Inštitut za slovensko literaturo in literarne vede ZRC SAZU in Odsek za tehnologije znanja IJS

e-Slomšek: arhiv razvoja



e-Slomšek: Arhiv debate o pretvorbi
Tomaž Erjavec

Vsebina

1. Arhiv

Ta stran vsebuje arhiv dokumentacije v zvezi s projektom: pisma, dele priročnika v nastajanju, ...

Za referenco je tule povezava na lokalno kopijo proporočil TEI P4 in povezava na mapo z gradivom projekta.

1.1. Hladnik, junij 2003: skeniranje rokopisov

----- Original Message -----
From: "Humanist Discussion Group (by way of Willard McCarty
<willard.mccarty@kcl.ac.uk>)" <willard@lists.village.virginia.edu>
To: <humanist@Princeton.EDU>
Sent: Thursday, June 19, 2003 11:19 AM
Subject: 17.102 image-enhancement


>                Humanist Discussion Group, Vol. 17, No. 102.
>        Centre for Computing in the Humanities, King's College London
>                    www.kcl.ac.uk/humanities/cch/humanist/
>                      Submit to: humanist@princeton.edu
>
>    [1]   From:    Melissa Terras <melslists@yahoo.com>               (155)
>          Subject: Re: 17.094 image-enhancement
>
>    [2]   From:    "Arianna Ciula" <ciula@media.unisi.it>              (20)
>          Subject: Re: 17.088 an image-enhancement manual?
>
>    [3]   From:    Suzana Sukovic <S.Sukovic@library.usyd.edu.au>      (41)
>          Subject: Re: 17.088 an image-enhancement manual?
>
>
> --[1]------------------------------------------------------------------
>          Date: Thu, 19 Jun 2003 10:05:18 +0100
>          From: Melissa Terras <melslists@yahoo.com>
>          Subject: Re: 17.094 image-enhancement
>
> Dear Willard et al
>
> At the risk of blowing my own trumpet (or publishing my own manuscript?) I
> recently completed my doctorate at the Department of Engineering Science
> and the Centre for the Study of Ancient Documents at Oxford, which looks
> into this problem, namely, how image processing techniques, and artificial
> intelligence techniques can be used to aid historians in reading damaged
> and deteriorated texts.
>
> Terras, Melissa (2002). Image to Interpretation: Towards an Intelligent
> System to Aid Historians in the Reading of the Vindolanda Texts. D.Phil
> Thesis, Engineering Science, Oxford University.
>
> However, to cover the main points:
>
> - Each document is different, and it is very difficult to come up with one
> technique that can be used on all. However, there are some points which
> become clear:
>
> 1. The most helpful technique developed so far uses infra red photography
> to make carbon ink visible. This is well documented in papers such as
>
> Bearman, B. H. and S. Spiro (1996).  "Archaeological Applications of
> Advanced Imaging Techniques." Biblical Archaeologist  59:1.
>
> Advanced techniques, such as those for the Vindolanda texts, can only be
> used prescriptively for documents with very defined and similar physical
> characteristics - see below for papers. There are various projects going
on
> at CSAD involving digital imaging and AI. The website, again, is
> www.csad.ox.ac.uk
>
> 2. There was a really interesting special edition of literary and
> linguistic computing that came out in 1997 which covered various
> techniques. ie Bagnall, R. S. (1997).  "Imaging of Papyri: A Strategic
> View."  Literary and Linguistic Computing 12(3): 153-154.
>
> Bowman, A. K., J. M. Brady, et al. (1997).  "Imaging Incised
> Documents."  Literary and Linguistic Computing 12(3): 169 - 176.
>
> etc
>
> 3. As far as individuals using photoshop... this is entirely anecdotal.
But
> with the experts I have worked with, it has become obvious that certain
> simple filters and techniques are the most helpful. a. reversal of
images -
> inversion of colour. this often brings out points the human eye could not
> see easily when the image was "the right way round". inversion of black to
> white is the most helpful. b. messing about with contrast and brightness.
> again, this can highlight areas which had been previously missed or
> overlooked. c. sharpening filters in photoshop are useful. Although it
> should be obvious that whenever filters are applied to a source image,
that
> image is distorted, and the user has to be careful to always look back to
> the original so that they dont see things which are "not there" in the
> first place. This is always an interesting problem when using imaging
> techniques and visualisation techniques in the humanities. d. reversing
