User talk:Hesperian/Archive 8


 * The following text is preserved as an archive of discussions at User talk:Hesperian. Please do not modify it. Subsequent comments should be made on User talk:Hesperian. No further edits should be made to this page.

Question about reprint of The Time Machine
Hi, since there is Index:The Time Machine.djvu and Index:The Time Machine (1st edition).djvu, which is the preferred one to work on, since they appear to be identical? Regards, Mattisse (talk) 15:14, 9 October 2010 (UTC)
 * The short answer is that I uploaded the latter because I'd much rather work on a first edition than on some random later edition.
 * The long answer is that I uploaded it so that I could page out and proof our mainspace The Time Machine, which is an old featured text. However I've since learnt that there are two completely different versions of The Time Machine. These djvu files are the 'Holt version', published by Henry Holt & Co in New York. The mainspace text is the 'Heinemann version', published by Heinemann in London. The Time Machine will need to become a versions page, the mainspace version moved to The Time Machine/Heinemann version, and one of these djvu's manually proofed and transcluded to The Time Machine/Holt version. You are welcome to work on this / take it over if you like; for me it is a background task amongst many other things I'm working on.
 * Hesperian 23:27, 9 October 2010 (UTC)
 * I've made the move to The Time Machine (Heinemann text), and set up a version page for The Time Machine (Holt text). Hesperian 02:07, 10 October 2010 (UTC)
 * I have been working on Index:The Time Machine (1st edition).djvu,  proofing the pages  per my understanding of what you said above. Is that worth doing? Should I stop?  I’m confused Best, Mattisse (talk) 23:29, 10 October 2010 (UTC)
 * Yes, that is what has to be done. I'll start transcluding them in a sec. Hesperian 23:31, 10 October 2010 (UTC)
 * The main page with contents, and the first two chapters, are now at The Time Machine (Holt text). :-) Hesperian 23:42, 10 October 2010 (UTC)
 * ok, thanks! I was confused about what was happening. Mattisse (talk) 23:52, 10 October 2010 (UTC)
 * Yeah, our proofreading setup is extremely confusing when you first start grappling with it. Hesperian 23:56, 10 October 2010 (UTC)

Manipulating with class=headertemplate
We used to be able to utilise the class headertemplate and have it apply the colour and borders in a controlled fashion, ie. within a table of a certain width. However it seems that the recent mods to common.css/.js no longer allow us to do so. Trying at Devonshire Characters and Strange Events to add the link to the index, though to differentiate it from what actually appeared in the ToC of the work, and now it makes it full width.

Of course, it may be that my brain is still just too tired after a busy week away and I am missing something (un)obvious. — billinghurst  sDrewth  00:14, 14 October 2010 (UTC)


 * Is that what you're after? Unfortunately "headertemplate" only works if applied to a table. You want to apply it only to the last row of a table. If you close the table you're in and open a new one, the class defaults will apply and you'll get full width on the new table. The trick is to leave your table open, and embed a new table within it. Hesperian 00:41, 14 October 2010 (UTC)
 * Thanks, yes, and when it was within prose I believe that it kept to the width. Thanks for fixing, as I said, brain is still not back into the mental gymnastics of css. — billinghurst  sDrewth  11:31, 14 October 2010 (UTC)

Thanks!
Thanks for reverting my page. (I didn’t figure out what was happening at first!) Regards, Mattisse (talk) 00:14, 19 October 2010 (UTC)
 * welcome. Hesperian 00:20, 19 October 2010 (UTC)

Houston book
Please see Featured_text_candidates WhisperToMe (talk) 09:17, 21 October 2010 (UTC)

Images
I've done a few of the pictures, but I am done for the night if you would like to continue working on them. Sorry for the edit conflicting earlier...I didn't notice. If you wouldn't mind, could you look over my pictures to see if they are rotated the right amount. I would love a second opinion. Thanks &mdash;Xxagile (talk) 01:30, 22 October 2010 (UTC)
 * No worries; it wasn't an edit conflict in the technical mediawiki sense. I cropped and rotated an image, but when I went to upload it I found that you had uploaded a cropped and rotated image a few minutes earlier. Okay, I'll have a look shortly. Hesperian 03:07, 23 October 2010 (UTC)

consideration
I considered it, or rather, I tested it … it works, what else is there to consider! :-)

Can I assume that useless strokes like cntrl-x (random page) can be overwritten in the same way? It is perhaps haphazard, or 'unrecommendable', to overwrite many of the keystrokes, but there is no reason to leave a Page in edit-mode (one risks losing data in some browsers).

