User talk:Alvonruff

From ISFDB

• User_talk:Alvonruff/Archive01
• User_talk:Alvonruff/Archive02

The Dread Tomato Addiction

We seem to have people verifying opposite ways: you for Essay, some for Shortfiction. Is there a standard for translating Contento's "Facetious Article" category? (I haven't got a copy, but it sounds such an intriguing title I might have to get one!) BLongley 16:18, 13 Oct 2007 (CDT)

Recent Integriations changes?

It looks like submission IDs are no longer hyperlinked from the Recent Integration pages with the exception of Author Updates that are still working fine. Was that "software rot", an accident or was there a problem with Publication Views? Ahasuerus 18:03, 15 Oct 2007 (CDT)

"ExPat Error" on two submissions

Al, when you get a chance could you zap these submissions? 865173 and 865185. I was copying-and-pasting an HTML note in a series of submissions, and mistakenly placed it into the wrong field, then stupidly submitted it. And to prove that I could not have made such a stupid mistake, I did it again. Right now the submission page looks like a mess. Thanks. Mhhutchins 15:39, 5 Nov 2007 (CST)

These were dumped long ago, but I did modify the submission list app to not look like a mess when an XML error occurs, and I also made it possible (for pv_update and pv_new for now) to look at mangled submissions and reject them. Alvonruff 11:58, 30 May 2008 (UTC)

Dark Horse vs. Dissembler

The other night I was cleaning up the Aliens series and noticed that Dissembler had failed to grab novels #13 (John Shirley's Steel Egg) and #14 (S. D. Perry's Criminal Enterprise) in the series. I then noticed that both books were published by "Dark Horse", a well known comics publisher, so I wondered if Dissembler was configured to ignore everything with the string "Dark Horse" in the publisher field. If so, then I suspect that we may be missing additional novels. At first any books published by "Dark Horse Comics" appeared as by their subsidiary DH Press, but apparently some of their recent titles have been cataloged by Amazon (and other online catalogs) as having been published by "Dark Horse".

Assuming my hypothesis is correct, the missing titles should be easy to identify since they have a complete catalog of their novels online, but we will probably want to fix Dissembler as well. Ahasuerus 19:48, 21 Dec 2007 (CST)

Dark Horse is definitely in the dissembler black list as a comic publisher. Is the fix to take Dark Horse off the list, or are there different imprints for novels vs comics? Alvonruff 06:11, 22 Dec 2007 (CST)

Well, the underlying problem appears to be that Dark Horse has now branched out into media-related SF novels. Although in theory all of their fiction is supposed to appear as by "DH Press", at least some of it is listed (by Amazon.com and other online sources) as by "Dark Horse", so there is no easy way of telling whether a particular title is a novel or a comic. Unfortunately, they still produce an awful lot of comics, so if we removed them from the black list we would end up with a lot of bad data in the database or at least a lot of bad Dissembler-generated submissions.

One reasonably reliable way of distinguishing between Dark Horse's comics and novels is the price. Their novels are, AFAIK, all mass market paperbacks and priced at $6.99, although I assume they will move on to $7.99+ eventually. Their comics, on the other hand, are usually $14.95+, sometimes $10.99. Would it be feasible to change the blacklisting condition to something like "If the Publisher contains the string 'Dark Horse' and the price is over $9.90, then do not create a submission"? If not, then we could keep the publisher blacklisted, but periodically check PH Press' online catalog. Ahasuerus 17:06, 22 Dec 2007 (CST)

Not a problem. Dissembler already uses price heuristics to weed out box displays. Alvonruff 06:42, 23 Dec 2007 (CST)

Excellent! Now we just need to get them to promise that they will never ever do trade paperbacks :) Ahasuerus 13:55, 23 Dec 2007 (CST)

EditBug 10092 back? Code reversion?

I removed all page numbers from a newly cloned Publication, but when I approved my submission, all page numbers were still there. This bug (EditBug 10092) is marked as "fixed" on the ISFDB Bug List, so I wonder if we may have accidentally gone back to a prior version of the software? Ahasuerus 17:47, 25 Dec 2007 (CST)

Was this during the cloning process or editing the clone once it existed? I've tried this while cloning numerous pubs at home and did one test case on the live system, and verified that the cloning tool doesn't generate cPage tags if the fields are blank (and also verified that empty cPage tags are generated when editing a pub). During moderation the cloning approval tool only adds page information if the cPage tag is present (which it wasn't), and the pub editing tool only modifies page information if a cPage tag is present (which it was). At this point I'm not sure how to reproduce the issue - perhaps something peculiar to that particular publication? Alvonruff 06:49, 26 Dec 2007 (CST)

I had cloned a much larger omnibus that needed to be Unmerged and pruned based in part on page numbers, so I didn't delete them during the cloning process. The problem is that I can't delete the page numbers from the resulting publication -- the submission looks OK on the Approval page with all the page numbers getting changed to "-", but when I approve it, the resulting SQL is only 3 lines long and the page numbers are not deleted. It's recreatable, give it a try and you will see. Ahasuerus 11:43, 26 Dec 2007 (CST)

Empty series deletion

One of the things that we can't do at this time is delete empty series. If I recall your explanation of the difficulties surrounding auto-deletion of empty series correctly, the problem was with nested series where the deletion of the last entry in a deeply nested series would suddenly make the whole series hierarchy subject to deletion, which wasn't easy to implement.

I assume that these considerations still apply, but I wonder if it would be possible to implement a palliative solution, i.e. a new option on the navbar that would allow editors to delete the displayed series. Naturally, the logic would first check to see whether the series was empty, but it wouldn't check whether the empty series' superseries. Thus, our editors would have to do a couple of submission cycles to get rid of empty nested series, but it would be a huge improvement over the current situation.

The reason I am asking is that we have accumulated quite a few empty series after multiple cleanup projects and sub-projects. At this point Dragonlance alone has a couple of empty series, so any relief would be welcome :) Ahasuerus 20:24, 25 Dec 2007 (CST)

Python error approving an e-book

I have received the following Python error when approving this submission:

SQL Updates:

    * insert into pubs(pub_title) values('xxx');
    * update pubs set pub_title='Gulliver of Mars' where pub_id=255979
    * update pubs set pub_tag='GLLVRFMRST1996' where pub_id=255979
    * update pubs set pub_year='1996-07-01' where pub_id=255979
    * update pubs set pub_ptype='ebook' where pub_id=255979
    * update pubs set pub_ctype='NOVEL' where pub_id=255979
    * update pubs set pub_isbn='#604' where pub_id=255979
    * update pubs set pub_price='$0.00' where pub_id=255979
    * update pubs set note_id='127941' where pub_id=255979
    * select publisher_id from publishers where publisher_name='Project Gutenberg';
    * update pubs set publisher_id='13741' where pub_id=255979
    * insert into pub_authors(pub_id, author_id) values('255979', '738'); --> -->
       
       
      AttributeError	Python 2.4.2: /usr/local/bin/python
      Tue Feb 12 14:11:52 2008