> images - back to front. it sometimes helps the human eye to see things.
> Papyrologists and palaeographers often like to read things backwards,
> funnily enough. e. obviously, zoom is really useful.
>
> It should be pretty simple to write a pared down manual of photoshop to
> cover these points. we tried to teach some of the experts to use layers
but
> generally no can do....
>
> 4. trivia hounds, the first use of computing/imaging and papyrology was
> actually back in the 1960s,with these two independent studies
>
> Levison (1965).  "The Siting of Fragments."  Computer Journal 7: 275 -
277.
>
> Ogden, J. A. (1969). The Siting of Papyrus Fragments: An Experimental
> Application of Digital Computers. Mathematics.  Glasgow, University of
> Glasgow.  Ph.D. Thesis.
>
> 5. Other interesting articles?
>
> Brady, M., X. Pan, et al. (Forthcoming, (2003)). Shadow Stereo, Image
> Filtering and Constraint Propagation.  Images and Artefacts of the Ancient
> World, London, British Academy.
>
> Brown, M. S. and W. B. Seales (2001).  3D Imaging and Processing of
Damaged
> Texts. ACH/ALLC, New York University.
>
> Kiernan, K. S. (1991).  "Digital Image Processing and the Beowulf
> Manuscript."  Literary and Linguistic Computing 6: 20-27.
>
> Lundberg, M. J. (2002).  "New Technologies: Reading Ancient Inscriptions
in
> Virtual Light", West Semitic Research Project, University of Southern
> California. 2002.
http://www.usc.edu/dept/LAS/wsrp/information/article.html
>
> Prescott, A. (1997).  "The Electronic Beowulf and Digital
> Restoration."  Literary and Linguistic Computing 12: 185-95.
>
> Schenk, V. U. B. (2001).  Visual Identification of Fine Surface
> Incisions.  Department of Engineering Science.  Oxford, Oxford
> University.  D.Phil Thesis.
>
> Seales, W. B., J. Griffioen, et al. (2000).  "The Digital Athenuem: New
> Technologies for Restoring and Preserving Old Documents." Computers In
> Libraries 20(2): 26-50.
>
> Stark, J. A. (1992).  Digital Image Processing Techniques With
Applications
> In Restoring Ancient Manuscripts.  Cambridge, Department of Engineering,
> University of Cambridge.  EIST Project Report.
>
> My thesis is in press at the moment. Well, its still being written into a
> book, but should be getting put together soon. when I get round to
> finishing it (and stop looking at the internet....)
>
> Hope this helps!
>
> Melissa
> ---------------
>
> Dr Melissa Terras
> Assistant Manager
> Engineering Policy
> Royal Academy of Engineering
> 29 Great Peter Street
> London
> SWIP 3LW
>
>    --- "Humanist Discussion Group (by way of Willard
> McCarty              <willard.mccarty@kcl.ac.uk>)"
> <willard@LISTS.VILLAGE.VIRGINIA.EDU> wrote: >
>          Humanist Discussion Group, Vol. 17,
>   > No. 94.
>   >        Centre for Computing in the Humanities,
>   > King's College London
>   >
>   > www.kcl.ac.uk/humanities/cch/humanist/
>   >                      Submit to:
>   > humanist@princeton.edu
>   >
>   >
>   >
>   >          Date: Tue, 17 Jun 2003 06:36:26 +0100
>   >          From: orlandi@rmcisadu.let.uniroma1.it
>   >           > manual?
>   >
>   > The best experiments, to my knowledge, have been
>   > done by
>   > Fotoscientifica Re.co.rd, a small very specialized
>   > firm
>   > based in Parma (Emilia, Italy), in collaboration
>   > with teams
>   > of scholars and archive people.
>   >
>   > You may consult their web page:
>   >
>   > http://www.fotoscientificarecord.com/
>   >
>   > which has also an English version.
>   >
>   > On the other hand, I think that what Chiara
>   > Faraggiana di
>   > Sarzana writes, prefacing the publication of one
>   > such
>   > experiment, is very true: [my translation] Our
>   > experiment
>   > has confirmed once more the peculiarity of the
>   > material
>   > of EACH palimpsest folio. It is manifestly
>   > impossible to
>   > propose one technique universally valid, and
>   > applicable
>   > with similar results to every kind of manuscript.
>   > (_Manoscritti Palinsesti Criptensi_, Ravenna-Parma
>   > 1998)
>   >
>   > Tito Orlandi
>   >
>   >
> -----------------------------------------------------------------
>   > Tito Orlandi
>   > orlandi@rmcisadu.let.uniroma1.it
>   > CISADU - Fac. di Lettere                     Tel.
>   > 39+06.4991-3936
>   > P.zale Aldo Moro, 5                          Fax
>   > 39+60.4991-3945
>   > 00185 Roma
>   > http://rmcisadu.let.uniroma1.it/~orlandi
>   >
> -----------------------------------------------------------------