[musing] This is, however, contra to the mnemonic cue of these, but I want the lazy left hand workin' and fancy adopting the gaming  (author, save, [but not] delete, watch)  and so on. I'm an instant fan of the rest of your scripting, the make ref and overflow gizmos are big time-savers. My work-flow is evidently different though, I hunt out the word breaking hyphens and new lines before the cleanup; this is possibly be too clever by half (knowing that there is a transition from things like folk-lore to folklore), because when I forget the compound usually turns up in the spell-check.

Secondly, is there a crack for pagination with roman numerals? I love those introductions, but need to add them as I go.

Thirdly, an issue woven from complaint, do I have to create a new skin to shed the useless buttons and suites of code given as default. I probably, for example, use redirect here more than the other place but I have to scroll to my 'Standard suite' to click that (having three areas of the screen for these things is bugging me). There are about 9 actions that I perform repetitively, 12-15 by most others I imagine, but we are lumbered with stuff we almost never use. I do see the sense in having the infrequently used stuff available in the drop-down menus, just not taking up my valuable screen space.

Parenthetically, cygnis insignis 15:39, 24 October 2010 (UTC)


 * Second things first... roman function. With this, you should just be able to change "(pagenum-20)" to "roman(pagenum-20)", and get roman numerals instead. Completely untested; pardon the hubris; let me know if it doesn't work, and I'll fix it. Hesperian 23:35, 24 October 2010 (UTC)
 * First things second... I tried clobbering some of these shortcuts with shortcuts of my own, and it didn't work. Either they are inherently unclobberable, or they get loaded last and clobber mine, or my browser just doesn't know what to do with the situation. Hesperian 23:42, 24 October 2010 (UTC)
 * Three: I don't think the useless buttons are under the control of your skin. Even if you edit under the "simple" skin you still get them. Have you tried unchecking the boxes in Special:Preferences | Editing | Beta features? Hesperian 23:48, 24 October 2010 (UTC)
 * I didn't follow the link, and I assumed that 'roman' was recognised like the index. Dunno why it doesn't work either, though I see a glimpse of how. cygnis insignis 13:14, 25 October 2010 (UTC)

my roman is working now; just a stupid coding bug. Hesperian 10:32, 28 October 2010 (UTC)
 * Neat. I discovered this url mix:  ! Perhaps these can be avoided; I also noticed you abbreviating them in yr spec. frmt, have you found any problem with that? cygnis insignis 11:14, 28 October 2010 (UTC)
 * Nah, you just need any unique string in the url, so in that case you could use "s_household_tales,_volume_2". Hesperian 11:45, 28 October 2010 (UTC)

Use page image vs page contains image
These two templates seem to be a variations without clarity of any difference in outcome and purpose between the two. Got an opinion that you would care to share? — billinghurst  sDrewth  01:32, 3 November 2010 (UTC)

Further note to Bot requests &hellip;

Discussion point. There are examples where use page image has been added to the page, however, the image has been added not in page sequence, instead it has been added with the appropriate part of the text, an example that I did is Page:Notes on the churches in the counties of Kent, Sussex, and Surrey.djvu/49, so the page itself is not problematic, at the same time this example doesn't have text, so is it proofread or not? — billinghurst  sDrewth  01:50, 3 November 2010 (UTC)

Personally this is what I would do. Feel free to revert;it just seemed easier to show than to tell. Hesperian 05:30, 3 November 2010 (UTC)


