<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <div class="moz-cite-prefix">Dear Thomas and All,<br>
      <br>
      [I'm a bit afraid that my message may bounce off the teiweblicht
      list, since I'm not a member, but let me try and count on the list
      administrators' "ok" for the attempt]<br>
      <br>
      It's taken me a while to go through this thread together with its
      branches into the Net via the URLs quoted. I'm taken up by the
      format-variant solution, it indeed seems the neatest of all for
      the purpose of processing. I've been dusting off some old TEI
      tickets in the meantime, to see how I could suggest a
      formalization of this on the TEI side, but this is of course a
      slightly different issue and task. A very new one, too, from what
      I can see, because indeed, apart from the quick round of applause
      after the introduction of MIME type, nothing much followed, at
      least in the official channels.<br>
      <br>
      I do not presume the authority to speak on behalf of the Standards
      Committee in this respect. In my own view, let me repeat, the
      solution seems brilliant. As far as the Standards Committee is
      concerned, and with apologies to those who have seen this message
      in the Standards mailing list, I would like to repeat my
      declaration of providing a proposal for a unified set of standards
      documents that could be advocated by CLARIN centres (and again I
      hasten to stress that this is not meant as a revolution, but
      rather as taking stock of the inventory and seeing what got
      obsolete beyond embarrassment and what has appeared on the scene
      in the time since the Short Guides and other proposals for
      standardization within CLARIN were created; Jan Odijk has in the
      meantime approached the issue from a slightly different but
      closely related angle, which gives me a good starting point, and I
      am happy to have received some important backchannel support for
      this initiative as well). Andreas Witt and I will present the
      proposal (we sometimes cautiously speak about it as a sketch) in
      Aix-en-Provence, and circulate it earlier among the Standards
      group and other interested parties. Naturally, we count on the
      support and advice of the Centres Committee in this endeavour.<br>
      <br>
      Also, the process of standardization is not about putting a stamp
      of one committee or another over a proposal, but rather it
      consists in recognition and promotion of existing good practices,
      so I would say, let's put this idea into practice and see if it
      works (I guess we're all pretty optimistic about that), and I will
      be happy to document it as a working practice that can constitute
      the basis for standardization. And while I do that I'll keep the
      TEI part of my brain and life in sync with that -- this is a
      pretty fortunate moment to speak of this, given the approaching
      TEI meeting, because it gives an opportunity to seed the TEI
      Technical Council's consciousness with these ideas, and hope that
      they ripen enough by the next release cycle to get reflected in
      the TEI documents as well.<br>
      <br>
      It seems like some exciting weeks may lie ahead. In the meantime,
      I wish everyone a good weekend (and some of us a good show on
      Sunday ;-)).<br>
      <br>
      Best regards,<br>
       <br>
       Piotr Banski<br>
      <br>
      <br>
      On 08/07/16 09:04, Thomas Schmidt wrote:<br>
    </div>
    <blockquote
cite="mid:CAD74COkhyq3At+++_7b5mbv4dw0AnopszfWc-CcB0AApE3M+Jg@mail.gmail.com"
      type="cite">
      <div dir="ltr">Dear all,
        <div><br>
        </div>
        <div>in the absence of further input from the standards
          committee and before we lose the momentum, I'd like to
          summarise our action plan according to the discussion so far:</div>
        <div><br>
        </div>
        <div>(1a) In WebLicht (in CLARIN in general?) ISO/TEI
          transcriptions of spoken language will be identified by the
          MIME type <span style="font-size:12.8px">text/tei+xml;</span><span
            style="font-size:12.8px">format-variant=tei-iso-spoken. A
            parameter "token=0/1" can be added to indicate whether (=1)
            or not (=0) the respective TEI file is tokenized (i.e. has
            <w> markup).</span></div>
        <div>
          <div>(1b) HZSK and myself will <span style="font-size:12.8px">adapt
              the respective web services accordingly</span><br>
          </div>
          <div><br>
          </div>
          <div>(2a) In WebLicht (in CLARIN in general?) DTA/TEI files
            will be identified by the MIME type <span
              style="font-size:12.8px">text/tei+xml;</span><span
              style="font-size:12.8px">format-variant=tei-dta. </span><span
              style="font-size:12.8px">A parameter "token=0/1" can be
              added to indicate whether (=1) or not (=0) the respective
              TEI file is tokenized (i.e. has <w> markup).</span></div>
        </div>
        <div><span style="font-size:12.8px">(2b) Bryan Jurish will adapt
            the respective web services at BBAW accordingly</span><br>
        </div>
        <div><br>
        </div>
        <div>(3a) In WebLicht (in CLARIN in general?), EXMARaLDA Basic
          Transcriptions will be identified by the MIME type <span
            style="font-size:12.8px">text/xml;
            format-variant=exmaralda-exb</span><span
            style="font-size:12.8px"><br>
          </span></div>
        <div>
          <div>(3b) In WebLicht (in CLARIN in general?),
            FOLKER/OrthoNormal transcription files will be identified by
            the MIME type <span style="font-size:12.8px">text/xml;
              format-variant=folker-fln</span><span
              style="font-size:12.8px"><br>