>
> ________________________________________________________________________
> Want to chat instantly with your online friends?  Get the FREE Yahoo!
> Messenger http://uk.messenger.yahoo.com/
>
>
>
>
> --[2]------------------------------------------------------------------
>          Date: Thu, 19 Jun 2003 10:07:45 +0100
>          From: "Arianna Ciula" <ciula@media.unisi.it>
>          Subject: Re: 17.088 an image-enhancement manual?
>
> I am working on digital reproductions of medieval manuscripts for a
research
> project of palaeographical analysis and what I have missed from the
> beginning of my research is such a manual.
>
> Surely everyone could experiment the best solution for his own case, but a
> directional methodology is still needed. What we have for physical
> restoration and conservation of old books, we do not have for treatment of
> digital images. This is valid not only for the general digital image
> processing (the features and opportunities regarding filters and the
> purposes of the digital edition), but even for the process of scanning of
> old books and charts as a whole (for instance: which product or scanner
fits
> better which document and which material conditions?).
>
> The link I might suggest is the web site that tells the story of another
> project I am involved with: a relational database on musical manuscript
>
> http://www.unisi.it/ricerca/centri/cislab/ritmi/ritmi.htm
>
> Actually, in this case the images are not really seen as primary
instruments
> of analysis (as it is the case of  the purposes of my Ph.D. research and
in
> general of the palaeographers' studies), but as terms of comparison and
> verification of the digital edition.
>
>
> Yours,
>
> Arianna Ciula
>
>
>
>
> --[3]------------------------------------------------------------------
>          Date: Thu, 19 Jun 2003 10:08:54 +0100
>          From: Suzana Sukovic <S.Sukovic@library.usyd.edu.au>
>          Subject: Re: 17.088 an image-enhancement manual?
>
> Dear Willard,
>
> Here are a few comments on your questions. I am glad that you started this
> discussion because I feel for quite some time that imaging experts and
> people who work on digitisation projects need to develop some
instructional
> material together.
>
>  >I am involved in a project to produce an electronic edition of a single
>  >medieval manuscript. This edition will of course provide high-definition
>  >images of the ms pages. Some of the pages in the manuscript are in poor
>  >condition, with among other things bleed-through of ink from the other
side
>  >of the leaf. We anticipate that users of the edition will want digitally
to
>  >enhance these pages or parts of them. If "enhancement" were a simple
matter
>  >then of course we would provide enhanced images, but it is not.
Deformative
>  >play of various sorts, yet to be discovered or problematic but fruitful,
>  >needs to be encouraged. It occurs to me that the best approach would be
to
>  >provide in the introductory material to this edition a description of
how
>  >particular filters, say in Photoshop, can be used under particular
>  >circumstances to bring out features of an image. I can imagine, for
>  >example, addressing the problem of a word originally written in silver
ink
>  >or paint for which the metal has mostly fallen off, leaving small bits
>  >behind. What filter, or what filters used in what sequence with what
>  >settings, would be best to show the remaining metallic bits?
>
> This one is very difficult to answer without seeing the images.
>
>  >Does such a image-manipulation manual for manuscript scholars exist? If
>  >not, would there be sufficient interest to motivate the collaborative
>  >production of such a manual?
>
> I think so. Most people develop their own ways of enhancing page images
but
> it would be very useful to have a proper instruction how to use various
> features of popular software packages. It would be great to have a regular
> review of a new Photoshop release that would list new features which could
> be used to enhance page images.
>
> Regards,
> Suzana
>
> _______________________________________________________
> Suzana Sukovic
> Rare Books and Special Collections Library
> University of Sydney Library, NSW 2006
> Australia
>
> tel:   (02) 9351 2992
> fax:  (02) 9351 2890
> e-mail: s.sukovic@library.usyd.edu.au
> http://www.library.usyd.edu.au/libraries/rare/index.html
>

