Language by field

[maieul]

I have read the manual of v 4.0, and I have found no things.

The langid field define the language of all the entry. But it's possible we have language which are not the same in differents fields. For example, in the case of a @incollection in english edited in a @collection in french : the language of the booktitle field should be french and title field should be english...

Should it be possible to add feature to allow to switch language/hyphenation between fields.

I imagine something like booktitle_langid, title_langid. In case of @crossref, the xxx_langid will follow the xxx source.

[maieul]

oups, sorry, post to fast. i will complete

[plk]

This is exactly what the experimental branch does - you'll need the biblatex and biber experimental branches and everything should be into the docs in those releases.

[maieul]

Ok, I have tried biblatex 4.0 + biber 3.0. Or maybe some thing is not clear in the handbook, or maybe some think is wrong, or maybe I am too tired to understand the handbook.

I have read in biblatex handbook to "2.2.2 Data Field Variants" and in biber handbook to "3.13. Extended BibTeX Datasource Field Format"

Consider this MWE

\documentclass[a4paper]{article}
\usepackage{polyglossia,xunicode,csquotes}
\setmainlanguage{french}
\setotherlanguage{italian}

\usepackage[citestyle=verbose]{biblatex}
\addbibresource{biblio.bib}

\begin{document}

\cite{Canart1982}

\end{document}

\end{document}

with the .bib file

@incollection{Canart1982,
    Address = {Milan},
    Author = {Paul Canart},
    Booksubtitle= {raccolta di studi in memoria di Agostino Pertusi},
    Booktitle= {Bisanzio e l'Italia},
    Number = {22},
    Pages = {19-28},
    Publisher = {Vita e pensiero},
    Series = {Pubblicazioni della Università Cattolica del Sacro Cuore. Scienze filologiche e letteratura },
    Title = {Cinq manuscrits transférés directement du monastère de Stoudios à celui de Grottaferata?},
    Year = {1982}}

I obtained the right result

Paul Canart. « Cinq manuscrits transférés directement du monastère de Stoudios à celui de Grottaferata? » In : Bisanzio e l’Italia. raccolta di studi in memoria di Agostino Pertusi. Pubblicazioni della Università Cattolica del Sacro Cuore. Scienze filologiche e letteratura 22. Milan : Vita e pensiero, 1982, p. 19–28

But I have not the right hyphen, because the book title etc are italien, not french. So I modify with

@incollection{Canart1982,
    Address = {Milan},
    Author = {Paul Canart},
    Booksubtitle_italian= {raccolta di studi in memoria di Agostino Pertusi},
    Booktitle_italian= {Bisanzio e l'Italia},
    Number = {22},
    Pages = {19-28},
    Publisher_italian = {Vita e pensiero},
    Series_italian = {Pubblicazioni della Università Cattolica del Sacro Cuore. Scienze filologiche e letteratura },
    Title = {Cinq manuscrits transférés directement du monastère de Stoudios à celui de Grottaferata?},
    Year = {1982}}

And despite the right entry in .bbl file

      \field[form=original,lang=italian]{booktitle}{Bisanzio e l'Italia}

These fields are not printed :

Paul Canart. « Cinq manuscrits transférés directement du monastère de Stoudios à celui de Grottaferata ? » In : 22. Milan, 1982, p. 19–28

Where was I wrong?

[plk]

Try this:

\documentclass[a4paper]{article}
\usepackage{polyglossia,csquotes}
\setmainlanguage{french}
\setotherlanguage{italian}

\usepackage[citestyle=verbose,autolangshort=hyphen]{biblatex}
\addbibresource{test2.bib}
\begin{document}
\vlang{italian}

\cite{Canart1982}

\printbibliography
\end{document}

What happens here is that the default lang is "french" as you want but then you switch the to "italian" variant but with default fallback to "french" as this is the global setting. The autolangshort option should do the right hyphenation for the language for the field that is printed.

The reason that the italian fields are not printed by default is that people may want to have one entry with many different translations of every field and then print several bibliographies, one french, one italian etc. With this new version, bibtex entries are more like databases where you have to select what you want to print from each entry - probably don't usually want to print everything in every language in every entry ...

[maieul]

Your example works. But I have some trouble with your explanation. What I understand is that the new feature you has added is to manage a same field in ≠ language. For example for a ancient greek text : latin (normalized title), greek (original title), french (translated title). That is very nice. The mechanism of \vlang, if I understand well, is to switch to the choosen language.

