Multilingual ...?! Mehrsprachig?! Posible multilingüe? Είναι πολύγλωσσο δυνατόν;

[MrPhil]

having multiple languages side by side isn't really going to be of interest to that many people.

In a support forum, I can see it being useful. I don't go browsing through non-English boards because I don't speak other languages. Once in a while, someone posts a non-English question to one of the regular (English) boards, and if it piques my curiosity, I might run it through Google Translate and see if I can understand it and help out*. Being able to browse other boards in, say, English, even if the posts were originally in another language, would be useful for supporting the community (i.e., the pool of people who can help is much, much larger).

* and when I Google translate back into the original post's language, I usually end up getting slapped down for such a bad translation...

We still have the problem of who will take the time to do a proper translation of a post to various other languages. You could pay a bunch of professional translators to do nothing but translate, or you can let the community do it, Wiki-style. I can't see burdening volunteer translators (such as SMF's language pack translators). Maybe the original translation could be done by machine, and then encourage the community to polish the translation (Wiki).

If you're not interested in actually doing (or checking) the translation, there's no point in showing them side-by-side if all you want to do is give an answer (in any language). Of course, if the translation is bad you may end up being told that your mother wears army boots or you may misunderstand the question and give a wrong answer. All translated posts should clearly indicate whether it's a machine translation, or if a professional/experienced volunteer, or N community members have gone over it.

[Irisado]

Surely it would be a huge burden to the translators if they had to translate it in real time?  Not doing it in real time would defeat the whole object of the exercise as far as I can see.

As for Google translate, it's too inaccurate to even act as a guide for translators in my opinion.

[sangham.net]

Thanks Irisado, thanks MrPhil for your hints and involvement.

Let me try to explain a vision of function in a general use (does not mean that all or a specific board would use all its possibilities).
Sample case: a mainly English board with many English, but also French, German and Spanish participator low participating of 10 other languages.

The board has a general setting of installed languages and a board language. If the option of multilingual post is enabled there would be an automatically produce of twins of the OP in the selected languages.
In the specific case (as there is general much interest of french, German and Spanish), this languages are enabled for the full use of the function.
This function can be upgraded for specific langauge board (if there are such). In or simulation case there is a subsection for Spanish. On this board is Spanish the default language and the default language of the main board is running here as well.
The member is able to set a second preferred language which allows him to see always a twinpost in the second language.

Settings of this board:
installed languages (en, es, fr, de, it, cz, and 10 others)
multiLP enabled languages (en, es, fr, de), mashineTwin is enabled
board default language is en
Subforum for Italian, mashineTwin is disabled (there is a much active translation work, or a translation board)

The general settings allow a automatically twin generation with machines or "empty twins". 

OP will be created with the language extension of the member default language (if the default language of the member is not enabled for mulitLP

* Member_A (en) opens a topic in the main board ...topic=192en.msg454en#msg454en
* machineTwin is enabled and so there is the creation of for the OP
...topic=192en.msg454es#msg454es
...topic=192en.msg454fr#msg454fr
...topic=192en.msg454de#msg454de
(generated post have the information of "machineTwin" original have the information of "Origin" an, translated "trans" and assigned as "assig". This information is visible in the Post as well the language of the "origin")

This language post are visible according to the settings of the member. Member_B has set Spanish as default and Italian as his second language.

So he will see the generated  ...topic=192en.msg454es#msg454es and the generated view of the post ...topic=192en.msg454es#msg454es in Italian in the folding second language window. (beside of the OP or translated posts it's always new generated, and not saved anywhere)

If you move a post for example form the Spanish sub board to the main board, missing twins (according to the settings of the board will be added as ) "machineTwin"
...topic=197es.msg454es#msg454es & ...topic=197es.msg454it#msg454it moved to main board as new Topic will generate
...topic=197es.msg454en#msg454en
...topic=197es.msg454de#msg454de
...topic=197es.msg454fr#msg454fr
...topic=197es.msg454es#msg454es

If there are any replay of people who have another language set as the default "machineTwin" the extension is coming to existence for this post as well and is available for others if wished.

In regard of arching and cleaning the database from time to time, there should be a task like: "clean all "machineTwin" OPs older then 2 month for languages with no replay" for example. That would solve the problems of unnecessary date spend as there are of course always topics which will not used much or are very specific to the certain language. There are of course more variable ideas how to keep it small.

