User talk:AndrewOfWyntoun

Welcome to Wikisource

Hello, AndrewOfWyntoun, and welcome to Wikisource! Thank you for joining the project. I hope you like the place and decide to stay. Here are a few good links for newcomers:

• Help pages, especially for proofreading
• Help:Beginner's guide to Wikisource
• Style guide
• Inclusion policy
• Wikisource:For Wikipedians

You may be interested in participating in

• Proofread of the Month
• Monthly Challenge
• Requested texts

Add the code {{active projects}}, {{PotM}} or {{Collaboration/MC}} to your page for current Wikisource projects.

You can put a brief description of your interests on your user page and contributions to another Wikimedia project, such as Wikipedia and Commons.

Have questions? Then please ask them at either

• Wikisource:Scriptorium; or
• Wikisource:Scriptorium/Help

I hope you enjoy contributing to Wikisource, the library that is free for everyone to use! In discussions, please "sign" your comments using four tildes (~~~~); this will automatically produce your username if you're logged in (or IP address if you are not) and the date. If you need help, ask me on my talk page, or ask your question here (click edit) and place {{helpme}} before your question.

Again, welcome! — billinghurst sDrewth 11:38, 26 March 2020 (UTC)

Encountering problems OCR proofreading

Hello,

I work for the National Library of Scotland and as part of our remote working at the moment we are correcting the OCR or some of our digitised texts. I have managed to proof 4 pages this morning, but when trying to save changes of the 5th I keep getting this error message: 'Error: This action has been automatically identified as harmful, and therefore disallowed. If you believe your action was constructive, please inform an administrator of what you were trying to do. A brief description of the abuse rule which your action matched is: Image abuser (import of global AF/219)'. This happens on the page 'Creating Page:Unparalleled sufferings and surprising adventures of Philip Quarle.pdf/5'.

Help please? AndrewOfWyntoun (talk) 11:49, 26 March 2020 (UTC)

Thanks for the ping. You should now be right to go. — billinghurst sDrewth 12:25, 26 March 2020 (UTC)

Index:The Laws and Acts of Parliament of Scotland.djvu

Thanks.. I responded to your question on the talk page, Don't worry about the images or drop initials.

This was a long-term work, and I appreciate you have other projects as well :). ShakespeareFan00 (talk) 07:40, 23 April 2020 (UTC)

@ShakespeareFan00: Thanks! I had a look at some of the other pages just to get an idea of how you were formatting things, but as you are aware I am very new to this so won't touch anything I'm not certain on. It's a really interesting work though, so hope to help out with even just a few pages a week. AndrewOfWyntoun (talk) 07:47, 23 April 2020 (UTC)

Catchwords and Quire Signatures

I have been placing catchwords and quire signatures in the footnote box. When it is simply a catchword I used align right using the right|WORD tag. When it is a combination of a catchword and quire signature I have been using the running header template so that it is properly aligned, e.g. rh||E2|WORD. I have seen instances of the use of the continues| tag, but can't seem to find any guidance about what exactly this does or when it should be used? Should the 'continued' tag be used instead of the 'right' or 'running header' tags for the purposes of quire signatures and catchwords?

If you can point to specific pages where this issue has arisen, I can more easily provide some guidance. Use the {{ping}} template to notify me when you've responded so that I will know there are examples to look at. --EncycloPetey (talk) 20:05, 8 May 2020 (UTC)

@EncycloPetey: An example of both quire signature and catchword is here. Most other pages in this work have catchwords, which I've just been using {{right}} template for. Thank you! AndrewOfWyntoun (talk) 10:34, 11 May 2020 (UTC)

OK, I see what you're dealing with now. Quire information is not part of the literary content, but is usually information used by the printer to assemble the work. Since it is not part of the author's work, and not relevant for most works, it is typically not transcribed. If the information is worth preserving for a specific work, then yes, you can use {{rh}} for these. --EncycloPetey (talk) 15:39, 11 May 2020 (UTC)

@EncycloPetey: Thanks for the info! AndrewOfWyntoun (talk) 07:11, 12 May 2020 (UTC)

You can also consider using {{continues}} under some circumstances. ShakespeareFan00 (talk) 19:28, 25 May 2020 (UTC)

Transclusion

@Billinghurst: Morning - I noticed that you've excluded some transcluded texts I've been doing from being indexed - is there a problem with the way I'm doing things? I've been following the guidelines, but did only just start yesterday, so appreciate that I might not have it correct just yet. Cheers for any help! AndrewOfWyntoun (talk) 07:19, 14 May 2020 (UTC)

@Billinghurst: Actually, you can probably definitely ignore that (I was looking at the one I was to proofread, rather than transclude) - I don't think I've had enough caffeine yet... I was trying to transclude a variety of texts to get to grips with the process and see any difficulties I encounter. I've added the ones I've done to my user page. I only asked before because I noticed that I could previously find the titles of the works I'd transcluded in the search box, but now upon searching I only get results for the index.pdf files of the transcluded works. Anyway, apologies if that came across as salty. Thanks for your help, time, and patience! Edit - any idea about the search box issue? AndrewOfWyntoun (talk) 07:27, 14 May 2020 (UTC)

Hi. I am guessing that you are referring to my addition of transcluded = ... in the templates {{index transcluded}} and/or {{index validated date}}, where index refers to the namespace name. These maintenance templates are simple markers for works that have or have not been transcluded, and are usually complete through proofreading. It is a manual process for checking, as we have no suitable ability to automate that checking, and we don't want to miss finished works, or accidental omissions of pages of works.

