Old-Time XML Standard

New York State
USA
Website: www.mort.net Contact: info@mort.net

Ken's Web Pages Brooklyn, New York, USA , Website: www.mort.net/users/krm Contact: ken@post6.mort.net

Note: I am not an XML expert and I will not try to explain XML fully. My main purpose here is to generate some discussion about XML as a means of cataloging old-time recordings that are avaliable on the internet. I have a brief description of XML which is not very good if you don't already know something about computers and databases. Also, it's been over two years since I read up on XML, so I probably don't have all the terms correct. This is really to begin some discussion about XML in the old-time music community. Please consider this document a rough draft of this topic. If interest developes then perhaps it will be refined and expanded. I can be contacted at the email address shown at the top and bottom of this page.

Old-Time XML Standard

It would be really nice to be able to search all the repositories of old-time recordings that are available on the internet. If a common XML definition were established then a search could be submitted to all the known repositories and a combined result could be returned.

Let's say a number of internet sites with on-line recordings agree on the following simple XML document definition. The definition has an element name followed by the type of data that it should contain. Some of the data types may be free-form text (anything may be entered), other types have to be a certain type (like a number or a date). Also, some data types may only contain one of a number of choices (shown here as text separated by vertical bars '|').

<COMMON-TITLE> any-free-form-text
<LOCAL-TITLE> any-free-form-text
<PRIMARY-PERFORMER-NAME> any-free-form-text
<PRIMARY-PERFORMER-BIRTHDATE> any-positive-date
<PRIMARY-PERFORMER-INSTRUMENT> value= fiddle | banjo | guitar | pipes
<RECORDING-DATE> any-positive-date
<STYLE-LOCATION-COUNTRY> value= standard-two-letter-country-code
<STYLE-LOCATION-STATE-REGION> value= standard-two-letter-state-code | Appalachia | Rocky Mountains
<STYLE-LOCATION-COMPASS> value= North | South | East | West | Central
<STYLE-LOCATION-COMPASS-SUFF> value= Eastern | Western
<SAMPLED-FORMAT-URL> URL_hyperlink
<SAMPLED-FORMAT-SIZE> positive-numeric-kilobytes
<DESCRIPTION-URL> URL_hyperlink
<MANDATORY-NOTES-COPYRIGHT> any-free-form-text-or-URL-hyperlink

Here is how the XML document for one of the songs on my web site would look.

<COMMON-TITLE> Rose Waltz
<LOCAL-TITLE> Frank Lowrey's Waltz 1
<PRIMARY-PERFORMER-NAME> Jim Bryner
<PRIMARY-PERFORMER-BIRTHDATE> 07/08/1905
<PRIMARY-PERFORMER-INSTRUMENT> fiddle
<RECORDING-DATE> 1982
<STYLE-LOCATION-COUNTRY> US
<STYLE-LOCATION-STATE-REGION> PA
<STYLE-LOCATION-COMPASS> South
<STYLE-LOCATION-COMPASS-SUFF> Western
<SAMPLED-FORMAT-URL> http://www.mort.net/bwf/krm/jbtapes/tape1/t1t35.MP3
<SAMPLED-FORMAT-SIZE> 1256
<DESCRIPTION-URL> http://www.mort.net\users\krm\otf\jbryner\tape1\index.shtml
<MANDATORY-NOTES-COPYRIGHT> This file can be shared with anyone in the world.
                            Please include a reference to the source or a link to
                            http://www.mort.net/bwf/krm/jbtapes/tape1/t1t35.MP3
                            when distributing this music.

Now if a number of old-time recording repositories have cataloged their music with the Old-Time XML definition, and a search form is developed at a web site that knows about a number of those repositories. An end user could search all of them using that single search form.

