MobileRead Forums - View Single Post - Start Conversation:

Cynosarges · 03-11-2017, 10:56 AM

Kovid,

My apologies for a posting that appears to have been taken as criticism. However you were incorrect - I had (partially) tried your suggestions, as follows.

I was not 'randomly downloading metadata'.

1) In my test set-up, I have recorded a small number of anthologies, added story authors, added a list of the stories in comments (noting that I would lose the book blurb from the book jacket) as you suggested and then thought (incorrectly) that a metadata download would fill null database elements (specifically Rating, which is not a Dublin metadata field) while leaving non-null values unchanged. A wrong conclusion, certainly. But given my current, limited, use of Calibre on a 600 ebook library, it matched my experience (Maybe Baen and TOR provide their metadata to Amazon?).

It was not a 'random' download. It was a purposed download, with the intent of adding metadata which was blank. My experience misled me about the download behaviour. After reading your reply, I discovered that the bulk download of metadata has a (previously unused) configure option where I can exclude individual fields from the download - which addresses the problem I caused/experienced. Mea Culpa.

2) Again, I tried to follow your suggestion. I tried to add the comments column to the book list, but could not find any example that allowed me to do this. I added a custom column, but could not find a way to insert a <newline> into the data (necessary if you are to record a list of fifteen books).

I could not prove that it was impossible to display book information over multiple lines,but was unable to find any document that made the statement that "columns in the book list never take up more than a single line". My only experience on that front is from "Polish books - Insert book jacket" where the comments text does spread over multiple lines. I generalised. Mea Culpa.

Being unable to positively determine the single-line limitation, I reverted to my profession as a business analyst. (To both my enjoyment, and to my regret, over the years I have migrated from programmer to systems analyst to business analyst, ever retreating from the writing of code.) I thought from the end-user viewpoint, and started from a user requirement of the three reasons for anthology information I stated. I then looked at the Calibre GUI, and thought how it could support those needs. My conclusion was that I would need multi-line display (to list 15 stories), and (unless I wanted to abandon book jackets) it would have to use a custom column. I did not prove that multi-line display was possible or impossible, but concluded that with single line display, the user would have to repeatedly enter the edit metadata screen, to see the comments field (apparently the only field with multi-line display), and considered the usability.

At the start of my emigration/retirement plan, I had consulted with the borough librarian primarily on the important issues involved in physically shipping books 6,000 miles, but also discussed library software and anthology recording. I used this discussion to develop my requirements for library recording. Before settling on Calibre for an integrated paper/ebook library solution, I had already determined an ugly, but usable, work-round for anthology data.

Calibre could record anthology data by setting up a second Calibre library, solely for anthologies. In the main library, use a tag "Anthology" to indicate inclusion in the anthology library. In the anthology library, each anthology book is represented by a series (Naming the series as "Editor - Title") You then create book records in this series for each book in the anthology, with author being story author, title being story title, and Series sequence being the order in which the story occurs in the anthology.

Ugly, and suffers from deficiencies. (No metadata download, all data has to be manually entered. The use of the series name is ugly. As far as I know, there is no integrated author names across two libraries) However the this approach does offer all stories by a given author, and all stories in a given anthology in an intuitive way for a Calibre user. The stories are not listed with non-anthologies (unless Calibre can integrate Catalogues across two distinct libraries - Not yet investigated) but it would allow me to compare what I own with the Encyclopedia of Science Fiction or an online bibliography.

03-11-2017, 10:56 AM	#12
Cynosarges Junior Member Posts: 8 Karma: 53566 Join Date: Mar 2017 Device: Kindle Touch	Kovid, My apologies for a posting that appears to have been taken as criticism. However you were incorrect - I had (partially) tried your suggestions, as follows. I was not 'randomly downloading metadata'. 1) In my test set-up, I have recorded a small number of anthologies, added story authors, added a list of the stories in comments (noting that I would lose the book blurb from the book jacket) as you suggested and then thought (incorrectly) that a metadata download would fill null database elements (specifically Rating, which is not a Dublin metadata field) while leaving non-null values unchanged. A wrong conclusion, certainly. But given my current, limited, use of Calibre on a 600 ebook library, it matched my experience (Maybe Baen and TOR provide their metadata to Amazon?). It was not a 'random' download. It was a purposed download, with the intent of adding metadata which was blank. My experience misled me about the download behaviour. After reading your reply, I discovered that the bulk download of metadata has a (previously unused) configure option where I can exclude individual fields from the download - which addresses the problem I caused/experienced. Mea Culpa. 2) Again, I tried to follow your suggestion. I tried to add the comments column to the book list, but could not find any example that allowed me to do this. I added a custom column, but could not find a way to insert a <newline> into the data (necessary if you are to record a list of fifteen books). I could not prove that it was impossible to display book information over multiple lines,but was unable to find any document that made the statement that "columns in the book list never take up more than a single line". My only experience on that front is from "Polish books - Insert book jacket" where the comments text does spread over multiple lines. I generalised. Mea Culpa. Being unable to positively determine the single-line limitation, I reverted to my profession as a business analyst. (To both my enjoyment, and to my regret, over the years I have migrated from programmer to systems analyst to business analyst, ever retreating from the writing of code.) I thought from the end-user viewpoint, and started from a user requirement of the three reasons for anthology information I stated. I then looked at the Calibre GUI, and thought how it could support those needs. My conclusion was that I would need multi-line display (to list 15 stories), and (unless I wanted to abandon book jackets) it would have to use a custom column. I did not prove that multi-line display was possible or impossible, but concluded that with single line display, the user would have to repeatedly enter the edit metadata screen, to see the comments field (apparently the only field with multi-line display), and considered the usability. At the start of my emigration/retirement plan, I had consulted with the borough librarian primarily on the important issues involved in physically shipping books 6,000 miles, but also discussed library software and anthology recording. I used this discussion to develop my requirements for library recording. Before settling on Calibre for an integrated paper/ebook library solution, I had already determined an ugly, but usable, work-round for anthology data. Calibre could record anthology data by setting up a second Calibre library, solely for anthologies. In the main library, use a tag "Anthology" to indicate inclusion in the anthology library. In the anthology library, each anthology book is represented by a series (Naming the series as "Editor - Title") You then create book records in this series for each book in the anthology, with author being story author, title being story title, and Series sequence being the order in which the story occurs in the anthology. Ugly, and suffers from deficiencies. (No metadata download, all data has to be manually entered. The use of the series name is ugly. As far as I know, there is no integrated author names across two libraries) However the this approach does offer all stories by a given author, and all stories in a given anthology in an intuitive way for a Calibre user. The stories are not listed with non-anthologies (unless Calibre can integrate Catalogues across two distinct libraries - Not yet investigated) but it would allow me to compare what I own with the Encyclopedia of Science Fiction or an online bibliography.