1.2. Matija, julij 2003: problemi v prvi redakciji pretvorbe

problemi I


PROBLEMI pri ozna?evanju ťTreh pridigŤ

Problem tistega drugega motta v II. pridigi sva z Jo?etom re?ila tako,
da sva ga obravnavala kot poseben vrinjen motto, enakovreden prvemu
mottu, ki je samostojno ?tevil?en, zato sva tudi tega drugega
?tevil?ila od 7 do 17. (besedilo pridige se za?ne z vrstico 18 Vvod);
ozna?iti pa bo treba ta moto kot avtorjev dodatek, vendar ne kot
nadpisan nad ostalo besedilo (<add>), ampak kar zraven, poleg
... Slom?ek je ta citat potem, ko je pridigo ?e napisal, prepisal s
konca pridige - (glej vrstice 342-346 po novem ?tevil?enju, v knjigi
330-335) - na za?etek pridige in ga malce variiral. Nov element !

V kriti?nem prepisu upo?tevati/neupo?tevati datum nastanka pridig,
naveden v izvirniku, glej npr. sic 2 (1825)

ťbesedilo uni?enoŤ - II, oba prepisa: 135, 227 (sic 10, 11, 18, 19)

tekst, ki je nadpisan, a samo kot navaden tekst: "to resnico bom
pri?al v 1.<sic>em </sic>deli," - II. 44, 50; v kriti?nem tekstu
kasneje vrst. 96 in 114 (v knjigi o?tev. 85 in 103) pa je to kar
razvezano : katerih [prva] ?ila je la?; v diplomati?nem pa je
nadpisano 1va, 2ga; trenutno vse ozna?eno kot <sic> .

v navadni vrstici je del teksta na levi strani, del pa na desni, vmes
prazno, prim: II. diplomati?ni, vrstici 1 in 2, (kaj storiti v
kriti?nem?)

ťprazneŤ vrstice - razmiki - ?

dodatek prepisovalca, ki ga ni v izvirniku: I. diplomati?ni 197:
tesh[avam] !? glej sic 1 in 3. (ni isto kot dodatek urednika; ?e ne
najdemo bolj?ega elementa, naj bo pa? <supplied>) Nov element !

Toma?: kaj pomeni <hi/> v II. diplomati?ni, v.59?

Jo?e: zapisovanje cifer pri svetopis. navedkih: kak?en je bil usus
dobe?

Mod. 1, 14 ali Mod. 1,14; Ps. 5. 7. ali Ps. 5.7.  ?

Toma?: Kaj pomeni <gap/> v vrstici 226, II. diplomati?ni? Prekopiral
sem ga tudi v 226, II. kriti?ni.

Toma?: III. kriti?ni: vrstica 291, glej <sic n="24"> Nov element !




TE?AVE <sic> n= od 1 do 24

I. diplomati?ni: 

<l n="197">	<note rend="left">Mat.</note>"klizhe Jes'u?: kateri ?te oblosheni s'revam ino tesh<sic n="1">avam</sic></l>


I. kriti?ni: 
<l n="1"><sic n="2">datum?!</sic></l>

<l n="197">kli?e Jezus, "kateri ste oblo?eni z revam ino te?<sic n="3">avam,</sic></l>


II. diplomati?ni: 

<l n="1"><sic n="4">1829 X</sic></l>
<l n="2"><sic n="5">1841</sic></l>

<l n="55">ino z'elo ?labo, ino to re?nizo bom prizhal v 1. <sic n="6">im</sic> deli, s'akaj</l>
<l n="61">vredno zhe?titi; to vam pokas'al bom v 2.<sic n="7">mo</sic> deli.</l>

<l n="96">1<sic n="8">va</sic>shila je <hi>lash</hi>. - V?igamogozhen ?tvarnik in moder, <del>je</del></l>

<l n="114">2<sic n="9">ga</sic>shila jes'ika hudobniga je <hi>ne?rezhno pohuj?hanje.</hi>

<l n="135">	< sic n="10">(besedilo uni&#x10D;eno)</sic>
<l n="227">	<sic n="11">(besedilo uni&#x10D;eno)</sic>


II. kriti?ni: 

<l n="1"><sic n="12">datum?!</sic></l>
<l n="2"><sic n="13">datum!?</sic>Godvanje sv. Janeza Napomucena.</l>

<l n="55">ino zelo slabo ino to resnico bom pri?al v 1.<sic n="14">em </sic>deli, zakaj</l>

<l n="61">vredno ?estiti; to vam pokazal bom v 2.<sic n="15">mo</sic> deli.</l>

<l n="96"><sic n="16">?</sic><supplied>prva</supplied> ?ila je la?. Vsegamogo?en Stvarnik in moder </l>

<l n="114"><sic n="17">?</sic><supplied>druga</supplied> ?ila jezika hudobnega je nesre?no pohuj?anje.</l>