      A problem occurred in a Python script. Here is the sequence of function calls leading up to the error, in the order they occurred.
       /www/isfdb.tamu.edu/root/cgi-bin/mod/ca_new.cgi
        492         print "<hr>"
        493 
        494         Record = DoSubmission(db, submission)
        495 
        496         print "<hr>"
      Record undefined, DoSubmission = <function DoSubmission>, db = <_mysql.connection open to 'localhost' at 836c00c>, submission = 928315
       /www/isfdb.tamu.edu/root/cgi-bin/mod/ca_new.cgi in DoSubmission(db=<_mysql.connection open to 'localhost' at 836c00c>, submission=928315)
        371                         artists = doc.getElementsByTagName('Artist')
        372                         for artist in artists:
        373                                 data = XMLunescape(artist.firstChild.data.encode('iso-8859-1'))
        374                                 addPubArtist(data, Record)
        375 
      data = 'Edwin L. Arnold', global XMLunescape = <function XMLunescape>, artist = <DOM Element: Artist>, artist.firstChild = None, artist.firstChild.data undefined

      AttributeError: 'NoneType' object has no attribute 'data'
            args = ("'NoneType' object has no attribute 'data'",)

Something about the artist field? The submission is currently on hold. Ahasuerus 14:15, 12 Feb 2008 (CST)

I just tried re-approving it and received the same error. The semi-approval seems to create orphan Publication records with no Title record. Ahasuerus 00:42, 15 Feb 2008 (CST)

Python Error on Recent Integrations

It seems to be with Submission 929307 - no doubt it'll drop off shortly, so not a major worry. BLongley 15:27, 14 Feb 2008 (CST) Here's what I saw though:

 
UnboundLocalError	Python 2.4.2: /usr/local/bin/python
Thu Feb 14 15:23:46 2008

A problem occurred in a Python script. Here is the sequence of function calls leading up to the error, in the order they occurred.
 /www/isfdb.tamu.edu/root/cgi-bin/mod/recent.cgi
  138         color = 0
  139         while record:
  140                 PrintRecord(record, color)
  141                 color = color ^ 1
  142                 record = result.fetch_row()
PrintRecord = <function PrintRecord>, record = ((929307L, 'I', 21L, '<?xml version="1.0" encoding="iso-8859-1" ?>\n<Is.../AwardLevel>\n </AwardUpdate>\n</IsfdbSubmission>\n', datetime.datetime(2008, 2, 14, 12, 46, 18), datetime.datetime(2008, 2, 14, 12, 49, 19), 2L, 2L, None, 0L),), color = 0
 /www/isfdb.tamu.edu/root/cgi-bin/mod/recent.cgi in PrintRecord(record=((929307L, 'I', 21L, '<?xml version="1.0" encoding="iso-8859-1" ?>\n<Is.../AwardLevel>\n </AwardUpdate>\n</IsfdbSubmission>\n', datetime.datetime(2008, 2, 14, 12, 46, 18), datetime.datetime(2008, 2, 14, 12, 49, 19), 2L, 2L, None, 0L),), eccolor=0)
   86                         submitter = 'unknown'
   87         approver = SQLgetUserName(record[0][7])
   88         print '<td><a href="http://isfdb.org/wiki/index.php/User:%s">%s</a></td>' % (submitter, submitter)
   89         print '<td><a href="http://isfdb.org/wiki/index.php/User:%s">%s</a></td>' % (approver, approver)
   90         print "<td><i>%s</i></td>" % (subject[:40])
submitter undefined

UnboundLocalError: local variable 'submitter' referenced before assignment
      args = ("local variable 'submitter' referenced before assignment",)

Are you still seeing this? Works fine for me. Disturbing because I fixed this problem (displaying recent AwardUpdate and AwardDelete) early this morning, but the submission was over lunch, and no changes have been made since. Alvonruff 19:17, 14 Feb 2008 (CST)

Yes, I'm still seeing it, but only for that submission. Which is 3 pages back now so no worry really. BLongley 13:00, 15 Feb 2008 (CST)

Potential enhancements to Coverart pages

I know there are some potential modifications to Editor and Coverart processing but I would like to know if a potential variant record method is technically feasible and consistent with potential changes. The quick story - some editors would like to document the fact that cover art is for a particular magazine story. One method would be to modify the Coverart record which is system generated upon original entry of the magazine. Some of us oppose that. The first paragraph of this entry has the technical details for another methodology which also has at least one negative display aspect. Are there any technical aspects we need to keep in mind for either method before we discuss this issue into infinity?--swfritter 20:28, 14 Feb 2008 (CST)

Thanks for the new tools!

I've no idea what you intend them for, but I've done a little bit of regularisation, e.g. there's no reason to record "St Albans" against Panther books, even though City/Town of publisher might be useful for foreign titles. It's also shown up a few misuses (IMO) of the field for printing numbers, variant titles, etc. If you have a long-term plan for publishers and imprints feel free to state it - if not, I'll go add known ISBN ranges, dates the imprint was used for, when a publisher became an imprint, who owned what in what years, etc, to publisher pages. Eventually - at the moment I'm just looking at the sub-zero temperatures predicted here for the next few days and thinking about finally USING the work lap-top in the nice warm bedroom... BLongley 19:07, 16 Feb 2008 (CST)

Date Ranges?

When you have a chance, please take a look at ISFDB:Community Portal#"Unknown, but after X" dates. Would implementing date ranges, either a limited special purpose (point in time -> later) range, or a more general full-function date range (for publication dates, and possibly other dates such as birth and death dates) be excessivly hard to do? -DES ^Talk 12:45, 25 Feb 2008 (CST)

Another Python Error

I suspected it was because I did a Publication search for "47", and one of those publications has a Tag of "472005", presumably as it was published in 2005. This could of course be interpreted as a record number rather than a Tag. But the error seems to occur on the NEXT record? (68940 rather than 68749 - 103581 seems normal, for those that think "Патруль Звездных Волков" is normal. BLongley 16:03, 4 Mar 2008 (CST)

Beginning of the record 68940 is displayed OK, but then ctype is null:

mysql> select pub_ctype from pubs where pub_id = 68940;
+-----------+
| pub_ctype |
+-----------+
| NULL      | 
+-----------+
1 row in set (0.00 sec)

No conversion from NULL to string? --Roglo 16:27, 4 Mar 2008 (CST)

I should mention, perhaps, that I was trying a publisher search first: "http://www.isfdb.org/cgi-bin/publisheryear.cgi?3473+2005", and clicking on the ISBN resulted in "pl.cgi: No such publication tag exists." error. I only got the big purple error later. BLongley 16:34, 4 Mar 2008 (CST)

Python error when trying to display a magazine with duplicate entries

A New Magazine submission resulted in a messed up pub record with all Contents level Titles duplicated. It also errors out when you try to look at it, but you can pull it up in the edit form and the Remove Title form. Ahasuerus 18:04, 8 Mar 2008 (CST)

I've straightened it out. Ahasuerus' message above about "Remove Title" sent me in the right direction. After removing the content that had no author (Waiting), the Python error disappeared. Then all I had to do was remove the duplicates and merge the title records that were duplicated by the submission. Thanks. MHHutchins 18:20, 8 Mar 2008 (CST)