Typically all pages are indexed, though your search settings may exclude certain namespaces (check special:search), or how your search completion is configured in your special:preferences. Usually search ahead will call up main namespace works first, so if you are wanting that work namespaces, start with Index: so it looks in that namespaces only. [Yes the digger you deep the more layers of complexity are exposed]. — billinghurst sDrewth 21:46, 14 May 2020 (UTC)

@Billinghurst: Yes, that's what I was referring to. Thanks for the extra info, that is good to know. Thanks also for sorting the Keys of Love text, I was aware something had went wrong there but for the life of me couldn't figure out what it was! AndrewOfWyntoun (talk) 08:26, 18 May 2020 (UTC)

I may have done or two before. Never an issue. — billinghurst sDrewth 09:40, 18 May 2020 (UTC)

Centre aligning side-by-side tables

@Billinghurst: Hello - would you please be able to offer some advice on the tables on pages 13 and 14 of this work? Magic oracle, or, Conjuror's guide I have probably done it wrong, but it took me long enough to get the tables side-by-side, but cannot work out how to align them to the centre. I've had a read through the guides, but am aware this is probably well beyond my capabilities. Thanks for any help you can offer AndrewOfWyntoun (talk) 15:17, 22 May 2020 (UTC)

There may be a html5 special way, whereas I am just pretty basic.

Guessing we are talking Page:Magic oracle, or, Conjuror's guide.pdf/13 … wrap a table (1r, 5c) around the outside and put each inner-table in a cell of the nesting table, then fiddle with some overarching tidying. Where I am doing that I often will use an html remark (<!-- -->) to comment for others to follow. I left /14 in case you wanted to play. and have that winning glow afterwards.

Also to note that digits in type faces are an en space in width (&ensp;) to pad if you don't wish to right align the column. If you have two digits, you can use an em space (&emsp;). — billinghurst sDrewth 00:58, 23 May 2020 (UTC)

Note about the title page graphics. As they are often not integral to the works, I can often wrap them into <noinclude></noinclude> so they don't show when trancluded, but they are still marked as needing work if possible. I wouldn't do it from images that are integral to the work, just those that are for display purposes. — billinghurst sDrewth 01:06, 23 May 2020 (UTC)

As a note please do look at Template:New texts and considering your works, even selectively, as you have not been getting the front page treatment. — billinghurst sDrewth 01:01, 23 May 2020 (UTC)

@Billinghurst: Thanks for the help with the table alignments - I will give that a go later on today. I look forward to that winning glow. I'll also check out the texts you've highlighted and make sure I'm getting the front pages of works correct before proceeding with any more transclusion. AndrewOfWyntoun (talk) 12:15, 25 May 2020 (UTC)

Transclusion (again)

@Billinghurst: Hello again! I was going to transclude some items this afternoon using our internal spreadsheet, but when I got to the items I noticed that these had already been done by another user. In the translucsion page there seems to be a set of letters/numbers in front of the title, such as 'NLS104186973'. We've not been told by the people running the project on our end to add these, and they said they've also not had any input from Wikisource regarding this. Do these letters and number sequences refer to anything in particular? The three texts I was going to transclude are:

• Index:History_of_two_brothers_misfortunes.pdf
• Index:Excellent_new_song_called_the_farmer's_glory.pdf
• Index:Jones'_address_to_a_sovereign.pdf

Thank you! AndrewOfWyntoun (talk) 12:59, 1 June 2020 (UTC)

Moved them to titles. There are some people who just have their own issues following the existing styles. <shrug> — billinghurst sDrewth 12:38, 2 June 2020 (UTC)

@Billinghurst: Thanks for sorting that. I notice that this user has also transcluded things in a different way, and has also used 'poem' tags rather than line breaks (as we've been instructed), should I go back and change these so that they are in line with what we've already been doing? AndrewOfWyntoun (talk) 12:54, 2 June 2020 (UTC)

Probably just called different, nothing right or wrong, and probably more identified as one's preferred methodology. Do as you need to, they are your project's contributions, and no one will mind. — billinghurst sDrewth 13:02, 2 June 2020 (UTC)

"Adventure of the Creeping Man" not working

Hi, I'm an anonymous user, and I'm trying to add "The Adventure of the Creeping Man" to "The Case-Book of Sherlock Holmes" (it is in the public domain now, after all), but I keep getting, "Error: This action has been automatically identified as harmful, and therefore disallowed. If you believe your action was constructive, please inform an administrator of what you were trying to do. A brief description of the abuse rule which your action matched is: Image abuser (import of global AF/219)". I'm copy-pasting it from this site: https://www.fadedpage.com/books/20160532/html.php#chap08. Please help!

