Tranlator's blues

1.

With a language that has about 3/4 of LT's text snippets translated, it is becoming more and more difficult to find and be able to reach--i.e. gain access to their translation edit pages--of the other LT's text snippets to be translated.

One method is to find a text snipped that has been translated into another language (recent translations listed at the Translation Main Page) and choose to edit those, changing the edit language. Though this technique is rather strange indeed, it works. I keep wondering how the other translator (translating into the other language) gained access to the edit page, though. There must have been a way--unknown to me so far.

2.

Some text snippets, containing some code, in particular html tags, are autoedited over and over again, and thus are to be corrected over and over again, too, ad infinitum.

A good example of this nasty behaviour is this one, taken from the User Profile page :
You can use some HTML tags, such as <b>, <i> and <a>.

Its edit field has to read (in translation) :
You can use some HTML tags, such as &lt;b&gt;, &lt;i&gt; and &lt;a&gt;.

The latter is than autoconverted to (in translation) :
You can use some HTML tags, such as <b>, <i> and <a>.

That's okay, because LT users are to see the latter.

However, when re-entering any edit page containing that text snippet (in its proper edit field), that edit field itself is containing the autoconverted text, too, thus :
You can use some HTML tags, such as <b>, <i> and <a>.

When saving whatever editing whereever on an edit page containing such text snippet containing any html tags as above, they are autoconverted again, i.e. converted into their meaning as html tags. Thus, italic and bold tags are opened, never to be closed, and so is an anchor tag. This is influencing whatever follows, because all of the latter is kept within the (unclosed) tags. Nasty.

Thus, the editor, becoming aware of what has happened, has to reenter the edit page and recorrect the entry and save. Until some other editing on a similar editing page is to be done. The entry is to be corrected over and over again, ad infinitum.

Possible solutions : LT had better cut such text snippets into smaller snippets, hard coding those tags, as in :
You can use some HTML tags, such as <b>, <i> and <a>.
substituting the hard coded snippets--in this case the html tags (I 've put them in bold now)--for editing purposes with place holders. The same technique is used elsewhere by LT, too.

Thus, the corresponding edit field should better read e.g.:
You can use some HTML tags, such as <SUB1>, <SUB2> and <SUB3>.

Alternatively, but this is no optimal solution--though I can understand them and why they did so (see above)--one is to opt for the German solution. The German language section seems to have got enough with it--and I do understand them--and halted the necessity for reediting over and over again at infinitum by translating :
You can use some HTML tags, such as <b>, <i> and <a>.
into:
Du kannst einige HTML-Tags verwenden, wie "b", "i" und "a".

Should other language opt for the same "solution" or is LT going to provide a real solution? A fitting method for the latter can be seen above.

Thanks. When I get back to the US I'll work on some of the problems, and add a "untranslated snippets" page.

>2 timspalding:

Thank you, too.

(I just recorrected the French version of You can use some HTML tags, such as <b>, <i> and <a>. Again.)

At the LT Blog, I read:

"Since then we've been making more pages transateable**. (...)

**Including the list of 500 MARC-specification languages. The system considers all "snippets" equally important, so that hit the percentages hard.
(LibraryThing in your language , excerpt.)

** The missing l in transLatable is no big deal.

Then again, I wonder how to reach the list of 500 MARC-specification languages in a way I (or whoever else) can edit and translate them.

Neither do I know how to reach the edit and translate page for e.g. the Touchstones explicatory paragraphs next to the (this) message edit & add field.

Problem is I discover lots of untranslated text, in yellow, but have no idea how to reach their applicable edit page. I must be a dumb. Oh, I am-- no single doubt about that--but that doesn't solve the question. =)

All this said, you will solve it when you get back--when you get back?--when you get back to the US.

In the mean time, I am quite curious about your exposé at the Frankfurter Buchmesse and about how that went. Will any streaming video of it be released for us to view? We are--speaking for myself--I am quite very curious to hear about that, you know. You know!

And while not back in the US yet, enjoy your journey!

Though I didn't make it to Frankfurt this time, I look forward to meeting you in person.