 * Thanks, though seems a lot of bother, and some potential for confusion, for not a lot of return. With ThomasV having updated the script where blank pages are transcluded and not displaying the page number, I have now just &lt;noinclude>'d a link to the image (with a note) and that allows a straight   run. Seems less confusing.  I do wish that one could utilise figure text without the surround border on images. C'est la vie. — billinghurst  sDrewth  11:40, 6 November 2010 (UTC)

User:Hesperian/Without text batch complete
I believe that the batch of images at User:Hesperian/Without text now has the requisite conversions of pages with images to Proofread. It would look as we need to revisit Help:Page Status and maybe some elements around the tag to update to our current reality. — billinghurst  sDrewth  10:41, 18 November 2010 (UTC)
 * I am gathering that you are busy with other things, however, did you manage to get further? Could we even flick to 2000-3000? BTW the page shows up at Commons. — billinghurst  sDrewth  02:31, 14 December 2010 (UTC)
 * Yeah, I have been very busy on other stuff. Likely to ease up a bit now, so maybe I'll be around here a little more. Thanks for the poke. I've posted the full gallery. MediaWiki can't handle it all on one page, so split into batches of a thousand. Hesperian 02:47, 14 December 2010 (UTC)
 * Cool. By the way, if you are again disappearing from where you run power searches over the Dec./Jan. period, would you be so kind to do an update on Validated pages needing to be typeset before you fly. Thanks. — billinghurst  sDrewth  03:48, 14 December 2010 (UTC)
 * Yes I am, for much of it. Wilco. Hesperian 04:02, 14 December 2010 (UTC)
 * Done. Hesperian 23:34, 14 December 2010 (UTC)
 * Like yeah and ouch. User:Hesperian/V‎ (diff | hist) . . (+77,978)  — billinghurst  sDrewth 
 * What will be the future means to track these pages? Keep track of the day when they were marked NO TEXT, and then watch the last 1000 of a batch? (presuming that once we have done the lot, we only need to keep a track of the last XXX edited? To note that I am converting them from PageQuality to &lt;pagequality> — billinghurst  sDrewth  14:45, 18 December 2010 (UTC)
 * When you ask the API to tell you the contents of a category, you have the facility to ask for timestamps indicating when each page was added to the category. I already use this facility elsewhere, so it should be a simple matter to restrict any updates to pages tagged into Category:Without text since last time the script was run. Hesperian 22:57, 18 December 2010 (UTC)
 * Ah, code for blackmagic. One day I will look at and try to understand the API. — billinghurst  sDrewth  00:07, 19 December 2010 (UTC)
 * Heh, when that day comes I'll be happy to talk you though it, donate my scripts, etc, whatever. Until then, I'm usually happy to script stuff if asked. Hesperian 01:00, 19 December 2010 (UTC)

I just implemented and ran an updated version of the script that only returns pages categorised as 'Without text' after a given time. I wasn't expecting much—since it has only been nine days since the last run—but there were about a thousand hits, apparently because your changes were touching the timestamp. It did, however, turn up six pages that had been pushed back to 'Without text' shortly after you promoted them to 'Proofread'. This demonstrates that my new version of the script works and should meet our needs, and also suggests that you may need to evangelise your changes somewhat. Hesperian 04:10, 23 December 2010 (UTC)
 * One can hope that the discussion that you started about indices will shake out a few components, and we will can paste it together into something linking from Style guide and Help:Proofread, give it a shortcut from there when we do the updates, we can have a link to it specifically. We may also be more obvious about the incorrectness if we revert a change from any of the proofread stages to without text. I also think that the text for the label without text may be part of the problem. — billinghurst  sDrewth  10:06, 23 December 2010 (UTC)
 * Yeah I have this notion of enhancing the script table with a column that says what the index status should be. That would depend on unwanted pages being tagged as "Not transcluded". Then the index status depends on the lowest status page not so tagged.
 * I'm finding lots and lots of index page histories with what almost amount to edit warring over status. I say 'almost' because the main culprit seems not to watch these pages nor check the page history, so is probably unaware that they have tagged the same page as 'Done' four or five times against a range of people demoting it.
 * I wouldn't be averse to adding documentation to the index page javascript itself. "Source file is an excerpt of a larger volume, or a mixture of several sources" is hell long, so there's no reason why "Done" can't be "Done: all transcludable pages are validated or without text".
 * Unfortunately "validated"—>"without text" can't be reverted, because "without text"—>"validated" is not permitted. But "proofread"—>"without text" can be reverted, and, as you say, probably should be. Hesperian 11:13, 23 December 2010 (UTC)