Hello, yes, as you can probably tell, I also had a similar problem. No idea what was causing it, but I got it fixed by contacting one of the Wikisource Admins (I am not an admin). List here: Wikisource:Administrators Good luck! AndrewOfWyntoun (talk) 09:41, 6 June 2020 (UTC)

added some maintenance Category:WikiProject NLS

Hi. I have added a couple of petscan: reports, that can be run, to the page, and also an intersection report for indexes in the transclusion check category. The first petscan and the DPL list should give the same output, just me not being trusting. If you were after some other checks, let me know and I can see what is possible. — billinghurst sDrewth 14:25, 8 June 2020 (UTC)

@Billinghurst: Hello, thank you, this is very interesting and useful! I'll pass it onto the others involved in the project as well. AndrewOfWyntoun (talk) 14:31, 8 June 2020 (UTC)

@Gweduni: Hi Gavin, thought you might be interested in the petscan reports that billinghurst has created on the Category:WikiProject NLS page. AndrewOfWyntoun (talk) 14:49, 8 June 2020 (UTC)

Section tags

Re Page:Old man's wish.pdf/6: If you do not have both a "start" and "end" section tag for every section (i.e. both 3 and 4), it will not transclude correctly. I did not remove them, I added a "start" tag for section "3" and an "end" for section "4".

If you are using the "Easy LST" gadget (which is on by default), you will see section tags that look like ## 3 ## rather than <section begin="3"/>...<section end="3"/>. Inductiveload—talk/contribs 12:32, 9 June 2020 (UTC)

@Inductiveload:Thanks - I didn't know that about every section needing the start/end. I had left the very last one blank because it was all on the one page so could just go from=8 to=8. I thought I had put tags in fully for the rest of the sections, I must have skipped over the other ones - apologies. I haven't used the Easy LST gadget, and was confused by the hash signs if I'm being honest - which is why I thought the tags had been removed! Thanks for sorting that though - appreciate it. AndrewOfWyntoun (talk) 12:35, 9 June 2020 (UTC)

@Inductiveload:Also, I did not realise it was you who had went in and fixed the tags - I failed to check the history, and I thought it was the same user who had transcluded that particular poem who had also changed the section tags. The transclusion looked strange, which is why I thought the new tags were incorrect. Thanks again! AndrewOfWyntoun (talk) 12:39, 9 June 2020 (UTC)

No worries. In fact, a quick check indicates it does actually work as you have it, it just stops "Easy LST" working, because Easy LST needs matching tags. I recently saw a case where missing "end" tag caused a transclusion failure (Wikisource:Scriptorium#Help_with_transclusion_required.), but you have the reverse case ("end" but no "start"), which apparently works (however it would not work if you tried to transclude section 3 from page 6 only, as there is no section start). I'd still say sticking with matching start/end tags for the common case is the more universally-recognised idiom.

Also, if you saw the hashes, you are in fact using "Easy LST", you have to turn it off in your gadget preferences if you don't want it (it's a default Gadget). But if you use it and use hashed sections, you are guaranteed to never mismatch any section tags, which can otherwise be a tricky thing to spot. Inductiveload—talk/contribs 12:56, 9 June 2020 (UTC)

Images for Old man's wish

Quick question: are you planning to extract the images for Old man's wish, or would you like me to do it? I'm happy to do it, but I don't want to step on your toes if you want to do it! :) Inductiveload—talk/contribs 14:00, 9 June 2020 (UTC)

Ooo... Honestly, I haven't thought about extracting images. I'm doing this proofreading/validating/transcluding work as part of a work-from-home project for the National Library of Scotland during lockdown. Our wikisource project page is here Wikisource:WikiProject_NLS. A lot of the chapbooks have cover images and woodcuts, less occasionally throughout as well. We've been told just to mark pages with images as problematic, but I'm not sure if there are plans to go back through an extract any images as needed. I've only done image extraction once and honestly it took me most of an evening... I'm very much new to all of this. I don't think you'd be stepping on anyone's toes if you were going to do the images, but I can check with the project co-ordinator if you'd like? I'd be very interested to know how you do it, because I feel like I spent a lot of time not doing it right. You can see my effort here Page:Buke of the Howlat.djvu/30 (constructive criticism always welcome) AndrewOfWyntoun (talk) 14:11, 9 June 2020 (UTC)

Here's an attempt at the title page image. This was actually quite a tricky one as the print was pretty light and the paper texture very strong. The main process I used, using GIMP (I do not have Photoshop, but maybe PS has better ways):

• Download the highest resolution image I can (in this case a 1290 x 2057px JPG). It is very important you do not use a PDF or, especially, DjVu as the source image if you can possibly avoid it, because images are generally very damaged by the compression, which is often heavy and optimised for text.
• Open in GIMP and crop and rotate the image as needed. Sometimes you might need to unwarp the page too, but that's quite rare.
• Add an alpha channel to the image and a white background layer so you can see what you are doing (or set your background to white, not checkerboard)
• Pre-process the image:
  • Use the paintbrush in Colour Erase mode to delete most of the paper colour from the top (image) layer
  • Remove the alpha channel to "reset" the colours (after Color Erase you often end up with very strong, but not very opaque colours, which confuse tools that are looking for "weak" colours).
  • Sometimes multiple rounds of Colour Erasing can help, but not in this case
  • In this case I then raised the Contrast to bring out the details and wash out the background. Sometimes "Levels" or "Curves" can work better here, it depends on the image. In this case I used Levels after the Contrast bump. Some images might not need anything.
• Now, is the fiddly bit. You have to scout around the image with the erase tool (usually a small, quite hard-edged, round brush works well for me) and erase splotches and defects from the top (image) layer. Sometimes you have to make a judgement about whether a dark area is image, or if it's background that has been darkened during processing. Generally, accidentally-enhanced background looks like a mid-grey or yellow/red tinted, soft-edged splodge, compared to darker-grey harder-edged areas that were originally ink. This shouldn't take too long, unless the image is very big, very damaged or very complex. Refer to the original when in doubt, sometimes it's then clear what is and isn't in the image.
• When you've tidied up all the defects, you can desaturate (Colour -> Desaturate -> Desaturate...) the image to clear out any residual colour casts and make it a nice neat greyscale (obviously only if it's not a colour image!)
• Next, flatten the image to turn the lacy mess that is the top layer into a nice flat greyscale image. Right click on the top Layer and "Flatten Image".
• Finally, Colour -> Colour to Alpha... using white as the colour and you get an image with transparency (this is important if you want to display the image on a coloured background - remember most ebook readers use an off-white background, even if the website background is white by default).
• Save as PNG (to retain the transparency and so as not to re-introduce compression artifacts: most graphics like this have hard edges, which means high-frequency image data, which means JPG compression will be quite obvious, even at high quality levels.
  • Saving as greyscale PNG (8bpc GREYA in the GIMP export dropdown) can save a few kB of image size, in this case, 122 -> 105kB.
  • Don't scale the image down and don't sweat the file size too much, Commons is not short of disk space, and the image will be automatically served at a appropriate resolution when embedded at smaller sizes.
• Upload to Commons. I like to make a a Category for the PDF/DjVu and the images, even if there only a handful of images.