But what I need in the present case is not that. There is only one language for one field BUT a same entry have different language for different fields. And I can't determine in advance which language is it for each field, so the \vlang fallback can't be use. For example I could have, in the text, this entry:

@incollection{Example,
  Address = {Milan},
  Author = {English People},
  Booksubtitle_italian= {raccolta di studi in memoria di Agostino Pertusi},
  Booktitle_italian= {Bisanzio e l'Italia},
  Number = {22},
  Pages = {10-12},
  Publisher_italian = {Vita e pensiero},
  Series_italian = {Pubblicazioni della Università Cattolica del Sacro Cuore. Scienze filologiche e letteratura },
  Title_english = {A title in English},
  Year = {1982}}

So what I hope biblatex will do is:

1. to switch to the language hyphen
2. to choose the field with is only value.

I can't see how I coul print this entry :

\vlang{english} will print only the title field \vlang{italian} will print only the booktitle field

And I don't speak of the final bibliography, which need also the @Canart1982 entry. but what I need is both : my entry is a "normal" @incollection entry, but the @collection is Italian, and tne @incollection is English…

[plk]

You are right, this is also a valid use case - let me think about it a bit ...

[maieul]

yes, of course!

[plk]

Please try biblatex and biber (both version 4.0) from the "experimental" folders on SF. There are now two "modes" for multiscript operation. Your use case is "entry" mode. So you can try just this:

\documentclass[a4paper]{article}
\usepackage{polyglossia,csquotes}
\setmainlanguage{french}
\setotherlanguage{italian}

\usepackage[citestyle=verbose,autolangshort=hyphen,msmode=entry]{biblatex}
\addbibresource{test2.bib}
\begin{document}

\cite{Canart1982}

\printbibliography
\end{document}

The two modes are documented in the PDF documentation. Mode can be switched per-entry too.

[maieul]

Thanks a lot,

I have made test and tried your example. It's work well. However, I have some suggestion/question.

a) In the handbook of biblatex, p. 59 "Selects the global multiscript mode. See secrefbib:fld:var." > bad crossref b) maybe the option could be also by entry. For example a @bookinbook should be in msmode=db (because we assume that is a primary source, where msmode=db is logical) while a @incollection should be in msmode=entry (because we assume that is a secundary source, where msmode=entry is logical.

There is, anyway, a case where a "mix" system could be useful. That is the case for example, in this case:

• an old source is published as a @bookinbook. So the title could be in three form:
  • translated, in the language of the main document (in my case, french)
  • normalized, in the scientific language (in my case : latin)
  • kept in original form (in my case : greek)
• this old source is published in a book which the only title is latin
• this old source is cited in a work in french (or english).

So concretly, the .bib file:

@bookinbook{BHG225,
  booktitle_latin = {Acta Apocryphorum Apostolorum},
  date      = {1903},
  location  = {Leipzig},
  publisher = {Hermann Mendelssohn},
  pages         = {292-302},
  title_original_greek ={Περίοδοι καὶ μαρτύριον τοῦ ἁγίου Βαρνάβα τοῦ ἀποστόλου},
  title_uniform_latin         = {Acta Barnabae},
  title_translated_french         = {Actes de Barnabé},
  volume        = {2.2},
}

In this case, I would like to profit of the msmode=db to switch between the original/uniform/translated form BUT profit of the msmode=entry to swith the hyphen of the booktitle.

[maieul]

my suggestion : msmode should make change only for the language variation, not the form variation.

[plk]

I don't think it's possible to mix msmode in the same entry - it needs a completely different code path. However, msmode can be set in the options of an entry which overrides the global per-entry:

OPTIONS = {msmode=entry}

can you get the behaviour you want by using related entries? That is, separate the bookinbook from book?

[plk]

Also, I think you can get what you want by settings the variant fallbacks so that regardless of what the main variant for your entry is in db mode, it will fall back to for the booktitle in latin if there is no matching variant.

[maieul]

hum, I am not sur to understand what do you mean by "can you get the behaviour you want by using related entries? That is, separate the bookinbook from book?"

I will try the vfallback.

[plk]

Ah, forget the related entries thing, that can't work.