The last post, if such function is installed, should also generate "machineTwin" in the default languages of the board. They would be also cleaned down to the default board language if there is not language replay in between, or for the languages where is no replay in between.
_____-
It seems that the "=0.4"- brain release beta version of the feature works already well and is able to be presented and tested well now.
Please feel invited to load it to your own simulation sections. If you have troubles to load it, try to empty the cache of temporary files first.
As there is no detail navigation support released please feel to request if any loading problem arouses or a special setting is not visible. Test it within your own board.

Of cause that would have a trembling effect on the whole board possibilities and culture. Worthy to shake?

[Arantor]

Even if the data is cleaned, there is *still* a fairly stiff performance overhead of having parallel data structures that will end up being queried. It's tantamount to all the discussions around threaded topics for its performance penalties.

Please can a moderator finally move this to somewhere useful? There is precisely zero chance this will be implemented in core SMF (this is the feature requests board after all) and if implemented on one specific site it might convey some small benefit in return for some serious performance management but it is technically infeasible to foist all users with that nonsense when it just won't help them.

[sangham.net]

Dear Arantor,

There is no parallel data structure, there is only are side datas. No impact on the whole system. The idea to save all of that in the extension is just as a sample. It could be stored within the content of the main data of the post as well.
I do not know which way would be better.

There is precisely zero chance this will be implemented in core SMF

Of course I do not know much, but I do not understand this definitely assuming claim.

(this is the feature requests board after all)

And that is the reason why it is here.

Let me ask you if you are used to speak or work in more as one language?  Maybe it more of an imaginary kind. It would not have impact on user who are not interested in such. The second window could be collapsed
*generated* posts could be set of not visible and even post from different language origin could be ignored.

I see a huge possibility for interaction and back flow as well as a huge possibility to reduce the datas in relation of general demand. Of course it could increase the participating generally to a high extend but that does not go hand in hand with the related data use out of bad organized system.

If I really have significant errors in my thinking out of not knowing important things, please be so compassionate to elaborate them. It does not help me (and maybe others as well) to say simply "It is not possible" or "not worthy".

Thanks for your patient and it would be great if you could explain the reason further.

Please can a moderator finally move this to somewhere useful?

It could be well that I did not explain it good as English is not my native language. As there is still more elaborating and explaining, as well as involvement, I guess it is good to keep here (put that is just a personal opinion). If I have over steeped a ruling position an structure within the decision makings here on board, please for give me, but let me know its ways at well.

Maybe we can continue the discussion here even it is not imaginable for all but that is the reason why we get involved in such at least or not?

[Arantor]

Dear Arantor,

There is no parallel data structure, there is only are side datas.

What does the word parallel mean again?

No impact on the whole system.

i'm sorry, you are mistaken, no matter how you might thnk you are not. Relational database like MySQL only work one way, and not how your brain does. There is absolutely no way to implement this in a relational database eithout it being parallel stored. Or REALLY screwing up performance.

The idea to save all of that in the extension is just as a sample. It could be stored within the content of the main data of the post as well.
I do not know which way would be better.

Like I said, other major issues. Accepting the reality in front of you would be a start.

There is precisely zero chance this will be implemented in core SMF

Of course I do not know much, but I do not understand this definitely assuming claim.

(this is the feature requests board after all)

And that is the reason why it is here.

Oh FFS. This board is for features for all SMF powered sites. This is NOT SUITABLE FOR ALL SITES.

Let me ask you if you are used to speak or work in more as one language?

Let me ask you if you are understanding anything I am saying. I understand what you're trying to do. I understand it will be convoluted to use and will make the entire site slower for every user without fail.

If I really have significant errors in my thinking out of not knowing important things, please be so compassionate to elaborate them. It does not help me (and maybe others as well) to say simply "It is not possible" or "not worthy".

I would if I didn't have to spend 3 or 4 attempts at trying to be understood. Unless I were to explain relational database theory, any explanation would be worthless, just like all the time I already wasted being ignored. It is possible but it has bad side effects but you're not listening to me.

Please can a moderator finally move this to somewhere useful?

It could be well that I did not explain it good as English is not my native language. As there is still more elaborating and explaining, as well as involvement, I guess it is good to keep here (put that is just a personal opinion). If I have over steeped a ruling position an structure within the decision makings here on board, please for give me, but let me know its ways at well.

I will explain it again, then. This board is for features for every SMF site. This is not suitable for all sites, who mostly use a single language and have no need of this and certainly do not need the performance troubles associated with this. As it is not suitable for most sites it should not be in this board and should be moved.