Newbery Medal winners

While obviously not an SF award, we have at least 9 winners of the Newbery Medal in the DB (although one is problably not SF) See hte list via this tag. Do we want to add this to our awards list? -DES ^Talk 10:58, 22 Mar 2008 (CDT)

Publisher Stuff

I see you're doing something interesting with "Top 100" at the moment, but when you get a chance can you look at my initial thoughts on the Publisher capabilities you added recently? If you can't read it all, then 1.8 and 8.1 are sort of possible Bug reports. BLongley 15:22, 27 Mar 2008 (CDT)

Nobody's really stated a long term goal for what we do with the tools you've given us, but I've fixed some obvious typos, standardized some names (and left other possibilities up for discussion) and we seem to have reduced the number of publishers from 9,820 to 9,467 - it's pretty slow though, so I (we?) might want some mass updates done. And of course there's lots of feature requests in there too, but feel free to ignore those unless you're thinking along the same lines anyway. BLongley 15:22, 27 Mar 2008 (CDT)

Wiki Links

I just tried to access the Wiki Biography & Bibliographic Comments pages from Philip José Farmers Author page to add this apparently useful site and we seem to have "incorrectly linked inter-language" problems. Is there a workaround for now, or will it be fixed on the next site, or should I just try to recall where bug reports go now? I'm not sure what the problem is as we cope with the "é" on the author page itself. (If that's the problem - bloody unicode again?) BLongley 15:17, 3 Apr 2008 (CDT)

Bloody unicode. Might have an impedance mismatch between MediaWiki, python, or MySQL. I'll poke around. Alvonruff 18:28, 3 Apr 2008 (CDT)

Firefox (at least my 2.0.0.13 version) has setting to switch between passing URLs as encoded on page (which is default) and UTF-8 (which seems to be expected by Wiki). Open new window; type in 'about:config' as URL; type in utf8 in Filter; double-click on netword.standard-url.encode-utf8 to change value to true. You don't have to restart Firefox. Now you should be able to click on '# Bibliographic Comments: Author:Philip José Farmer' and see Internal consistency pass done. Ahasuerus 21:36, 19 Jun 2006 (CDT). Obviously you have to double-click on netword.standard-url.encode-utf8 again to make links to Philip José Farmer work in ISFDB. Wiki page should still work, once you got there.

Wikipedia somehow converts URLs on the fly; they send new location:

$ wget -d 'http://en.wikipedia.org/wiki/Philip_Jos%E9_Farmer'
[...]
---request begin---
GET /wiki/Philip_Jos%E9_Farmer HTTP/1.0
User-Agent: Wget/1.10.2 (Red Hat modified)
Accept: */*
Host: en.wikipedia.org
Connection: Keep-Alive

---request end---
HTTP request sent, awaiting response... 
---response begin---
HTTP/1.0 301 Moved Permanently
Date: Sat, 05 Apr 2008 20:16:29 GMT
Server: Apache
X-Powered-By: PHP/5.2.1
Cache-Control: private, s-maxage=0, max-age=0, must-revalidate
Vary: Accept-Encoding,Cookie
X-Vary-Options: Accept-Encoding;list-contains=gzip,Cookie;string-contains=enwikiToken;string-contains=enwikiLoggedOut;string-contains=enwiki_session
Last-Modified: Sat, 05 Apr 2008 20:16:29 GMT
Location: http://en.wikipedia.org/wiki/Philip_Jos%C3%A9_Farmer
Content-Length: 0
Content-Type: text/html; charset=utf-8
X-Cache: MISS from sq28.wikimedia.org
X-Cache-Lookup: MISS from sq28.wikimedia.org:3128
X-Cache: MISS from knsq23.knams.wikimedia.org
X-Cache-Lookup: MISS from knsq23.knams.wikimedia.org:3128
X-Cache: MISS from knsq29.knams.wikimedia.org
X-Cache-Lookup: MISS from knsq29.knams.wikimedia.org:80
Via: 1.0 sq28.wikimedia.org:3128 (squid/2.6.STABLE18), 1.0 knsq23.knams.wikimedia.org:3128 (squid/2.6.STABLE18), 1.0 knsq29.knams.wikimedia.org:80 (squid/2.6.STABLE18)
Connection: keep-alive

---response end---
301 Moved Permanently
Registered socket 3 for persistent reuse.
Location: http://en.wikipedia.org/wiki/Philip_Jos%C3%A9_Farmer [following]
[...]

--Roglo 15:25, 5 Apr 2008 (CDT)

Hm, however author's summary doesn't show that the page exists. --Roglo 15:29, 5 Apr 2008 (CDT)

Publisher field?

Is the Publisher field supposed to be "deletable from the pub record" after the last round of changes? I am trying to get rid of "Arab Record" in this Arabic translation of Andre Norton's Shadow Hawk, but it won't go away. Ahasuerus 00:08, 9 Apr 2008 (CDT)

.htaccess to rewrite isfdb.org

Hi Al - can you put the .htaccess rewrite back in for www.isfdb.org? TIA. Marc Kupper (talk) 07:54, 14 April 2008 (UTC)

.htaccess isn't working as expected, which is why this is taking a while... Alvonruff 15:35, 15 April 2008 (UTC)

yet another minor display bug

I mention this hoping it's a quick & easy fix:

In pub, or at least magazine, listings, it seems that if the author's name is a pseudonym, the display of any series gets lost. For example, consider this pub, specifically the editorial, "A Place for the Subconscious". It's by John W. Campbell, Jr., as by John W. Campbell. The series [Editorial (Analog)] isn't displayed. This is probably because it's the parent (JWCJr) that actually has the series in the title record. But if you click on the editorial link, the child (no Jr) record shows the series as though it were in that title, & of course the parent is being read to get the "as by". Don't know how easy it would be to fix, though. Thanks Dave (davecat) 14:18, 15 April 2008 (UTC)

Python error during make varient

 /var/www/cgi-bin/edit/submitmkvar1.cgi in () 
   48                 sys.exit(0)
 
   49 
 
   50         if int(title_id) == int(parent_id):
 
   51                 print '<div id="WarningBox">'
 
   52                 print "<h3>Error: Title record can not be a variant to itself.</h3>"
 
builtin int = <type 'int'>, title_id = '107917', parent_id = '?575173' 

<type 'exceptions.ValueError'>: invalid literal for int() with base 10: '?575173' 

The above happens if a user is trying to make a title a varient of an existing record, and copies one character too many from the address bar of the browser window where the target record is displayed, thus prefixing the record number with a question mark (?). Apparently no harm is done, you can jsut hit back and enter the correct number, but the error message is confusing at best. -DES ^Talk 01:32, 4 May 2008 (UTC)

Leading/trailing spaces in tags?