<l n="135">	<sic n="18">(besedilo uni?eno)</sic></l> 
<l n="227">	<sic n="19">(besedilo uni?eno)</sic></l>




III. kriti?ni: 

<l n="1">Dol?nost svoj jezik spo?tovati<sic n="20">*</sic></l>
<l n="27"><sic n="21">* Pridiga, povedana v Mo?burzi na Koro?kem binko?tni pone-</sic></l>

<l n="192">veliko premo?enje zgubijo, ki se ve? povrniti ne da.<sic n="22">*)</sic> Koli-</l>
<l n="210">	<sic n="23">*Tako nerodno delajo Slovenci, kteri za nem?ko mejo</sic></l>
<l n="291"><sic n="24">u?enem</sic>kraljestvi zdru?il. -</l> [Toma?: tu je treba najti nov element za popravek: namesto 'u?enem' mora priti 've?nem'; to ni isto kakor <supplied> ! 

1.3. TEI-l, avgust 2003: TEI označevanje poporavkov

Tomaz Erjavec -> TEI-L@LISTSERV.BROWN.EDU

>Dear manuscript & text critical gurus,
>
>we are encoding a book containing 19th cent. sermons in facsimile,
>as well as diplomatic and critical transcriptions. Photographs of one
>particularly nasty page in the three versions are at
>http://nl.ijs.si/et/tmp/slomsek/
>
>The diplomatic transcription has deletions and additions, which
>reflect the author's changes in the manuscript.
>
>The critical transcription takes these corrections already into
>account (i.e. they are gone), but introduces deletions and additions
>by the editor/transcriber, correcting the author's mistakes.
>
>Right now I've used <add> and <del> for this, and distinguish the
>author/transcriber by the hand attribute. I can see no problems with
>this in the diplomatic, but in the critical I am much less certain.
>
>First, the 'deletions' are in the book indicated by [...], i.e. the
>original text is not there any more - so, the critical <del>s are
>always empty.
>
>One is almost tempted to use <gap>, but we already use that for
>missing and illegible parts of the manuscript, also, there is a nice
>symetry between add and del.  I am also somewhat unsure if 'hand' is
>in fact meant to be used for editorial corrections. Still, for a while
>there it seemed I could get away with:
>
>Diplomatic:
><add hand="AUT">
><del hand="AUT">
>
>Critical:
><add hand="TRA">
><del hand="TRA">
>
>But, as happens, there now turns out to be exactly one place in the
>critical tr., where a mistake has been *corrected* (and for good
>reasons).  Of course I thought I'd just use <corr> and be done with it
>- but <corr> doesn't have the hand attribute, and uses resp for the
>agent of the correction.
>
>It seems ugly to have <corr resp="TRA"> because the semantics of this
>"resp" attribute is identical to "hand" in <add> and <del>. On the
>other hand (khm), it also seems somewhat strange to use "resp" in
><add> and <del> - we are in fact talking of the person who made the
>changes, and not of the one who identified them.
>
>So, what now? I can either modify the TEI to add "hand" to <corr>, or,
>it occured to, maybe because of the heat, is to use <add> and <del>
>only in the diplomatic tr., to mark authors changes, but in the
>critical use only <corr>. So:
>
>add:  <corr resp="TRA" sic="">foo</corr>
>del:  <corr resp="TRA"></corr>
>corr: <corr resp="TRA" sic="fu">foo</corr>
>
>Anyway, any advice on which path to follow are most welcome!
>
>Best,
>Tomaz
>  
>

From: "Daniel O'Donnell" <daniel.odonnell@uleth.ca>
To: Tomaz Erjavec <tomaz.erjavec@IJS.SI>, TEI-L@LISTSERV.BROWN.EDU
Subject: Re: add/del/corr conundrum
Date: Thu, 14 Aug 2003 09:19:43 -0600

A while ago we discussed another way of understanding CORR, ADD, and 
DEL. That is to see CORR as the name for a process and ADD and DEL as 
the name for states. In my project, we consider all corrections to be 
processes of addition and/or deletion.  Hence an addition would be tagged:
<corr>
    <add>foo</add>
</corr>

a deletion is tagged:
<corr>
    <del>fi</del>
</corr>

and the replacement of one form for another:
<corr>
    <del>fi</del>
    <add>foo</add>
</corr>.