[Kindred]

Johann,

we are understanding you quite clearly. We understand what you want and even why you want it.
We are syaing that it is not realistically  possible to implement from a tehcnical (code and server resource) or a manpower resource standpoint.

You keep arguing the same thing and you don't appear to be listening to us.

And no...  this most definitely is **NOT** something that would be a reasonable "feature" for anything other than 1 or 2 sites in the world. So no... this no longer belongs in the feature requests board.

[sangham.net]

Ohh, I have seen there have been two post (I have not read them now, so please forgive me. I will read them after this)

I try to explain my perception of storing and handling within the content of the post:

Post and content are produced within the information of the default language of the poster. Just as a sample:

<en>What do you think about this idea<en>>"orig" is the content of the original post.

That will be generated (as it is the OP) into a content of:

<es>Qué te parece esta idea?<es>"gen"
<fr>Que pensez-vous de cette idée?<fr>"gen"
<de>Was wissen Sie über dieser Idee?<de>"gen"

So the content of the post ...topic=192.msg454#msg454 as OP of a topic after sending looks like that:

<en>What do you think about this idea<en>>"orig"
<es>Qué te parece esta idea?<es>"gen"
<fr>Que pensez-vous de cette idée?<fr>"gen"
<de>Was wissen Sie über dieser Idee?<de>"gen"

With the settings of the user he can chose the visibility. As soon as one would be translated or aproved it would look like that:

<en>What do you think about this idea<en>>"orig"
<es>Qué te parece esta idea?<es>"trans"
<fr>Que pensez-vous de cette idée?<fr>"prov"
<de>Was wissen Sie über dieser Idee?<de>"gen"

The replay would be maybe on the French one: ...topic=192.msg455#msg455 with such a content:

<fr>pas bon!<fr>"orig"

If a other user (es) would reply to the post, having seen the "mashine translation" it would create the content:

...topic=192.msg456#msg456

<es>no es bueno<es>"gen" recite the quote (incl, the origin of it, like used now) and this new content
<es>por qué?<es>"orig"

The possibility of machine translation could be always visible under the op (or fold-out).

For the post where is no "orig", "prov", or "orig" available in your language, "gen" could be as visible set on default or you would not see the content at all in your language stream.

Important posts could be selected from moderators and forced to be "gen" (as for example the final post)


Of course also very naive explained, but the function or system should be visible. The settings and possibilities are nearly the same as the model of creating additional language posts with extension.

Personally I thing that the "extension solution" is easier to develop later on, as you do not need to go into the content of datas. Also in regard of cleaning, I guess the "extension solution" would be more secure even it would crate more data fields in the first place.

Seeing that in <de>Was wissen Sie über dieser Idee?<de>"gen" in the post topic=192.msg454#msg454 I "D" would need to go into the content of this post and make it to
<de>Was halten Sie von dieser Idee?<de>"tran"

Here maybe useful a quick visualisation how it could look like

[sangham.net]

Johann,

we are understanding you quite clearly. We understand what you want and even why you want it.
We are syaing that it is not realistically  possible to implement from a tehcnical (code and server resource) or a manpower resource standpoint.

You keep arguing the same thing and you don't appear to be listening to us.

And no...  this most definitely is **NOT** something that would be a reasonable "feature" for anything other than 1 or 2 sites in the world. So no... this no longer belongs in the feature requests board.

Dear Kindred, dear Arantor,

I understand your views and I also have my assuming how they come to be. If I would be not convinced that this is a very needed stuff (english and european languages are actually a small marked today, even we might feel secure) and that it would open a back flow and not like now a merely flow from a great team down dependent also on it, I would not have introduce it. As well as I am convinced that it will reduce the amount of data over all and has no need to be disturbing for board runner who do not like to use it at all.

How ever, thanks for your patient and interests. Maybe you like to apprciate the idea it self a little and put the topic on a forum where developers of mods are very active, as I think that a light version is even develop able within such stuff like BBC codes and some small other codes within the structure.

I feel a little shocked when I see the topic next to "Add Read Members PM'S"

Lat but not least it should not be understood as a reduce of my gratitude that SMF exists like it is and so many people share there great work I (we) can benefit, actually it's meant as the opposite.

So maybe somebody likes to pic it up, maybe even out of gratitude for the shares he/she had got as well.

[Arantor]

as I think that a light version is even develop able

Except that any 'light' version will have all the same performance struggles as a 'big' version.

As well as I am convinced that it will reduce the amount of data over all and has no need to be disturbing for board runner who do not like to use it at all.

The reality does not agree with you on this.