            </span></div>
        </div>
        <div>
          <div>(3c) In WebLicht (in CLARIN in general?), Transcriber
            transcription files will be identified by the MIME type <span
              style="font-size:12.8px">text/xml;
              format-variant=transcriber-trs</span><span
              style="font-size:12.8px"><br>
            </span></div>
        </div>
        <div><span style="font-size:12.8px">(3d) </span>HZSK and myself
          will <span style="font-size:12.8px">adapt the respective web
            services accordingly</span></div>
        <div><span style="font-size:12.8px"><br>
          </span></div>
        <div><span style="font-size:12.8px">(4a) It would have to be
            checked (note the passive, I don't know who could be in
            charge of this) whether competing MIME types for these file
            types are already registered somewhere. I know that WebLicht
            already seems to have two variants of EXMARaLDA
            transcriptions. The mechanims specifying those would
            probably have to be deprecated. Transcriber is also not
            unlikely to have been given some kind of mimetype elsewhere
            in CLARIN.</span></div>
        <div><span style="font-size:12.8px">(4b) Further relevant
            formats will be ELAN/EAF, CLAN/CHA and PRAAT/TextGrid (the
            latter two being text, not XML formats). All three of them
            are also likely to have been registered somewhere already,
            so "someone" (again, I wouldn't know who) should check if
            mime types have been defined for those.</span></div>
        <div><span style="font-size:12.8px"><br>
          </span></div>
        <div><span style="font-size:12.8px">I guess that this is as good
            an answer as we can currently give to address points 1-3 in
            Marie Hinrich's list. @Marie: can you confirm that this is
            suffient for you, also to address point 4 in your list? In
            my understanding, whatever works for WebLicht in this
            respect should also be a suitable basis for a larger context
            (the SwitchBoard in particular?). </span></div>
        <div><span style="font-size:12.8px"><br>
          </span></div>
        <div><span style="font-size:12.8px">In my eyes, it remains
            crucial, however, that such standardization "decisions" are
            centrally documented (including the information </span><span
            style="font-size:12.8px">Tomaž suggested).</span><span
            style="font-size:12.8px"> The CLARIN standards pages as they
            are now (<a moz-do-not-send="true"
              href="https://www.clarin.eu/content/standard-recommendations">https://www.clarin.eu/content/standard-recommendations</a>
            / <a moz-do-not-send="true"
              href="http://clarin.ids-mannheim.de/standards/index.xq">http://clarin.ids-mannheim.de/standards/index.xq</a>
            are the ones I know) are, IMHO, incomplete, inconistent and
            outdated, and they certainly do not provide accurate
            information on the mime types. Any input from the standard
            committee on this question would therefore still be much
            appreciated.</span></div>
        <div><span style="font-size:12.8px"><br>
          </span></div>
        <div><span style="font-size:12.8px">Best,</span></div>
        <div><span style="font-size:12.8px"><br>
          </span></div>
        <div><span style="font-size:12.8px">Thomas</span></div>
        <div><span style="font-size:12.8px"><br>
          </span></div>
        <div><span style="font-size:12.8px"><br>
          </span></div>
        <div><span style="font-size:12.8px"><br>
          </span></div>
      </div>
      <div class="gmail_extra"><br>
        <div class="gmail_quote">On Thu, Jun 23, 2016 at 10:56 AM, Marie
          Hinrichs <span dir="ltr"><<a moz-do-not-send="true"
              href="mailto:marie.hinrichs@uni-tuebingen.de"
              target="_blank">marie.hinrichs@uni-tuebingen.de</a>></span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div style="word-wrap:break-word">Hi All,
              <div><br>
              </div>
              <div>Thanks to all of you for all the work you’ve done so
                far to get TEI processing integrated into WebLicht.</div>
              <div><br>
              </div>
              <div>From WebLicht’s side, there are several places where
                some work/coordination needs to happen:</div>
              <div><br>
              </div>
              <div>1. TCF: agree on the textsource.type attribute and
                make sure that the encoder services set it properly</div>
              <div>2. Agree on type names (i.e. text/tei+xml or
                text/x-tei-dta-xml)</div>
              <div>3. Make sure the CMDI for encoder and decoder
                services reflect outcomes of 1 and 2</div>
              <div>4. Add new mappings to WebLicht for TEI.</div>
              <div><br>
              </div>
              <div>Steps 1-3 are being worked out here on the mailing
                list and whichever solution/conventions you agree on are
                fine with us.</div>
              <div><br>
              </div>
              <div>Step 4 requires some changes to the WebLicht code -
                in particular to the component that we call the
                “profiler”. When a user uploads a file, the profiler
                tries to figure out what it is and if any of the
                WebLicht services can process it. The contentType of the
                uploaded file, in combination with standard libraries
                for file type recognition are used for this. But
                sometimes more digging is necessary, as in the case with
                tcf - which is recognized as xml, but it needs a closer
                look to see if it is tcf.  The profiler will have to be
