Index talk:Imperialism (Lenin).djvu

Percy Reginald Stephensen

Note that the Australian Percy Reginald Stephensen claimed credit for the translation and that it was published originally 1925/26. 1917 is the date of the original. MarkLSteadman (talk) 00:03, 21 December 2021 (UTC)

Good info. Sounds like it should be localised then, based on UK publication and author death in 1965. Inductiveload—talk/contribs 09:14, 21 December 2021 (UTC)

Quality of proofreading

Casually reading these pages I am coming across typos, missing words, wrongly inserted words on several pages already marked as proofread. I feel that before marking a page as proofread those really need to be picked up and corrected. The point here, surely, is to be accurate rather than swift, and leave few—and hopefully no—errors for validators to check. I've not updated the status to validated even where I think the page is accurate as I would only expect to validate a page to which I had not made a significant change. I'm new here btw so apologies for any clumsiness and I've no intention of slighting others' work on this which is probably considerable; but as an editor in a previous job I feel a need to raise this.

@Fairflow: Hello, first of all thank you for working on this index. On talk pages, it is best to sign one's own messages by typing four tildes: ~~~~. @Tylopous: Thanks for the reminder! I haven't got the habit yet.

As regards the proofreading process, it is important to consult Help:Beginner's_guide_to_proofreading, where it is encouraged to mark a page as proofread when one has corrected every error that one can find. Of course the ideal case would be that one sees every error, but this is sometimes hard to achieve for a single contributor. This is why the validation process is important, which has to be done by a second contributor independently. For a general discussion of the proofreading process, perhaps the best place would be the Scriptorium: Wikisource:Scriptorium. Tylopous (talk) 12:11, 16 January 2022 (UTC)

Thanks for this guidance too; I'm trying to learn the standards at the same time as learning the tools and techniques but I do hope I can be of service. I'll have a thorough read of the Scriptorium. Fairflow (talk) 13:43, 16 January 2022 (UTC)

@Fairflow See points below. Hopefully I am not too blunt, but I might be (I make no promises). Also, I have never been an editor in a past profession... And also, somewhat belatedly, welcome to Wikisource.

First, it is probably worth remembering that people on Wikisource are volunteers. I will go through a page twice when proofreading, once quickly, once slowly, no more. If it helps, remember there are 30000 pages on the monthly challenge and very few significant contributors (say 20-30).

Second, as to your comment "The point here, surely, is to be accurate rather than swift, and leave few—and hopefully no—errors for validators to check." -> How can any single person know the point of Wikisource when it is a community collaboration (also I believe Wikisource "declares" its point somewhere on the website if you would like to have a look for it)? Also, if there are no errors for validators, then why have pages be validated at all, as it would then be a "waste" of the validator's time (if there sole aim were error correction)? Admittedly, I agree it would be nice for there to be no errors, but personally, it is unlikely to happen. Even once a page goes live (is transcluded) it is easy enough for anyone reading to correct anything they think is an error (whenever I see these things on live pages I go back to the main document and correct if need be, so things "can" have more than two passes even once transcluded), if that helps set your mind at ease. Finally (for this point), I agree that no errors pre-validation would be nice (but sadly not realistic).

Third, if you are new at Wikisource, I can promise you that one edit per page or two (what you seem to have observed, with a very small sample size mind you) when going from proofreading from validating is good (if not great proofreading). It may be worth setting the expectation at three or four changes per page for any text when the print quality isn't perfect. Maybe try some proofreading yourself and see how you go, to set the right expectation. (And yes, I realise you say "I've no intention of slighting others' work", but really that is just like saying no offense, before saying something that might well be construed as offensive).

Fourth, and importantly for this text (and maybe one of the most important points, as lots of proofreading relies on the OCR to transcribe the text), but there are lots of lines/scribble/nonsense all over the text. These screw up the OCR and force things to be typed by hand. You will notice that at least two of your errors were located on text which was underlined/scribbled over (meaning that I couldn't use the OCR). Even rereading these, it is really easy to miss things. Probably the only times when there are no errors going from proofreading->validating a text is when the OCR yields a near perfect text on the proofreading stage. If you do continue to validate this text, please focus on these areas.

Fifth, there are also things like tables and the like in this work, which are also very time consuming.

Sixth, I second @Tylopous's point, it can be really hard to see some of your own errors when proofreading, particularly when going through lots of pages. And as I said, there are lots of pages in the Monthly Challenge, and I dread to think how few we would get through if we required a perfect proofreading stage (I would argue that it would take more than twice as long for one person to proofread twice properly, than for two people to proofread and validate).

Seventh, in case it need be said, I am well aware that I am not perfect. But, and you might be surprised from your original comments, my track record seems to be one of the better ones when it comes to errors in proofreading.

Also, just to hijack my own message, but @Ashawley, please stop changing the style of the single and double quotes when validating, unless you plan to validate every page in the text (else it will be inconsistent). This is doubly important, as when inserting some, one was not inserted facing the correct way. I am switching them all back. Also, out of curiosity, why did you convert the paragraphs back to single lines? As far as I am aware this is work for no gain? The same is true when adding hws, or hwe, it is just another chance to introduce an error, without any gain in the final result? Or were you not aware there is no difference, if the word isn't "naturally" hyphenated? TeysaKarlov (talk) 06:08, 19 January 2022 (UTC)

Sorry for the confusion. Unicode single and double quotes and multi-line paragraphs makes it easier for me to proofread the original. Also, these are the default output of Transcribe text button when editing. The point of the {{hws}} and {{hwe}} is explained on Template:Hyphenated word start. --Ashawley (talk) 16:12, 19 January 2022 (UTC)

@Ashawley I am not sure what exactly you mean by default output of Transcribe text, but the style of quotes seems to greatly differ depending on what text is being transcribed (sometimes it generates two single quotes for a double quote, sometimes, even two double quotes for one double quote, sometimes straight, sometimes unicode). I agree that the unicode style are more often observed for this text, but in general, I would prefer consistency based on whatever has previously been proofread, rather than the default of Transcribe.

I was also thinking more about the mutli-line paragraphs and realised they could be more of an issue than I first thought. For example, if I start italicizing text with two single quotes (' '), and then there is a line break before the italicized text finishes, and the closing italicizaton quotes appear (' '), then the text will display incorrectly (it will italicize everything to the right of the starting quotes on both lines, rather than everything between the opening and closing sets). Hence, it is better to use single paragraphs of text for single paragraphs in the original.

As for hws and hwe, I was merely referring to the yellow information warning at the top of said link on the template hyphenated word page, which reads: "For simple hyphenations, use of this template is not necessary since the transclusion software connects the hyphenated word automatically. See also H:HYPHEN."

Hope this helps, and sorry for any confusion in the original message. TeysaKarlov (talk) 22:35, 20 January 2022 (UTC)

Ok, I wasn't aware about hyphenation being automatically handled by transclusion. I am aware of certain markup syntax not working well around multi-lines paragraphs. --Ashawley (talk) 23:41, 20 January 2022 (UTC)