I approved and massaged a non-genre Dissembler submission a few minutes ago and then noticed that we seem to have two identical tags ("strong female characters") on Jennifer Roberson's page. A closer look at the data revealed that tag 3131 has a leading space. Do we want to strip leading and trailing spaces at file time? Ahasuerus 03:35, 8 May 2008 (UTC)

Probably. Sorry about the dissembler leftover - I'm methodically going through Locus Books Received, but needed some example XML for the Web API article, and didn't go back to clean up that submissions. Alvonruff 11:26, 8 May 2008 (UTC)

No worries, it has now found a loving home in the Non-genre section of Jennifer Roberson's bibliography :-) Ahasuerus 14:38, 8 May 2008 (UTC)

Shortfiction-only series

I noticed two commented-out lines in ea.cgi (aka biblio/ea.py)

 #bib.loadShortSeriesData() 
 #bib.printShortSeries() 

One way to fix the out-of-chronology display on Summary bibliography would be to use again these functions to get 'short fiction only series' section at the beginning of short fiction listing (and fix them so that they won't display an empty series where there is a book-size work in the series). These would cause the series to 'jump' up the page when a collection or a novel are published.

Another way would be to add or titles.title_ttype='SHORTFICTION' to the SQLgetNCSeriesData(authorec). These would add short fiction only series to the Fiction Series. Is there a reason not to do it? (I'm still trying to work out the way series are displayed, so there may be side effects I don't see yet.) --Roglo 11:05, 10 May 2008 (UTC)

I added some notes about the fiction series display on Summary bibliography. --Roglo 14:32, 10 May 2008 (UTC)

Python error on advanced search

The following occured whil trying to search for titles containing "whispers" but not "cover" -DES ^Talk 20:25, 10 May 2008 (UTC)

QUERY: select titles.* from titles where titles.title_title like '%Whispers%' ANDNOT titles.title_title like '%cover%' order by titles.title_title limit 100 --> -->


<class '_mysql_exceptions.ProgrammingError'> Python 2.5: /usr/bin/python Sat May 10 15:22:14 2008

A problem occurred in a Python script. Here is the sequence of function calls leading up to the error, in the order they occurred.

/var/www/cgi-bin/edit/tp_search.cgi in () 
 174         db = dbConnect()

 175         db.select_db(DBASE)

 176         db.query(query)

 177         result = db.store_result()

 178         num = result.num_rows()

<class '_mysql_exceptions.ProgrammingError'>: (1064, "You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'ANDNOT titles.title_title like '%cover%' order by titles.title_title limit 100' at line 1") —The preceding unsigned comment was added by DESiegel60 (talk • contribs) .

There are so many known issues with Advanced Search that I believe Al is planning a complete overhaul in the foreseeable future. Ahasuerus 23:15, 10 May 2008 (UTC)

I see. Thanks. -DES ^Talk 00:07, 11 May 2008 (UTC)

Image Uploading

Al, I've tested the waters by uploading the cover of the most recent Asimov's, trying to keep the same file naming scheme that you've started, and the same dimensions as well. It's looks pretty good, so I think I did it correctly. Let me know if there's a problem. Thanks. MHHutchins 02:10, 14 May 2008 (UTC)

Looks good. Alvonruff 02:15, 14 May 2008 (UTC)

I got the first problem then? "Error creating thumbnail: /var/www/html/wiki//bin/ulimit4.sh: line 4: /usr/local/bin/convert: No such file or directory". It's the tallest image I've tried so far so that might be the cause (1912 image at latest, apparently, so hopefully well out of copyright and therefore not subject to 500 pixel rules?). Image:Frontispiece.jpg BLongley 20:13, 14 May 2008 (UTC)

Uploads going well so far otherwise, although looking at the samples so far (not sure how to Wiki-Link to it, but this works for now) makes me realise we will need naming standards fast, and I've already got two categories DES hasn't covered. Off to try the templates. BLongley 20:13, 14 May 2008 (UTC)

I thought that the 500 pixel standard was at least in part to reduce server & backup load. As to tempaltes, what additional cases are needed? I see logos and signature blowups in your uplaods. -DES ^Talk 14:50, 15 May 2008 (UTC)

The only 500 pixel limit I recall is Marc Kupper mentioning that's what Amazon resize to: but I'll take guidance on server and backup stresses. The new daily backups (just tested, seem fine so far) look like we're avoiding backup stress with partial backups though? BLongley 19:02, 15 May 2008 (UTC)

As to templates - well, I was hoping for ones that would allow positioning or resizing of images on Wiki pages, or automatically including title/artist/date/etc. : i.e. for USE of the image rather than ABOUT the image. But I'm not sure what CAN be done with templates so I'm open to suggestions. BLongley 19:02, 15 May 2008 (UTC)

Images can be displayed on wiki pages, and can be positioned and sized there. Existing wiki syntax will do most of that, but I can easily create templates to standardize and simplify doing that. I wasn't aware that there were plans to display significant numbers of images on wiki pages. In what circumstances do you plan such displays? i can probably create better templates with more info about their purpose or purposes. -DES ^Talk 19:54, 15 May 2008 (UTC)

The plans for the use of the Wiki are always a little vague as we're never quite clear how much Al is intending to put into the database side instead. I know with the latest publisher changes that some of my activity on the Wiki side is now redundant and/or needs to be moved: I don't need to add Wikipedia links on the Wiki Publisher pages now they're recorded in the database. And some of my shorter Wiki Publisher entries could have been done in the new Publisher Notes. I see no signs of publisher/imprint hierarchies in the database yet though, so those I won't move yet, and am trying to improve links for those in the Wiki. (As you may have noticed from my previous "How do I do X?" questions.) On the topic of images (which I will try to stay on), I've always been quite glad of the ability to add coverart images to pubs, and been a bit annoyed at the use of unstable URLs. What I would like to add here on ISFDB are all the things I've had to do with workarounds: like hosting my own images to ask "does the signature look like this?" (here) or ask should "THIS cover be considered the same as THAT cover?" like here. Those are examples of where I would like to upload sample signatures, and where I would like to have better control of how to USE artwork so it doesn't blast out at people. The other immediate thought (from working on Publishers recently) is that a change in a publisher's logo often dates a particular publication even if someone hasn't entered details from the copyright page, or if the copyright page is missing, or even if the book predates such niceties. "A picture is worth a thousand words" isn't always true - think of toilet signs where one picture effectively says "Men" and the other says "Women" and that's all - and I own original art where the picture CAUSES a thousand words of questions and explains nothing - but I think we can usefully use images more than we currently do. BLongley 22:34, 15 May 2008 (UTC)

As to automatically picking up info, I'm not aware of any way for the wiki page to pick up and use info from an ISFDB data record, but there may be one.-DES ^Talk 19:54, 15 May 2008 (UTC)

Al is working on Web APIs to provide ISFDB data, I'm not sure if Wikimedia could use them. There's a lot of Web 2.0 buzz about Web Service providers and consumers but I've no idea if they're going to work together anytime soon, and I personally wouldn't worry about it unless we can see a major bonus. BLongley 22:50, 15 May 2008 (UTC)

However, if we are to host and display book cover images, most of which are under copyright and which we can therefor only use under "Fair use" I think we need something like the templates I designed. Perhaps the amount of info recorded can be cut down, but i think at least some of it ought to be there. That is, ultimately, a policy decision -- I modeled my templates on the ones currently used for book covers on Wikipedia -- in fact I blatantly copied colors and wording to a significant degree -- but if we jointly decide (or if Al decrees) otherwise, I'll be happy to reword the templates to whatever standard we think is best.