[MrPhil]

There is no parallel data structure, there is only are side datas. No impact on the whole system.

I think I understand what you mean by 'side data'... think of a topic's original/primary posts being in a chain (ordered), and one or more posts could each have a side chain of multiple translations. Well, you still have to store those translated posts in the database, and their language and translation status, as well as all the additional pointers linking them together. If a topic or post is erased or moved, you need to pick up all the translations, too. Then you need to modify at least the themes, so that they can display side-by-side for translation purposes, or show only topics/posts a given language (if at least one post exists in your desired language), and collapse untranslated posts, and various buttons to show different languages the post is available in and single/side-by-side mode, and offer Wiki-style editing (anyone or only approved translators), etc.

As cool as this would be to have, I have to agree with the others that it would be a huge architectural change to SMF, and probably of limited enough use to not make it worthwhile for the base SMF product. It might even be too pervasive (hitting too many things) to make it a mod. If you think it would be really useful, consider making a fork of SMF that features this. Maybe you could get some people to work on it with you (I'm willing to comment on design and interface/usability ideas, but don't have time to do the coding).

[sangham.net]

as I think that a light version is even develop able

Except that any 'light' version will have all the same performance struggles as a 'big' version.

Yes Arantor, every desire will lead to struggle. The more skillful the desire the more it leads to spheres where desire decays.

SMF would not exist if there was not strong desire developed at the beginning.

With light version I thought of functions like enable of disable smiles. If there is a smart code system for translated text parts it would have no impact on much. That is just another thought bridge for a solution in the frame of my understanding of all functions.

If there is such a code like <en>content<en> it could could be enabled or disabled and a small side application could generate such texts with machine translation.

As well as I am convinced that it will reduce the amount of data over all and has no need to be disturbing for board runner who do not like to use it at all.

The reality does not agree with you on this.

Disagreement is always the reaction if assuming goes against "reality". Or should it be, as long as there is assuming, there is no reality. Reality does not have a point of view. There is just "with this arises that" and from "with this decays that", no opinion at all.

I really appreciate your warnings and I guess you know well why. And I am sure there is big good will behind.



See, I really have not much intention to make consumer easier consume. It should be more a possibility to give those who like to put an end on consuming a possibility and they dwell within every community.

There is no parallel data structure, there is only are side datas. No impact on the whole system.

I think I understand what you mean by 'side data'... think of a topic's original/primary posts being in a chain (ordered), and one or more posts could each have a side chain of multiple translations. Well, you still have to store those translated posts in the database, and their language and translation status, as well as all the additional pointers linking them together. If a topic or post is erased or moved, you need to pick up all the translations, too. Then you need to modify at least the themes, so that they can display side-by-side for translation purposes, or show only topics/posts a given language (if at least one post exists in your desired language), and collapse untranslated posts, and various buttons to show different languages the post is available in and single/side-by-side mode, and offer Wiki-style editing (anyone or only approved translators), etc.

As cool as this would be to have, I have to agree with the others that it would be a huge architectural change to SMF, and probably of limited enough use to not make it worthwhile for the base SMF product. It might even be too pervasive (hitting too many things) to make it a mod. If you think it would be really useful, consider making a fork of SMF that features this. Maybe you could get some people to work on it with you (I'm willing to comment on design and interface/usability ideas, but don't have time to do the coding).

Dear MrPhil,

thanks for your hints. "consider making a fork of SMF that features this" what does that mean "a fork"?
"I'm willing to comment on design and interface/usability ideas" Thanks for your generous offering, lets see if my own desire is strong enough to even fall into the ways of learning about all this issues or I (we) can win gurus and professionals to lead or co_work on this.
It's great to have even mental co-support and interest. Even a good wish is great support. I keep your offering in mind and its already very useful and supportive.

[Arantor]

If there is such a code like <en>content<en> it could could be enabled or disabled and a small side application could generate such texts with machine translation.

And this will lead to other, very serious, performance issues.

No matter how you do it, you cannot avoid performance troubles. I do this sort of thing for a living, figuring out how to do the ridiculous and the insane and make it work. I know SMF's code intimately, I also know computer science extremely well and know just how complicated this is to make.

Please stop trying to convince us otherwise, because all you're doing is pissing us (me, mostly) off.

You cannot change reality just because it seems inconvenient to you. The laws of mathematics and logic are not magically flexible. There simply is no way to do what you're asking for in a convenient and efficient way. The only ways are all slow, impractical and unlikely to work well on any site.

I really appreciate your warnings and I guess you know well why. And I am sure there is big good will behind.

