User talk:Magnus Manske

}}

If you can't think of any particular corners to improve on Wikisource, how about taking a look at Portal:Religious texts, Portal:Wars or Portal:Texts by Country for some ideas? Don't forget to list your contributions on those pages as well so others will find and read them in the future!

Yann (talk) 19:44, 23 July 2009 (UTC)

Hello
Hi Magnus, nice to see you here. I created Clint, George (DNB00) so you can see how the transclusion syntax works, the header, the hidden category for "qv" (in theory the DNB here will be completed one day as a piece of hypertext!). And I added the format, and author template, on the page you proof-read. Charles Matthews (talk) 21:11, 23 July 2009 (UTC)
 * I tried my luck at Clint, Scipio (DNB00). Two issues:
 * Scipio doesn't have an article in Wikipedia. Set up a link anyway? --Magnus Manske (talk) 22:26, 23 July 2009 (UTC)
 * Why does Clinton, Charles (DNB00) transclude two sections?

I have applied the usual conventions. Clint, Scipio (DNB00) is now in the category of articles without enWP equivalents; and Clinton, Charles (DNB00) has been moved to include dates in the title, which is the standard method for disambiguation. Charles Matthews (talk) 08:18, 24 July 2009 (UTC)
 * Thanks. Clint, Scipio (DNB00) now shows an ugly broken pseudo-link to Wikipedia. Any way to fix that? --Magnus Manske (talk) 08:21, 24 July 2009 (UTC)


 * Probably! I had stopped noticing it was ugly. Charles Matthews (talk) 08:35, 24 July 2009 (UTC)
 * In the Charles Clinton article there was an example of an author name needing small caps, i.e. . Charles Matthews (talk) 08:41, 24 July 2009 (UTC)
 * I've added your articles to Dictionary of National Biography, 1885-1900/Vol 11 Clater - Condell. Currently there aren't many full volume listings on the site, for the DNB, so articles have to be listed by hand. Also, unfortunately, the exact titling conventions haven't been agreed or written down, so redlinks on index pages may not work immediately. New articles should also be added to Author pages: most of the DNB author pages do now exist. Charles Matthews (talk) 12:40, 24 July 2009 (UTC)
 * That's a lot of links to keep updated manually. It seems to distract quite a lot from the OCR-checking. Is there noone running bots here? Or generating to-do-lists, at the very least? Should I? --Magnus Manske (talk) 12:44, 24 July 2009 (UTC)


 * Judging by Bot requests, the supply of bot programmers is less than demand. I know little about the general situation. For the DNB there are several tasks such as adding scanned text; but there are old problems too (see the "Glitchiness" section on my userpage. (I have added author links for the Clint and Clinton articles.) Charles Matthews (talk) 12:50, 24 July 2009 (UTC)

New tool
I've been able to remove numerous anomalies, of different kinds. The remaining hits under "biography => author links" are mainly because of general wikification of biographies (I think all are where there is postfix (DNB00), the other cases being the compilation articles which are anyway in need of attention). Is it possible to separate out the hits, so that the cases where biography => author is via a template are listed in one place? Charles Matthews (talk) 13:05, 25 July 2009 (UTC)
 * MediaWiki doesn't store the "source" of a link. But I can make a list of articles with "DNB XXX" templates, or a subset thereof. --Magnus Manske (talk) 19:06, 25 July 2009 (UTC)


 * Like this, comparing two lists? For each "DNB XXX" template, list the backlinks from Template:DNB XXX that have postfix (DNB**); and on the author page for that template find the links ending (DNB**). Then those lists should be the same, so collate all the differences to be found, running over Category:Dictionary of National Biography templates. Maybe that is what you meant. Charles Matthews (talk) 16:13, 30 July 2009 (UTC)

DNB volume 11
The text was uploaded by Mjbot, and unfortunately the scan used (from http://www.archive.org/details/dictionarynatio03stepgoog, according to the index of the pages on Commons) was not the right choice. In other words the text at http://www.archive.org/details/dictionaryofnati11stepuoft is going to be better (for almost any page, I think). I changed the text for you on p. 473, where obviously the existing text is pretty bad. As I am not a bot, I don't myself add text to many pages at once. But in general it is worth also knowing what different scans there are - see the project page about this which contains a copy of what I have put up in my enWP userspace. Charles Matthews (talk) 14:43, 3 August 2009 (UTC)
 * Thanks! Where did you get the OCR text - the plain text from archive.org? I tried the DJVU viewer, but it doesn't have plaintext. --Magnus Manske (talk) 15:08, 3 August 2009 (UTC)


 * It's the "full text" option in the viewing options box, or in the index under "HTTP" it is the "djvu.txt" file. Charles Matthews (talk) 15:30, 3 August 2009 (UTC)

DNB and dates
Past discussion is at Wikisource talk:WikiProject DNB. There is not yet a full "manual page" on titling, and things have been a little casual, so far. My understanding is that


 * 1) dates are the preferred form of disambiguation;
 * 2) dates are not included where disambiguation is not needed within (DNB00) pages;
 * 3) for floruit or death-only dates no space is used - like (fl.1800) or (d.1800).

