Feedback Topics
The current release is a ‘beta version’, and of the 500 lines for which scholia have so far been compiled from many witnesses, only a little over 50 lines have been put into the XML edition format. Before I proceed with formatting the other 450 lines (some of which task should be accomplished by the end of 2010), it is important to know whether modifications should be made in the XML structure or whether some entirely different format or tool should be used instead.
Feedback is welcome on any aspect of this site, but I also offer the following list of aspects or questions.
Please send feedback by email, using in the subject line the phrase ‘scholia feedback’. Use the email address that can be located by using the link on my name in the page footer.
For general users:
When commenting on the following, please mention what OS and what browser you are using and, if you know, which Greek font is being used by your browser.
- navigation: can you find what you want?
- format of the pages of the site
- format of scholia edition: are the different types of information adequately labelled and distinguished by features of typography and paragraph formatting and spacing?
- available views: what other views of the scholia would you be most likely to want available, whether in respect to inclusion or exclusion of sections of the accompanying material or in respect to classifications of content?
For those experienced in textual editing:
- any comments about the goals of the project?
- is the apparatus criticus clear enough in indicating to what words in the text of a scholion a particular item applies?
- available views: what other views of the scholia would you be most likely to want available, whether in respect to inclusion or exclusion of sections of the accompanying material or in respect to classifications of content?
- are there additional refinements in the classification of types and subtypes of scholia that you would like to see?
- are there other desiderata that this edition seems to neglect?
- do you agree with the following policy adopted for reporting witnesses of glosses? When witness A has two words that are alternative glosses (written at the same time by the same hand, so far as can be determined), and witnesses BCD have one of those words as the lone gloss, I am reporting the shared word as a gloss present in ABCD, while I am reporting the alternative gloss in A as a separate gloss in A only. That is, I am not using the other possible policy: reporting as a separate glosses A’s pairing of the two words and the single-word gloss (with A not included among its witnesses).
For those who are technically savvy:
- any comments about the XML structure?
- any comments about the degree of utilization of various features available in TEI P5 (particularly, my decision not to use its method of tagging variant readings)?
- any recommendation of tools or formats that would be superior?
- does my format strike you as creating a ‘data silo’?
- any suggestions about more technologically adept methods of reprocessing the file for the different views? I suspect this could be done with CGI and php, but have not investigated this yet and am loath to rely on server-side processing if that can be avoided. For one thing, alterations in the host server environment can necessitate reprogramming. And in general, I like all programming to be such that I can understand it and upgrade it (and preferably test it on an off-line desktop computer), without dependence on external experts who may be available at one time but not in the future.
- I am aware of the following apparent bugs or confusions and would be grateful if anyone happens to know anything about them.
- The XSLT/XPATH expression that generates an item separator (vertical bar with spaces before and after) between separate items of the apparatus criticus (and separate items of some other sections) is supposed to place this after each item except the last one. Most of the time this works, but you will note that this item separator sometimes occurs after the last item as well (and I have not yet taken the time to remove the unwanted ones manually from the HTML). The code used is
<xsl:when test="not(position()=last())"><xsl:value-of select="$itemSeparator"/></xsl:when>. I can’t detect anything in the XML file that differs at the points where this test fails and the item separator is added to an item that is in fact the last. Perhaps this is a bug in the Saxon processor, but this seems unlikely, since this is a standard piece of code recommended in XSLT books, including the one written by the author of the Saxon processor.
- Most of the pages of this site validate as XHTML 1.1, but the scholia documents are at this time HTML 4.01 (strict). When I experimented with setting xsl:output for xhtml, a whitespace was introduced between the superscript number that indicates the sentence number within scholia with length of more than one sentence. A couple of attempts to use whitespace settings in the XSLT did not change this, and I gave up without spending more time trying to understand the whitespace settings. If anyone can enlighten me, I’ll be grateful.
- I’m also curious as to why Saxon processing creates a lot of xlmns attributes as part of certain tags in the HTML output, attributes that are declared to be illegal when I validate the HTML document in BBEdit 9. Currently, I remove these in seconds with a global search and replace, and then the document validates. Is BBEdit’s validation wrong about this? Similarly, the code for the Creative Commons link obtained from the Creative Commons site contains attributes that BBEdit declares to be illegal, but I have left those in the document.