Part 1:The Report
For Astounding Science Fiction of 1949,
5 of 6 pieces from January have been reprinted
5 " 7 " February "
4 " 6 " March "
3 " 6 " April "
3 " 5 " May "
4 " 6 " June "
4 " 7 " July "
3 " 5 " August "
3 " 5 " September "
6 " 7 " October "
6 " 6 " November "
4 " 5 " December "

The most quickly reprinted short story was Katherine MacLean's "Defense Mechanism" from the October issue, published 10 months later in Groff Conklin's _Big Book of Science Fiction_. The one that has taken longest to be recognized (so far) is Poul Anderson's "Entity" from June, published in _The Complete Psychotechnic League volume 3_ in 2018, a full 829 months after first appearance. The serialized novels _Needle_ and _Seetee Shock_ also came out in 1950, but in unknown months.

Part 2: What the Heck?
I was chasing down a story through ISFDB. Alas, it was only ever published in an issue of Analog, but that was an issue that one of the database editors had compiled the Analytical Laboratory ("Anlab") results for and put into a note. (Hey, the magazine editor had gone to the trouble to quantify it and publish it, why not collect it?) I could take solace in the fact that it was the fourth-most liked item in its issue, and, of the three items above it, only the serialized novel would come out later in book form. This led to the question: of these stories that were good enough to print, how many were good enough to reprint? It seems simple enough (I really should learn databases some day) to send some bot to gather that info on each issue, and then we can study the changes over years, decades, or editorial reigns, and between different magazines. Before sending Mycroft to find out what is "funny once" and what is "funny always," I thought I'd process a year by hand; to find out what data to gather, so we wouldn't have to send it back for what we forgot, to find out any issues with interpreting the data, and maybe the report will stir interest in the idea. So, picking a year I hadn't already been searching through, and not entirely at random...

Part 3: Thoughts and problems
The process seems simple: for each magazine, go to each issue, and for each item tagged with a fiction category (short story, novella, novelette, serial, novel, excerpt?) check its entry for having more than one entry in Publications. One problem I knew of is that there is an Astounding/Analog UK edition that gets a separate entry for some stories. I see in this epoch it is published three times a year with about 6 stories per issue. (The last two of 1949 and the first of '50 are mostly selections from '49.) So, it is a bit more selective than Galaxy's UK edition of a few years later -which is most of its US issue but later- but I don't feel comfortable counting it as a reprint of the story. If I knew for sure it was sold by subscription, with a buyer blind to the stories contained, I would firmly settle against. What about "Best from..." collections? Another ploy to get more reprint credits from me? I would say 1)a book is even more selective than 1/4 the yearly output, and 2)the book is through a publisher separate from the magazine publisher (except Destinies and Omni?) who has to be convinced that the best is good enough.

Of the 71 total items in the 12 issues, 12 are parts of six serials. This is good enough on an issue-by-issue basis, but does it throw things off if aggregated over a year or decade? One two-parter has not been reprinted, so should 50/71 fiction items be corrected to 45/65 fiction pieces? That's a ratio of 0.704 vs. 0.692, would a +/-2% cover the question? Or would that grow larger with larger groupings?

Serials are also likely to overlap whatever division we make in our larger
aggregations; whatever allowances we make for those that span chronological or editorial divisions, there's one case I know that spans magazines. _The Gods Themselves_ had its middle part in IF and its beginning and end in Galaxy in order to get it all out before the book version. (ISFDB has the book published in March, the cover date of its last installment, and a nominal 0 wait for reprint. Due to actual practices, the magazine would have been taken off the rack before the book was put on the shelves. [1])

One other thing I had in mind before beginning was copyright status. Take the story "Special Jobbery" by H.B. Fyfe: it was published in September, in one of the UK issues, andthen in 2015 in _The 24th Golden Age of Science Fiction Megapack_, an e-book collection, which certainly sounds like something filled with public domain stories as cheap to produce. This suggests to me the accidental property of its quality is secondary to the essential property of it being free to distribute. Or am I being a snob? Something that only shows up in a collection titled "The Complete..." is just as much playing on properties other than its own strength.

