Goodreads Librarians Group discussion

I'm updating a batch of 10,000 books from ingram right now. This will only operate on the following fields at the moment: isbn, isbn13, title, and author.


If you guys don't notice any quirks in the data i will run the rest (100,000+ books) later on this weekend.


Let me know if you see any unusual data associated with the data source 'ingram'.

I think this might be a little late (or a LOT late) but... I was looking at some of my books and adding series and saw these:
http://www.goodreads.com/book/show/28...
ingram updated the book By Time I Get Nashvil by Lionel Fenn
title: 'By the Time I Get to Nashville' to 'By Time I Get Nashvil'
http://www.goodreads.com/book/show/27...
ingram updated the book Time: Semi-Final Front by Lionel Fenn
title: 'Time: The Semi-Final Frontier' to 'Time: Semi-Final Front'

Of course, the other ten Lionel Fenn books look fine (at a quick glance) but it does at least raise the question. I kinda wish I'd seen the post a little earlier, but who knows whether these fell in the first 10K or the last 100K.
(shrugs)

Just confused by all this:
http://www.goodreads.com/book/show/42...

This was updated by ingram as the book (except for cover and number of pages) is now, but all the reviews are for My Man Blue. Looks like a dupe ISBN? I have a copy of The Eagle and the Raven on hand, and it has this IBSN on the copyright page in the LOC info. There is a second ISBN for the Canadian edition given as well.

Data wrote: "Just confused by all this:
http://www.goodreads.com/book/show/42...

This was updated by ingram as the book (except for cover and number of pages) is now, but all the reviews are for My Man Blue. Looks like a dupe ISBN?"

Apparently so:

http://www.worldcat.org/search?qt=wor...+

Since the book is combined with another version of My Man Blue, that was probably the info GR had before the data changes. I would change the info to match My Man Blue. Then an Alternate Cover Edition can be created for The Eagle and The Raven. With the new data imports, this is going to happen from time to time. However, my understanding is that once edited by a librarian, it should not get overwritten again.

"Let me know if you see any unusual data associated with the data source 'ingram'."

Well, change #24975403 @ Feb 08, 2012 12:33am from ingram
=> title: 'Berlioz: Roméo et Juliette (Cambridge Music Handbooks)' to 'Berlioz: ROM O Et Juliette
was not a particularly useful change.

Thanks for reporting this, we'll try and investigate before importing more ingram data.

"Let me know if you see any unusual data associated with the data source 'ingram'."

change #24460531 @ Feb 07, 2012 09:39pm from ingram
=> title: 'More and Harder' to 'Idol More and Harder'

Once again the title field from ingram is worse than the original title. Ingram includes the name of the publisher's series as the first word of the title, for some unknown reason.

Just noticed a "7" added to a title by ingram: http://www.goodreads.com/book/show/59...

(I already reversed it before seeing this thread.) Have noticed several numbers in title lately but this is the first time I checked the change log.

I doubt we'll be able to filter those. There are too many cases where a number like that would be valid. (Number as part of title or subtitle, series indicator, etc.) Feel free to fix them, though!

One more instance, if data will help.
http://www.goodreads.com/book/edits/1...
title: 'The Mirror Cracked from Side to Side Book and MP3 Pack (Penguin Reader Level 4)'
to 'The Mirror Crack'd from Side to Side. Based on the Story by Agatha Christie'

Sounds like it was trash to begin with and was changed to different trash. :-|

Extraneous (and now removed) information is hardly "trash" The title is mostly correct; and author, publisher, etc. are fine.

Andy wrote: "...if data will help..."

Sorry... Should we still be posting problem records or just fix them? I mean in most of the cases I've seen the fixes would be pretty obvious.

Unless they are examples of overall problems (like the dates being imported incorrectly because they were listed elsewhere as 199-), it's best just to fix them.

Just in case this is still an ongoing problem and/or can be fixed automatically, I just fixed a number of ebooks with missing formats and this in their edit logs:

ingram updated the book xxx
format: 'ebook' to 'Open Ebook'
Feb 08, 2012

ingram updated the book xxx
format: 'Open Ebook' to ''
Feb 14, 2012

All of these show ebook to Open Ebook on Feb 8th and then Open Ebook to empty string on Feb 14th. (They're also all published by Harlequin, which I doubt makes a difference locally but might have something to do with the incoming data from ingram.)

Examples:
http://www.goodreads.com/book/edits/6...
http://www.goodreads.com/book/edits/6...
http://www.goodreads.com/book/edits/7...
http://www.goodreads.com/book/edits/8...
http://www.goodreads.com/book/edits/8...
http://www.goodreads.com/book/edits/1...
http://www.goodreads.com/book/edits/1...

That's the result of a cleanup script that didn't work quite right. It's been fixed for the future, but it has been decided not to run in again. feel free to set all of those to format ebook, though.

Ingram is overwriting existing publishing dates, for what I assume is a new printing of the same edition.

On Jan 16, this edition of How the Grinch Stole Christmas had its publication date changed from 2011 to 2012. I assume this is a new printing of the same edition. However, in the past we have always used the date of the first printing of a given edition. It will get very confusing if you consider that there are people who probably reviewed that edition, and now it looks like it hasn't even been published yet! :)

ETA, Here's another one:
http://www.goodreads.com/book/show/37...

It shows that onix firebrand changed the date to 2012 back in December, but attributes my name to the currently listed date. I know I did not change this to a 2012 date; it's a book I already own! (Unfortunately, I do not have a time machine :( so I didn't buy it in the future).

The really weird part of all of this, is that the dates appear to have been changed months ago based on the change log. However, these editions were not showing up in my books with these dates until today.

Pretty sure we're not going to be able to change dates like that from importing, but I'll ask.

vicki_girl wrote: "The really weird part of all of this, is that the dates appear to have been changed months ago based on the change log. However, these editions were not showing up in my books with these dates until today."

That's weird. Links to edition-specific changelogs?

Here's some:

http://www.goodreads.com/book/edits/3...
http://www.goodreads.com/book/edits/6...
http://www.goodreads.com/book/edits/1...
http://www.goodreads.com/book/edits/3...
http://www.goodreads.com/book/edits/9...

To be perfectly clear, I didn't see these before today. However, I have been reorganizing my books, sorting them by publication date, over the last several weeks. Today was the first time I noticed this.

(1st and 4th link are the same.)

Other than the last one, those are all Firebrand Onix updates. The dates in the logs match the dates that was first run, I'm pretty sure. The last one is RH's Onix, and again the date makes sense.

Can't even guess why you might not have seen their changed pub dates before today. I'm not aware of anything that we did in the last week that should have affected librarian logs retroactively, or anything along those lines.

Well, there is at least on error with this one:

http://www.goodreads.com/book/edits/3...

It shows onix changed the date, but shows my name as the date source (bottom of page). Again, it's a book I own and I know I did not change the date to something in the future.

Should I go ahead and fix these? If I do will the change stick? I don't care overly much that it is wrongly attributing me as the source, I just want the data to be right, and stay right. :)

vicki_girl wrote: "It shows onix changed the date, but shows my name as the date source (bottom of page)."

Definitely weird, but I think it's because it was from our test imports, and the scripts Brian used to source fields ran AFTER those imports.

Yes, go ahead and fix them. I fairly certain the changes should stick.