Talk:Reading/Web/PDF Functionality/2017/09

About giving feedback

Please read Reading/Web/PDF Functionality and comment on the plans we lay out there, to tell us what you need from the PDF service. We're especially interested in what you need in the future that doesn't exist in the plans laid out there – if there's a bug with something that should work right now (e.g. you get an error message when you try to create a PDF), we need to fix it, of course, but that would have been on the agenda anyway.

Editing the Table of Contents

Will the function for editing layout and titles allow editing the title of articles as they appear in the Table of Contents? This would be important to me, especially for including the lifespan dates next to the names of individuals in the contents menu.

I appreciate the chance to give feedback. Thanks. JohnGardner (talk) 01:12, 1 September 2017 (UTC)

Artifacts in Indic Script rendering

Specifically for Sinhala. Adaptive accents do not change/place where they should be. It may be the case for other types of scripts too. Amilapsn (talk) 10:51, 1 September 2017 (UTC)

@Amilapsn - thank you for your report! Could you give us some more details on where you're observing the issue - browser, OS, a sample article to look at? OVasileva (WMF) (talk) 14:44, 5 September 2017 (UTC)

I tried to recreate the issue. The problem occurred when I tried to export the wiki to PDF (two column). Now there's only one column option and it's fine and no artifacts can be found. @OVasileva (WMF) Amilapsn (talk) 17:37, 6 October 2017 (UTC)

cut and paste documents

I work often with Linux - for instance with KNOPPIX 7.7.1

Do not know, which tool is used to create pdf's.

For comparing I did for the article For example for articel https://de.wikipedia.org/wiki/PDF/A

a pdf creation (over my KNOPPIX ) while using - "Print Preview" - Print to File

                                                    File size: 50.3 kIB  and text can be marked and copied

and another pdf by using the offered dowload pdf function

                                                    File size: 78.0 kIB  and text can   n o t   be marked and copied.

Besides I think, copying would be possible from the original wiki page (without print preview) so there is alway a pollibillity to copy - but not perfect formated for printing ;) 77.190.210.181 (talk) 09:51, 2 September 2017 (UTC)

What software are you using to view the pdf and to copy and mark the text ? For me both the 1 column and the 2 column version when opened on the Mac using Preview.app can be copied and marked. —TheDJ (Not WMF) (talk • contribs) 14:10, 2 September 2017 (UTC)

other pdf-creating tools - maybe i Linux

I do not know with which operating sy<stem wikipedia works - Linux or Windows???

Maybe this is interesting https://wiki.ubuntuusers.de/xsane2sandwich/.

Maybe you/or the german wiki project can get helb from http://exactcode.com/opensource/exactimage/. They seem to make some pdf-code and offer some open source tools (maybe they can help - when they have free timeI) Ifound it while I read the first obove link - there was an info about using "exactimage" so I found the company.

Much luck and thanks for your wiki work.

Greetings M 77.190.210.181 (talk) 10:00, 2 September 2017 (UTC)

downloading PDf went smoothly thanks Diana

downloading PDF went smoothly. Thanks Diana 110.143.225.72 (talk) 06:58, 3 September 2017 (UTC)

Thank you for giving feedback Diana ! —TheDJ (Not WMF) (talk • contribs) 18:12, 5 September 2017 (UTC)

Download of PDF went beautifully. Thanks for your hard work. 2601:642:100:DBB0:74CE:49F3:4080:3089 (talk) 18:46, 5 September 2017 (UTC)

perfect!

downloaded quickly and perfectly! 2607:FEA8:875F:EF98:5C4:6775:E4C6:8907 (talk) 18:30, 3 September 2017 (UTC)

Book generator skips some formulas and the diagram

Error report: In https://de.wikipedia.org/wiki/Metrischer_Raum the pdf print is fine, but the eBook generator skips content. Three formulas on first page missing, section "Formale Definition", (1) Positive Definitheit, (2) Symmetrie, (3) Dreiecksungleichung. Diagram missing on last page, in section "Einordnung in die Hierarchie mathematischer Strukturen" 2A04:4540:1108:5001:EC93:6467:943D:41DA (talk) 09:30, 5 September 2017 (UTC)

