Real world experience with data sets and lawyers - Our company lost a legal battle with a competitor over how we presented automotive color data. Vehicles use a colored primer under the topcoat color. Very often the primer color is a shade of gray. We indexed topcoat colors to the coresponding primer gray color in our database that is distributed to end users. Our competitor sued us claiming indexing topcoat colors to gray primer shades on a part number by part number basis was their IP. We lost and had to remove the index. The lawsuit wasn't about the fact  we had both collected the same data but rather over the way we had both indexed and presented it the same - topcoat color/primer gray color as one extended part number. Apparently, the combination part number method was considered IP. We still have all the same data in our database, just now the user has to make an intermediate reference (2 step process) instead of the primer color being shown alongside the topcoat color (1 step process). I don't know the specifics of the case and am not a lawyer but I did learn that IP cases sometimes defy common sense. Best to engage legal services before you invest tons of effort rather than afterwards in a court.

Sad to hear that the MR index is no more, and that the folks at Kalmbach could not find the effort to make it current. Time is money, and I am sure the business decision was made that the level of effort for them was too great considering the volume of back issues.

Converting what MR calls the old DOS code to a web accessible database would have taken many man-hours to generate, most likely it would be easier to just build a new database from scratch, yet still, that is a heck of a lot of data entry.

Just off the top of my head, you are looking at the following simple formula for factoring the number of data entries: # of years published X number of issues per year X number of articles in each issue. Then each record would require several fields to be populated, each record represents an article title, a author, date, issue number, page listing, then cross referencing indexes also. Indeed, a major project taking a huge number of coding hours!

Hope MR will be willing to turn over their resources for public use.

 I have to disagree with what Pieter said above. Taking the "clean Room" approach will not protect you from copyright infringement.

If you have something that is copyrightable, and someone produces a copy of that,the law does not care about how that copy was produced, only that it exists in violation of the original copyright.

As regards copyright on anything we produce I would suggest we take the Wikipedia  approach and say that all contributions are made under some sort of creative commons share-alike licence.

Yes, I agree - the final index should be delivered using a Creative Commons - non-commercial license.

This means it must remain forever free and someone can't create a derivative and then sell it for money.

some body mentioned reproducing Harry Potter in it's whole WRONG.

in that I have never read Harry potter this is an example!!the Key word is INDEX

Chap 1 Harry goes to wizard school ---page 2

Chap 2 Harry gets his first wand ---Page 13

Chap 3 Harry rides a broom ----page 22

This is not the whole story or even 1/10 of the book just the Index That's all you will get not the articles!!!

The Index for all the years of these magazines will explain what you will find in what year and month and page of each magazine and unless you own these magazines you will NOT get anything from the Index other than subject matter, chapter, Tittles, as to where it can be found.

The Index is Not the reproduction of any single magazine!!!

It is the collective Indexes and only the indexes of 70 plus years of all Model Railroad Magazines.

The Index will not have a single article that can be read unless you own the magazine the Index references.

SO NO COMPLETE MAGAZINE will be in the Index Just the Index so, don't think we want every publication printed in their entirety on line that can't Happen as these companies still sell old copies of these Magazines..

Dan

Exactly! If an index fell under copyright law then library card catalogs would be a violation of the law! 

Ken L.

in their entirety on line that can't Happen as these companies still sell old copies of these Magazines..

Actually Dan, the entire contents of Model Railroader from the last 10 years is indexed online through my local library. The entire article text is searchable. I can download PDFs of articles I search up. I have to use my library PIN to use it. I forget the name of the service, I will post back if anyone cares. Its a general periodical search and does not go back much past 2000 (when magazine content started becoming digital..). So, I think its more of a technical issue. Perhaps there would be probelms getting content indexed from magazines whose ownsership is murky,

Chris

I don't want to go too far into a legal discussion, but a database of information like an index IS protected by copyright. You can read from some testimony by the Office of Copyright here:

http://www.copyright.gov/docs/regstat092303.html

or go by the salient comment (underline added for emphasis) excerpted below:

"On the question of the scope of protection afforded to compilations, there was somewhat greater uniformity in the case law. In compilation cases, regardless of the theoretical framework adopted to justify copyright protection, once the plaintiff's work was determined to be copyrightable, courts generally held a defendant to have infringed whenever material was copied from the plaintiff's work. Typically, there was no inquiry as to whether the particular material copied was protected by the plaintiff's copyright. To avoid infringement, a second-comer was required to go to the original sources and compile the material independently, without reference to the earlier work. ⁽⁷⁾ A common thread running through many of these decisions was the court's desire to prevent the copier from competing unfairly with the compiler by appropriating the fruits of the compiler's efforts or creativity. In this sense, courts treated copyright protection for compilations much like a branch of unfair competition law."

A similar issue in strictly model railroad terms would be to assert that you could legally copy a decal for a prototype car, because it is a representation of the real car and not an "original work". Microscale will strongly disagree with you on that assertion. You can draw your own artwork (or compile your own index), but you cannot copy another's work even if that work is itself a representation of another existing "thing".

Pieter Roos

I don't think anyone is debating that the various magazines own their content. I do think it is possible to get permission to use it for indexing if an owner can be found. That is why I gave the example of the library index. Clearly Kalmbach, in example, is willing to make this content available under some level of license to indexes even to the point of allowing article download. As much as people bash them.. they have always shown a willingness to make their content available for non-commercial use. For instance, its very easy to get permission to use their track plans in LDSIG articles.

Chances are its not free for the downloading part of the index service I used. The library likely subscribes as a non-profit.. but the interest here is in indexing not article download. It is only in the interest of the magazine for back issues they may sell to be readily indexed so that potential customers may find they have a need..

I still think this is more of a technical issue. Its a LOT of data.. and doing that kind of text searching requires a LOT of power.. more then may be available to a non-profit attempt to make this available. The advantage to keywords and TOC indexing is it is a relatively finite and targeted search. This makes the hosting side of things less expensive.

Just my two cents.. SQL server has embedded full text indexing capability. Its a little awkward but I have been able to write ranked searches in it. Third party engines ($$!) are also available. However, this index needs to be usable by many users concurrently and would have much more data to search then my customers did.. well with one exception but I can't really discuss that. Google search may be a possibility but then the text needs to be available for google to crawl and thay could violate an agreement.

Chris

If we generate the material (by everyone contributing a couple year's worth of data) using a pre-determined copyright (e.g. Creative Commons (http://www.creativecommons.org), there'll be no problems even if (unlikely) Kalmbach tried to assert restrictive ownership on the current index data.

The software itself will be governed by an Open Source license (maybe Java Community Process terms, maybe Apache, don't quite know what Clint will decide yet.)

dave