My (possibly over-simplistic) suggestion would be to restrict ourselves to having limits that are ALL Fair-Use and have ONE justification for each type of use. I know that I'm not going to add 2,500 coverart images HERE if I have to justify them all individually. But I can support further justifications for other usages like the logos and artist sigs and any other uses people want. BLongley 22:50, 15 May 2008 (UTC)

I fear it is slightly over simplistic. For oen thing, some iamges will be PD, and we really ought to indicate which are which. But the most important thing is that valid fair use generally requires proper attribution and credit. The point of the template was so that you don't have to justify each iamge -- the justification is bolierplate that is repeated automatically -- you only need to provide identification and attribution info, Pershps some of the info in my first draft tempaltes could be dropped to make the process easier. I'm already using more bolierplate and less individual justification than Wikipedia does, because the situations for our iamges will be less varied. -DES ^Talk 05:28, 16 May 2008 (UTC)

As for size limits, there was some mention in the dialog on the Community portal, i think -- not being involved with the server side of things myself, i disclaim any responsibility for saying what would or would not affect server space or backup times. On legal issues, using a version that is at a reduced resolution helps with the "effect on the market" prong of the fair use test, but there is no absolute standard that images use under fair use must be reduced to a particular size, or percentage, or indeed at all. If you can say that the reduction is enough that the image could not be printed out at the original size and quality, that is all that matters for fair use, and even that is not required. -DES ^Talk 19:54, 15 May 2008 (UTC)

Misattributed submissions?

We seem to have at least one mis-attributed submission as discussed here. I hope we don't get other fields mixed up or else we may end up with The War of the Worlds by Robert A. Heinlein! Ahasuerus 03:14, 16 May 2008 (UTC)

Verified Pubs, F&SF 2007

Added cover images to your verfied pubs Magazine of F&SF, 2007 - 01, 02--Rkihara 17:14, 23 May 2008 (UTC)

Two Reviews in one magazine for same book under two pseudonyms

One of your verified pubs Analog July 1965 has two reviews for "The Issue at Hand" one by James Blish and one by William Atheling. Analog May 1971 only has one, but it's by William Atheling, Jr - where we have the pseudonym in the database but not the title under that pseudonym. Given your comment "the authors aren't of much value if the names are mangled", what would you suggest for the way forward? BLongley 18:14, 25 May 2008 (UTC)

Ouch. It looks like I probably added the second one (at a guess, from a note I left earlier on this (Al's) page). Looking at the pub (in the ISFDB data), I see that the first review lists before the entry for "The Reference Library", which suggests that I may just have not seen it. <sigh> I'm pretty sure that I entered the one that says it's by Blish. Since this isn't a review with a normal heading (buried in Miller's initial essay), there's a slight ambiguity; but I'd understand it to say he's reviewing a book by Blish containing (but not limited to) columns originally published as by "William Atheling, Jr.". So my own take would be that the "Atheling" entry is just wrong here. (But I won't nuke that unless someone asks me to; I'm just putting in my $0.02. If the decision is to remove one of them, I can certainly remove whichever one is chosen, of course.) -- Dave (davecat) 20:51, 25 May 2008 (UTC)

I Own a copy of The issue at Hand. I think, but would have to check, that all its contetns were originally published under the name of "William Atheling, Jr.". However, the colelction had Blish's name on it, and soem of the contents had i think been revised for book publication. Some of the essays in the later The Tale that Wags the God were, i think originally published under Blish's own name, and some under "William Atheling, Jr.". -DES ^Talk 20:59, 25 May 2008 (UTC)

Oh, there's no blame here, Davecat, just a question: this sort of review linking is only hours old and is throwing up new questions all the time. Whether we want to link to an existing title/author only, or use the reviewee and canonical title (if we have the title and author separately) is something to ponder. DES - for links to the title I don't think the content attributions matter, but if it's a "James Blish writing as William Atheling" (Jr. ?) then we'd want to get that right. Although it's perfectly possible we have all 3 author variants for the pub(s), and all 3 as reviewees too, and not necessarily linked the same way... BLongley 22:22, 25 May 2008 (UTC)

The Issue at Hand, at least in my copy, is credited as a book to "Atheling, jr" but said to be "edited and with an introduction" by Blish, and in the intro Blish makes it clear that he is Atheling, and says that most but not all the essays originally appeared over the Atheling signature. I have entered the essay titles, and started to enter the reviews, from my copy. (By the way this includes the essay where "Atheling" reviewd, at soem length and fairly bitingly, the magazine version of "A Case of Conscience" by Blish). -DES ^Talk 15:13, 26 May 2008 (UTC)

Use of uploaded images

Was the uploading facility turned on only for testing purposes? Or is it now ok to use images in our wiki as cover images on pubs in the live db? -DES ^Talk 05:36, 26 May 2008 (UTC)

Python error making submission containing bad HTML

I failed to put in the clsoe quote on an HTMLlink tag, and got:

--------------------------------------------------------------------------------
<?xml version="1.0" encoding="iso-8859-1" ?>
<IsfdbSubmission>
<NewPub>
<Submitter>DESiegel60</Submitter>
<Subject>Uncle Silas: A Tale of Bartram-Haugh</Subject>
<Parent>885686</Parent>
<Title>Uncle Silas: A Tale of Bartram-Haugh</Title>
<Year>1926-00-00</Year>
<Publisher>Oxford University Press</Publisher>
<Pages>441</Pages>
<Binding>hc</Binding>
<PubType>NOVEL</PubType>
<Isbn>#306 (World's classics)</Isbn>
<Note><a HREF="http://worldcat.org/oclc/56253138>OCLC: 56253138</a>.
 
Entry based on data from OCLC/Worldcat, see link. (DES May 2008)</Note>
<Authors>
<Author>Joseph Sheridan Le Fanu</Author>
</Authors>
<Content>
</Content>
</NewPub>
</IsfdbSubmission>
--> --> 


<class '_mysql_exceptions.ProgrammingError'> Python 2.5: /usr/bin/python
Tue May 27 18:05:57 2008 

/var/www/cgi-bin/edit/submitpub.cgi in () 
 210         else:

 211                 update = "insert into submissions(sub_state, sub_type, sub_data, sub_time, sub_submitter) values('N', %d, '%s', NOW(), %d)" % (MOD_PUB_UPDATE, update_string, submitter_id)

 212         db.query(update)

 213 

 214         PrintPostSearch(0, 0, 0, 0, 0)

db = <_mysql.connection open to 'localhost' at 945ae1c>, db.query = <built-in method query of Connection object at 0x945ae1c>, update = "insert into submissions(sub_state, sub_type, sub...\\n </NewPub>\n</IsfdbSubmission>\n', NOW(), 13372)" 

<class '_mysql_exceptions.ProgrammingError'>: (1064, "You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 's classics)</Isbn>\n <Note><a HREF="http://worldcat.org/oclc/56253138&' at line 1") 