Thanks for reporting. Johan (WMF) (talk) 13:52, 9 September 2017 (UTC)

Downloaded pdf of some svg files and it converted them to bitmap

This made them not useable to me and i still had to get all 115 images one by one. would be nice if the pdf kept the original formating or at least kept vector drawings as curves not pixels. 82.16.213.83 (talk) 14:29, 5 September 2017 (UTC)

We don't have pages with SVGs. All our original SVGs are always rendered as PNGs (because some of the SVGs are several megabytes) and as such will also wind up in the PDF as PNG images. There is a separate improvement task on this topic, but it is not specific to PDF print functionality.

This problem is tracked here: https://phabricator.wikimedia.org/T5593

Thank you for giving feedback ! —TheDJ (Not WMF) (talk • contribs) 18:12, 5 September 2017 (UTC)

PDF Pages Odd Size

The pdf created has an odd size of 8.26 x 11.69 inches. Most common sizes are 8.5 x 11 inches (letter) and 8.5 x 14 inches (legal). 107.77.197.173 (talk) 16:48, 6 September 2017 (UTC)

That is the metric A4 paper size, standard almost everywhere outside of the U.S.

Is it possible to get the user's default paper size from the browser? Ideally there needs to be an option for the user to choose paper size, even if the user's default is read somehow. Gpc62 (talk) 17:09, 6 September 2017 (UTC)

"Is it possible to get the user's default paper size from the browser".

That's not possible. —TheDJ (Not WMF) (talk • contribs) 06:45, 7 September 2017 (UTC)

Empty page at end of file.

I retrieve PDF from "https://en.wikipedia.org/w/index.php?title=Work-conserving_scheduler&oldid=784337814" and it has one empty page. Qin-nz (talk) 10:28, 8 September 2017 (UTC)

@Qin-nz, what web browser and operating system are you using? I can't see the same problem. CKoerner (WMF) (talk) 15:56, 8 September 2017 (UTC)

@CKoerner (WMF) I am using chrome 60 on macOS 10.12. And select single column. You can get the pdf file here https://hdy.blob.core.windows.net/public/Work-conserving_scheduler.pdf Qin-nz (talk) 08:42, 12 September 2017 (UTC)

There is no extension when downloading

When you click download, the file does not have an extension，manually adding extensions does not affect opening files.

By the way, I'm in PRC 于璐铭 (talk) 12:06, 8 September 2017 (UTC)

@于璐铭, I'm on a Mac so when I download I get the extension. Can you tell me a little more about which browser and operating system you are using? CKoerner (WMF) (talk) 15:54, 8 September 2017 (UTC)

I'm using Windows 7 system and a Chinese browser called 360 broswer. 于璐铭 (talk) 14:00, 14 September 2017 (UTC)

not able to download book

The "download" button is dark and their seems no remedy to an inability to download the book that I created. (lot of time wasted) I also received an error message when I downloaded to PediaPress. 2606:6000:D6D0:BD00:9066:4A57:5F5B:D42F (talk) 06:16, 9 September 2017 (UTC)

+ Var-sasha (talk) 04:04, 19 September 2017 (UTC)

Thanks for reporting the issue. Could you tells us roughly what you tried to download, which broswer you're using (and version) and on what OS (and version)? Johan (WMF) (talk) 13:52, 9 September 2017 (UTC)

+ Var-sasha (talk) 04:04, 19 September 2017 (UTC)

Hi, i have same issue, can i fix this ?

Google Chrom 61.0.3163.79

Win 10 - 1703 (15063.608)

https://ibb.co/n5p7Ok Vick416 (talk) 11:59, 14 September 2017 (UTC)

+ Var-sasha (talk) 04:04, 19 September 2017 (UTC)

This is a disaster , how are we suppose to download if the download button is not working)😠 L.z.motloung55 (talk) 19:04, 17 September 2017 (UTC)

+ Var-sasha (talk) 04:04, 19 September 2017 (UTC)

(Please don't drop) Editable formats: ODT, plain text

As the PDFs are huge and not editable, I tried the "print version" which unfortunately still shows the wiki navigation column rather than being a single-paged html version of the book.