[maieul]

Hum, I have tried with the following file:

\documentclass[a4paper]{article}
\usepackage{polyglossia,csquotes}
\setmainfont{Linux Libertine O}
\setmainlanguage{french}
\setotherlanguage{latin}
\setotherlanguage[variant=ancient]{greek}

\usepackage[citestyle=verbose,autolangshort=hyphen,msmode=db]{biblatex}

\DeclareVariantFallbacks{ 
  \vfallback[form=original] 
  \vfallback[form=original, lang=latin]
  \vfallback[form=original, lang=greek]
  \vfallback[form=romanised]
}

\addbibresource{test2.bib}
\begin{document}

\cite{BHG225}

\end{document}

and the following file

@bookinbook{BHG225,
  booktitle_latin = {Acta Apocryphorum Apostolorum},
  date      = {1903},
  location  = {Leipzig},
  publisher = {Hermann Mendelssohn},
  pages         = {292-302},
  title_original_greek ={Περίοδοι καὶ μαρτύριον τοῦ ἁγίου Βαρνάβα τοῦ ἀποστόλου},
  title_uniform_latin         = {Acta Barnabae},
  title_translated_french         = {Actes de Barnabé},
  volume        = {2.2},
}

and nor the booktitle nor the title field are printed. So or I don't understand the \DeclareVariantFallbacks handbook, or there is a bug. Notes that if I change to

@bookinbook{BHG225,
  booktitle = {Acta Apocryphorum Apostolorum},
  date      = {1903},
  location  = {Leipzig},
  publisher = {Hermann Mendelssohn},
  pages         = {292-302},
  title_original ={Περίοδοι καὶ μαρτύριον τοῦ ἁγίου Βαρνάβα τοῦ ἀποστόλου},
  title_uniform_latin         = {Acta Barnabae},
  title_translated_french         = {Actes de Barnabé},
  volume        = {2.2},
}

that workings, but I have problem with hyphenation/language switching.

In any case, there something that I don't understand : why mixing the problem of the form and the problem of the language? For an entry, a translated / uniform / original has, in many case, only one language (there is only one original title, only one translated title etc.). So it should be more useful if the variant choice of biblatex concern only the form, and not the language. Because that will imply to explain each original/language tuple allowing in a database if our choice is to use the original title.

I can reexplain my b) suggestion. In many case the choice between msmode=entry / msmode=db is depending of the entry type. So instead of set it for each entry of a same type, a global choice by entrytype could be useful. In other word, p. 267 of the handbook, it will be nice if the case msmode/per-type will be with a bullet.

[plk]

Hmm, let me think about this, it's not obvious how to mix db and entry mode in the same entry. I think we have to leave variants as form+lang because people might want to store multiple languages per form (for translated forms, for example).

I can make msmode per-entrytype no problem if necessary.

[maieul]

I understand. A simple (for user) and simplier (for you) way should to provide a option (by a command) which make biblatex goes to the only language of the field for a specific variant. That make db/entry mode obsolete AND should solve my problem with BHG225 entry.

Explanation : with this option set to true,

@bookinbook{BHG225,
  booktitle_latin = {Acta Apocryphorum Apostolorum},
  date      = {1903},
  location  = {Leipzig},
  publisher = {Hermann Mendelssohn},
  pages         = {292-302},
  title_original_greek ={Περίοδοι καὶ μαρτύριον τοῦ ἁγίου Βαρνάβα τοῦ ἀποστόλου},
  title_uniform_latin         = {Acta Barnabae},
  title_translated_french         = {Actes de Barnabé},
  volume        = {2.2},
}

with vform=original will choose "Περίοδοι καὶ μαρτύριον τοῦ ἁγίου Βαρνάβα τοῦ ἀποστόλου" as title and "Acta Apocryphorum Apostolorum" as booktitle. With vform=uniform, with choose Acta Apocryphorum Apostolorum because of the \DeclareVariantFallbacks setting BUT Acta Barnabae as title. And with vform=translated, will choose Actes de Barnabé as title. In any word : juste choose the form, and get the only language available.

[plk]

Actually, it's not so clear in the documentation but \DeclareVariantFallbacks only changes certain things that biber does. Fallbacks for biblatex just means falling back to the global vform/vlang setting.

[maieul]

Ok, but I still don't understand how to use it. This entry:

@bookinbook{BHG225,
  booktitle_latin = {Acta Apocryphorum Apostolorum},
  date      = {1903},
  location  = {Leipzig},
  publisher = {Hermann Mendelssohn},
  pages         = {292-302},
  title_original_greek ={Περίοδοι καὶ μαρτύριον τοῦ ἁγίου Βαρνάβα τοῦ ἀποστόλου},
  title_uniform_latin         = {Acta Barnabae},
  title_translated_french         = {Actes de Barnabé},
  volume        = {2.2},
}

msmode =db.