This would partially solve your problem, because it frees up a HAND 
attribute for every correction if necessary (it is on the child ADD 
and/or DEL tag) and gives you the RESP tag on CORR to imply whether the 
change is a correction or emendation (i.e. using AUT or TRA as you 
suggest). (It is not clear to me from your discussion how the "necessary 
correction" in the critical transcription differs from the things you 
are using ADD and DEL for, but I think this method clears up enough 
attributes for you anyway).  It also has the advantage of avoiding 
having to put potentially tagged information on the SIC attribute in the 
CORR element (or more correctly, strip this information away, since it 
is not supposed to be used), since the "sic" reading (when present in 
the original) is always available in the child DEL tag.

The main problem with this solution, however, is that it fails to 
indicate explicitly whether your "correction" is a "correction" (i.e. a 
change made in the contents of the text's primary witness at the time of 
its production by people responsible for its production) or an 
"emendation" (i.e. a change made in the contents of an idealised text 
based on a witness by subsequent editors who are working metatextually). 
This information is implied by the value you assign to the RESP 
attribute, but only if users (or your rendering software or user agent) 
recognises the significance of each value. This could be complicated if 
there are many different people responsible for corrections and 
emendations. Since there is no TYPE available for CORR, you can't 
classify the correction generically as editorial or 
authorial/compositor. I've never been able to figure out why this isn't 
possible in the TEI since the distinction between correction and 
emendation is commonly observed in type: i.e. we use square brackets for 
emendations and (usually) other devices such as italics, or brackets of 
various shapes for corrections.

In our case, we gave up and made an EMEND element identical to CORR in 
every way except application: we use it for changes to the ideal text 
and reserve CORR for alterations to the contents of original witnesses 
by producers.

Hope this helps; let me know if I've missed the point.
-dan

-- 
Daniel Paul O'Donnell, PhD
Associate Professor of English
University of Lethbridge
Lethbridge AB T1K 3M4
Tel. (403) 329-2377
Fax. (403) 382-7191
E-mail <daniel.odonnell@uleth.ca>
Home Page <http://people.uleth.ca/~daniel.odonnell/>



 
From: "Daniel O'Donnell" <daniel.odonnell@uleth.ca>
To: Tomaz Erjavec <tomaz.erjavec@ijs.si>
Cc: TEI-L@LISTSERV.BROWN.EDU
Subject: Re: add/del/corr conundrum
Date: Sat, 16 Aug 2003 08:44:44 -0600

> Tomaz Erjavec wrote:
>
>The only part of corr/del|add I don't like is the complication in the
>processing: when you transform to, say, html you have to always check
>if add/del have a sibling del/add, as of course the rendering is
>different if it is a true blue addition or a correction.. 
>

That's an interesting problem.  I guess in my case, I didn't really 
care: I treated addition and/or deletion per se as independent states 
and used the corr/emend tags to describe the process of substitution 
when it occurred; I also didn't handle the display or HTML tagging of 
additions or deletions differently when they occurred in pairs. Every 
deletion has a value for the rend attribute, and every addition uses 
either place or place and rend.  Some of these imply a corresponding 
deletion or addition (e.g. rend="overwriting"), and all show up 
symbolically in the HTML in my diplomatic transcriptions; but I never 
used a test for a siblings.  

I can think of four possible solutions for this.

1) Use the rend attribute on add/del or even the parent correction tag 
to indicate "substitution" (I'd avoid using the word "correction" for 
this: it really is a substitution, since it can also happen to the ideal 
text during emendation, and since a deletion or addition by itself can 
also commonly <del>be</del> be understood as a correction).  I suppose 
if you thought hard enough you could think of a way of really describing 
this in terms of rendition so as not to commit tag (or attribute) abuse.
2) Use the rend tag on del and/or the place tag on add to indicate how 
the deletion was done and where the addition was found (again trying to 
avoid attribute abuse)
3) Use the id and corresp. attributes on add and del (or both) to link 
the two explicitly together, and, implicitly, indicate a substitution is 
occurring.
4) Test for children of corr/emend instead of siblings and handle your 
HTML at that level.

I've never had any problem checking siblings, BTW.  It works fine in 
CSS2, as far as I remember, and I depend on it for my apparatus files 
when I convert to HTML using Multidoc (I still don't use XSL, but can't 
imagine it is very difficult there, given how easy it is in the other 
two methods).

-- 
Daniel Paul O'Donnell, PhD
Associate Professor of English
University of Lethbridge
Lethbridge AB T1K 3M4
Tel. (403) 329-2377
Fax. (403) 382-7191
E-mail <daniel.odonnell@uleth.ca>
Home Page <http://people.uleth.ca/~daniel.odonnell/>

1.4. Tomaz, september 2003: komentirano poglavje druge redakcije

V drugi ročni redakciji se je pod drobnogled vzelo primere za katere ni bilo jasnega označevanja. Proces razjasnjevanja teh napak in izjem je s seboj prinesel tudi dosti revizij uporabe že sprejetih elementov, tako, da je šele v tej fazi nastal dokončen opis uporabe posameznih oznak.

