Yesterday a message in theTest board in my forum (http://www.aeromodelling.gr/ForumS/index.php?topic=44.msg2697#msg2697) contained polytonic Greek. It appears equally well in a SMF 1.03 forum with a patch by [Uknown], after spiros' s testing and report in http://www.simplemachines.org/community/index.php?topic=19572.0.
[Uknown]' s patch can be found in http://www.simplemachines.org/community/index.php?topic=19572.msg160840#msg160840.
I started used this patch on May 2005 after a suggestion by spiros in Character codepage settings for multilingual selection (http://www.simplemachines.org/community/index.php?topic=34799.0) and I thank him very much for this.
I don' t want to argue with this against UTF-8.
I want to report that a working multilingual solution exists in a previous SMF version without using UTF-8. With this solution:
1. By cheating in codepage one can have correct display for a non latin letters language and menus in other languages unless there is a codepage conflict.
2. It accepts and displays other languages correctly (http://www.aeromodelling.gr/ForumS/index.php?topic=44.msg2540#msg2540), including polytonic Greek.
3. The search in SMF works OK for other languages and polytonic Greek. I can' t tell about Web search.
4. The main problem is in special character displaying when using a non latin letters codepage and this is overcomed by the patch.
You can test the above without registering in my Test Board (http://www.aeromodelling.gr/ForumS/index.php?language=english&board=58.0). I give a link that will go directly to english menu.
1,2 and 3 (I am not sute about polytonic Greek) work with standard SMF and ISO-8859-1. 4 needs a patch.
I tried the patch in 1.1 RC1 but I can't make it work.
With a similar patch for 1.1 RC2 and above my multilingual needs are fulfilled.
The use of standard Greek codepage (ISO-8869-7 or windows-1253) is almost 100% in Greek PCs and over 90% in Greek webpages. I believe a similar situation exists in other countries with non latin letters language.
I think this solution would be enough and maybe preferable if someone doesn' t want to use UTF-8.
SMF' s team should consider all the trends and possibilities. Compuart' s message (http://www.simplemachines.org/community/index.php?topic=63235.msg443044#msg443044) shows they take care of all this.
Please Compuart, as you are focused on multilingual, take a look at this solution too and [Uknown]' s patch. Spiros' s help would be appreciated in this topic too.
It's a big problem with SMF, but I try convert it to Vietnamese with without UTF-8 and I've just succeed with NCR-Decimal code, can you trying it now ?
elpvn the above proposed solution is a bit European oriented. I don't know how Asian languages are coded and I certainly can' t check the results. It might be a good solution for Asian languages but unfortunately I can' t help as I lack knowledge.
The UTF-8 approach is more multilingual oriented. You can find a topic about UTF-8 and SMF 1.1 RC2 in http://www.simplemachines.org/community/index.php?topic=63235.0
NCR-Decimal code and entities should be kept to a minimal necessary.
What is the trick?
You CHEAT IN CODEPAGE. That means that "windows-1253" is used for Greek, English, French, German,Italian, etc. This is not right in theory but in practice this way you can have English menu and also see correctly the messages written in Greek by a member using Greek lanquage. A cheat is needed for the special characters to display and handle correctly. In this way Greek letters and the basic Latin letters are stored directly. The entities are a negligible percentage.
One can argue that it is not a true multilingual approach. I will agree but I will add that it is more practical for my needs, while involving with UTF-8 will add some unecessary work in my whole project.
Who might be interested in this solution?
Someone who has an SMF forum with a non latin language and the main purpose of adding multilingual approach (menus) for latin letter languages. If the majority of messages uses only this language and the latin alphabet and doesn' t want to be involved with UTF-8 it is a simple, quick and working method (It already works in SMF 1.0x).
A plus benefit of this method.
It can be very easily and quickly applied to an already working forum with this codepage without requiring too much skill.
It's an old topic but I should write what happens.
SMF RC1 and RC2 don' t need a patch to work this way in a Unix or Linux server.