in 2 minutes.
As to the En:wiktionary, there are translations that are hopelessly wrong and nothing is done about it.
When I read articles like [[Korean]] in en:wiktionary I find they are unusable. Many of the Romanian translations of many words have numbers in them . I hope that this is not because of a consensus.
You may have noticed that the discussion stopped on the page that you refer to. This is not because I do not cherish the good points that were made and it is not because I do not welcome constructive criticism.
At present I am fleshing out the words, the format, the translations. I am adding _loads_ of translations thanks to Timwi. When “we” find that the system does not work and will not work, there are bots that can change messages to the locally required texts so what I do is not in vein.
I also have done mainly nouns, no idioms, no sayings, so there is still a lot to be done. However, the words for languages are central to being able to translate. That is the function I am focussing on.
NB I have no idea where the "we" will be. But I hope and am certain it will be a bigger discussion than just nl:WikiWoordenboek.
Thanks,
Gerard
ds have numbers in them . I hope that this is not because of a consensus.
You may have noticed that the discussion stopped on the page that you refer to. This is not because I do not cherish the good points that were made and it is not because I do not welcome constructive criticism.
At present I am fleshing out the words, the format, the translations. I am adding _loads_ of translations thanks to Timwi. When “we” find that the system does not work and will not work, th
Ray,
There are issues. And as I have mentioned this is a system being developped. Nothing is carved into stone yet. You state that it is my "pet project" and indeed I among others try to nurture it to become better. As to your criticism, I do acknowledge that several issues have to be resolved. This is consistent with something that is under construction.
It would however be foolish not to respond to questions of why things need _many_ messages. That is exactly why we have mailinglists; to inform about what is what and why and how things are evolving. Again, en:wiktionary is not the only one. The issues that can be resolved should be resolved. But it is not a zero sum situation; the benefits must be weighed against drawbacks. At this moment people can cooperate and look at what this way of working produces. I am aware of issues and I am working towords resolutions.
Only stating that this system is not finished/ready is not helpfull as it is not a finished product yet.
Gerard
-----Original Message-----
From: wikitech-l-bounces(a)Wikipedia.org
[mailto:wikitech-l-bounces@Wikipedia.org]On Behalf Of Ray Saintonge
Sent: woensdag 16 juni 2004 10:38
To: Wikimedia developers
Subject: Re: [Wikitech-l] Why use so many (small) messages)
Gerard.Meijssen wrote:
> LS,
> Tim Starling wondered why we want to use so many messages in
> wiktionary. The reason is that the same article can be copied to
> another wiktionary (that contains the native content for the same
> messages). This will result in a correct article for the word. As an
> example the word [[Nederlands]], [[Engels]], [[English]] can be found
> in the Japanese wiktionary.
>
> Another example is when a translation has been added to a
> participating wiktionary, this translation can be added to all other
> participating wiktonaries. Yes, a bot needs to be written to do that,
> but that is the intention. Example; I have added _loads_ of new
> translations for the word [[English]] and [[Dutch]] on the
> en:wiktionary. (Originating from Timwi's website and then nl:wiktionary)
>
This system of messages is Gerard's pet project with which he has
carried on despite some serious questioning at Wiktionary talk:MediaWiki
custom messages
<http://en.wiktionary.org/w/wiki.phtml?title=Wiktionary_talk:MediaWiki_custo…>
We don't need all those messages. The potential for Gerard's scheme
interfering with the efforts of others is too high. A bot would only
make things worse.
Ec
_______________________________________________
Wikitech-l mailing list
Wikitech-l(a)Wikipedia.org
http://mail.wikipedia.org/mailman/listinfo/wikitech-l
LS,
Tim Starling wondered why we want to use so many messages in wiktionary. The reason is that the same article can be copied to another wiktionary (that contains the native content for the same messages). This will result in a correct article for the word. As an example the word [[Nederlands]], [[Engels]], [[English]] can be found in the Japanese wiktionary.