Again? Billinghurst (talk) 12:44, 14 March 2011 (UTC)
 * Wilco. In the morning, if I remember.... Hesperian 14:16, 14 March 2011 (UTC)
 * Done. 1955 pages. Hesperian 00:21, 15 March 2011 (UTC)

Second opinion User talk:The Tetrast
Has added an excerpt and not within WS:WWI, am I being too nasty to push on "plan and typeset for all"? — billinghurst  sDrewth  02:19, 20 December 2010 (UTC)
 * The way I see it, if we had a textual framework in place for the Century Dictionary, and The Tetrast had been able to submit this at the title The Century Dictionary/Supplement/Pragmatism, then we'd be saying thankyou very much, and we certainly wouldn't be calling it an excerpt. Are you comfortable telling The Tetrast that Century Dictionary articles are unwelcome here until someone puts in place a textual framework for the entire multi-volume work? I'm not. Hesperian 02:59, 20 December 2010 (UTC)
 * Comfortable? Not the choice of word, though I have done it. — billinghurst  sDrewth  10:28, 20 December 2010 (UTC)

Incomplete indexes.
Hi Hesperian. I just glanced at your Scriptorium post regarding Index:1947SydneyHailstorm.djvu and aside from the contents of your post, I would gladly clean and upload the missing images. All I would need is the source of a higher definition image (like .JP2) to download.Ineuw 22:19, 21 December 2010 (UTC)
 * Thanks for the offer. I've been doing quite a lot of this kind of thing lately, and this was on my list of things to do. Normally I'd be happy to divide the labour, but...
 * The file page says that it is based on a private PDF, so apparently no online source exists. The only way to obtain images is to decode them from the DjVu file, which, fortunately, is quite high resolution, though the scan quality is low. But the time I've decoded the DjVu file, cropped and balanced the images, and uploaded them to Commons, that's the job 95% done, so you may as well leave it with me...
 * Unless you want to download, install, and learn to drive djvulibre and imagemagick for yourself, in which case I'm happy to leave it with you.
 * Hesperian 00:31, 22 December 2010 (UTC)

Manar Group
I know you hang out here more, I left a note on W:User talk:Hesperian. Jeepday (talk) 19:36, 24 December 2010 (UTC)
 * Thanks mate. Hesperian 03:28, 25 December 2010 (UTC)

Linking to authors/people where they are the subject
I need somewhere to throw about a thought, especially as you someone prompted the work that has come about in the Portal: ns. I am looking at how we designate our linking, and trying to get the right emphasis


 * Where a complete work is about an author, it is fairly cruisy, we can link to the author page, and the author page links to the work, though I wonder whether the linkage is suitably highlighted.
 * Where the component of a work is a small, especially encyclopaedic, eg. DNB, SBDEL, ..., then we can fairly easily link using traditional wikilink at the top of the article.
 * Passing mention, again we can wikilink.

What I am not so certain or clear is when we have chapters, or parts of chapters. We can easily add sister links in the header through the adaption/amalgamation of plain sister. However, we seem incapable of/unthinking towards the linking to our own works by a similar clean and obvious method. Backlinking seems functional in that we can link

Similarly, the portions around indexes and portals are butt ugly, and maybe usable for how we meld into other works, they are however, butt ugly in implementation (IMNSHO). We also haven't done as well in working out how to do the Portal linking for significant people and we are still linking through to WP through plain sister, or direct wikilink. And/or maybe we need to do more in the disambiguation space to develop pages that collect such information, however, again, we don't have the ready means to point to them internally. Cyg identifies that the use of similar often passes him by.