Q & A:

  1. V kriti?nem prepisu upo?tevati/neupo?tevati datum nastanka pridig, naveden v izvirniku, glej npr. sic 2 (1825) Matija: Za zdaj naj bo tisti "datum" v obeh prepisih obeh pridig oznacen, kakor pravis ti - <date>. Razresiti bom moral uganko o dejanskem nastanku pridig. Takrat bom vstavil tja zraven stvarno opombo s pojasnilom, kaj tisti datum pomenijo. To velja pravzaprav za vse tri pridige. Tomaz, premisli prosim, kako naj bi taksna opomba tehnicno izgledala. To spada ze k 'znanstvenemu aparatu', ki ga bomo postopoma dodajali, na zacetku samo par nujnih opomb. ET: se strinjam, da je <date> v kriticnih prepisih - opombo o tem, kaj ta "datum" pomeni bi jaz dal kar v glavo TEI, pri opisu nase uporabe elementa <date> - takole:
       <tagUsage gi="date" occurs="3">Date is used in the critical
       editions to describe the putative date(s) that appear at the head
       of sermons I and II. Sermon I. contains one such <q>date</q>,
       i.e. <q>1825. XIII</q>, while sermon II. has two years, one written
       above the other, i.e. <q>1829/1841</q>.</tagUsage>
    
  2. dodatek prepisovalca, ki ga ni v izvirniku: I. diplomati?ni 197: tesh[avam] !? glej sic 1 in 3. (ni isto kot dodatek urednika; ?e ne najdemo bolj?ega elementa, naj bo pa? <supplied>) Nov element ! ET: pri vseh ostalih velja da se hand="JFA" pojavi samo v kritičnem prepisu - tu se pač enkrat v diplomatskem? Matija: dodatek prepisovalca: hand=JFA se sme pojaviti VEDNO samo v KRITIC. prepisu - DIPLOM. naj ostane samo reprodukcija originala. Jaz sem te spravil pri tem na napacno sled. ET: ampak je res dodan [avam] v diplomaticnem! samo da ni oznacen. jaz sem sedaj ta sic spremenil v:
    ino tesh<add hand="JFA">avam</add>
    (sic!)
  3. Toma?: III. kriti?ni: vrstica 291, glej <sic n="24"> Nov element ! ET: ta, edini primer je povzrocil celo filozofski problem (glej debato na to temo) Sedaj je zapisano takole:
    <l id="sl3k.291" n="291">učen<corr><del hand="JFA">i</del><add hand="JFA">e</add></corr>m kraljestvi združil. –</l> 
    
  4. ‘prazne’ vrstice; pridige imajo dosti notranje strukture, ki je bila v konverziji povsem zanemarjena. Postavi se vprašanje, ali vztrajati na strukturi, ki temelji na vrsticah, <l> ali pa preiti na vrstice kot prelome <lb/> in vpeljati v prepis, bodisi samo kritični, ali pa tudi v diplomatičnega, retorično strukturo, t.j. <div>, <head>, <p>, <list> in <item> Način ispisa bi se (do neke mere) lahko uravnavalo z atributom "rend", npr. <head rend="centered +1">, ali, verjetno bolje <head rend="centered"><head rend="+1">...</head></head>. Zanimiva možnost je, da v diplomoatičnih prepisih ohranimo strukturo vrstic, in tudi razmike obravnavamo kot (prazne) vrstice. Zato pa kritičnem prepisu uvedemo polno retorično strukturo; povezave ohranjamo preko ID/IDREF na <head> in <lb/>. Matija: 10. Odlocila sva se tako: v diplomaticnem naj ne bo 'praznih vrstic', ker razmiki v originalu niso dovolj prepoznavni in dosledni; v kriticnem pa bomo vstavljali prazne vrstice, da bo razvidna retoricna struktura. Pri tem naj se ostevilcenje ne spreminja - ce je to mogoce. Drugo, kar tam omenjas, mi bos moral malce komentirati ... ET: torej: v diplomatičnem ni praznih vrstic, razmike na označujemo. V diplomatičnem uvedemo (počasi) retirčno strukturo in razmike 'naračunamo' (t.j. v XSL) iz nje.
  5. Problem tistega drugega motta v II. pridigi sva z Jo?etom re?ila tako, da sva ga obravnavala kot poseben vrinjen motto, enakovreden prvemu mottu, ki je samostojno ?tevil?en, zato sva tudi tega drugega ?tevil?ila od 7 do 17. (besedilo pridige se za?ne z vrstico 18 Vvod); ozna?iti pa bo treba ta moto kot avtorjev dodatek, vendar ne kot nadpisan nad ostalo besedilo (<add>), ampak kar zraven, poleg ... Slom?ek je ta citat potem, ko je pridigo ?e napisal, prepisal s konca pridige - (glej vrstice 342-346 po novem ?tevil?enju, v knjigi 330-335) - na za?etek pridige in ga malce variiral. Nov element ! Matija: 11. kako bo po novem oznacen ta dodatek? Joze misli, da ga ja treba postaviti na desno, ne na levo. Glede tega citata bom dodal spet stvarno opombo z razlago, kakor v 4.