                updated in a similar way to recognize TEI, and hopefully
                there is even some straightforward way of distinguishing
                between the DTA and the spoken variants. Finally,
                mappings need to be established between the results of
                the profiler and the service input types so that the
                right services are offered to the user for selection.</div>
              <div><br>
              </div>
              <div>Also note that WebLicht chains can be called from the
                command-line or programmatically using WebLicht as a
                Service (WaaS) - see instructions here: <a
                  moz-do-not-send="true"
                  href="https://weblicht.sfs.uni-tuebingen.de/WaaS/"
                  target="_blank"><a class="moz-txt-link-freetext" href="https://weblicht.sfs.uni-tuebingen.de/WaaS/">https://weblicht.sfs.uni-tuebingen.de/WaaS/</a></a> This
                is useful for larger inputs and avoids timeout issues
                that arise when using the web interface.</div>
              <div><br>
              </div>
              <div>Best Regards,</div>
              <div>Marie</div>
              <div>
                <div class="h5">
                  <div><br>
                  </div>
                  <div><br>
                  </div>
                  <div>
                    <blockquote type="cite">
                      <div>On 21.06.2016, at 14:28, Tomaž Erjavec <<a
                          moz-do-not-send="true"
                          href="mailto:Tomaz.Erjavec@ijs.si"
                          target="_blank"><a class="moz-txt-link-abbreviated" href="mailto:Tomaz.Erjavec@ijs.si">Tomaz.Erjavec@ijs.si</a></a>>
                        wrote:</div>
                      <br>
                      <div>
                        <div bgcolor="#FFFFFF" text="#000000">
                          <p>Hi,</p>
                          <p>as regards <br>
                          </p>
                          <p>> these format-related specifications
                            (in this case: the name and possible<br>
                            > values of attributes which are used in
                            addition to a mime type) would<br>
                            > need to be documented and made known at
                            a central place. <br>
                          </p>
                          I'd say the documentation for each would need
                          to be accompanied by its TEI schema, i.e. the
                          TEI ODD file and the derived (probably)
                          RelaxNG schema. Then it would be a simple
                          matter to check if a document conforms to the
                          mime type.<br>
                          <br>
                          Best,<br>
                          Tomaž<br>
                          <br>
                          <div>Bryan Jurish je 21/06/2016 ob
                            14:22 napisal:<br>
                          </div>
                          <blockquote type="cite">
                            <div dir="ltr">morning all,
                              <div><br>
                              </div>
                              <div>sounds good to me.</div>
                              <div><br>
                              </div>
                              <div>@Marie: can you give an estimation of
                                how well this might work for WebLicht?</div>
                              <div><br>
                              </div>
                              <div>I'll add the "format-variant=tei-dta"
                                parameter to the DTA TEI<->TCF web
                                service in the next few days, so we can
                                see how that at least works out.</div>
                              <div><br>
                              </div>
                              <div>marmosets,</div>
                              <div>  Bryan</div>
                            </div>
                            <div class="gmail_extra"><br>
                              <div class="gmail_quote">On Tue, Jun 21,
                                2016 at 12:32 PM, Thomas Schmidt <span
                                  dir="ltr"><<a
                                    moz-do-not-send="true"
                                    href="mailto:thomas.schmidt@ids-mannheim.de"
                                    target="_blank"><a class="moz-txt-link-abbreviated" href="mailto:thomas.schmidt@ids-mannheim.de">thomas.schmidt@ids-mannheim.de</a></a>></span>
                                wrote:<br>
                                <blockquote class="gmail_quote"
                                  style="margin:0 0 0
                                  .8ex;border-left:1px #ccc
                                  solid;padding-left:1ex">Dear all,<br>
                                  <br>
                                  revising my suggestions from the
                                  teiweblicht list according to Bryan's<br>
                                  proposal to use official mime-types
                                  plus parameters (instead of<br>
                                  x-extended custom mime types) would
                                  mean that:<br>
                                  <br>
                                  "text/x-tei-isospoken+xml" could
                                  become "text/tei+xml;<br>
                                  format-variant=tei-iso-spoken" (+
                                  tokenized=0/1)<br>
                                  "text/x-tei-dta+xml" could become
                                  "text/tei+xml;<br>
                                  format-variant=tei-dta" (+
                                  tokenized=0/1)<br>