Apparently you don't because you keep trying to convince me otherwise.

With light version I thought of functions like enable of disable smiles. If there is a smart code system for translated text parts it would have no impact on much. That is just another thought bridge for a solution in the frame of my understanding of all functions.

Please stop making assumptions about things you clearly have no knowledge of. That is not how it works. Smileys being disabled doesn't prevent the smileys in the content, nor does it hide them, it simply doesn't run the extra effort of trying to convert text to smiley images. The raw code is still firmly present and it has a material effect, though usually very small.

Please. I have worked on the aspects towards this. I even have a system with a functional [lang=code] tag that works much as you describe and I understand only too well what problems there are with it in reality in terms of performance, in terms of how it can actually *break* a community in some respects.

I do understand what you're trying to achieve but years of programming experience and years of using software simply have shown too many cases of similar ideas in different contexts and why they are unworkable.

thanks for your hints. "consider making a fork of SMF that features this" what does that mean "a fork"?

A fork is a branch of SMF's code. It is to take SMF, give it a new name and change some or all of its code to suit a specific and different purpose.

I (we) can win gurus and professionals to lead or co_work on this.

This seems unlikely. You have people here, who do this sort of thing for a living, whose job it is to write code, understand databases and so on, telling you why this is a bad idea.

[agridoc]

Johann B I will propose a test to do. Before this let's clear some things.

- I see that you come back again and again to machine translation. Try to understand that this can't work in a forum. With two languages there will be many misunderstandings with more languages there will be a mess. less or more doesn't matter.

- You started your topic for bilingual so let's stay on this as a start.

This is a test that I propose

- Find some good cooperators that can translate, have time to monitor the forum and do some moderation.

- In a working forum that you already have or in a new one. Install as or convert to UTF-8.

- Install the two languages you intend to use.

- Make a board with a title "Test for bilingual" or what you like best.

- Add your cooperators as moderators for this board.

- They will have to monitor translate titles and messages to the other language. Users should be allowed and respected to write both versions, if they can.

- I believe that both language versions should be visible. This way translators can be evaluated by users and as it's usually said "the original is always best than a translation" (although that might not always be true in a forum  )

- Invite people to participate.

- Evaluate the results.

With this simple test you will see in praxis how it goes an/or you might see that it's enough for your needs.

After quite some time you can come back and inform about the results.

Please don't give a reply to me, if you want just take my 2 cents.

[sangham.net]

If there is such a code like <en>content<en> it could could be enabled or disabled and a small side application could generate such texts with machine translation.

And this will lead to other, very serious, performance issues.

No matter how you do it, you cannot avoid performance troubles. I do this sort of thing for a living, figuring out how to do the ridiculous and the insane and make it work. I know SMF's code intimately, I also know computer science extremely well and know just how complicated this is to make.

Please stop trying to convince us otherwise, because all you're doing is pissing us (me, mostly) off.

You cannot change reality just because it seems inconvenient to you. The laws of mathematics and logic are not magically flexible. There simply is no way to do what you're asking for in a convenient and efficient way. The only ways are all slow, impractical and unlikely to work well on any site.

Where is a will there is a way and I thinks it's just the missing desire and facing much work imbalance in the right direction. I am pretty sure that you professionals are able to do it.

Sometimes we see a great challenge as a offense against us, but that feeling of offense has just to do on the fact that we dislike to go into a challenge.

Confidence is most important, one would not even open the refrigerator if there is no confidence that there could be food inside. Of course is would be blind believe if we think that we find food to eat when we open the laptop. If I give you the feeling that I try to make you believe that you can find food if you simply open the laptop, that please sorry that is not my intention. I love insight but to gain it we need confidence.

Over all I feel more than honored that such a qualified person participates and does not simply ignore it. That would be in fact a sign that it is not worthy to develop.

I really appreciate your warnings and I guess you know well why. And I am sure there is big good will behind.

Apparently you don't because you keep trying to convince me otherwise.

Maybe challenging, but not to make your believe or to put you devalue you or others, but to increase all your possibilities. I know that the potential is sleeping behind. You could not image how grateful I am on what all of you did here and how grateful I am that I can use all your work. So please do not misunderstand my word and give them an intention which would be not like that.

The fact that a person like me, who has not a slight knowledge of all this specific programmings, is able to build up a great forum and possibility without the need to ask somebody else, is a great sign that Simlpe machine is a simple machine in the meaning of clear, structured and great performance. Everybody with basic but good understanding of mathematics and logic is able to use it as a great opportunity, a great and solid basement. And I understand that there is much desire and need that this fact will not be disturbed by macerating it with individual wishes.