1.5. Tomaz, oktober 2003: zapis problematičnih zakov v Unicode

V XML so znaki zapisani po standardu Unicode, za evropske pismenke običajno v kodiranju UTF-8. V delovnih verzijah gradiva uporabljamo kar UTF-8 kodiranje neposredno, v končni obliki pa XML znakovne entitete (npr. &#x017E; za ž). Z uporabo teh entitet dosežemo 7bitni zapis gradiva, ki je še vedno boljši za izmenjavo.

Poleg običajnih znakov za slovenščino pa najdemo v gradivu (v diplomatičnih prepisih) še naslednje problematične znake:

Primer tako zapisanega besedila:

‘ Poglejte ſtare duhovnike, kolko ſe trudijo, kolko puſtijo
ſvoje s’rozhene predrage ovzhizhe s’velizhat, predrage ovzhi〟
ze, katere vezhnim pogublenju odtet je ſam boshji ʂin
nebeſe s’apuſtil, priſhel na ſvet, vſe reve, nadloge ter〟
pet, na krishu vmret, in tok vſe is’velizhat. – Ne ſmeje
... ’

1.6. Michael Beddow, Oct 2003: epcEdit

epcEdit:

From: Michael Beddow <mbteil@mbeddow.net>
Sender: "TEI (Text Encoding Initiative) public discussion list" <TEI-L@LISTSERV.BROWN.EDU>
To: TEI-L@LISTSERV.BROWN.EDU
Subject:      Re: TEI authoring with epcEdit editor
Date:         Tue, 14 Oct 2003 00:58:51 +0100

> I am planning to use epcEdit editor to write my TEI documents and I'm
> interested if someone can share his/her experiences and have some ready
> stylesheet suited for TEI DTD?
>

We use it a lot (Win32 and Linux) and there isn't really much to share. It
just works (all the P4 files come bundled with it and it knows what to do
with them). It does like a lot of memory, though. I'd advise 1G for Win2K/XP
(don't even try with Win 9x or ME) and 512K for Linux. It's stable enough if
memory grows tight, but can then get frustratingly slow (most XML editors
that aspire to user-friendliness have this drawback).

The authors are responsive and take requests for features seriously, though
the rather long release cycle means it can be some time before a new or
modified feature gets implemented (bugfixes however, where necessary, come
through quickly and there's a self-patching utility for licensed users that
runs across the Net). It's fully scriptable if you are reasonably
comfortable with writing tcl. In terms of core editing features,
TEI-compatibility (meaning really just proper XML conformance) pricing and
support, its pretty similar toXygen (though the latter has a few more bells
and whistles: epcEdit, for example doesn't attempt to integrate XSLT
transformations, but I'd rather use Komodo for those anyway because of its
debugging facilities). I personally favour epcEdit over oXygen because it
doesn't need Java, and I prefer to avoid using (and worse still, having to
support others using) Java apps on Win32 platforms.

We don't have any epcEdit style sheets worth passing on. We tend to use the
styling features just to mark out things we are especially interested in
during a particular editing session (e.g. to display <pb/> milestones in
bright green if we want to check or adjust their location.) though it has
the potential to achieve a fair degree of WYSIWYG if you use the epcEdit
styles to mirror the presentational aspects of external XSLT sheets of your
own which you intend to apply to your documents. You have to author and
apply those sheets outside of epcEdit, though.

Michael Beddow

1.7. Matija Ogrin, oktober 2003: zadnji detalji



Datum: Avtor: Tomaž Erjavec .
Copyright ZRC/IJS 2004