                                  "text/x-exmaralda-exb+xml" could
                                  become "text/xml;
                                  format-variant=exmaralda-exb"<br>
                                  ... and so forth (for other TEI oder
                                  XML based formats)<br>
                                  <br>
                                  Wouldn't that be a solomonic solution?
                                  What do the WebLicht developers<br>
                                  say? And independently of that, I
                                  think that Hanna is right that these<br>
                                  format-related specifications (in this
                                  case: the name and possible<br>
                                  values of attributes which are used in
                                  addition to a mime type) would<br>
                                  need to be documented and made known
                                  at a central place. I guess it<br>
                                  would be up to the standards committee
                                  to decide on that?<br>
                                  <br>
                                  Best regards,<br>
                                  <br>
                                  Thomas<br>
                                  <div>
                                    <div><br>
                                      <br>
                                      <br>
                                      <br>
                                      <br>
                                      On Sat, Jun 18, 2016 at 10:56 AM,
                                      Bryan Jurish <<a
                                        moz-do-not-send="true"
                                        href="mailto:jurish@bbaw.de"
                                        target="_blank"><a class="moz-txt-link-abbreviated" href="mailto:jurish@bbaw.de">jurish@bbaw.de</a></a>>

                                      wrote:<br>
                                      > moin all,<br>
                                      ><br>
                                      > fwiw, I agree with Dieter
                                      that we need to differentiate
                                      between "proper"<br>
                                      > MIME types (i.e. standardized
                                      conventions registered with IANA)
                                      and<br>
                                      > CLARIN-internal (rsp.
                                      WebLicht-internal) conventions. 
                                      We have been using<br>
                                      > MIME types as the basis of
                                      the WebLicht textSource/@type
                                      attribute,<br>
                                      > analogous to the HTTP
                                      "ContentType" header, cf.<br>
                                      > <a moz-do-not-send="true"
                                        href="https://tools.ietf.org/html/rfc2045#section-5.1"
                                        rel="noreferrer" target="_blank">https://tools.ietf.org/html/rfc2045#section-5.1</a>
                                      .  At the risk of repeating<br>
                                      > what I've already said on the
                                      tei-weblicht list, use of the
                                      ContentType<br>
                                      > syntax allows us to have our
                                      cake and eat it too: we can go
                                      ahead and use<br>
                                      > "official" IANA-sanctioned
                                      "true" MIME types and specify
                                      variants<br>
                                      > ("dialects", "flavors") using
                                      parameters.  The DTA
                                      TEI<->TCF converter is<br>
                                      > already doing this, setting
                                      textSource/@type to either
                                      "text/tei+xml;<br>
                                      > tokenized=0" or
                                      "text/tei+xml; tokenized=1",
                                      depending on the relevant<br>
                                      > properties of the input
                                      document.<br>
                                      ><br>
                                      > just my €0.02.<br>
                                      ><br>
                                      > marmosets,<br>
                                      >   Bryan<br>
                                      ><br>
                                      ><br>
                                      > On Fri, Jun 17, 2016 at 1:43
                                      PM, Dieter Van Uytvanck <<a
                                        moz-do-not-send="true"
                                        href="mailto:dieter@clarin.eu"
                                        target="_blank"><a class="moz-txt-link-abbreviated" href="mailto:dieter@clarin.eu">dieter@clarin.eu</a></a>><br>
                                      > wrote:<br>
                                      >><br>
                                      >> On 17/06/16 12:59, Sander
                                      Maijers wrote:<br>
                                      >> > After all, you would
                                      want a<br>
                                      >> > resource's metadata
                                      to be completely descriptive of
                                      such elementary<br>