Another example is when a translation has been added to a participating wiktionary, this translation can be added to all other participating wiktonaries. Yes, a bot needs to be written to do that, but that is the intention. Example; I have added _loads_ of new translations for the word [[English]] and [[Dutch]] on the en:wiktionary. (Originating from Timwi's website and then nl:wiktionary)
Thanks,
GerardM
T
LS,
In the nl:wiktionary (WikiWoordenboek) we are experimenting with the use of messages to indicate languages, wordtypes, genders and all sorts of other things that go with words. To indicate a language, we use the ISO 639 code.
Mr Suisui, who can be found on both the nl:- and the ja:wiktionary took an intrest and translated, together with some more people on ja:wiktionary, the templates into Japanese.
Because of this I was able to make an entry in ja:wiktionary for “Nederlands” (you English call it Dutch :). I was also able to create the Japanese word for Nederlands and I added all translations from the nl:wiktonary there.
The one thing I cannot do is give meaning to the word in Japanese. And at present I cannot create the Japanese word in the nl:wiktionary due to technical reasons.
At the same time Timwi informed us of the existence of his site with the words for languages in other languages http://timwi.dyndns.org:8989/cgi-bin/a.cgi?r=s&l=ISO%20639 This is a real treasure trove. The only thing missing is an indication for capitalization and genders. But it has Armenian and Farsi and and :) .. !!
Timwi wants to discontinue his site and wants to incorporate it in the wiktionaries. Now the question: were will we enter all those words first?
The nl:wiktionary uses messages throughout so that makes it easy for migration to other wiktionaries.
The en:wiktionary has more English words for languages but, you can use the messages there..
A happy GerardM
LS,
I understand from Brian and from Timwi (who placed the monobook.js and monobook.css on nl:wiktionary) that this is where these things must be fixed. I am happy, this does indicate which way to go.
Brian mentioned that he is going to revisit the “template inclusion”. I hope that he will fix the inclusion of headers from within templates; the problem is that no TOC is generated from headers entered from within templates. This would help me a lot in making the standardised presentation (as I envision it) acceptable.
In his mail Brian mentions: “English and German are exceptions to this rule because of special-case work done by the respective contributors.” I have several words ready for inclusion in the nl:wiktionary. I cannot create them for the current restrictions. The idea of wiktionary is that it not only provides the local language words but also the words in other languages. Therefore it is part and parcel of what wiktionary stands for.
The "Enhanced recent changes (not for all browsers)" works again as specified. Thanks!
Brian is right where he says that he does not want to publish everywhere what he is doing. Consequently I am not subscribed to wikitech-l. He mentions that I cannot expect a minimum level of service. This is true up to a point. When a massive exercise like the “1.3 upgrade” is undertaken, there should be a proper conversion with all the necessary steps taken. This does mean that the sysops as a group undertake this. All steps should be taken for all environments. Doing it this way ensures that you provide a quality upgrade and, it prevents all kinds of nastiness down the line. It is in the interest of the sysops to ensure a standardized environment!!
Practically: the {{msg:-nl-}} should be converted to {{-nl-}} there is a script to do it. I cannot run it. It should be part of finishing up the upgrade.
Brian does not think much of en:wiktionary. He thinks that it needs many new features. In the nl:wiktionary, I am adding words for languages with a ISO 639 identification. I have a standardized way to indicate the translations (using ISO 639 to indicate the language) this should enable the quick translation from words from one wiktionary to other wikionaries that support the messages that are used.
I also have a 8878 word glossary (public domain) with mainly English botanical terms with a Dutch explanation. Those are the things I work on.
Concluding; Wikimedia is a success because of all the people that work on it. We all play our part and we do what we can, what we want. However, the aim is to produce something that stands up to scrutiny. It is a long hard road and traveling it together makes it enjoyable.
Thanks,
Gerard
o {{-nl-}} there is a script to do it. I cannot run it. It should be part of finishing up the upgrade.
Brian doe