Without \DeclareVariantFallbacks:, the .bbl file contain

    \entry{BHG225}{bookinbook}{}
      \list[form=original,lang=french]{location}{1}{%
        {Leipzig}%
      }
      \list[form=original,lang=french]{publisher}{1}{%
        {Hermann Mendelssohn}%
      }
      \field{sortinit}{Π}
      \field{sortinithash}{9056cb6813a3c373de7ee34a24f2f7ae}
      \field{labeltitle}{Περίοδοι καὶ μαρτύριον τοῦ ἁγίου Βαρνάβα τοῦ ἀποστόλου}
      \field{labeltitlesourcefield}{title}
      \field{labeltitlesourceform}{original}
      \field{labeltitlesourcelang}{french}
      \field[form=original,lang=latin]{booktitle}{Acta Apocryphorum Apostolorum}
      \field[form=original,lang=greek]{title}{Περίοδοι καὶ μαρτύριον τοῦ ἁγίου Βαρνάβα τοῦ ἀποστόλου}
      \field[form=translated,lang=french]{title}{Actes de Barnabé}
      \field[form=uniform,lang=latin]{title}{Acta Barnabae}
      \field{volume}{2.2}
      \field{year}{1903}
      \field{pages}{292\bibrangedash 302}
    \endentry

And with

\DeclareVariantFallbacks{ 
  \vfallback[form=original] 
  \vfallback[form=original,lang=latin]
  \vfallback[form=original,lang=greek]
  \vfallback[form=romanised]
}

I still obtain the same file... and no title or booktitle is printed

[maieul]

sorry, message send to fast. Please wait 10 minutes before read it.

[plk]

My comment wasn't supposed to solve the issue, I'm still looking into it - it was just for information.

[maieul]

Ok. take time... no emergency and thanks again.

[plk]

Try the 4.0 versions of biblatex and biber again. msmode is gone and there is a new setting for vlang. Setting vlang to * should enable the behaviour you want - it auto-selects the only language for the currently active form.

[maieul]

Ok, thats work very fine. Thanks a lot ! However, there is a typo p. 20 of the biblatex handbook "TITITLE".

Most of that, I have a question on the relationsheep between the \vform and \vfallback. If I understand well the handbook (which is clear) in this point:

• vform/vlang concerns the way th field is printed in biblatex
• vfallback concern the way biber:
  • add a label
  • sort fields

My first remarks is that \vfallback can be define only in the preamble, but vform can be change in the main stream. So the way fields are sorted can be different the way they are sorted. Suppose, as in your example, we decide to 19, we switch from original form to translated form : we will have to change the order of the entries, if not, it will be not sorting by author/title/year...

There is a second point, close to the first:

with this entry:

@bookinbook{BHG225,
  booktitle_latin = {Acta Apocryphorum Apostolorum},
  date      = {1903},
  location  = {Leipzig},
  publisher = {Hermann Mendelssohn},
  pages         = {292-302},
  title_original_greek ={Περίοδοι καὶ μαρτύριον τοῦ ἁγίου Βαρνάβα τοῦ ἀποστόλου},
  title_uniform_latin         = {Acta Barnabae},
  title_translated_french         = {Actes de Barnabé},
  volume        = {2.2},
}

biber doesn't generate a labeltitle field, which is problematic, because many styles use it. For example, with the citestyle=verbose, try:

\cite{Canart1982}

\cite{BHG225}

\cite{Canart1982}

\cite{BHG225}

The second cite of BHG225 will be empty... except if I do

\DeclareVariantFallbacks{
  \vfallback[form=original] 
  \vfallback[lang=greek] 
  \vfallback[form=romanised]
}

but that implies to do it for each potential language I my bibliographie (and it's not clear : what happens if I have a translated title in greek?)

[plk]

Please pull biber again. Now you can specify lang=* in all places which pass through to biber. So:

\DeclareLabeltitle{%
  \field{shorttitle}
  \field[lang=*]{title}
}

Will select a labeltitle using the global vform and whatever lang is available. Note that \vform and \vlang commands are document commands and can't influence what is passed to biber at all. biber always gets the global vform and vlang setting defaults.

[maieul]

I will try two days. What about the problem on the dissonance between the way a field is printed and the way it's sorted?