After trying several times to re-enter with the link syntqax corrected, and getting the same error, i re-ented without the link, and still got an error, as follows:

<?xml version="1.0" encoding="iso-8859-1" ?>
<IsfdbSubmission>
<NewPub>
<Submitter>DESiegel60</Submitter>
<Subject>Uncle Silas: A Tale of Bartram-Haugh</Subject>
<Parent>885686</Parent>
<Title>Uncle Silas: A Tale of Bartram-Haugh</Title>
<Year>1926-00-00</Year>
<Publisher>Oxford University Press</Publisher>
<Pages>441</Pages>
<Binding>hc</Binding>
<PubType>NOVEL</PubType>
<Isbn>#306 (World's classics)</Isbn>
<Note>OCLC: 56253138 
 

Entry based on data from OCLC/Worldcat. (DES May 2008)</Note>
<Authors>
<Author>Joseph Sheridan Le Fanu</Author>
</Authors>
<Content>
</Content>
</NewPub>
</IsfdbSubmission>
--> --> 


<class '_mysql_exceptions.ProgrammingError'> Python 2.5: /usr/bin/python
Wed May 28 09:46:11 2008 

A problem occurred in a Python script. Here is the sequence of function calls leading up to the error, in the order they occurred.

/var/www/cgi-bin/edit/submitpub.cgi in () 
 210         else:

 211                 update = "insert into submissions(sub_state, sub_type, sub_data, sub_time, sub_submitter) values('N', %d, '%s', NOW(), %d)" % (MOD_PUB_UPDATE, update_string, submitter_id)

 212         db.query(update)

 213 

 214         PrintPostSearch(0, 0, 0, 0, 0)

db = <_mysql.connection open to 'localhost' at 8223e1c>, db.query = <built-in method query of Connection object at 0x8223e1c>, update = "insert into submissions(sub_state, sub_type, sub...\\n </NewPub>\n</IsfdbSubmission>\n', NOW(), 13372)" 

<class '_mysql_exceptions.ProgrammingError'>: (1064, "You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 's classics)</Isbn>\n <Note>OCLC: 56253138 <br>\r\n\r\nEntry based on data fr' at line 1") 

It loooks to me as if the problem was not due to the link at all. -DES ^Talk 14:50, 28 May 2008 (UTC)

I'll have a look this evening, but I'd guess that the problem is probably the single quote in the ISBN field. Alvonruff 16:59, 28 May 2008 (UTC)

That makes sense. When i retried it with that in the notes it went through OK. I was thinking that it might be the parens or the space in that field, which is why i re-tried. -DES ^Talk 18:59, 28 May 2008 (UTC)

Approval warning

See this test submisison. The pub is marked as being verified agaisnt Locus, but the yellow warning alerts that it has been primary verified. Either the warnin text should change, or the code the triggers it. Ideally, it would show what source a pub had been verified agaisnt, and by what editor but that may be too much for a bug fix. -DES ^Talk 22:01, 28 May 2008 (UTC)

The first part is fixed - the warning only shows up for primary verified. The other ideas are good, but I can't implement them this morning. Alvonruff 11:45, 29 May 2008 (UTC)

I figured as much, I think there is a feature request already on the list for them. Thanks. -DES ^Talk 12:54, 29 May 2008 (UTC)

Okay - the part about displaying the verifier when a verification warning comes up is now fixed and online. Alvonruff 23:44, 30 May 2008 (UTC)

Wonderful!! -DES ^Talk 23:55, 30 May 2008 (UTC)

Out of the Wild and Dissembler?

Just wondering why Dissembler caught both editions of Sarah Beth Durst's Into the Wild, but missed the sequel, Out of the Wild, which I added manually a few minutes ago? Amazon.com seems to have it labeled correctly. Ahasuerus 03:21, 1 June 2008 (UTC)

Incorrect award link

The 1977 Locus Award for The Best of John W. Campbell is linked to the wrong title record. It should be the 1976 collection, not the 1973 UK collection with the same title. Thanks. MHHutchins 00:49, 2 June 2008 (UTC)

Fixed. Alvonruff 02:08, 2 June 2008 (UTC)

SVG plugins

There are several plugins available for IE that support SVG, such as the one at the http://www.examotion.com/ site. I downloaded and installed that one and could see your just-added svg charts with no problem. Or SVGs can be created off-line and uploaded as images, and if SVG support has been installed in our copy of the wiki software, they will be rendered as PNGs, accodign to the notes at Mediawiki. -DES ^Talk 16:46, 6 June 2008 (UTC)

Automatic Attribution for Galactic Central

At this time I've probably linked to ~500 cover images from Galactic Central. I've added a credit to the bottom of every magazine publication page that uses these images, but it would be good to automatically add a credit under the image, as is now done with images from Visco.--Rkihara 17:54, 6 June 2008 (UTC)

Import / Export

Al there is no Export or Import content "Button", I had to use the "Enter" key. This may be a bit confusing for some editors.  :-)Kraang 03:47, 15 June 2008 (UTC)

There is a "Submit" button at the bottom of the page, but you may have to scroll down a bit. BTW, Al, Ernest Bramah has just called and asked me to thank you :-) Ahasuerus 03:54, 15 June 2008 (UTC)

On the first page where you enter the tag or pub # there is no "Submit" button on this page for me at least. The page it generates with the cloned info. has the "Submit" button at the bottom. My page shows no "errors", would the problem come from me using XP Windows?Kraang 12:28, 15 June 2008 (UTC)

This problem only occurred on Windows Explorer. Now Fixed - button appears. Alvonruff 12:58, 15 June 2008 (UTC)

I see it now, Stupid Windows! Thanks!Kraang 15:08, 15 June 2008 (UTC)

Error dusplaying image categories

When a wiki category that includes image fiels is displayed, the following error msg appears in plce of each image thumbnail:

thumbnail: /var/www/html/wiki//bin/ulimit4.sh: line 4: /usr/local/bin/convert: No such file or directory

I looks to me as if some part of the path used by default to strore dynamically create thumbnail images does not exist. -DES ^Talk 14:09, 16 June 2008 (UTC) For examples, see Category:Test Cat 1 and Category:Fair use images. -DES ^Talk 14:10, 16 June 2008 (UTC)

This requires installing ImageMagick, which has dependencies requiring the installation of 11 other packages, each of which has some unknown number of additional dependencies, and so on. Which translates to: I haven't had a few hours to burn on this one yet. Alvonruff 23:24, 16 June 2008 (UTC)

Oh really? That's a lot more work than I would have guessed. Sorry. Can we turn off thumbnaling in cats and jsut display the file/page name until/unless this is installed, or is that a lot of work too? -DES ^Talk 23:51, 16 June 2008 (UTC)

According to this media wiki page, setting $wgCategoryMagicGallery to off should avoid the error msgs. -DES ^Talk 23:59, 16 June 2008 (UTC)

Well - it's interesting. Not what I expected on your test page (prints an h3 'I'). Alvonruff 00:56, 17 June 2008 (UTC)

That is normal. I set up the test in a very quick and dirty way, and did not provide a pipe alias for the apges. Thus each of them is indexed under therei full page name, which begins "Image:..." so all the pages are sorted under "I" for "Image". Look at Category:Fair use images. for a better view. I'll change the links that feed Category:Test Cat 1 to make it look more reasonable. -DES ^Talk 18:23, 17 June 2008 (UTC)

Have a look at Category:Test Cat 1 now. -DES ^Talk 18:27, 17 June 2008 (UTC)

That looks a lot more usable to me! Not sure how well it would work with a few thousand images though. But we definitely don't want thumbnails until there's a manageable number of selections to choose from. BLongley 21:25, 17 June 2008 (UTC)

You may not be aware that wiki categories always limit any given display page to no more than 200 items, with links to "next" and "previous" pages as needed. -DES ^Talk 23:51, 17 June 2008 (UTC)

Web API and bad submissions

I was playing with the Web API yesterday night and one of my submissions, 993590, was accepted, but then it caused problems in the submission queue and I had to reject it.

