Clint has been busy "scraping bits" from the Internet Wayback Machine and other sources, as well as software updates. This is moving along quite well, we're starting to work with relatively complete datasets, as well as the transition from a development environment to hosting presumably by NMRA. We're not ready to go public (my home internet won't carry the load), but I wanted to assure everyone that we're making very good progress on this.  

Our query approach is to emphasize how we observe people using Google, i.e. adding keywords until the number/accuracy of query results approaches a reasonable number, rather than instituting a complex query language (although our software will support boolean operations, we don't expect that's how people will use it.) As a result, we don't have a fixed set of keywords for describing articles, although we anticipate collecting and advocating a default set of keywords.

And of course, our approach is based on 'crowdsourcing' data entry and updates. Any authorized user ('how authorized' is something that we'd expect NMRA to work out when they host the facility) will be able to edit, and any user will be able to mark an entry as incorrect/incomplete. In this respect the facility will work like Wikipedia, albeit with the requirement to get an editor's account before editing any pages.

dave

Hi

I've just been following this index thread and mostly you guys have got some great ideas for setting up an index.

A couple of thoughts though.

Near the start, it was stated that it was a general railway & model railway index and to be made available to everyone, including those residing outside the USA & also magazines etc from outside the USA. Does that still hold true?

If so, some way of recognising where the source of any refence is based. Now the easiest way to do this would be to use the standard 2 letter country code as a suffix, thus making it quite clear.

The codes are these

http://www.theodora.com/country_digraphs.html

An example being for the top of page 19 in the thread where there is a list of magazine codes (many of which are unknown or there is some doubt, due to reuse). The assumption is that are all US titles, but are they? Putting the Country Code at the end would confirm this.

On the magazine codes, it is a most confusing list, with letters varying from 1 to 4 letters. My suggestion is to make a standard 4 character minimun (more if required) followed by the Country Code. Then everyone ought to know that MODR(US) is the Model Railroader magazine from the US and MODR(UK) is the Model Rail magazine from the UK. Both are large selling mainstream magazines.

Thinking about what I have just written, I suspect more than 4 letters will be required. At least with modern computers we don't have to worry about minimising hard disk space.

I have got some more ideas, but I'll wait for some feedback on the above.

Kevin

 One might choose to store values in the database in some abbreviation (e.g. MODR(US) vice MODR(UK) ), but that should not be visible to the user.  

In our current implementation, the full (proper) name of the magazine is stored and used for retrieval.  We use some 'keyword abbreviations' that the software -expands- to a full keyword when data is stored into the database; those abbreviations are only for data entry convenience.  

We're in a limited beta test right now, along with discussions with NMRA to host the system.  But I have to admit, the state of the discussions with NMRA at this point are not encouraging!  I would hope their basis would be not "if" but "when/how", and we're not there yet.

dave

Trains.com has posted that the magazine index will be back in Janurary!!!!!

I have been sending them alot of emails and so have others telling them that

we need that index.

Here is the links:

http://trc.trains.com/modeltrainindex

Bob D.

While it's nice to see it's apparently going to be back, one of the things I noted in a first press release that was passed around, and reinforced by the "For print magazines..." line in this one, is that E-Zines don't get to play here.

Yes, it appears that Kalmbach has decided not to include whatever might need to be included in this database to support indexing E-Zines such as MRH.

I would suggest instead that everyone take a look at http://www.rrmagindex.org which is already available in limited beta and open to all.

I assumed that once "they" realized where a non-Kalmbach controlled index was heading, that they would be quick to do an about face.

It's stuff like this that makes those so called, "ridiculous evil plot" conspiracy theories, seem a lot less ridiculous, IMO.

I thought the notion of Kalmbach giving control of the index to the NMRA was a "pipe dream" from the get go.

Thanks Kalmbach for doing exactly what I expected!

Thanks for the link to the index that is in Beta. I would rather support this index as I see any reference to this online magazine, "there", is quickly removed by "Them"..

I did not consider the posibility of "them" not including MRH.

Rich

Thanks for the link to the index that is in Beta. I would rather support this index as I see any reference to this online magazine, "there", is quickly removed by "Them"..  I did not consider the possibility of "them" not including MRH.

Rich

I know.  It seems like a fairly childish stance to take, IMO.

I was not aware that "they" were still removing MRH references.  I wonder why they are seemingly so threatened?

The one thing that concerns me about the Beta index, is the exclusion of MRH.  I looked at the form for submitting data and I see that mention is made of adding data from paper mags and that there is no option in the drop down list of publications, to add MRH data.  What's up with that?

I just signed up with the Beta index and will be attempting to fix that missing category in the next few days as I have the time.

I just signed up with the Beta index and will be attempting to fix that missing category in the next few days as I have the time.

Jeff Shultz

Great!  Thanks Jeff.