Plain text schould please not contain any other line breaks than the ones separating paragraphs so that text can be reflowed in e-readers (especially useful on small screens / mobile devices).

Last, the functions should preferably available w/o JavaScript (accessibility, security).

Thank you! 2.247.249.142 (talk) 10:03, 9 September 2017 (UTC)

Thank you for you comments. Johan (WMF) (talk) 13:50, 9 September 2017 (UTC)

Some fonts are missing

eg. in "Brahmi script", pdf file showed boxes instead of letters. 71.207.150.75 (talk) 17:40, 9 September 2017 (UTC)

Different use of text colors

There are inconsistencies within the one-column layout and a variety differences with the two-page layout:

1) Hyperlinks in "Selected Articles" are not shown by different color 2) The notes in the "Notes" section also have a strange color layout: the note itself (which is a hyperlink) is normal black but the additional information, if any, is printed in grey. 3) In the "Further Reading" section: here as well there is no difference between regular text and hyperlink. And some other color differences, most involving hyperlinks.

Article: https://en.wikipedia.org/wiki/Stuart_Kauffman 78V2z (talk) 21:25, 9 September 2017 (UTC)

We will be implementing some styling improvements as a part of the new PDFs, which will fix these issues. However, the new styles will not be identical to the 2-column layout - we've made some improvements to fonts, hyperlinks, and overall look and feel. An example of the updated styles can be seen in the design section of the functionality requirements. OVasileva (WMF) (talk) 15:55, 12 September 2017 (UTC)

Very small page margins and no page number or paragraph indicator in header / footer

Compared with the two-column layout I see very small page margins, especially left, right, and bottom. There's also not page numbering or paragraph indicator in the header (or footer).

Article: https://en.wikipedia.org/wiki/Stuart_Kauffman 78V2z (talk) 21:28, 9 September 2017 (UTC)

Formulas: bold and sometimes too large

Not sure how Latex is being implemented but there's something wrong with the formulas; and I see this on a variety of pages where math formulas are used.

(1) Sometimes all the formulas in the article are the correct size but they all are in bold (and they shouldn't).

(2) Sometimes the integral sign and the rest of the integral aren't on the same "line", meaning they're not aligned.

(3) Sometimes some parts of the formula are bold and some aren't but none should.

(4) Sometimes part of the formulas are in different size that other parts, e.g. integral sign is too large compared with what's behind it.

P.S. The two-page layout has similar issues. 78V2z (talk) 21:41, 9 September 2017 (UTC)

Thanks. Johan (WMF) (talk) 17:58, 10 September 2017 (UTC)

Seems like glyph fallback issues, are the math fonts installed on the rendering machines ? —TheDJ (Not WMF) (talk • contribs) 06:17, 12 September 2017 (UTC)

@78V2z - the bolding I believe is due to the browser itself, but I can double check. Could you give us example articles of the other issues (2-4)? OVasileva (WMF) (talk) 15:51, 12 September 2017 (UTC)

@OVasileva (WMF) - I don't think the browser has anything to do with it. The PDFs are generated on the server side. However, I have tried both Firefox and chrome and several tests yielded exactly the same documents.

I had a nice example of errors 2 to 4 but can't find it anymore.

I did find an example for 4 but it's on the two-column generated PDF. Anyway, it's an interesting example that shows some very severe issues where the Latex code isn't being resolved at all.

https://en.wikipedia.org/wiki/Maxwell%27s_equations

An example of the alignment error (#4) can be seen in the paragraph "Charge conservation" on PDF page 5, bottom right.

Hope this helps.

When I run into another of these issues I'll post them here. 73.97.162.120 (talk) 04:17, 13 September 2017 (UTC)

@OVasileva (WMF) - Actually, the 1-column PDF of Maxwell's Equations shows most formulas in bold but some not. See an example in the paragraph "Formulation in Gaussian units convention" on PDF page 3, at the bottom, in the text.