One statistic I thought might be useful would be a count of the number of places a story had been anthologized, but I can't think of any use for it, and didn't count on my trawl through the data. However, the first story I checked, "Private Eye" by Lewis Padgett, was first reprinted in _The Best Science Fiction Stories: 1950_, which gave me the idea that dating them would be entertaining, see above. It also has me thinking that any two data points within twelve months of each other should probably be considered a tie -- I'm assuming their story would be as well-liked no matter when in the year it appeared, like Simak's "Eternity Lost" which was picked for the book from the July issue. This would be convenient, as the month of publication for some books is unknown, marked in the data as "00".

There is the case where the magazine appearance is the reprint. Amazing Stories did this a lot in its early numbers, and at least once in a later number,
http://www.isfdb.org/cgi-bin/pl.cgi?56511
looking at that, the one reprint stands out for having the year of first publication after the title. Good, I don't have to invent some filter based on a getting a negative number when subtracting the magazine date from the first date not from the magazine (and worrying about a false positive from a nominal negative if a reprint that comes out the same year as the magazine, but in month "00"). Or should we not ignore them? I'm thinking of how Pohl pulled "Scanners Live in Vain" and Cordwainer Smith from an obscure magazine; is there any lore of a magazine editor going to bat for an obscure book writer? Could we find any in the data? Quicker and more frequent reprinting after a magazine appearance than before? We have to be careful not to misinterpret, certainly Verne has had more reprintings since his Amazing Stories appearances than before.

[1] ISFDB uses the cover date of magazines for the month in the publication data, even though the issues are issued month or months before. This is practical, and, given that January issues copyright their stories for the year following their publication [2], official enough. We need to remember in comparing that differences are useful but ratios are not. In the example in Part 1, we can say one story took 68 years and three months longer to get reprinted, but we cannot say it waited almost 83 times as long; the actual ratio is at least 830/11 and could be 831/12 or only 68 times as long.

[2] How, legally, does that work? It seems to me there was a window where the stories were available, but not *yet* copyrighted. Was the law depending on a certain amount of lag time in the 20th century? I mean, one would have to be quick to grab a magazine, reset its stories in type, print out and distribute it to legally pirate ...a cheap pulp.

--
-Jack

In article <07adcae3-a1f4-4c3d-b745-ae4139a76f30n@googlegroups.com>,
Jack Bohn <jack.bohn64@gmail.com> wrote:
>Part 1:The Report
>For Astounding Science Fiction of 1949,
>5 of 6 pieces from January have been reprinted
>5 " 7 " February "
>4 " 6 " March "
>3 " 6 " April "
>3 " 5 " May "
>4 " 6 " June "
>4 " 7 " July "
>3 " 5 " August "
>3 " 5 " September "
>6 " 7 " October "
>6 " 6 " November "
>4 " 5 " December "
>
>The most quickly reprinted short story was Katherine MacLean's "Defense
>Mechanism" from the October issue, published 10 months later in Groff
>Conklin's _Big Book of Science Fiction_. The one that has taken longest
>to be recognized (so far) is Poul Anderson's "Entity" from June,
>published in _The Complete Psychotechnic League volume 3_ in 2018, a
>full 829 months after first appearance. The serialized novels _Needle_
>and _Seetee Shock_ also came out in 1950, but in unknown months.
>
>Part 2: What the Heck?
>I was chasing down a story through ISFDB. Alas, it was only ever
>published in an issue of Analog, but that was an issue that one of the
>database editors had compiled the Analytical Laboratory ("Anlab")
>results for and put into a note. (Hey, the magazine editor had gone to
>the trouble to quantify it and publish it, why not collect it?) I

I don't have any real thoughts on your database issues, but I have found
recently that many full issues of SF pulp magazines are now available
digitally on archive.org. The format is usually PDFs of page images,
so not the best for pull-quotes etc, but if you just want to read it,
it may be there. (You don't mention the story/issue for your "chasing"
item..)
--
columbiaclosings.com
What's not in Columbia anymore..