<publicity block>
Tip: There is a big book fair over here at Antwerp in short time: 70e Boekenbeurs, Antwerp Expo, November 1-12, 2006. This is a yearly major event in Low Countries' book life. It is open to the grand public and attracts some 180.000 visitors.

You might know Antwerp is the home of the famous Officina Plantiniana (printers since 1555), too. It is the very last Renaissance and Baroque printers' and publishers' left intact. Nowadays, it is open to the public and know as the Museum Plantin-Moretus/Prentenkabinet, being i.a. the only museum in the world recognised by UNESCO as World Heritage.
They print on demand; e.g. Christophorus Plantyn's "Le Bonheur de ce Monde", at one of the oldest printing printing presses of the world : theirs--i.e. those of the house--for this is the only place where that old printing presses still exist, dating back from ca. 1600 A.D. and in mint condition--of course.
<nagging>
Frankfurt am Main to Antwerp is 400 km. Were you ever that close?
</nagging>
</publicity block>

Envious to hear your trip story.

As I pointed out elsewhere, pretty much everything that is switched in through script is unreachable fortranslation as of now, that include the MARC list, the touchstone text, the green stuff for data taken for otehr books and the poweredit text, amongst otehr stuff.

EDIT: Nevermind that, the list is now fully available, have fun.

>4 circeus:

> As I pointed out elsewhere, (...)

Thank you. I 've read that, too, and I agreed and agree--in full--of course. Then again, of course, the blues with this is no less, for it is obstructing the translation process, if not halting it.

I think there is no doubt you, too, are enviously looking forward to reach some 100% translated stage. We do read each other, don't we?

> EDIT: Nevermind that, the list is now fully available, have fun.

Thank you again, but, hmm, but this is not exactly the list, I think.

Tim said :
"Since then we've been making more pages transateable**. (...)
**Including the list of 500 MARC-specification languages.

The list you are linking to--and which I 'd already found, too--is counting rather some 150 languages. Even if they were 200--and they are not--thit is not, i.e. this cannot and must not be the list of 500 MARC-specification languages list Tim was talking about. Methinks.

Then again, as I am writing these words, there is the untranslatable Touchstones explanation aside. Untranslatable for now, because neither of us knows how to reach and edit it.

You get the feeling.

Persiste et signe,
Translator's blues...

3023787817 is the "see all languages" variant of translate.php. It only lists those languages that have at least one book in LT.

A quick-and-dirty hack would be for someone at LT to add a request parameter to the PHP that removes that check (or substitues a different SQL query or whatever) so that the "green box" lists all known codes.

>6 MMcM:

That's why we--(and) LT--need some lists. No matter whether all entries in those are actually in use now.

Some lists that should be available for translation (and number of entries in them) :

1) days of the week (7);
2) month's names (12);
3) language names (see e.g. MARC) (+/- 500)
4) numbers : 0-99, 100, 200, 300,...900, 1000, 2000, 3000, ... 10,000, (...) (+/- 130?)
5) ...
...

(Please add as you feel necessary.)

Because, though maybe not immediately--i.e. now--LT will need them, all or at least most of those entries and their translations--soon. Thus, LT 'd better keep them at hands.

Keep in mind, too, that LT translation hype is now. Thus, best moment to get those lists and all of their entries translated in full is now too. That will make them immediately available whenever one or some of their entries are wanted--or rather needed : now!

Disclaimer : Sorry for trying to think along & ahead.

>7 boekerij:

Some of those lists have become available for translation now--in the "odds and ends" section--though not in full and/or with problems :

1. Days of the week :

OK, except for translators don't know for sure whether capitalisation is obligatory--or not.

2. Month's names :

problems :

2.1. In English, month's names get capitals, while in some (most?) other languages, they do not. For the source language is English, LT should rather make clear whether plain month's names are wanted or rather month's names with capitalisation.

2.2. There is only one single "May" entry. It is used as well as a month's name AND as the month's name abbreviation. This is no good, for though in English both are the same, in many other languages this month's name is longer than three characters. At present, translators cannot know whether they 'd use the former or rather the later. Worse : whatever they choose, the entry is always wrong, either in the former or in the latter way. Solution : split this entry please.