                                      >> > aspects as internal
                                      structure and content of the TEI
                                      files, and not<br>
                                      >> > dependent on system
                                      configuration (served as custom
                                      media type x or y,<br>
                                      >> > as long as the
                                      server remains so configured).<br>
                                      >><br>
                                      >> Hi Sander,<br>
                                      >><br>
                                      >> Thank you for sharing
                                      your opinion.<br>
                                      >><br>
                                      >> One side note: we are
                                      talking about detecting the
                                      mimetype as indicated<br>
                                      >> in the CMDI ResourceProxy
                                      attribute, see:<br>
                                      >><br>
                                      >><br>
                                      >> <a
                                        moz-do-not-send="true"
href="https://www.clarin.eu/faq/how-can-i-specify-additional-details-about-resourceproxy"
                                        rel="noreferrer" target="_blank"><a class="moz-txt-link-freetext" href="https://www.clarin.eu/faq/how-can-i-specify-additional-details-about-resourceproxy">https://www.clarin.eu/faq/how-can-i-specify-additional-details-about-resourceproxy</a></a><br>
                                      >><br>
                                      >> So for the scenario VLO
                                      -> LR switchboard ->
                                      processing application<br>
                                      >><br>
                                      >> the system configuration
                                      would not be relevant, since the
                                      mimetype is<br>
                                      >> explicitly mentioned in
                                      the metadata. The key is to find
                                      agreement about<br>
                                      >> a simple and light-weight
                                      way of designating the variants of
                                      TEI.<br>
                                      >><br>
                                      >> best,<br>
                                      >><br>
                                      >> --<br>
                                      >> Dieter Van Uytvanck<br>
                                      >> Technical Director CLARIN
                                      ERIC<br>
                                      >> <a
                                        moz-do-not-send="true"
                                        href="http://www.clarin.eu/"
                                        rel="noreferrer" target="_blank"><a class="moz-txt-link-abbreviated" href="http://www.clarin.eu">www.clarin.eu</a></a>
                                      | tel. <a moz-do-not-send="true"
href="tel:%2B31-%280%29850091363" value="+31850091363" target="_blank">+31-(0)850091363</a>
                                      | skype: dietervu.mpi<br>
                                      >>
                                      _______________________________________________<br>
                                      >> Teiweblicht mailing list<br>
                                      >> <a
                                        moz-do-not-send="true"
                                        href="mailto:Teiweblicht@lists.informatik.uni-leipzig.de"
                                        target="_blank"><a class="moz-txt-link-abbreviated" href="mailto:Teiweblicht@lists.informatik.uni-leipzig.de">Teiweblicht@lists.informatik.uni-leipzig.de</a></a><br>
                                      >> <a
                                        moz-do-not-send="true"
href="http://lists.informatik.uni-leipzig.de/mailman/listinfo/teiweblicht"
                                        rel="noreferrer" target="_blank"><a class="moz-txt-link-freetext" href="http://lists.informatik.uni-leipzig.de/mailman/listinfo/teiweblicht">http://lists.informatik.uni-leipzig.de/mailman/listinfo/teiweblicht</a></a><br>
                                      >><br>
                                      ><br>
                                      ><br>
                                      ><br>
                                      > --<br>
                                      >
                                      ***************************************************<br>
                                      > Bryan Jurish<br>
                                      > Deutsches Textarchiv<br>
                                      > Digitales Wörterbuch der
                                      deutschen Sprache<br>
                                      > Berlin-Brandenburgische
                                      Akademie der Wissenschaften<br>
                                      ><br>
                                      > Jägerstr. 22/23<br>
                                      > 10117 Berlin<br>
                                      ><br>
                                      > Tel.:     <a
                                        moz-do-not-send="true"
                                        href="tel:%2B49%20%280%2930%2020370%20539"
                                        value="+493020370539"
                                        target="_blank">+49 (0)30 20370
                                        539</a><br>
                                      > E-Mail:   <a
                                        moz-do-not-send="true"