If you look at the XML dump that I linked above, the problem was with the ISBN field or rather with the fact that the first ISBN tag was malformed. You'd think that the Web API would have rejected a submission with a bad tag, but it was accepted twice. As far as I was able to determine, all tags were properly formed in the "problem" submission except that there was no space before the <Isbn> tag, which apparently confused the Web API. Is this enough information to diagnose the problem or should I experiment with other submission permutations tonight? Thanks! Ahasuerus 20:30, 17 June 2008 (UTC)

In dump linked to above, the first few tags are as follows:

<Submitter>Ahasuerus</Submitter>Isbn>1575661292</Isbn><Subject>Lady of the Glen: A Novel of 17th-Century Scotland and the Massacre of Glencoe</Subject>

The Isbn tag is lacking the first angle bracket (<), so I presume it would not have been recognized as a tag at all. I don't know if this was also true in the source. -DES ^Talk 20:40, 17 June 2008 (UTC)

Al's code checks for malformed XML in submissions and rejects them out of hand -- as I know all too well after sending more than my fair share of bad submissions yesterday :) The fact that this submission got through and the fact that (as far as I could tell) it had the first angle bracket on my side suggests that there is something wrong when processing Isbn tags. Also, when there was a space before (<Isbn), the submission was properly created, so it seems to point to a problem with the Web API. As an aside, it would be great if the Web API generated more specific error messages when rejecting submissions, but I assume that it may take additional work. Ahasuerus 20:52, 17 June 2008 (UTC)

I DREAM of meaningful error messages! At work, we've just given up on Internet Explorer 6 not working with the latest upgrade to our Web-App. After two months of testing, we've still not got a meaningful error message from it, it just freezes: and every other browser works fine. So IE6 goes into the sin-bin along with IE4 and Netscape 2 and other dinosaurs and we'll go back to actually enhancing our system rather than waste time figuring out what broke IE6. Of course, this takes us back into dealing with systems sending us XML that doesn't conform to the XSD or DTD they claim to comply with (the validation performance is so bad they switch it off, and don't even turn it back on for testing of later versions it seems), but I'm getting used to rewriting XSDs to match reality. I may even try the ISFDB Web API soon, but I'll let you Alpha testers do a bit more first. ;-) BLongley 21:39, 17 June 2008 (UTC)

This one is my error. The code that removes the license key (so others can't see it in the submission queue) had an off by one error, removing an extra character. My submissions have newlines and spaces, so the missing character wasn't missed. FIXED. Alvonruff 23:40, 17 June 2008 (UTC)

Excellent! Fixing the multiverse one bug at a time :) Ahasuerus 23:54, 17 June 2008 (UTC)

Merged Story Indicator

Love all the new stuff, especially the linked reviews and the efficient manner in which they were attached in batch mode but there is one subject which has created a constant, continuing headache for us all - merged titles which are changed universally when changing a single entry. It would really be a help if there were an indicator in Publication Editor mode and, perhaps more importantly, on the submission screen indicating that editing the title information (title, author, entry type, or length) or approving the submission will result in changes in multiple publications. An asterisk next to the title line with a note at the bottom of the screen might be an idea. Thanks.--swfritter 17:30, 19 June 2008 (UTC)

I agree. Occasionally this is desired, but all to often it is an error, with potentially far-reaching consequences. -DES ^Talk 17:35, 19 June 2008 (UTC)

WorldCat and legal issues

Al, you may want to take a look at this discussion if you haven't done so already. I doubt a boring copyright case would trump your grandfather's exploits during the Prohibition, but still... Ahasuerus 03:15, 26 June 2008 (UTC)

recent changes to Template:A

I noted you were using Template:A on your user page just now. You may not have noticed, i modifed this template so the user no longer needs to supply underscores for spaces, the template now makes that conversion automatically. -DES ^Talk 13:54, 3 July 2008 (UTC)

Thanks! Alvonruff 12:07, 5 July 2008 (UTC)

Erroring submissions

There's nasty Python errors on this for instance. Something to do with changes from two authors (both pseudonyms) to one, I guess. BLongley 20:43, 4 July 2008 (UTC)

This is another example of deleting an author with the space key. The space allows the author string to be accepted, but the leading/trailing spaces are deleted later, resulting in a NULL author. FIXED. Alvonruff 12:06, 5 July 2008 (UTC)

Carrion Comfort and Phases of Gravity

Another question about the Warner Books edition of Carrion Comfort, which you re-verified last month. Is it a true first printing, i.e. does the number line under the words "First Warner Books Printing: October 1990" go from 10 to 1, or does it stop at a higher number? For example, mine stops at 4, which makes it a fourth printing. Ahasuerus 00:44, 7 July 2008 (UTC)

It goes from 10 to 1, and states above the number line: "First Warner Books Printing: October, 1990" Alvonruff 00:57, 8 July 2008 (UTC)

Ah, so it's a true first printing, nice! Mine also says "First Warner Books Printing: October, 1990", but the number line stops at 4. I am sure publishers have all kinds of excellent reasons to keep the date of the first printing on the copyright page of subsequent printings, but it can easily confuse the rest of us :( Ahasuerus 02:35, 8 July 2008 (UTC)

P.S. After entering and verifying my fourth printing version of Carrion Comfort, I realized that I had created a duplicate of my previously verified pub and deleted it. However, my "Top Verifier" score remained the same, which suggests that deleting a Verified publication has no effect on that counter. Is that by design or do we want to decrement the verifier's counter when one of his verified publications is deleted? Ahasuerus 00:53, 7 July 2008 (UTC)

It's a product of evolution, not intelligent design... I'll patch it up (and some others) when I put in support for multiple verifications. Alvonruff 00:57, 8 July 2008 (UTC)

Multiple verification = Very Good!! Ahasuerus 02:35, 8 July 2008 (UTC)

P.P.S. Re: Phases of Gravity, does your copy have a new afterword by Simmons? Also, the only date stated in my third printing of the Bantam Spectra edition is "May 1989", i.e. the date of the first printing of the Bantam Spectra edition. Did you take the date, 1991-03-00, from the Locus Index and, if so, does it mean that your copy is a second printing as per their description? Ahasuerus 01:05, 7 July 2008 (UTC)