A blowfly on this ice-cream is that with the developing use of  eliminates many of the internal embedding of our linking template, and we are almost needing to look to linking/compiling outside of header

So in summation, it seems that we need a neat and stylish way to better link to internal works. — billinghurst  sDrewth  13:32, 28 December 2010 (UTC)

Warning about abuse
You have no right to issue "warnings" in a manner as you did here. After asking multiple people in IRC, not one found anything resembling your claims in my words. I have not even come close to violating our standards, but many people were aghast in your initial incredibly hostile use of the privacy policy in a very misleading manner. Accusing others inappropriately of incivility while making personal attacks and such is very inappropriate, especially for an admin. Maybe when you have gotten some sleep you will realize the error in your approach and manner. Ottava Rima (talk) 06:41, 30 December 2010 (UTC)
 * Didn't even read this. You are warned for the last time, you will be taking an involuntary wikibreak if this continues. cygnis insignis 06:43, 30 December 2010 (UTC)

Your edit contained url
Your edit still had the url, so being lazy (and about to out for a meal) I have just blocked the whole comment, rather than resurrect the witty component. — billinghurst  sDrewth  07:59, 31 December 2010 (UTC)
 * No dramas. I saw Cygnis reverted it, and couldn't be bothered saying it again. ;-) Hesperian 08:16, 31 December 2010 (UTC)

Kinda contradictory
There is a glaring ethical contradiction between the happiness with which you attack my name and the respect you'd like accorded yours. Both of us began on Wikipedia under our names, so that's no different either. I could easily say, "Ah, but it was your choice to edit under your name, and you can't take that back!" Guess what, that's exactly what English Wikipedia's ArbCom said when they outed me. That's how this whole mess started. So how about everyone redact, I won't out anyone here, and we move on? If you wanna go tit-for-tat, you can kiss your pseudonymity goodbye, because I won't put up with being treated this way. 24.18.132.13 06:48, 2 January 2011 (UTC)
 * You want your name redacted? Sure. I figured you didn't mind, since you already announced it yourself. Hesperian 06:51, 2 January 2011 (UTC)
 * And that's why I stopped using it, because nasty unaccountable sysops decided to attack it. If this were really a civil environment, it wouldn't be a problem. Running around accusing people of crimes (e.g. "stalking", see my usertalk) and trying to get them blocked or banned ain't civil.24.18.132.13 07:08, 2 January 2011 (UTC)

Talkback
prod and poke — billinghurst  sDrewth  13:49, 3 January 2011 (UTC)

Abaout tl|left margin
You wrote: "(divs need to go on a separate line, else mediawiki rendering can be unusual)". Thanks, very interesting. But... precisely, when? why? --Alex brollo (talk) 15:31, 3 January 2011 (UTC)
 * See Scriptorium/Archives/2008-12. That's about all I know on the subject. Hesperian 00:11, 4 January 2011 (UTC)
 * I'll never rest, since I'll understand it fully. I found by experience that these kind of fuzzy behaviour are the key to understand deeply how wiki software runs. I too have been extremely frustrated by broken paragraphs at the end of a nsPage .... but I found the when and the why. :-)
 * See recent User talk:Inductiveload (under the title "You are invited :-)" for details. --Alex brollo (talk) 13:49, 4 January 2011 (UTC)

Can the discussion look to begin?
I would like for the discussion around Poetlister and his ancillary bits to begin. Is there information that you would like from me? I think that waiting longer means that the matter will lose impetus. I think that discussions at other places will just have to wax and wane, and no longer be a distraction to what we need to accomplish. — billinghurst  sDrewth  09:49, 4 January 2011 (UTC)
 * Wilco. Wait out. Hesperian 11:10, 4 January 2011 (UTC)
 * My offer (made at Meta last night) to assist with drafting was for real. What can I do to help? Beeswaxcandle (talk) 06:08, 5 January 2011 (UTC)
 * Send me an email and I'll cc you into the discussion. Hesperian 06:38, 5 January 2011 (UTC)