It looks like a long process but it doesn't actually take too long the second time around! Sometimes you have to undo a step and try again. GIMP has an excellent "Undo History" window.

This is the hard way, and maybe there are better/quicker/easier ones. I also very often use an excellent little GIMP script called "Leon's Reset Background to White". The original source seems to have vanished from the Internet, but I've put a copy at User:Inductiveload/leon-paper-white.scm. It can be a bit over-aggressive and it doesn't always get all the defects, but it's always worth a try. You still need to do the colour to alpha step to get a transparent image.

If you're really stuck, the Graphics Lab at Commons might be able to assist. Inductiveload—talk/contribs 16:01, 9 June 2020 (UTC)

In terms of co-ordination, I have done the two images in Old man's wish, because there were only two to do to complete the work, and I already did one as the example. Obviously, feel free to tweak, improve or replace the images if you think they can be better: I'm not an expert at image restoration! :) Inductiveload—talk/contribs 17:17, 9 June 2020 (UTC)

@Inductiveload: Thank you so much for taking the time to go into all this detail, I really appreciate it. I will definitely give some images a go now if I'm feeling a little braver (it was equal parts fun and frustrating before!). I think the images you've done for Old Man's Wish look great, and, as you say, there's really only so much you can get out of older material - especially since chapbooks were printed on paper which was not of the best quality to begin with. I've marked the pages as validated now. Cheers! AndrewOfWyntoun (talk) 07:10, 10 June 2020 (UTC)

Glad it was helpful. Looks like it's working for you - I see some very tidy-looking images! :-) Eventually I'll get around to writing this up as an illustrated tutorial.

By the way, I found a really handy thing if you are using GIMP, for the fiddly bit of chasing the defects with the eraser tool. Under View -> Display Filters..., select "Contrast" - this applies an overlay filter that makes it extremely obvious where defects are: white areas with a thin black border are generally "good" ink, wide black borders are fuzzy edges (might be bad) and isolated grey/off-grey splotches are junk. Even the faintest smear in the background shows up pretty vividly. You can use the eraser with the filter overlay on, the overlay doesn't change the image, just affects how it is displayed, then turn it off to check. Note: it needs the white background to "see" the contrast, so it's only useful before the final colour-to-alpha step. Inductiveload—talk/contribs 09:10, 11 June 2020 (UTC)

@Inductiveload: Thanks for the extra tips! So far I've only managed to make images with white backgrounds, as I was finding that transparent images were either far too fiddly for a temperamental laptop mouspad, or that I would somehow make the image in a very over-exposed contrast. I will give them another go once I get an image that isn't so daunting... On another note - do you know anything about using {{overfloat image}} for adding framed pages? I tried to do it here: Theatrical speaker, and it looks ok when rendered on a device with a big enough screen, but on a mobile device, or when viewing the validated text as a scrolled image, it doesn't render properly. I know I've done something wrong - because I generally am not familiar with this stuff at all - but can't quite work out what's gone wrong, any ideas? AndrewOfWyntoun (talk) 08:20, 15 June 2020 (UTC)

Re transparent: I work with them with a white background to the very last moment, so I can see the image, then finally Color To Alpha and export. There is an option in Preferences -> Display to set the transparency background to white (default is mid-grey checks) too, but then it's hard for me to see when it is and isn't transparent. It's not personally necessary for me, actually, as my e-reader automatically removes white image backgrounds in ebooks.

Re overfloat: In general, it is not possible to format an image border like that and have it work on mobile. What the construction you have there does is place an image of a fixed pixel size (550px) on the page and then float some text over the top of it in a box the right size.