Mine is the third printing as well. This was verified 4 days after the verification feature went online, and I don't think we had settled into tracking different printings with the same ISBN yet. Feel free to modify at will to bring up to current standards. Alvonruff 00:57, 8 July 2008 (UTC)

Ah, I see! OK, I'll reshuffle them in a moment, thanks! Ahasuerus 02:35, 8 July 2008 (UTC)

Chap(ter)books and some basic design issues

Al, could you please take a look at my summary of this discussion of Chap(ter)books and associated fundamental database design issues when you have a chance? A number of bandaid solutions have been proposed, but we probably need to make some basic decisions about the direction that we want to take the application in before we do anything. Thanks! Ahasuerus 23:27, 23 July 2008 (UTC)

Change to your verified pub Analog Science Fiction and Fact, January/February 2007

I added cover scans to this issue. Tpi 09:09, 24 July 2008 (UTC)

Change to your verified pub Analog Science Fiction and Fact, April 2007

I added cover scan to this issue. Tpi 10:15, 25 July 2008 (UTC)

Asimov's, September 2007

I think there's a stray editor title in there? BLongley 20:05, 25 July 2008 (UTC)

2007 	EDITOR 	 	Reflections: Saddam Wasn't the Worst 	Robert Silverberg
2007 	ESSAY 		Reflections: Saddam Wasn't the Worst 	Robert Silverberg

FIXED. Alvonruff 22:55, 25 July 2008 (UTC)

Cheers! I get less and less confident about magazines as time goes on, but big yellow warnings still help. BLongley 23:04, 25 July 2008 (UTC)

Problem with editing the verification table

There seems to be a problem with editing the verification list (http://www.isfdb.org/cgi-bin/mod/editrefs.cgi). I was attempting to add Bleiler78, in accord with ISFDB:Moderator noticeboard#Adding Bleiler's Checklist to the Verification Source. I could not make a new line stick. there were error msgs about missing UIRLs for earlier entries, so i created and entered relevant Reference pages for thsoe that did not have outside URLs, and converted old tamu.edu URLs to www.isfdb.org. Erro msgas stopped, but still no luck. Also "Primary (Transient)" is ID #17, and there are no IDs 14, 15, or 16 listed. I can't get a URL for 17 to stick, nor an entry for 18 at all. And by the way, what is the functiion of the URLs, nothing seems to display or use them? Also, what is the function of the Pub Id field? None of the lines currently have a non-blank Pub Id. Help please? -DES ^Talk 05:31, 17 August 2008 (UTC)

Darwin's Radio

Bluesman has submitted an update to your verified Darwin's Radio to change

• Catalog # from #02066 to #03066
• Note:
  SFBC #02066. No price and ISBN.
  SFBC #03066. ISBN same as publisher's edition.

I'd approve it but am puzzled you could not find an ISBN and Bluesman seems to have done so.

I suspect the best place to reply for this issue is at User talk:Bluesman#Darwin's_Radio Marc Kupper (talk) 05:14, 5 September 2008 (UTC)

I think the way SFBC books are entered has evolved over time to include the ISBN, but I've personally only been entering SFBC catalog numbers, so I don't think the original note is mine. Nonetheless, it's quite clear that the book is catalog #03066 and DOES have an ISBN - there's an ISBN box on the back cover. So, can't explain the record's current state, but the change sounds like a good one. Alvonruff 11:38, 9 September 2008 (UTC)

Problem with XML output from wiki

I had some problems using the AutoWikiBrowser tool, which I reported at this wikipedia page. The response was: "Your MW API install is borked: it prepends newline to XML output, scaring the .NET's XML parser to death." I don't know what if anythign you can do about this, nor how hard it might be. -DES ^Talk 01:19, 8 September 2008 (UTC)

Hmmm. Seeing as how I know nothing about AutoWikiBrowser, MW API, or borking, should be fairly trivial, no? Can't promise any quick responses, but I'll look at it as time permits. Alvonruff 11:33, 9 September 2008 (UTC)

Figured I'd report it in case the newline was a trivial fix, but this is not vital. AWB is a tool, not an essential -- it basically allows doing wiki edits in a semi-automated form -- and I can still use it as things are, there is merely one convenient feature that doesn't work. If you have limited time, i'd frankly rahter you concentreated on the verification table isue mentiuons a couple of sections up. -DES ^Talk 00:23, 10 September 2008 (UTC)

Dissembler omission?

I read in Lunnaya Pravda... er, I mean I have just read a review of Ekaterina Sedia's The Alchemy Of Stone in Locus. Amazon.com has it listed under both Science Fiction and Fantasy, so I wonder why Dissembler missed it earlier this year? Ahasuerus 22:16, 1 October 2008 (UTC)

Dissembler also missed all of Piers Anthony's ChroMagic books in 2003-2008. Ahasuerus 16:50, 28 October 2008 (UTC)

Oh, I see, no browse nodes! That's very sloppy of Amazon, but let me see if I can come up with a sub-400 Power search on Subjects and years... Ahasuerus 17:10, 31 October 2008 (UTC)

OK, Powersearch should work for our purposes if you use a subject line like "fantasy" in conjunction with "pubdate:during MM-YYYY" since it's very unlikely that you will get more than 4,000 hits per month. However, Amazon's US (and only US, apparently) Webservice is currently broken: all pages higher than 100 return the same text, so you are effectively limited to 1,000 hits and that's not enough for traditionally high volume months like September. I suppose you could (ab)use "binding" to get a smaller subset, but the list of allowed values is not published and even if was, plugging "Leather Bound", "Library Binding" and "Loose Leaf" in the query is kludgey. Oh well, at least it should get us 90%+ of the missing pubs.

Will you have the time/inclination to update Dissembler in the foreseeable future or should I throw something together on my end and use the Web interface to submit the missing pubs? Recreating the anti-comic/RPG logic would be the only really painful part, I think. Ahasuerus 00:16, 1 November 2008 (UTC)

Spam Filter

Please take a look at User talk:DESiegel60#Spam filtered. Thank you. -DES ^Talk 19:12, 2 October 2008 (UTC)

Oddity with import

I had added two essay that a reliable secondary source (the author's web site) told me were present to the 1st printing of a book (The Seven Altars of Dusarra by Lawrence Watt-Evans actually.) The same source told me those essays were also included in subsequent printings (and other editions) for which we already had records, but records that did not include the essays. So i used the import function to import the content from the first printing to the second. It copied the records for the essays, alright, but it also imported a second copy of the record for the Novel. And remove titles won't remove the "main" title, the one that is the title reference. So i figured I'd change this to a different type (Poem), remove one of the two records, and change it back. Not so fast. When i changed it, remove listed the two poem records, but when i removed one, both went. So i added a new version of the novel, and merged the two titles. All is OK. But it seems to me that import shouldn't import a 2nd copy of a record that is identical to an existing copy, and remove should allow deletion of one of two identical records. -DES ^Talk 05:05, 30 October 2008 (UTC)

Singularity Sky

Added $C price to your verified SNGLRTYSKY2004. How's the de-bugging going?--Bluesman 19:19, 9 November 2008 (UTC)