I guess I understand your point well even trough I am still convinced that there are possibilities. It would be more than overbearing to come as a newcomer and care about foundation or even criticize them. No, no, that is not my intention. That would be of course more then shameless, foolish and harmful.

I guess I can estimate my possibilities on a health level and I know my position well. I move all this upward from a very low position and it's even a honor to be able to do even this.

With light version I thought of functions like enable of disable smiles. If there is a smart code system for translated text parts it would have no impact on much. That is just another thought bridge for a solution in the frame of my understanding of all functions.

Please stop making assumptions about things you clearly have no knowledge of. That is not how it works. Smileys being disabled doesn't prevent the smileys in the content, nor does it hide them, it simply doesn't run the extra effort of trying to convert text to smiley images. The raw code is still firmly present and it has a material effect, though usually very small.

Please. I have worked on the aspects towards this. I even have a system with a functional [lang=code] tag that works much as you describe and I understand only too well what problems there are with it in reality in terms of performance, in terms of how it can actually *break* a community in some respects.

I do understand what you're trying to achieve but years of programming experience and years of using software simply have shown too many cases of similar ideas in different contexts and why they are unworkable.

[/quote]
Of course smiles was maybe a bad sample, but you suddenly got the message what I meant.

One aspect of why most people do not like more languages is that they confuse a little, they are not used to. Its like if you sit in a community where different languages are spoken. If you are not able to understand them you feel insecure and you what them to be turned of.

So this function could within the content and as a "hide" function could solve the problem for those how feel disturbed by different things they are not used to and give the possibility to use it for those who seek to broaden there ways of communication.

thanks for your hints. "consider making a fork of SMF that features this" what does that mean "a fork"?

A fork is a branch of SMF's code. It is to take SMF, give it a new name and change some or all of its code to suit a specific and different purpose.

I see, now actually not really an interests. That is somehow against the fundamental spirit of SMF, even if the share of this opportunity is at fully within such a great intention.
I have been giving so much and I would find it not proper just to use and perform it simply for my interests. But I can value this gift of possibility. A great Sadhu! (very good, excellent, worthy of praise) toward the developers and people in charge on SMF for such gifts.

I (we) can win gurus and professionals to lead or co_work on this.

This seems unlikely. You have people here, who do this sort of thing for a living, whose job it is to write code, understand databases and so on, telling you why this is a bad idea.

In the sense of Forest Gump " stupid is as stupid does" - "Bad idea is as bad ideas does" or it is just a matter of how to come to it. I don't think that multilangage. language accesibility use and cross conversation and the possibility of back flow in a multilingual developing community is a bad idea.

Of course its ways and performing can the stupid, that is why it is important to have rebukes from professionals always on side and this is more than valuable.

So what? Do you like to search for reasons why it is not good, or to search for reasons how to implement. If we are in charge of both, to value the idea and to value the ways of how to perform, we have a lot of struggle, so if there are general doubts on the idea it self, we should work out this before. Otherwise we would have suddenly opponent in our mind as soon as we face problems. Fist there must be the confidence that this aim is useful. If we have doubt on the idea, we will find 1001 reasons why it is difficult and a bad idea. That is not a fruitful base and we will talk for years for things they might have been done in some weeks or even moth.

In usually live money would solve that problem, to lay aside the doubt. Here we have no such bait. I like honest working out and I would not doubt a little that you do it as well. I guess that is the reason why we have so much attrition at the starting point.

The google for Arantor gave me great sentence "Innovate, not Imitate". As you have spoken for "a live time", even you have done real great work already you are not that old it's even not the first third of a average lifetime. Do you think that we generally know all already not even on the half of our ways? I know the danger of thinking that we understand everything already well and that is most dangerous at a point where we are very successful.

To secure that this would not happen Forums and communities with good friends are a gift.

[sangham.net]

Johann B I will propose a test to do. Before this let's clear some things.

- I see that you come back again and again to machine translation. Try to understand that this can't work in a forum. With two languages there will be many misunderstandings with more languages there will be a mess. less or more doesn't matter.

- You started your topic for bilingual so let's stay on this as a start.

This is a test that I propose

- Find some good cooperators that can translate, have time to monitor the forum and do some moderation.

- In a working forum that you already have or in a new one. Install as or convert to UTF-8.

- Install the two languages you intend to use.

- Make a board with a title "Test for bilingual" or what you like best.