                                        href="mailto:jurish@bbaw.de"
                                        target="_blank"><a class="moz-txt-link-abbreviated" href="mailto:jurish@bbaw.de">jurish@bbaw.de</a></a><br>
                                      >
                                      ***************************************************<br>
                                      ><br>
                                      >
                                      _______________________________________________<br>
                                      > Teiweblicht mailing list<br>
                                      > <a moz-do-not-send="true"
                                        href="mailto:Teiweblicht@lists.informatik.uni-leipzig.de"
                                        target="_blank">Teiweblicht@lists.informatik.uni-leipzig.de</a><br>
                                      > <a moz-do-not-send="true"
href="http://lists.informatik.uni-leipzig.de/mailman/listinfo/teiweblicht"
                                        rel="noreferrer" target="_blank">http://lists.informatik.uni-leipzig.de/mailman/listinfo/teiweblicht</a><br>
                                      ><br>
                                      <br>
                                      <br>
                                      <br>
                                      --<br>
                                    </div>
                                  </div>
                                  <div>
                                    <div>Thomas Schmidt<br>
                                      IDS Mannheim<br>
                                      R5, 6-13<br>
                                      D-68161 Mannheim<br>
                                      Tel.: <a moz-do-not-send="true"
                                        href="tel:%2B49%20%28621%29%201581-313"
                                        value="+496211581313"
                                        target="_blank">+49 (621)
                                        1581-313</a><br>
                                      <a moz-do-not-send="true"
                                        href="http://agd.ids-mannheim.de/index.shtml"
                                        rel="noreferrer" target="_blank">http://agd.ids-mannheim.de/index.shtml</a><br>
                                      <a moz-do-not-send="true"
                                        href="http://www.exmaralda.org/"
                                        rel="noreferrer" target="_blank">http://www.exmaralda.org</a><br>
                                      <br>
                                    </div>
                                  </div>
                                </blockquote>
                              </div>
                              <br>
                              <br clear="all">
                              <div><br>
                              </div>
                              -- <br>
                              <div data-smartmail="gmail_signature">
                                <div dir="ltr">***************************************************<br>
                                  Bryan Jurish<br>
                                  Deutsches Textarchiv
                                  <div>Digitales Wörterbuch der
                                    deutschen Sprache
                                    <div>
                                      <div>Berlin-Brandenburgische
                                        Akademie der Wissenschaften<br>
                                        <br>
                                        Jägerstr. 22/23<br>
                                        10117 Berlin<br>
                                        <br>
                                        Tel.:     <a
                                          moz-do-not-send="true"
                                          href="tel:%2B49%20%280%2930%2020370%20539"
                                          value="+493020370539"
                                          target="_blank">+49 (0)30
                                          20370 539</a><br>
                                        E-Mail:   <a
                                          moz-do-not-send="true"
                                          href="mailto:jurish@bbaw.de"
                                          target="_blank"><a class="moz-txt-link-abbreviated" href="mailto:jurish@bbaw.de">jurish@bbaw.de</a></a><br>
***************************************************</div>
                                    </div>
                                  </div>
                                </div>
                              </div>
                            </div>
                          </blockquote>
                          <br>
                        </div>
                      </div>
                    </blockquote>
                  </div>
                  <br>
                </div>
              </div>
            </div>
          </blockquote>
        </div>
        <br>
        <br clear="all">
        <div><br>
        </div>
        -- <br>
        <div class="gmail_signature" data-smartmail="gmail_signature">
          <div dir="ltr">
            <div>Thomas Schmidt<br>
              IDS Mannheim<br>
              R5, 6-13<br>
              D-68161 Mannheim<br>
              Tel.: +49 (621) 1581-313<br>
              <a moz-do-not-send="true"
                href="http://agd.ids-mannheim.de/index.shtml"
                target="_blank">http://agd.ids-mannheim.de/index.shtml</a><br>
              <a moz-do-not-send="true" href="http://www.exmaralda.org"
                target="_blank">http://www.exmaralda.org</a></div>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
    <br>
    <pre class="moz-signature" cols="72">-- 
Piotr Bański, Ph.D.
Senior Researcher,
Institut für Deutsche Sprache,
R5 6-13
68-161 Mannheim, Germany</pre>
  </body>
</html>