Article: https://en.wikipedia.org/wiki/Maxwell%27s_equations 78V2z (talk) 04:28, 13 September 2017 (UTC)

That's intentional isn't it? The section even mentions: "Symbols in bold represent vector quantities, and symbols in italics represent scalar quantities, unless otherwise indicated." —TheDJ (Not WMF) (talk • contribs) 14:15, 13 September 2017 (UTC)

@TheDJ - You're right 73.109.63.144 (talk) 17:41, 13 September 2017 (UTC)

@TheDJ - I think part of the confusion is that on one-column PDFs the difference between bold and not bold isn't as outspoken as on the two-column PDFs. Not sure why that is but it strongly affects readability. Take a look at the two formulas on page 9 of the one-column PDF, right after the sentence "...by splitting the total electric charge and current as follows:". It's as if normal became bold and bold became super bold...

I actually just noticed that the two tables at the beginning of the article didn't make it to the two-column PDF. I'll open a new thread for this issue. 78V2z (talk) 18:06, 13 September 2017 (UTC)

@78V2z - Yes, the fonts are too heavy in displayed equations in one-column format. Looking at the Maxwell's equations page: In the one-column pdf, the normal-weight characters in the displayed equations are about as heavy as the bold characters in the body text. Then the displayed bold characters are even heavier. That makes it harder to read which are bold and which are normal.

The type in the 2-column pdf is much clearer. In 2-column, the fonts in displayed equations seem to match those used in the body text. In 1-column format, the fonts don't match, which can be confusing.

On the other hand, look at what a mess 2-column format makes of some of the displayed equations! eg, the equation just above the section header "5 vacuum equations..." on page 5. The version in 1-column format (bottom of page 7) is much better, despite the too-heavy fonts. Gpc62 (talk) 19:21, 13 September 2017 (UTC)

@Gpc62 - Exactly. The formula you mentioned is indeed a mess; the line break is at the wrong location and the integral sign isn't aligned vertically. I've already mentioned this in another thread. The biggest thing with the 2-column layout though is that all the large tables are omitted - they're simply not there. I've started a separate thread on this issue as well. But when it comes to readability the 2-column layout wins hands down, every time. 78V2z (talk) 03:59, 14 September 2017 (UTC)

"Sometimes part of the formulas are in different size that other parts, e.g. integral sign is too large compared with what's behind it." This might have to do with the fact that at several points, the integrals are actually images, because it's a notation that is not properly supported in Latex.

See also: https://en.wikipedia.org/wiki/Template:Oiint —TheDJ (Not WMF) (talk • contribs) 14:18, 13 September 2017 (UTC)

5.11.35.19 frameware 70.74.132.64 (talk) 18:54, 24 September 2017 (UTC)

A Couple / Few Observations

First, Thanks to y'all for your valiant efforts creating this functionality

1) Both single and double-column generation worked fine for me (p.s., I like the 2-column version, makes for easier reading)

2) I noticed, when generating pdf's for the article at WikiSource (https://en.wikisource.org/wiki/Occasional_Discourse_on_the_Negro_Question) the Author was missing in the pdf's.

3) The pdf's font size is a bit small for my ageing eyes; I hafta zoom in a couple of notches on my 17" laptop to read it (yes, with my reading glasses on :-) )

4) When checking functionality of hyperlinks in the pdf's, most seemed to work OK, but I happened upon a link to works @ "Internet Archive" that didn't work as expected – it loaded the Internet Archive website but the search field was filled with the likes of "%28%28subject%3A%22Carlyle%2C%20Thomas%22%20OR%20subject%3A%22Thomas%20Carlyle...". The link in the original Wikipedia article works OK) ref: https://en.wikipedia.org/wiki/Thomas_Carlyle 72.173.49.17 (talk) 15:11, 10 September 2017 (UTC)

Thanks for your feedback. Johan (WMF) (talk) 17:58, 10 September 2017 (UTC)