- Add your cooperators as moderators for this board.

- They will have to monitor translate titles and messages to the other language. Users should be allowed and respected to write both versions, if they can.

- I believe that both language versions should be visible. This way translators can be evaluated by users and as it's usually said "the original is always best than a translation" (although that might not always be true in a forum  )

- Invite people to participate.

- Evaluate the results.

With this simple test you will see in praxis how it goes an/or you might see that it's enough for your needs.

After quite some time you can come back and inform about the results.

Please don't give a reply to me, if you want just take my 2 cents.

Thanks a lot for your share!

Actually I do that already here (even just started some weeks ago and a topic which has general not much people who are interested) and I know the difficulty and problems already, I had spend the last 2,5 year mostly on translating to share knowledge to communities who have not much possibilities to access other languages and I know how hard it is to get people out of their simply consuming mood and to open other aspects. Its a walk against usual and will only reach others slowly (most people would not even know their neighbors, now think of getting in contact with different languages of different cultures) much fear and conceit is the ruler. But to break this boarder a little it need possibilities which are proper for people at certain levels. Just a stairway, no need to walk. Someone which more language skills will not have the problem at all and if he is a simply consumer he has a much greater field to feed.

There are many languages which will have not much fruitful places in the futures accept we are able to build drains between them. Otherwise they will all be lost soon, and it will not be english, that will be the main field of getting much food. That is what we need to consider very strong. A wall around a nice place is fine only for a limited time. Then it grows to a prison.

I usually do not step forward if I have not proved all I am able to prove before. So the whole ideas is not simply an imagination but has all its reasons and experiences behind.

Thanks for your ideas and your care. More than praise worthy! I am already aware of the difficulties as it is for sure some ahead of the times, so no follow the times and moods invention.

To get an impression of the urgency of language accessibility, please look here: People across the globe are coming online faster than ever before – and bringing their languages with them. Below are industry stats related to the emerging global internet. 

I Asia and third world people with language skills are the most powerful people, they have the key to the food people like. They mostly do not have a global team spirit but know how to feed form the global desire for communities.

[MrPhil]

Johann, don't let Arantor get you too upset. He may have great technical skills, but he has very poor people skills and comes across as a Grumpy Old Man. He enjoys picking apart everything and explaining in great detail why it's stupid and won't work. I know; he does it to me all the time.

That said, I have to agree that your proposal is probably a non-starter for SMF, at least for the main development. It's a major development effort and will probably be a major performance hit, and likely would not be used by many forums. If you think it would be useful, by all means start your own fork. I think SMF 2.1 is on GitHub, and once you've signed up and have your own account and project, I think it's like one button click to copy it over to your project. Then you can invite others to work on your fork, implementing your concept. If it proves workable and performs well, SMF might even consider folding it back in to the main SMF effort.

P.S. If a post copy is the original machine translation, I would force the side-by-side display of original language and the translation, just to emphasize that this is a very rough machine translation and not something that's been polished by a human. You also need to think about whether to translate all posts to a common language (e.g., English) automatically, or just do it upon request by a member (and save the translated post). You would want to automatically translate at least the subject line, so speakers of other languages can see if a post might be of interest to them. You also need to mark a post as needing retranslation if the original one is edited (assume that new information is added only to the original, by its author or a moderator).

[Arantor]

He may have great technical skills, but he has very poor people skills and comes across as a Grumpy Old Man. He enjoys picking apart everything and explaining in great detail why it's stupid and won't work. I know; he does it to me all the time.

I see more as encouraging you to avoid wasting your time. As for you specifically, I get a bit annoyed with the 'high and mighty attitude' you demonstrate when you're blatantly incorrect. But you won't be told, even when you are wrong, especially when you have been *demonstrated* to be wrong.

[Irisado]

Sample case: a mainly English board with many English, but also French, German and Spanish participator low participating of 10 other languages.

If the main language is English, why not just have a separate category for other language sections/boards?  You would save yourself a lot of work, and those who are bilingual can participate in both the English, and relevant other language, boards on the forum in question.

The board has a general setting of installed languages and a board language. If the option of multilingual post is enabled there would be an automatically produce of twins of the OP in the selected languages.

We've highlighted the problems with this though.  When you say 'produced automatically', what do you mean?  You can't just get an automatic translation, and expect it to be accurate.  There are foreign language translators who work on SMF for a reason.  If it were so easy, there wouldn't be foreign language packs, developers, translators et al.

This leads us back to the point I made earlier that you would actually need people to carry out live translation, and you haven't been able to explain to me how you could make this work with an active forum.