On Wednesday, October 27, 2021 at 4:21:06 AM UTC-4, jack....@gmail.com wrote:
> [snip-snip]
> This led to the question: of these stories that were good enough
> to print, how many were good enough to reprint? It seems simple enough
> (I really should learn databases some day) to send some bot to gather that
> info on each issue, and then we can study the changes over years, decades,
> or editorial reigns, and between different magazines.
[snip]

You don't need to send a bot out since our backups are publicly available. Just
follow the instructions posted at
http://www.isfdb.org/wiki/index.php/ISFDB_Downloads#Installation_and_Setup
and you can have a copy of the ISFDB database running on your home PC within
hours. Then, assuming that you do learn databases at some point, you can run
SQL queries and create any number of reports.

As to why the ISFDB doesn't have this information available on the "Statistics and
Top Lists" page (http://www.isfdb.org/cgi-bin/stats-and-tops.cgi), it's mostly due
to the ambiguities outlined in your post. Should a public domain reprint count?
Should a simultaneous publication in a library binding count? Should a magazine
which appears on paper as well as electronically count as two printings? Etc.

Ted Nolan <tednolan> wrote:
>
> I don't have any real thoughts on your database issues, but I have found
> recently that many full issues of SF pulp magazines are now available
> digitally on archive.org. The format is usually PDFs of page images,
> so not the best for pull-quotes etc, but if you just want to read it,
> it may be there. (You don't mention the story/issue for your "chasing"
> item..)

My dear sir! I am *studying* science fiction, not *reading* it! :)

I can admit that my not-entirely-random reason for choosing ASF 1949 was that after getting through the year I would reward myself by finally taking a look at the "predicted issue," November, 1949, and the letter of comment on it that got sent to the November 1948 issue. (I do wonder if the intent of the letter was a spoof on the type of opaque comments that only make sense to one who remembers the original issue or has it at hand.)

Please note EVERY story from that issue went on to secure positions. I thought I would have had at least one issue with no reprints; imagine my surprise at having no issue with less than half the stories reprinted. But then, I have nothing to compare it to, which is one of my points.

--
-Jack

Ahasuerus wrote:
>
> follow the instructions posted at
> http://www.isfdb.org/wiki/index.php/ISFDB_Downloads#Installation_and_Setup
> and you can have a copy of the ISFDB database running on your home PC within
> hours. Then, assuming that you do learn databases at some point, you can run
> SQL queries and create any number of reports.

SQL OK

> As to why the ISFDB doesn't have this information available on the "Statistics and
> Top Lists" page (http://www.isfdb.org/cgi-bin/stats-and-tops.cgi), it's mostly due
> to the ambiguities outlined in your post.

What a page!
"Percent of Publications by Format by Year" A graphic presentation of various Ages, I suppose the rise and fall have more to do with commerce than art. 2007-2015: The Age of Magazine as Digital Audio Download: 1) I didn't realize 2) what?!?
"Most Viewed Short Fiction" I see that "There Shall Come Soft Rains" is number 6 today. I don't suppose that has anything to do with the recent YASID, although I'll try to remember to check during the next YASID to see if titles bump up.

--
-Jack

On Thursday, October 28, 2021 at 12:17:19 PM UTC-4, jack....@gmail.com wrote:
> Ahasuerus wrote:
> [snip-snip]
> > "Statistics and Top Lists" page (http://www.isfdb.org/cgi-bin/stats-and-tops.cgi)
[snip-snip]
> "Most Viewed Short Fiction" I see that "There Shall Come Soft Rains" is number
> 6 today. I don't suppose that has anything to do with the recent YASID, although
> I'll try to remember to check during the next YASID to see if titles bump up.

Our reports are regenerated daily, but the displayed numbers are for "all views
since 2006", so they change very slowly.

Ahasuerus wrote:
> On Thursday, October 28, 2021 at 12:17:19 PM UTC-4, jack....@gmail.com wrote:
> > Ahasuerus wrote:
> > [snip-snip]
> > > "Statistics and Top Lists" page (http://www.isfdb.org/cgi-bin/stats-and-tops.cgi)
> [snip-snip]
> > "Most Viewed Short Fiction" I see that "There Shall Come Soft Rains" is number
> > 6 today. I don't suppose that has anything to do with the recent YASID, although
> > I'll try to remember to check during the next YASID to see if titles bump up.
> Our reports are regenerated daily, but the displayed numbers are for "all views
> since 2006", so they change very slowly.

Oh! I'd gathered the awards accumulated over the decades.
Revising downwards my view of the busy-ness of the site.

--
-Jack