3. Language names & false statements

Though part of language names is available for translation now, most of them are not.

This is quite frustrating, for : a) "international users" have to fall back to the language names in English; b) translators are to be at their "qui vive" (attentive) and are to check and recheck whether a--one single or some handful at most--"new" language might be added yet (and thus become available for translation). There is no "permanent" solution now. The availability of the list of 500 MARC-specification languages--in full, i.e. : "unabridged"--at least, might bring some relief.

Worse : more than a month ago, Tim said it were available (see above AND Tim's "LibraryThing in your language" LT Blog entry). It was not and it is not--and we all know this.

Possible solutions :

Update ASAP--i.e. : as soon as possible or without any more delay (read : now)-- the "LibraryThing in your language" LT Blog entry (by Tim at Thursday, October 5, 2006) by striking i.a. its statement : "**Including the list of 500 MARC-specification languages." for it was never met and thus--though this is a pity indeed--until now, is false.

Better solution (highly preferred) :

Meet the statement. This can be and is to be done by "**Including the list of 500 MARC-specification languages." indeed.

I am looking forward to reading an update as e.g. :
"We are sorry there has been some delay in this--and thus your frustration in this matter--but eventually, we 've been able to meet our own statement. Thanks for your kind patience. The unabridged list of all 500 MARC-specification languages is available for translation now."

It might be a good idea to include the latter list (in full) in "Odds and ends", perhaps in its own subsection.

Note :

I do know that there are real people at the other side. There are this side, too, and we know many of them, either side indeed, are investing lots of their time and best efforts in trying to ameliorate LibraryThing.

Important :

Then again, I do not think it is any good, nor can it be in LibraryThing's interest, to make false statements and not correcting them--at once.

LT does not meet just expectations by making any, worse : by its even not addressing and correcting false statements. This is a Bad Thing indeed. This very problem was mentioned more than a month ago (see above). It has not been corrected yet. It should have been corrected more than a month ago. This has not been done yet. It should MUST be addressed and corrected. Now.

FYI : Whomever might think this is personal, is wrong. This is about LT's very interest. Whomever dissents should say so.

4. Numbers

Some time ago, the "six" entry was made available for translation. Some days ago, the "seven" entry was added. The "eight" entry might be expected in less than a month. This method is no use, nor is it a good thing, I think, for, as we all know, higher numbers too will be needed quite soon. Of course, the latter is a Good Thing. =)

Then again, the present method is cumbersome. The latter might be solved in an easy as well as an elegant and handy way by addressing proposal 7.4 (see above), i.e. : by having and keeping those numbers (and their translations) at hands. In other words : let LT's "Odds and ends" section have a "Numbers" subsection. We all know LT will need them--some of them rather sooner than later. Thus, LT 'd better have those translations at hands for off the shelve use whenever it sees fit. Please do not delay providing for those translations.

5. "See all translated"

Last week, the different languages' "See all translated" pages ("For manaical translators") went blank. (cf. "Translating LibraryThing? (General Talk) : "See all untranslated!" inaccessible?")

They are not up and running again, yet.

This is a pity, because they have been around for about a month--see i.a. Message 2 above and thank you very much indeed for having provided them-- and are were a handy translators's tool. Without those pages, newly added text snippets to be translated are to be discovered more or less by accident, or by time consuming continuous wading through each and every single LT page to try and detect whether--perhaps--new text snippets might be added. This is not quite fun--and rather discouraging indeed.

Please have the "See all translated" sections up and running again soon.

6. Untranslatable translatables : homographs (to be split!)

This one might be a rather puzzling header, but it is pointing at a rather real and quite frustrating problem : (English language) homographs combined as one text snippet to be translated. The latter must go wrong.