• In the main namespace, on desktop, when the text isn't too long this works, as there is space for the image to be that pixel width and the box you provided is big enough for all the text.
• In the main namespace, on desktop, when the text is too long (or wide), you get a scrolled box, which is better than hiding the contents. Note that you, the Wikisource editor, do not have full control over this: the user could have any font, or use any font size (e.g. if they are visually impaired) and there's nothing you can, or should, do about it. This is doubly true for e-readers, where the font size is often a slider control that can be changed at will.
• In the page namespace, the side-by-side layout means there simply isn't the horizontal width on the page for the fized-pixel-width image and the scan, so the image overlaps the scan.
• On mobile, there are some extra CSS rules applied (Mediawiki:Mobile.css + CSS from the mobile skin, a.k.a. Minerva). One of the effects of that is that images are constrained to be at most the width of the screen to avoid them "escaping" off the right hand border, which looks terrible and is user-unfriendly. However, this means that your carefully-constructed overfloat box, which is assuming a fixed width of 550px is now too big - the image is only as wide as the mobile screen, which on most phones is in the region of 350–400px. On top of that, the mobile skin applies a 400px "container" div, so even on big screens in mobile view, it breaks.

The cardinal rule of formatting for mobile is no fixed widths, because you cannot have any confidence of the screen size. Technically this rule goes for desktop too, but since computer monitors are generally big enough, it goes un-noticed most of the time (e.g. you do notice in the page namespace, but no-one cares as it's not a "Presentation" namespace, so if it looks a bit rubbish, so what.

Now, things you can do to fix this, in no particular order:

• Give up, and just use a simple border like "3px double black" or "10px solid grey". Pros: independent of the screen size; easy. Cons: A cop-out; a pity when the actual border lends to the feel of the work.
• Use a "proper" CSS image border: Newes from the Dead. Pros: independent of the screen size, works on desktop and mobile; will never cause a scrollbar as the border can extend as needed. Cons: A faff, needs a special image at Commons and a CSS file to be created; doesn't work on export as the CSS refers to an image by URL and this isn't available in most e-readers; needs "surgery" on the image and there are limitations. In this case, the two "pips" in the centre of the top/bottom sides image would be tricky to replicate as CSS image borders work in terms of corners and repeating units for the sides.
• Continue using overfloat left, but recast all your sizes in terms of percentages. This means the inner box will continue to match even when the image shrinks by command of the mobile CSS. Pros: you get the real image, probably can be made to work 90% of the time, and you get a scrollbar the rest of the time. Cons: the template hard-codes the use of px, so it's not actually possible right now. I have no real handle on how this might look on an e-reader.

In a nutshell: it's a mess. For the moment, I would say you're best off doing what you're doing, and I'll try to look at fixing {{overfloat image}} to at least give a fighting chance of it working on mobile (and, indeed in Layout 2, where it also doesn't work properly). At least when using OI like this, your intentions are clear, so it's can be found and improved later. Inductiveload—talk/contribs 09:46, 15 June 2020 (UTC)

A new thread, mainly to avoid hijacking the previous thread :)

[ I started to type this in the images thread above, but didn't want to hijack that discussion. This is kinda sorta a tangent sprung from that though.]

Incidentally... Who are you coordinating with on the wiki side? I see Beeswaxcandle is mentioned sometimes, but they've not been active here since April. Do you have a point of contact on Wikisource and/or WMF? Is there any overall information about your project anywhere (beyond Wikisource:WikiProject NLS)?

I'm wondering about things like what you're doing in terms of metadata for the files on Commons (some institutions have institution-specific templates linking the scans back to their catalog etc.), or what the strategy for bibliographic data at Wikidata is. How are you categorising works here, if at all, in relation to your project and as part of our general category system? That sort of thing…

And as Inductiveload mentions above, it would be useful to know what part of this project you would like the community to help with and which parts you'd really rather have full control over yourself. I'm very happy, of course, that the project sees value in Wikisource's approach, but I'm a little worried that it's a little bit of a one-way street and that we're all missing opportunities for synergy.

I would also love to see us establish an ongoing relationship even beyond your immediate project (the 3000 Scottish chapbooks). I feel certain there are lots of things that would benefit both NLS and Wikisource on an ongoing basis, even if a lot more low-key than the current project (3000 works is a pretty massive effort, especially over such a short period of time). I know I often want to tear my hair out about the bibliographic data in library catalogs and archives and the inability to fix it. Perhaps crowdsourced corrections to your catalog would be of benefit to the NLS? We do quite a lot of detective work to make sure our bibliographic data is correct (including editions, authorship, etc.) and it would be nice if others could benefit from that effort. We similarly put a lot of effort into determining copyright status of individual works, the results of which might let archives and libraries make more nuanced decisions than the single bulk cutoff dates (that have to be ultra-conservative because they're one-size-fits-all) a lot of them operate under (I'm looking at you here, HathiTrust!).

On our end we sometimes have trouble getting hold of decent-quality scans of works (Google Books' high volume of lousy quality scans is the bane of my existence!), and having a point of contact for requesting scanning of specific items in your collections would be of great value to us (Don't even get me started on the time I was writing an article on w:James MacLehose and Sons and spent 3 days wheedling (unsuccessfully) with an archivist at the Glasgow City Archives at The Mitchell Library to get a scan of a photo from their collection to illustrate the article. I even offered to pay out-of-pocket for the digitisation to no avail. Argh!). I imagine it would also be possible to device some kind of process where the NLS was notified if the community on its own initiative had proofread and validated such a work so you could get the proofread text back. We also sometimes could use provenance data that libraries and archives sometimes have for their holdings, but don't necessarily publish in their online catalogues.