Post and content are produced within the information of the default language of the poster.

Automatically, or by a translator?

<en>What do you think about this idea<en>>"orig" is the content of the original post.

That will be generated (as it is the OP) into a content of:

<es>Qué te parece esta idea?<es>"gen"

Just to say that the Spanish is incorrect, as the inverted question mark ¿ is missing from the translation.

<es>no es bueno<es>"gen" recite the quote (incl, the origin of it, like used now) and this new content

If referring back to the noun idea this translation is also incorrect, because the adjective should be buena to agree with the gender of the noun.

<es>por qué?<es>"orig"

Again, the inverted question mark is missing.

Assuming this is automated translation conducted by a machine, can you see why there are going to problems now?  A human translator would notice these things, but you cannot expect someone to translate forum posts in real time, as they could be multiple paragraphs in length, and there could be multiple threads and replies being posted at once.

I still cannot understand how you propose to make this work I'm afraid.

[sangham.net]

¡Hola Irisado, ¿cómo está

Sample case: a mainly English board with many English, but also French, German and Spanish participator low participating of 10 other languages.

If the main language is English, why not just have a separate category for other language sections/boards?  You would save yourself a lot of work, and those who are bilingual can participate in both the English, and relevant other language, boards on the forum in question.

That would be the usual way.

The board has a general setting of installed languages and a board language. If the option of multilingual post is enabled there would be an automatically produce of twins of the OP in the selected languages.

We've highlighted the problems with this though.  When you say 'produced automatically', what do you mean?  You can't just get an automatic translation, and expect it to be accurate.  There are foreign language translators who work on SMF for a reason.  If it were so easy, there wouldn't be foreign language packs, developers, translators et al.

"You can't just get an automatic translation, and expect it to be accurate." of course and I am sure, that if it hits the attention, the problem would be solved as soon as one gets interested in the topic.
Did you ever had an interest on reading what is posted on the Arabic forum here? It never meats your intention. As for important topic (taking SMF as sample) it is clear that the specific translators would work on it. (if it is machine, translated or approved; all are able to see). It is even imaginable that you could upgrade the automatic translation for sensible topics.

This leads us back to the point I made earlier that you would actually need people to carry out live translation, and you haven't been able to explain to me how you could make this work with an active forum.

That depends on the intentions of the forum team and of course of the availability of people like to get involved. It would be of course nonsense if you enable 50 languages and there is actually just in two languages a good intention as we well as support. That is a thing that will change with people like to engage are coming and going.

Post and content are produced within the information of the default language of the poster.

Automatically, or by a translator?

Automatically would be the first step for the OP, like told above.

<en>What do you think about this idea<en>>"orig" is the content of the original post.

That will be generated (as it is the OP) into a content of:

<es>Qué te parece esta idea?<es>"gen"

Just to say that the Spanish is incorrect, as the inverted question mark ¿ is missing from the translation.

That is right, but in that case it was my ("translators") fault, I thought that is an error this ¿. As it looks I just have learned some spanish specials. DO you think that you would correct it immediately if you see such errors in your language? See, and exactly this "Uhh! Wrong... " is the reason why such things will work and increase the involvement. A board where everybody agrees will be silent. This psychological thing is how such contents like wiki are able to grow. People love to be wiser, better and smarter which is at least not that bad. We learn from each other.

<es>no es bueno<es>"gen" recite the quote (incl, the origin of it, like used now) and this new content

If referring back to the noun idea this translation is also incorrect, because the adjective should be buena to agree with the gender of the noun.

I had shown the case with memeber D and his correction on the German translation. Thats the way it is, image that 70% of people depend on googe-translate, google themselves and will never have even the possibility to have somebody aside to say "Uhh! Wrong... "

<es>por qué?<es>"orig"

Again, the inverted question mark is missing.

Assuming this is automated translation conducted by a machine, can you see why there are going to problems now?  A human translator would notice these things, but you cannot expect someone to translate forum posts in real time, as they could be multiple paragraphs in length, and there could be multiple threads and replies being posted at once.

I still cannot understand how you propose to make this work I'm afraid.

You would had mad it already. That is the magic with it. and people will increase joy with it.

An German proverb "Wer schimpft der kauft" ¿Quién regaña compra translated right¿ 

You would love it. Love an hate is on the same coin, that turns the world and that is why discussion board exist. Give them the possibility to learn in this struggle, not just feed them like pets.

Thanks for your hints and involvement.  Gracias por sus consejos y la participación