7. Changes to the skaldic database
This is not currently part of the peer-reviewed material of the project. Do not cite as a research publication.
Some of you may already have noticed some changes that have been made to the skaldic database, particularly with the web entry for verses. There are various technical reasons for the change which will be outlined below, but the main practical implications are as follows:
a. The links which take you to the forms for editing data (e.g., relating to skalds, poems, stanzas and so on) appear automatically as icons at the top and bottom of each page when you're logged in. The pencil icon is the standard link to forms for editing the data. Where more complex forms exist (e.g. for verses), a pencil-on-page icon will take you directly to the web forms; they are also accessible via the 'more editing forms' tab if you are looking at the standard forms.
b. There are now two sets of forms for some data sets (notably verses) in rows of 'tabs': a top row and a second row which appears when you click on 'more editing forms'. In general, use the lower set where available. Some forms, such as for entering and editing notes and variants, only occur in the top set. (The top set of forms are created by a simpler method and their content and behaviour is edited through the web interface; the lower set, however, is usually more user-friendly and can update data in more complex ways.)
c. If you want to add an edition of a stanza to the database, you will have to clear the old Skj stanza first. The word index and edition are now essentially the same. In order to make sure all the verses are indexed, the Skj text of stanzas which have not yet been entered have been automatically added. There's a simple function for clearing the old text, and a link on the first form you will encounter.
d. The forms for adding the prose order, translation and kennings; and for reordering, splitting and deleting words function more or less the same as before. The form for editing the base text, however, is quite different: a list of words with separate boxes for text and punctuation. It may take a little time to get used to the new behaviour.
e. Most problems have been worked out with the new interface, but others may arise as the database develops. One ongoing problem is that inverted kennings are sometimes not added correctly by the forms for adding the translation and kennings. You can get around this by using the kenning form in the top set of tabs; but in general, if you have problems, let Tarrin know.
f. All the data from the old format has been converted to the new format, with the old format still stored in the database. There have been some problems with the conversion process, but most of these have been fixed. If you entered material in the database before July, you might want to check that it converted properly.
g. On the upside, the whole interface for entering the edition works much faster, both in updating and viewing (you may have noticed this already). Some things should be easier and simpler to use, but others may be a bit different. The ‘split words’ function has been extended to make it more useful, and in general the new format is less likely to produce odd or unpredictable behaviour when it comes to display and printing.
The reasons for the change are largely technical. The General Editors have been discussing future plans for a new Lexicon Poeticum and index of kennings. The production of such material concurrently with the edition will be considerably aided by integrating the word index and kenning index into the actual edition: they had previously been separate from the actual code of the edition, updated when the edited text was entered or updated. No major changes to the database could be made, however, which would risk the progress of Vol. VII, so the changeover had to wait until Vol. VII was exported from the database in June. Since then, following feedback from those using the database, almost all of the initial teething problems have been resolved.
The changes have addressed a couple of other issues with both the data structure and interface, namely, that the data was stored in two different formats (SQL and XML) and the interface was written in two different programming languages (PHP and Perl) for different parts. The new format removes the old XML format and the Perl interface for editing it; this has made development easier. The interface is also much faster using PHP. More importantly for the longer term, interfaces can now be developed for the reference material on lexis and diction. Some of these have already been developed, e.g. the 'lemmata' form in the verse edition interface, which allows words to be linked to dictionary headwords.
a. The links which take you to the forms for editing data (e.g., relating to skalds, poems, stanzas and so on) appear automatically as icons at the top and bottom of each page when you're logged in. The pencil icon is the standard link to forms for editing the data. Where more complex forms exist (e.g. for verses), a pencil-on-page icon will take you directly to the web forms; they are also accessible via the 'more editing forms' tab if you are looking at the standard forms.
b. There are now two sets of forms for some data sets (notably verses) in rows of 'tabs': a top row and a second row which appears when you click on 'more editing forms'. In general, use the lower set where available. Some forms, such as for entering and editing notes and variants, only occur in the top set. (The top set of forms are created by a simpler method and their content and behaviour is edited through the web interface; the lower set, however, is usually more user-friendly and can update data in more complex ways.)
c. If you want to add an edition of a stanza to the database, you will have to clear the old Skj stanza first. The word index and edition are now essentially the same. In order to make sure all the verses are indexed, the Skj text of stanzas which have not yet been entered have been automatically added. There's a simple function for clearing the old text, and a link on the first form you will encounter.
d. The forms for adding the prose order, translation and kennings; and for reordering, splitting and deleting words function more or less the same as before. The form for editing the base text, however, is quite different: a list of words with separate boxes for text and punctuation. It may take a little time to get used to the new behaviour.
e. Most problems have been worked out with the new interface, but others may arise as the database develops. One ongoing problem is that inverted kennings are sometimes not added correctly by the forms for adding the translation and kennings. You can get around this by using the kenning form in the top set of tabs; but in general, if you have problems, let Tarrin know.
f. All the data from the old format has been converted to the new format, with the old format still stored in the database. There have been some problems with the conversion process, but most of these have been fixed. If you entered material in the database before July, you might want to check that it converted properly.
g. On the upside, the whole interface for entering the edition works much faster, both in updating and viewing (you may have noticed this already). Some things should be easier and simpler to use, but others may be a bit different. The ‘split words’ function has been extended to make it more useful, and in general the new format is less likely to produce odd or unpredictable behaviour when it comes to display and printing.
The reasons for the change are largely technical. The General Editors have been discussing future plans for a new Lexicon Poeticum and index of kennings. The production of such material concurrently with the edition will be considerably aided by integrating the word index and kenning index into the actual edition: they had previously been separate from the actual code of the edition, updated when the edited text was entered or updated. No major changes to the database could be made, however, which would risk the progress of Vol. VII, so the changeover had to wait until Vol. VII was exported from the database in June. Since then, following feedback from those using the database, almost all of the initial teething problems have been resolved.
The changes have addressed a couple of other issues with both the data structure and interface, namely, that the data was stored in two different formats (SQL and XML) and the interface was written in two different programming languages (PHP and Perl) for different parts. The new format removes the old XML format and the Perl interface for editing it; this has made development easier. The interface is also much faster using PHP. More importantly for the longer term, interfaces can now be developed for the reference material on lexis and diction. Some of these have already been developed, e.g. the 'lemmata' form in the verse edition interface, which allows words to be linked to dictionary headwords.