In short, I'm thinking there is potential for lots of both smaller and bigger benefits medium and long term, and would love to see such a relationship explored. --Xover (talk) 17:04, 9 June 2020 (UTC)

@Xover: Hello, and thank you for all these very interesting questions! First of all, on a personal level I agree with a lot of what you're saying and can only begin to imagine the kinds of opportunities this could bring. On an institutional level - I'm not sure how to answer a lot of your questions, as it's probably a lot more complicated than even I know, especially in terms of copyright and metadata, and these questions span quite a few of our various departments.

The lead on our end of the project is Gweduni, and we've mostly had assistance (and a lot of patience) on the wiki side from Beeswaxcandle, billinghurst, EncycloPetey, and ShakespeareFan00. Most of the other questions you've asked would probably need to go to Gweduni.

I should note that in general the Library is trying to make material available under less restrictive licences, especially material which is out of copyright and available online through our digital galleries/Flickr photostream. Some of the material in the digital gallery is covered by third-party copyright, but our general copyright page might be of interest if you're ever trying to narrow down material which can be included on wiki pages: NLS Copyright.

So, over to...

AndrewOfWyntoun (talk) 08:10, 10 June 2020 (UTC)

@Xover: Wondering whether Wikisource talk:WikiProject NLS may be a better place for the broader conversation. — billinghurst [[user talk:|sDrewth]] 22:50, 10 June 2020 (UTC)

Morning @AndrewOfWyntoun: and thanks so much for starting this discussion. I wholeheartedly agree with pretty much everything you've said above and would love to set up a proper relationship between NLS and the Wikisource community - there's so much potential for great things to happen! Over the years we've had a bit of engagement (we had the first Scottish Wikimedian in Residence back in 2014 I believe), one of our staff was on the Wikimedia board for a while and I've done a bit of work in my current role and when working at Edinburgh Uni on Wikipedia, 1lib1ref, Wiki Commons and Wikisource. Thanks to this chapbooks project there is also an army of 70-odd new Wikimedians at the Library (about a fifth of the entire staff) who could potentially get further involved. We're at the stage as an organisation that we need to have a proper conversation about how we want to strategically approach our engagement with Wikimedia over the next 5 years or so and I would need to get input from colleagues in several teams before we can take anything forward. There is definitely a feeling that we would like to do more, and we'd like to get more stuff into Wiki Commons (my dream would be to make this an automatic function of our digitisation workflow), add more to Wikisource, add more links to NLS collections to Wikipedia and explore wiki data in more detail (this is the platform I'm least familiar with). Doing all this while engaging with the wider wiki community is crucial (as you say, our current project is a bit "one way", and we'd love to make it more interactive - one thought I had on that was to identify 10 books we'd like the wider wiki community to help us proofread and see what the uptake is. I'll put them in the wiki project and alert you when done - what do you think?). For now, I need to speak to colleagues about our possible wikipedia strategy - and will get back to you in the next month if that's ok? Gweduni (talk) 08:21, 16 June 2020 (UTC)

@Whatamidoing (WMF), @Quiddity (WMF): ^^^^ waving and pointing to my favourite people for your team—as I never know to whom I should be pointing. — billinghurst sDrewth 08:52, 16 June 2020 (UTC)

Hi again - list of 10 books we'd love the wikisource community to help us with: https://en.wikisource.org/wiki/Wikisource:WikiProject_NLS#Items_for_Proofreading_by_the_Wikisource_Community @Xover:, @Billinghurst:. They're all pretty grisly (theme of "murder") and there are plenty more where they came from :) Gweduni (talk) 08:57, 16 June 2020 (UTC)

@Gweduni: I had a look at one of these (Index:Remarkable family adventure of Saunders Watson (1).pdf, now transcluded at A Remarkable Family Adventure of Saunders Watson). These pamphlets, by virtue of being short, are in that sense probably good candidates for soliciting contributions from the community. But I think doing it well requires some level of familiarity with Scots: for large swathes I ended up essentially just transcribing letter by letter, and picking up typos and OCR errors when you're not familiar with the language is a lot harder. Especially so when the copy is damaged or the scan of poor technical quality (it was hard to tell which of these was applicable in this case) because missing or partially missing words can't be reconstructed from context. In this case, as luck would have it, the Thomas Fisher Rare Book Library's chapbooks collection also has a copy of this, in much better condition, and had uploaded a scan to the Internet Archive. With that for reference it was possible to fill in the gaps, but it was pure luck to find an alternate copy and fairly tedious work.

Based on that I'm not certain straight up proofreading is where the community will be best able to assist. We may have a community of Scots speakers who would be interested and qualified that I'm not aware of, but the community in general will likely struggle like I did with these. Nobody here has any shortage of projects to tackle (personally I have several hundred editions of Shakespeare and related scholarship I would like to get proofread eventually), and we are a relatively small community to begin with (at its peak the NLS contributions dwarfed all other contributions). So for proofreading I think it would have to be for works that have some particular set of challenges that an experienced Wikisourcer could add some value to; for example by dint of knowing where to look for alternate scans to reconstruct missing text, or constructing complex formatting.

But in general I think we would need to look for other points of opportunity. I've noticed some of the NLS contributors sometimes struggle with "ideomatic" formatting, use of templates, and other such more technical aspects. Perhaps we could come up with some way for the community to assist in that area? Anything from mentorships, through a dedicated "help desk" for NLS contributors (like Wikisource:Scriptorium/Help but specifically for NLS), to a general channel for feedback on issues observed with their contributions (to avoid potentially discouraging feedback directly to individual contributors). Perhaps some kind of review mechanism where finished works can be checked for formatting and other issues (something less comprehensive than actually validating the work, as that's pretty labour intensive). There are probably also technical things like templates where we could help.