"The pdf's font size is a bit small for my ageing eyes; I hafta zoom in a couple of notches on my 17" laptop to read it"

Please note that sometimes PDF reading programs open the page at 75% size or something like that by default. —TheDJ (Not WMF) (talk • contribs) 06:15, 12 September 2017 (UTC)

m unable to download my book Siddharthjarse (talk) 04:00, 24 September 2017 (UTC)

please help Siddharthjarse (talk) 04:00, 24 September 2017 (UTC)

two column !!!!!!!!

for two column you use >>> internalinstanceid="7"   for on e column >>> internalinstanceid="3" 116.58.232.128 (talk) 15:30, 12 September 2017 (UTC)

Could you help explain a little more? I'm sorry I don't quite understand. Is the suggestion to make them the same? If so, which do you desire? CKoerner (WMF) (talk) 17:26, 12 September 2017 (UTC)

1-column PDF doesn't number sections as on the 2-column PDF

The numbers are incredibly useful and it would be great to see this feature from the 2-column PDF as well on the 1-column PDF. - Thanks!

Article: https://en.wikipedia.org/wiki/Maxwell%27s_equations 78V2z (talk) 04:22, 13 September 2017 (UTC)

Latex code for formulas isn't resolved at all!

This is on the one-column PDF of the following article: https://en.wikipedia.org/wiki/Maxwell%27s_equations (The two-column PDF does it correctly.)

Instead of the formula it shows an "image" showing part of the Latex code. 78V2z (talk) 04:25, 13 September 2017 (UTC)

Contibutors on 2-column-document

The section with the contributors on the last page of a 2-column document expands to the whole width of the document instead of also being 2 half-size columns. 62.225.102.92 (talk) 09:06, 13 September 2017 (UTC)

Date and url, 'this page downloaded [date] from [URL]'?

Will the new version retain "Date and url, 'this page downloaded [date] from [URL]'", at least as an option?

That's listed on Reading/Web/PDF Functionality#Update After Consultation but not in Reading/Web/PDF Functionality#Differences between current and future implementation nor Reading/Web/PDF Functionality#Single Articles. The former suggests it is being retained but it's not listed in the latter. I'm confused.

I find it very valuable to have the date and URL:

• I sometimes print an article, mark up the paper version, make the changes, then reprint. Usually the reprint is a few days later, so I can easily see which version is later by checking the date.
• I also sometimes distribute copies of article(s) at meeting(s). With the URL, anyone can find that article on the web later, and the date eliminates confusion about different versions.

Thanks, ~ DavidMCEddy (talk) 15:03, 13 September 2017 (UTC)

Tables didn't make it to the 2-column PDF

The one-column PDF does show the tables, the two-column PDF does not. Not sure how to squeeze a table into the small columns though. I think back in the Latex days that it for that one table locally the one-column mode enabled. It's been a while though.

Article:

https://en.wikipedia.org/wiki/Maxwell%27s_equations 78V2z (talk) 18:30, 13 September 2017 (UTC)

Sections --> Bookmarks

The two-column PDF has a very nice feature that turns the sections into bookmarks. (It would even have been better if the bookmarks also included the section numbering.) The one-column PDF doesn't have this feature. It's a very useful feature, especially for long articles. 78V2z (talk) 18:35, 13 September 2017 (UTC)

Minor glitch in One Column PDF

Just downloaded a one-page PDF. No links are underlined or live, which is fine. I understand you are working on it. One thing IS underlined right now, however.

It is the phonetic alphabet characters used for pronunciation. In the online version, clicking them gives an idea about their pronunciation. In the PDF version, one is directed to a page describing all the IPA characters. Too Much Information for me. The page is https://en.wikipedia.org/wiki/Help:IPA/English Lou Sander (talk) 20:56, 13 September 2017 (UTC)

Cannot copy or save the document for use later

There is no capability to save the document for later use. the save button is inactive. This is a huge issue for continuing to use this site. 184.97.199.18 (talk) 18:36, 14 September 2017 (UTC)

+ Var-sasha (talk) 04:05, 19 September 2017 (UTC)

I have had the same problem. The print book links take me to the book publishing page, etc. and seem to work just fine. The download page, however, shows a greyed out, non-functioning link.