WYSIWIG djvu layer project
While thinking and testing my ideas about this fashinating topic, I'll write down anything here: User:Alex brolloBot/WYSIWYG djvu project. Yesterday in the evening I tested them on a bundled copy of file:Horse shoes and horse shoeing.djvu, a very large file; the idea runs, but scripts are a mess so far, I'll post them as soon as I'll be moderately satisfied about them... and I'll turn as soon as possible into a different approach, using an indirect version of the file, so allowing a much faster "try and learn". Consider that I didn't unzip so far your scripts... what a shame :-( . --Alex brollo (talk) 08:31, 19 January 2011 (UTC)
 * While exploring deeper and deeper the djvu text layer stuff (and your scripts: thanks!), a question: did any of you write a general XSLT file to transform XML coming from djvutoxml.exe into HTML? I can't write such a file so far... but I guess, it will be very interesting. And: wouldn't it be interesting to add book metadata to djvu text layer? And ... did you ever try to insert into djvu layer elements of wiki markup? They run perfectly, and the " character is the only one that needs care/escaping! Dsed format is very tolerant about special characters, much more than xml. I see interesting opportunities coming out. --Alex brollo (talk) 01:10, 25 January 2011 (UTC)

and don't play fair
Gday Hesperian. When dropinitialis used within block center (and I believe &lt;poem> is in the mix) one can need to force the width of block center to stop line wrapping. My lack of knowledge in the space doesn't give me a clue of which is the cause, and which is the effect. Can you help, or is that not possible at the moment? — billinghurst  sDrewth  22:45, 24 January 2011 (UTC)

Djvu text layer
First of all, please throw away all the scripts and files I sent you! Now things are different - very different. I'll send you a new zip, if you like.

Second: I successfully uploaded running wiki markup into a page, so that extracting the text layer by proofread extension returns a perfectly formatted page! Plase try to create it:Pagina:Indice delle prove.djvu/4; don's save it, take olny a look to preview, so other users can experiment the trick. --Alex brollo (talk) 09:42, 27 January 2011 (UTC)
 * I'm in touch with Jayvdb since there is that unexpected result from my layman tries - a promise of a wikirecaptcha. Well, my script can select and send to a rechaptcha procedure the doubtful OCR words - where a ^ character is usually added - but it would be great if even the words found by your algorithm could be sent to a recaptcha procedure. Would you like to come into the running talk? I shared my running scripts into a shared DropBox folder "as they are" (really it's not a backup folder, I work into it in my pc). I realized that I shared also the Hesperian folder, can I keep it there, or you sent me your files privately for personal use? Can I invite you too to the shared dropbox folder? --Alex brollo (talk) 17:36, 1 February 2011 (UTC)

re:edit wars
I was only aware of some of them, I've put Index:La Fontaine - The Original Fables Of, 1913.djvu back to be proofread as there are images still missing and I won't interfere anymore with Problematic Pages. --kathleen wright5 (talk) 07:55, 7 February 2011 (UTC)

Zodiac letters
I had started a discussion at Possible copyright violations before I saw that you removed the tag. I may have a faulty understanding of the copyright law.

If someone had a letter from George Washington, previously unpublished, and they published it in a new book then the letter would be protected by the copyright of the book, correct?

With the Zodiac materials, my questions were first pertaining to the channels by which they were published (i.e. newspapers) but secondly, what is the source of the digital files. If they are published via someone else's website...would that not be a copyvio? Thank you, ⋙–Berean–Hunter—► ((⊕)) 01:00, 23 February 2011 (UTC)

Scriptorium
Can I be bold to specifically point you towards Scriptorium. You have done scripting for your personal js file that we may be able to utilise around the odd and even page numbers, and other bits if you think that you can help get bits organised for ThomasV. Thanks. — billinghurst  sDrewth  12:48, 5 March 2011 (UTC)


 * ''The above text is preserved as an archive of discussions at User talk:Hesperian. Please do not modify it. Subsequent comments should be made on User talk:Hesperian. No further edits should be made to this page.