That all, however, is immediate term stuff, and I see the biggest potential in long-term areas. Cross-connecting our systems and databases to some extent as a central pillar: as a minimum having a static link from a work here (both the scan on Commons and the text on Wikisource) directly to that work in your collection, and a similar link from your database to the work here. If we could come up with a suitable schema for bibliographic data it should be possible for crowd-sourced corrections on the Wikimedia side to be automatically fed back to the NLS (subject to approval or rejection by your people, I presume). There are some possibilities for improved tooling in this area that might be of benefit to all involved (hey Sam, I'm having trouble figuring out the right User: … (WMF) to ping for this. Any suggestions? Keywords are GLAM, outreach, partnerships, and tooling around media and bibliographic metadata.).

In any case, once you've started talking internally, I hope we can start a conversation here (but, as Billinghurst suggests above, we should probably move that to Wikisource talk:WikiProject NLS). Donations of media and data can usually be handled by a GLAM institution deciding on a strategy and with minimal and mostly technical help from the WMF. But an ongoing relationship that actually leverages the community (and culture) of the Wikimedia projects will, I think, require some actual discussions and probably some trial and error. As completely volunteer driven and consensus-based we can never make any firm commitments ahead of time; we can just develop frameworks and facilitate, and hopefully the community will be interested in contributing.

In any case, I am definitely interested in this conversation, and would be happy to help facilitate where I can. --Xover (talk) 07:37, 22 June 2020 (UTC)

@Xover: Re "review" - perhaps it could be an aspect of the same hypothetical process which leads to "Good for export" (which implies "proofread" and "technically well-formed" as prerequisites for "will work as an ebook"). Most NLS works appear to be exportable without major issues. "Non-ideomatic" (to use your well-chosen phrase which I think conveys an absence of criticism well) works would likely generally get flagged up as part of that process.

just to flag that we have started discussions internally and there is definitely an appetite for closer collaboration with the WS community and Wikimedia more generally. Unfortunately we can be a bit of a slow-moving beast, and all energy is currently focused on resumption planning (hoping the Library will be reopening at the end of July). One positive thing is that I have started discussions and am setting up a Wikimedia "Community of Interest" in the Library, which will help us to crystalise the organisation's approach to this area. Planning another meeting for end of July and we will take things slowly from there. Agreed we should move this discussion over to the project page where I'll post a monthly update Gweduni (talk) 12:58, 26 June 2020 (UTC)

Inductiveload—talk/contribs 14:21, 22 June 2020 (UTC)

Bits

Hi. Generally we would do a number of things only at the top level of a work, eg. year of work, or categorise when the subparts are all the same. We would separately categorise the subparts where different. Similarly for year in {{header}} it is either a year, or a decade, or a century, and we do have a circa if you want a bit tighter. For a title on subpages, we would do [[../]] which is the relative to point up a level. — billinghurst sDrewth 05:45, 11 June 2020 (UTC)

@Billinghurst: Ah, is this in reference to Berkshire Lady's garland (2)? Sorry about that, I'd seen its use elsewhere but clearly haven't understood its use properly! Thanks for the clarity on the dates, I'll make sure to use those going forward. Does your circa have to be within five years either side of the circa date, or is it not as prescriptive as that? I wasn't sure about the categories for each work within a wider work and whether or not the categories should apply to all the works therein, but I'll make sure to separately categorise the subparts from now on. Cheers! AndrewOfWyntoun (talk) 08:30, 11 June 2020 (UTC)

Thumbs up, and no need to be sorry, this place has its quirks, and getting all the right info under eyes is always going to be an issue—I have left comments in to others in places.. Also fixed some though no way got all of them, you guyes are way to busy as a collective. Our use of circa is not prescriptive. — billinghurst sDrewth 08:37, 11 June 2020 (UTC)

Authors linked between Wikidata and Wikisource

{{help me}}Hello, It's just me again, your friendly neighbourhood pest. I thought I'd have a go of linking authors, specifically of a record I created on Wikidata, but I've obviously done something wrong. The most obvious issue is that when the author is linked to Wikisource he is listed as a 'living author' in the categories section. I don't have a definite death date for him, just a flourit date, so was wondering what is the correct thing to put to show the 'status' of the author properly? The page is Allan Boyd (Wikidata) and Allan Boyd (Wikisource) - thanks for any help! AndrewOfWyntoun (talk) 11:07, 30 June 2020 (UTC)

Solved by User:Inductiveload Thanks! AndrewOfWyntoun (talk) 11:45, 30 June 2020 (UTC)

Based on wikidata:Property_talk:P570#Living_people_and_no_value (which appears to have no consensus), I gave him a death date of "19th century" with a "presumably" qualifier. At least that works for enWS auto-categorisation and I guess if Wikidata doesn't like it, they'll have to deal with it by coming up with an answer! I tried to find a SPARQL query to isolate similar cases, but I'm not very good at that. Inductiveload—talk/contribs 11:53, 30 June 2020 (UTC)

Where to put the transclusion templates in the Index?