[maieul]

I have take time to test. The label title is ok. In my sense, the lang=* (for both biblatex and biber) should be the default setting.

However, there are something very strange.

Suppose the two followings entries:

@bookinbook{BHG225,
   booktitle_latin = {Acta Apocryphorum Apostolorum},
   date      = {1903},
   location  = {Leipzig},
   publisher = {Hermann Mendelssohn},
   pages         = {292-302},
   title_original_greek ={Περίοδοι καὶ μαρτύριον τοῦ ἁγίου Βαρνάβα τοῦ 
ἀποστόλου},
   title_uniform_latin         = {Acta Barnabae},
   title_translated_french         = {Actes de Barnabé},
   volume        = {2.2},
}

@incollection{Canart1982,
     Address = {Milan},
     Booksubtitle_italian= {raccolta di studi in memoria di Agostino 
Pertusi},
     Booktitle_italian= {Bisanzio e l'Italia},
     Number = {22},
     Pages = {19-28},
     Publisher_italian = {Vita e pensiero},
     Series_italian = {Pubblicazioni della Università Cattolica del Sacro 
Cuore. Scienze filologiche e letteratura },
     Title = {Cinq manuscrits transférés directement du monastère de 
Stoudios à celui de Grottaferata?},
     Year = {1982}}

(I have volontary suppress the author of Canart1982)

and

This file:

\documentclass[a4paper]{article}
\usepackage{polyglossia,csquotes}
\setmainfont{Linux Libertine O}
\setmainlanguage{french}
\setotherlanguage{latin}
\setotherlanguage[variant=ancient]{greek}

\usepackage[bibstyle=verbose,  
citestyle=verbose,autolangshort=hyphen]{biblatex}
\vlang{*}
\vform{original}

\DeclareLabeltitle{
   \field{shorttitle}
   \field[lang=*]{title}
}
\addbibresource{test2.bib}
\begin{document}

\cite{Canart1982}

\cite{BHG225}

\cite{Canart1982}

\cite{BHG225}

\printbibliography[sorting=nty]
\end{document}

The BHG225 is printed before the Canart1982.

I have modified it by doing

\DeclareSortingScheme{nty}{
   \sort{
     \field{presort}
   }
   \sort[final]{
     \field{sortkey}
   }
   \sort{
     \field{sortname}
     \field{author}
     \field{editor}
     \field{translator}
     \field{sorttitle}
     \field[lang=*]{title}
   }
   \sort{
     \field{sorttitle}
     \field[lang=*]{title}
   }
   \sort{
     \field{sortyear}
     \field{year}
   }
   \sort{
     \field[padside=left,padwidth=4,padchar=0]{volume}
     \literal{0000}
   }
}

but it's very contre-intuitive. That why I plead for a connexion the way biblatex work and the way biber work... or at least for a use of lang=* in all the \field or as default value... in any case, i think the * should also be included in the handbook of sorting system p. *.

I have also tried

\DeclareVariantFallbacks{
\vfallback[form=original]
\vfallback[lang=*]
\vfallback[form=romanised]
}

and the \vfallback[lang=*] seems to work, in the meaning of the field title_original_greek is used for translation. That should be documented. But in any case two questions remains

a) what exactly does \DeclareVariantFallbacks make ? it seems to influente the sorting but not the labelling, despite the p. 20 which say it influences labelling b) if \DeclareVariantFallbacks is to be used to influente the sorting, it should be used not only in the preamble. If not, it should be clarified in the handbook and the way to influente the sorting should be explained (eg. a link should be made between p. 19 and p. 182)

In any case, thanks for your work. It's a very complex problem, and I biblatex can do it, Zotero will be out for a long time.

[maieul]

I have also thing to an other problem : sometime we have a BDD which mixt entries with only one language by title and with multiple language by title (book published with 2 titles for example). So \vlang{} is problematic, because we don't know which entry will be selected. So the best solution should be to define a list of language to choose, and, after it, the default one. For example \vlang{english,latin,}.

[maieul]

example of such case

@proceedings{Amsler2008,
   Editor={Frédéric Amsler and Albert Frey and Charlotte Touati},
   Title_french={Nouvelles intrigues pseudo-clémentines},
   Title_english={Plots in the Pseudo-Clementine Romance},
   Publisher={Éditions du Zèbre},
   Location={Prahin},
   Year={2008}
}

in the same work that the BHG225 entry. I should prefer the french form when I write in French, but if I must write in English...