These are the conventions I'm used to: they don't even answer all questions. I think titles of nobility probably should not be included (no Sir, certainly); post-nominal initials such as M.D. should not be included. But some of the existing lists of redlinks don't follow these rules, because they were created before many articles. (I did find one case where name plus date of death wasn't enough to distinguish two DNB entries.)

There needs to be some more work done before these conventions are fully formalised!

Charles Matthews (talk) 20:43, 21 August 2009 (UTC)

Ability to bulk transfer Images?
Gday MM. Is there an easier way to bulk transfer images like exist at Category:Pigling Bland to Commons:? Thanks. -- billinghurst (talk) 13:20, 28 September 2009 (UTC)
 * I could probably write a wrapper around CommonsHelper to transfer the entire category, or maybe a specialized bulk upload tool. Come to think of it, the latter might be better in this case, as the pigling images don't give a source, which could lead to problems on Commons; a generic mask would be better to enforce the Information template. I'll write here when I come up with something. --Magnus Manske (talk) 10:19, 2 October 2009 (UTC)
 * Thanks MM, though do note, that my chosen example was not well chosen, as these files need to remain locally, as not being out of copyright and therefore ineligible for Commons. Oops. A separate mask would be very useful as we may wish to apply category for Commons, and/or apply a better copyright tag to go with transfer.  Thanks muchly. -- billinghurst (talk) 11:23, 2 October 2009 (UTC)

/DNB11
I have placed a cleaned-up copy of this list of yours on Dictionary of National Biography, 1885-1900/Vol 11 Clater - Condell. The conventions on titles are now more solid: in almost all cases we are using minimal titles. So I have moved some pages.

Further changes mean that some author pages now have subpages listing DNB articles. This doesn't exactly break the tool you wrote for checking linkage. It does mean that if article P correctly links to Author:A, and Author:A/DNB correctly links to P, the detailed readout is somewhat harder to understand. Charles Matthews (talk) 15:09, 1 February 2010 (UTC)

Brought overcss image crop
Gday. Brought over the documentation (move the files) and updated to include. When using this, and I found out that we have some discrepancy in an image being displayed. In the end we found that phe had some personal css settings that were disturbing the image, and wondered whether you had thoughts on whether there was a means to circumvent personal settings.

Plus have an issue that the positioning paremeter (left/center/right) isn't working for me within the template, though I was able to float the picture to circumvent the issue.

Also, with a bit of trial and error I was able to get an image with boundaries required, though wondered there was some science to work out the required components rather than pure guess. Thanks. — billinghurst  sDrewth  07:02, 11 April 2010 (UTC)
 * For my problem, I setup this page to show the problem without relying on my own css, the two call to css image crop are identical, rendering is different because the image is shifted by the around the second template call, I'm unusure but I suspect it comes from Mediawiki:common.css ".lefttext p { text-indent: 2em; }" Phe (talk) 10:04, 11 April 2010 (UTC)

Greeting from Italy
Hi Magnus, we met into that talk about BackToScroll tool. As you certainly imagine, I like to "play" a little with templates & tricks, but I'm far from a decent programmer! I'm only a curious fellow that knows something about programming in general and about python.

Nevertheless... could be that one or the other from many ideas I test have something good inside, even if I can't develop them as they deserve. Can I submit them to you? Feel absolutely free to tell me "Please, no, I lack time and will". :-) --Alex brollo (talk) 09:56, 18 August 2010 (UTC)
 * Please go ahead, I don't lack the will. Time, on the other hand, can be an issue... --Magnus Manske (talk) 10:26, 18 August 2010 (UTC)
 * Ok!First of all, I've to mention two interesting it.source templates (only the second is mine)

it.source template §
it:Template:§ coupled with this css row:  span.Citazione:target {background-color:#DEF;} 

Its code: 

It's used to seed an invisible anchor named  that highlights the optional text passed as  with a nice blue background when it is reached by a link. If no parameter 2 is passed... well, this is the trick I added to the code. ;-)

I developed too a "cross-link" from this code, that runs (with the same code in the two linked texts) both as an anchor and a link to the second linked text. Now, I'm thinking about this trich do build a DIY system for references.

It.source template Pt
It's something like a generalization of Hyphenated word, that I built without any knowledge of the latter. Its code (in an old, very simple but running version): This allows to pass into ns Pagina: (your Page:) anything in two formats: one to be visualized into Pagina:, the second into ns0. An hyphenated word too... but lots of other stuff.

Next time, something much more interesting and promising.... our template Ns0.--Alex brollo (talk) 12:30, 18 August 2010 (UTC)

CommonsHelper doesn't like transferring with book template
Gday Magnus. Hope that all is well for you.

I was cleaning up here at enWS, moving compliant files to Commons using CommonsHelper. These days, the djvu files are preferably wrapped in book rather than information, however, CH just decided to wrap it in "information", take out the paired fields, and put the remainder into the description field, rather than take it as is. When you are next tinkling with the script, it would be great if there was a check filter for "book". To note that the template structures are paired between Commons and enWS. Having "book" is really helpful as it allows a direct port of the data into the Index: ns, and hopefully one day will all be paired nicely with WD. Thanks. — billinghurst  sDrewth  01:24, 28 June 2014 (UTC)