An example of those is (i.a.) "read != read" (cf. e.g. "Translating LibraryThing? (General Talk) : read != read", mentioned Oct 23, repeated Nov 1)

English speakers will know that the homographs "read" and "read" are having a different meaning--AND a different pronounciation (/ri:t/ vs. /rEt/) for that matter. Thus, having the single text snippet "read" only, one cannot decide what meaning (and pronounciation) is to be chosen. Worse : in translation, those are no more homographs. Worse still : LT is using both meanings combined into one single text snippet to be translated. The latter is not possible : choosing the one meaning, the translation will be right in one place (and meaning), but wrong in another place (and meaning), while choosing the other meaning, it's all the same the other way round.

Please split this entry, for the combining into one single "read" is no good--unless for permanent error and frustration, that is.

Addendum

I hope everyone reading this will understand that I am not trying to bash anyone--and I think I am not, either--thus no one is to feel hurt.

Though "de mortuis nil nisi bono"--LibraryThing is not dead--it would be a pity if it were--nor are the remarks as above personal. I do not think we 'd be helping LibraryThing by applauding only, even when we see acute problems, difficulties, errors and even false statements.

While trying to help LT--in our own time indeed, that is--I think that is our duty to mention--and even remention and not to give up--acute errors etc.

Of course, there are many more--and I know of many more, too, as without any doubt do many other thinkers-along--but those few quite acute ones might do for a start--for now.

I feel sorry I thought this addendum might be needed. I think it shouldn't be questioned at all, for it was obvious, in my view at least, and most probably in the views of the many users having invested and investing many dozens of hours--of their proper time indeed--and best efforts in trying to help and ameliorate LT. I think the LT community as such is doing a great job.

Then again, in Dutch we 've got a saying : "Zachte heelmeesters maken stinkende wonden"--literaly : "soft doctors call for stinking wounds"--~ "desperate cases call for desparate remedies"). I think LT is no desparate case at all. It is great indeed. Don't take my word for this. For words might be cheap. Better take a look at the many users--LT members living wherever in the world--offering and investing lots of their time and efforts into LT. Those are not cheap words. Those members are volunteering by putting their time and efforts where their mouth mood is. It might be mute, but the word is loud and clear. It is crying louder than any words might do. It states and proves : we believe in this project and we go for it. Not by cheap words, but rather by acts. I do not think anyone could imagine nor want any stronger proof. This is value, I think.

Thus, even though some comments might have been felt, by some people, as being rather provocative--perhaps they were--I do not think :

Didn't your parents teach you that if you don't have something nice to say, don't say anything at all.

was the right answer. For :

There are real people on the other end of these message boards.

indeed. And they care about LT, too. Don't take their word for it. Look at proof of their efforts and acts.

What's more : if all of them (us) would conform to :

Didn't your parents teach you that if you don't have something nice to say, don't say anything at all.

I think this 'd be the worst service they could provide to LT. It 'd be plain disaster, for it 'd mean they 'd refrained from caring about LT.

There is good news too, very good news indeed. They didn't and they don't. They say what they think they have to say. Their honesty is their proud. I think LibraryThing should be can be is proud of this, too. And right so.

Courage, persévérence, and keep up with the good work. All of you.

FYI : the latter is personal.

boekerij wrote Nov 10, 2006 among other things:
"2.2. There is only one single "May" entry. It is used as well as a month's name AND as the month's name abbreviation. This is no good, for though in English both are the same, in many other languages this month's name is longer than three characters. At present, translators cannot know whether they 'd use the former or rather the later. Worse : whatever they choose, the entry is always wrong, either in the former or in the latter way. Solution : split this entry please."

The problem still exist. I just tried to translate the abbreviated May, but it changed the full form, too. Maybe it is not possible to translate LibraryThing properly. What to do?

> 9

I've just bumped into the same phenomenon in Hungarian, so it still exists.

I need to find where this is used. One solution would be to translate everything in the long form, no?

Re 11: Dates (times, months, weekdays, ...) should be written with PHP's locale functions.

> 11

here it asks for the longer form:
http://hu.librarything.com/translate.php?page=3329435732

and here it asks for the shorter:
http://hu.librarything.com/translate.php?page=2692807616

Hope this helps!