[plk]

Ok, another attempt - please pull 4.0 of biblatex and biber again. Now, vlang=* is removed and there is a separate option autovlang=true which is the default. This is passed through to biber. See how this works for you.

[maieul]

Ok,

a) the labelfield is now automatically good. b) but the dissonnance between the field_language used for printing and for sorting is still here. Try for example with these two entries:

@bookinbook{BHG225,
  booktitle_latin = {Acta Apocryphorum Apostolorum},
  date      = {1903},
  location  = {Leipzig},
  publisher = {Hermann Mendelssohn},
  pages         = {292-302},
  title_original_greek ={Περίοδοι καὶ μαρτύριον τοῦ ἁγίου Βαρνάβα τοῦ ἀποστόλου},
  title_uniform_latin         = {Acta Barnabae},
  title_translated_french         = {Actes de Barnabé},
  volume        = {2.2},
}

@incollection{Canart1982,
    Address = {Milan},
    Booksubtitle_italian= {raccolta di studi in memoria di Agostino Pertusi},
    Booktitle_italian= {Bisanzio e l'Italia},
    Number = {22},
    Pages = {19-28},
    Publisher_italian = {Vita e pensiero},
    Series_italian = {Pubblicazioni della Università Cattolica del Sacro Cuore. Scienze filologiche e letteratura },
    Title = {Cinq manuscrits transférés directement du monastère de Stoudios à celui de Grottaferata?},
    Year = {1982}}

and with

\usepackage[bibstyle=verbose,  citestyle=verbose,autolangshort=hyphen,autovlang=true]{biblatex}

\vform{uniform}

The problem of the choice between two language remains, for example in the case of

@proceedings{Amsler2008,
    Editor={Frédéric Amsler and Albert Frey and Charlotte Touati},
    Title_french={Nouvelles intrigues pseudo-clémentines},
    Title_english={Plots in the Pseudo-Clementine Romance},
    Publisher={Éditions du Zèbre},
    Location={Prahin},
    Year={2008}}

mixted with a case where the autovlang=true. That why I was thinking that define a order of language should be better french,english,greek,latin,default. There is one other problem in this case : the second run of xelatex said

! Missing \endcsname inserted.
<to be read again> 
                   \abx@name@editor 
l.30 \cite{Amsler2008}

[plk]

1. The first issue isn't really solvable - \vform is a document command which switches the default form during document processing and so biber never sees it. The only way to deal with this is to define multiple sorting/label schemes and use them in the right places.
2. I can't see much use in this sort of case - you will often not be able to predict which language for a field you will actually print with a complex bibliography. You can always just set autovlang=false and then switch with \vlang?
3. The error should be fixed now.

[maieul]

I make answer in the reversed order.

1. I will try
2. It's not possible to switch with \vlang, because you can have, in a same bibliography, a entry which the lang is unique (for the entry), but this lang is not common to all the entries (like my BHG225), and, in the same bibliography, and entries whith multiple language (like my Amsler2008) Why not propose to the author to define an order list of language to choose. For example : \vlang{french, english, italian, latin, *} ? In this case : a) For "Amsler2008" Biblatex will choose the french, as it the language in the top of the list b) For "BHG225", biblatex will choose the latin, as the * is mentionned in my list.
3. That is what I don't understand. Why having splitted the way an entry is processed by latex and the way it is processed by biber? My suggestion will be a) or to mention it explicitly in the handbook. b) or, better, create a command which switch the ways both biblatex prints and biber sorts. Because, actually, your example in the documentation is not very useful, because it's create an unordonned list.

[plk]

I think you are right - I'm looking into redesigning the system to address these points - it probably means changes in biblatex dev branch too first though.

[maieul]

hi,

do you have some news for this issue?

[plk]

Yes, I have refactored a lot of code in biber and I have to a bit more in biblatex still. It's a lot of work as I had to rethink the whole system.

[maieul]

ok, don't hesitate to ask me for test

[maieul]

hi, do you think to have time to work in it, or is it a dead project?

[plk]

I had done most of the biber work but I have currently no time to look into the biblatex part - I think we really need to find an active biblatex developer to help on this if you have any ideas ...

[maieul]

unfortunatly, I have no idea :-(. I have also a lot of thing to do in reledmac/reledpar, and I can understand you have no time. but I see noone.

[plk]

I will close this for now - this whole thing needs re-thinking.

[maieul]

I understand you close. All the conversation is quite complex. Maybe we could just open an issue with the conclusion on the last comments?