I've just started exploring the book function and think it's a great idea. Ingenuity Arts (talk) 19:50, 14 September 2017 (UTC)

+ Var-sasha (talk) 04:05, 19 September 2017 (UTC)

Printable html

Can we have the simplified clean html layout "printable" button back? More useful than PDF. Ross.woods1954 (talk) 02:34, 16 September 2017 (UTC)

+ Var-sasha (talk) 04:05, 19 September 2017 (UTC)

The download button is greyed out in Vivaldi

I am unable to create a book from collected articles in Vivaldi browser (Chrome-based) because the download button is greyed out and nonfunctional. 2606:A000:4740:A00:A53E:A307:64FA:B63D (talk) 05:45, 18 September 2017 (UTC)

+ Var-sasha (talk) 04:03, 19 September 2017 (UTC)

Under Firefox, the situation is even worse -- the book creator interface doesn't even show up. 2606:A000:4740:A00:A53E:A307:64FA:B63D (talk) 05:56, 18 September 2017 (UTC)

+ Var-sasha (talk) 04:03, 19 September 2017 (UTC)

Numbering pages

it would be nice to have the page numbers Ypirétis (talk) 09:17, 18 September 2017 (UTC)

agree Var-sasha (talk) 04:03, 19 September 2017 (UTC)

Very good

Thank you 176.58.100.116 (talk) 15:59, 21 September 2017 (UTC)

Soy tecnico en Reparacion y Mantenimiento de Computadoras; pero el curso lo hice en 2001. Con el tiempo estuve de a poco actualizando mis conocimientos, ya que tanto el software como el hardware han cambiado y bastante. Digamos que a falta de recursos economicos me he visto impedido de seguir especializandome en lo que me gusta. Del material expuesto aqui en esta pagina, solo puedo decir que es uno de los mas completos que he visto, bastante actualizado y muy interesante en sus explicaciones. Mi opinion es: Muy buen material, complet, muy util y bien explicado. 152.156.216.114 (talk) 06:09, 23 September 2017 (UTC)

Single Column STILL Hangs or Takes VERY Long Time...

'Nuff Said. Are You EVER Going To Fix This Ridiculous Issue?????? 68.98.170.156 (talk) 15:15, 25 September 2017 (UTC)

It is indeed our plan to make it work as well as possible. Johan (WMF) (talk) 15:18, 25 September 2017 (UTC)

ممتاز 196.82.47.125 (talk) 21:53, 25 September 2017 (UTC)

Appreciate!!

I truly appreciate the ability to change the webpage into a .pdf!! Instead of the old highlight, copy, paste. etc... It may not be perfect yet, but it is a long ways from where we used to be!! 2601:603:5100:2C3C:FC6A:411:12D2:48AC (talk) 03:03, 26 September 2017 (UTC)

Thanks. Johan (WMF) (talk) 17:52, 26 September 2017 (UTC)

Limit line length?

How about limiting the line length? According to Line length, the optimal length for printed text is 66 characters per line. Florian Blaschke (talk) 19:29, 27 September 2017 (UTC)

overlapping English characters at mywiki PDF using Electron

Before Electron, converting PDF with OCG at Burmese Wikipedia (mywiki) result in unreadable boxes. When converting with Electron, Burmese characters on PDF are displayed correctly but not English characters. The English characters are overlapped on converted PDF. I think it is due to old Padauk font that used in PDF. Current version of Paduak Font is 3.003. Is there is anyway to upgrade to latest Padauk font or using other Burmese Unicode compatible fonts like Noto Sans Myanmar? Please see sample PDF of main page of mywik here. Ninja✮Strikers ^«☎» 09:59, 28 September 2017 (UTC)

Thank you for reporting. Johan (WMF) (talk) 12:18, 28 September 2017 (UTC)