Hi, as long as the template is in one of the three fields Pages, Volumes, or Table of Contents, it doesn't matter. My own preference is to use Pages when there is no other content in either of the other two fields. Beeswaxcandle (talk) 09:38, 21 July 2020 (UTC)

@Beeswaxcandle: I knew my confusion wouldn't go unnoticed for long... Thanks for the info! :) AndrewOfWyntoun (talk) 09:42, 21 July 2020 (UTC)

Whereas I keep everything out of Pages except for <pagelist> as I think that it just adds to complexity and too easy to eff up. — billinghurst sDrewth 00:45, 5 August 2020 (UTC)

Author:Janet Hamilton

Question. The data from Wikidata says that the person died 1696, yet the work says published in 1753. Can you help on this one? — billinghurst sDrewth 00:43, 5 August 2020 (UTC)

@Billinghurst: What I am taking it to mean is that it is a later publication of the work long after her death, which seems to be fairly common in the Chapbooks with their cheap re-prints. I was taking the bibliographic data from our catalogue record: https://search.nls.uk/permalink/f/hha760/44NLS_ALMA51586226730004341. The ODNB article for her husband, who is mentioned in the title page reads:

'Gordon, Alexander, of Earlston (1650–1726), covenanter and conspirator, was the eldest son of William Gordon of Earlston (1614–1679) and Mary (d. 1697), second daughter of Sir John Hope of Craighall, Fife. He had twelve siblings, most of whom died young with the exception of William, of Afton, whom Queen Anne made a baronet, John, a surgeon of Carleton, and Margaret, who married Sir James Holborn of Menstrie in 1682. On 16 November 1676 he married Janet (1653–1697), eldest daughter of Sir Thomas Hamilton of Preston and his wife, Anna; at this time he inherited his father's estate, including debts totalling 14,200 merks. Their children included Ann (b. 1679), William, Sir Thomas (26 Oct 1685–23 March 1769), Margaret (b. 1687), Robert (1688–1750), Archibald (1691–1754), Hope (d. 1736), Mary (b. 1689), and Jane.'

The bibliographic information from the ODNB seems to match up with the wikidata item in question, which is why I linked them.AndrewOfWyntoun (talk) 08:43, 5 August 2020 (UTC)

edit - What I'm trying to say is... I think it's just been published in many forms through the ages, and like with any other author whose work is republished long after their death, they are still credited as the author - happy to change this one though if you'd prefer AndrewOfWyntoun (talk) 08:44, 5 August 2020 (UTC)

Thanks, just wanted to check that we had the detail right. — billinghurst sDrewth 10:57, 5 August 2020 (UTC)

Index:The Normans in European History.pdf

{{helpme}} Hello, I recently uploaded this text to commons and created a wikisource page. I realised, however, that I meant to upload it as djvu file using the IA Upload tool, but clicked the wrong button. The quality of the scans as the PDF are mostly ok, but a bit fuzzy in some places. Is there a way I can upload a djvu version of the file to commons? I want to know what I'm doing before mindlessly pressing some buttons and causing havoc... AndrewOfWyntoun (talk)

You will need to go back to IA upload tool, and do it again though this time for a djvu. There is no ready means to do a conversion. We can do a tidy up afterwards. — billinghurst sDrewth 21:48, 6 August 2020 (UTC)

@Billinghurst: Thank you, I'll request a deletion of the file on wikicommons, as the IA Upload tool won't let me upload a file from the same source if it's already on commons. Once that's done, would it just be a case of moving the .pdf index to the .djvu index? AndrewOfWyntoun (talk) 05:33, 7 August 2020 (UTC)

Give me five minutes. I have rights there, and it is easier if I do it. — billinghurst sDrewth 05:36, 7 August 2020 (UTC)

You should be right now, and I have moved over those pages that you had already done and the Index, and they should appear when you have the file in place (only difference is .pdf -> .djvu). Any problems, then please let me know. — billinghurst sDrewth 05:41, 7 August 2020 (UTC)

@Billinghurst: Just uploaded the file as djvu, and all looks good on wikisource. Thanks for sorting that! I'll be more careful next time... AndrewOfWyntoun (talk) 06:00, 7 August 2020 (UTC)

Meh! You don't think that even someone with my experience gets it right all the time? Too easy to assist, so it is never a major issue it is more a coordination matter, and why the tools are available, and one of my original reasons for getting tool access at Commons back in the dark ages. — billinghurst sDrewth 06:14, 7 August 2020 (UTC)

Jscript tool

(Sorry if this is overstepping my bounds), just saw your comment and wanted to offer you the response I got on my own talkpage from Inductive when I asked him what I think is the same question last week, and his response was

"Add to your Special:MyPage/common.js:
mw.loader.load("//en.wikisource.org/w/index.php?title=User:Inductiveload/index preview.js&action=raw&ctype=text/javascript");

It worked for me. Hopefully that helps? If I misunderstood, I apologize. Peace.salam.shalom (talk) 21:56, 11 December 2020 (UTC)

@Peace.salam.shalom: Thanks! Should have it working now AndrewOfWyntoun (talk) 10:32, 14 December 2020 (UTC)

Using px vs Em Values...

Here Page:Famous history of the learned Friar Bacon (1).pdf/1, absolute pixel values have been used. Whilst this works for screen display, it's typically better to use em based values, which scale when the text is re-rendered on other devices such as e-book readers or mobile devices. See Help:Preparing_for_export#PXWIDTH ShakespeareFan00 (talk) 14:51, 13 February 2021 (UTC)