If someone was searching for songs where the primary instrument was fiddle and the style was typical of Southern Pennsylvania.
(PRIMARY-PERFORMER-INSTRUMENT=fiddle)
(STYLE-LOCATION-COMPASS=South)
(STYLE-LOCATION-STATE-REGION=PA)
the results would include most of the music at my Jim 'Ike' Bryner Tapes page.

Let's say someone is searching for early recordings of fiddle music. The search site could let the person search by recording date (RECORDING-DATE) or birthdate of primary performer (PRIMARY-PERFORMER-BIRTHDATE).
Sometimes a repository site might not know the exact recording or birth date. Additional definitions on those elements could be defined to produce intervals. Let's say we think a recording took place in 1982, but it might have been up to five years earlier but no more than one year after. We could have the following.

<RECORDING-DATE>      value= 03/25/1982
<RECORDING-DATE-LOW>  value= -5
<RECORDING-DATE-HIGH> value= 1

In a seach for recordings prior to 1980, the above recording could be included as a possibility.

Ideally the search site would present a simple flexible way to enter search criterion. Also, it should allow the user to specify what output they wish or to return summary statistics instead. Because the search site is requesting XML documents from the repository sites, it can format (sort, create summary statisticts, ...) the result any way it (or the user) wants.

The search site would not include the actual song file that contains the recording. This should be handled by displaying a URL hyperlink (SAMPLED-FORMAT-URL) instead. This way the user could click on the link to retrieve the song from the original site directly.
Also, the search site should include any information that the repository site considers mandatory. This may include a copyright notice or other notes on usage/distribution of the files.

Some definition of a 'Collection' of songs might also be useful (and collections of collections). This could be useful for persons searching for songs in a general catagory (not specific song titles). Since XML documents are hierarchial in nature, this would be very easy to do without duplicating at the individual song level, since the songs could inherit their parent's (the collection's) descriptions. An individual song could also then override or add to a parent's description.

How Do We Start This Project?

Define the elements in the Old-Time XML document: The elements (their names, values and attributes) need to be agreed upon and defined. It would probably be best to begin with a simple set of elements and add to them after some experience is aquired. Hopefully there are some persons in the old-time music community who have experience with XML. Some investigation of current XML standards for music cataloging may turn up something already in place. Maybe an existing standard that is deficient for Old-Time music could be extended to better serve the Old-Time community.
Have some repositories of old-time music catalog their music using the Old-Time XML definitions: Any repositories of old-time music that wish to be included in searches would have to catalog their music using the Old-Time XML definitions. It's also possible for someone else to catalog a respository and provide the XML document (with the permission of the respository).
I am willing make a committment to catalog all the old-time music I have on my web site for such a project.
A web site will need to host a "Seach Page" for the project: This would be the web page that users come to in order to search for songs and collections. The web site would need to be XML enabled if the requesting and processing of the documents is to be done on the server side. If the whole process can be done on the client (browser) side, then any web site could host the "Search Page".
The web serving running my site is not XML enabled. I plan to do that some time in the future however, I cannot make any committment as to when this will happen.
If the entire process can be done on the client side (using javascript or other client means), then I am willing to commit to putting the "Search Page" on my web site.
Sit back and listen to all that Old-Time music you have been looking for!: Aaaaaaaah. The work is done, you can find all the stuff that you've been looking for, even stuff you never dreamed existed.

How To Contact Me

Send me email at the address shown at the top or bottom of this page.

Brief Description of XML

XML stands for eXtended Markup Language. Over the last few years there has been a great push in the technology industry to begin using XML to 'describe' data. When two or more entities (persons, organizations, computers,...) agree on a given set of descriptions for their data, they can exchange that data more efficiently. XML documents are very similar to hierarchial databases, however they are 'text' (human readable) based documents so they can be exchanged between different computer systems even if the computer systems uses a different internal representation of the data. XML documents can also contain 'binary' (computer only readable) data (ie. digital music, images, ...) , however the use of this is discouraged if the data can be resonably represented in text format.
(Maybe I'll add some more to this section someday)