Template talk:Encyclopaedia of Islam, Second Edition

Latest comment: 4 years ago by Srnec in topic "Search result"

Edition

edit

@Cplakidas: Please change "New Edition" to "Second Edition", since New Edition means 3rd. Thanks. AhmadLX-)¯\_(ツ)_/¯) 20:46, 21 July 2019 (UTC)Reply

AhmadLX, you are correct, but that the name under which the 2nd edition of the EI was published. EI3 on the other hand labels itself as "Encyclopaedia of Islam Three". Constantine 07:12, 22 July 2019 (UTC)Reply
Yes, but if you check RS, in few places is it cited as "EI New Edition", almost always as "EI2" or as "Encylopedia of Islam 2nd Ed." AhmadLX-)¯\_(ツ)_/¯) 17:28, 22 July 2019 (UTC)Reply
Also note that Brill uses "Encyclopedia of Islam, Second Edition" on the article pages. "Cite this page '“Muḥammad”, in: Encyclopaedia of Islam, Second Edition, ...'" AhmadLX-)¯\_(ツ)_/¯) 17:39, 22 July 2019 (UTC)Reply

proposed changes to this template

edit

cs1|2 automatically sets each template's anchor ID to the selected name list (first four) and the value assigned to |year= or |date=. Order of name list selection is |contributorn=|authorn=|editorn=. This template is a wrapper around {{cite encyclopedia}}.

When en.wiki editors use short-form citations, {{harv}} or {{sfn}} templates, Module:Footnotes attempts to read the citation templates in the article so that it can do error messaging and categorization. Wrapper templates like {{EI2}} hide the pertinent details so Module:Footnotes emits false positive error messages. There are ways around that. Module:Footnotes has a whitelist of anchor IDs that quash the false error messages. What the short-cite templates (via Module:Footnotes) do is detailed at Category:Harv and Sfn template errors.

I have come to this template from a discussion at Module talk:Footnotes § Encyclopaedia of Islam, New Edition whitelist where we are discussing how to improve whitelisting for this template. The live version of this template causes cs1|2 to create an anchor ID that looks like CITEREFEI2 which can be linked to with {{sfn|EI2}}. That sort of anchor ID is cryptic to readers: the short-form citation renders as a linked EI2 which cannot be found in any rendering produced by this template. This is aggravated because readers will expect that the text 'EI2' will be found at the left-most end of a rendered citation (where contributors', authors', or editors' names are to be found). Yes, the links from EI2 to the EI2 citation do work but, not when article pages are printed so the link and the pale blue highlight cannot be relied upon.

So, I have created {{Encyclopaedia of Islam, New Edition/sandbox}}. Here is a list of things that I have changed:

  1. sandbox uses Module:Template wrapper to render the citation. This allows the template to use any and all parameters supported by {{cite encyclopedia}} without the need to directly include those parameters in this template's code.
  2. added #default= statements to |editor-<first|last|link>1=|editor-<first|last|link>5=. These account for the collective editors named by Brill at their website.
  3. |editor-<first|last|link>5= parameter presets use code that is different from the code used by the other editor preset because #default= adds editor 5 parameters to volumes that don't have editor parameters for editors four and five.
  4. added ' (search results)' static text to the value assigned to {{cite encyclopedia}}'s |title= value when |url= and |doi= do not have assigned values and when |title= or |article= do have assigned values because under those conditions, this template creates a url to Brill's search engine. This may not be the best place for this annotation this because ' (search results)' will end up in the citation's metadata where it does not belong.
  5. modified |url= preset:
    when the template's |url= parameter is set, use that value; else
    when the template's |doi= parameter is set, do not preset |url=; else
    when the template's |volume= is set and:
    when |title= or |article= is set, create a url to Brill's search engine or
    when |title= and |article= are not set, create a url to Brill's EI2 'homepage'; else
    set |url= to empty; I could not find urls for online access to An Historical Atlas of Islam and Islamic Desk Reference: Compiled from The Encyclopaedia of Islam; if these exist, the code must be changed to accommodate their urls
  6. deleted all of the author-related parameters because they are not required when using Module:Template wrapper
  7. deleted the |lastauthoramp=y preset because that is a preference parameter that should only be set when all other cs1|2 templates in an article also use that preference
  8. deleted the |ref= preset to discontinue creation of CITEREFEI2; anchor IDs are created by cs1|2 from author parameters (if provided) or from the editor parameter presets; the collective rendering (the '12 volumes') illustrated at the top of the template's doc page can be rendered by setting |display-editors=0 otherwise the collective editor name-list is rendered:
    {{Encyclopaedia of Islam, New Edition/sandbox |display-editors=0}}
    The Encyclopaedia of Islam, New Edition (12 vols.). Leiden: E. J. Brill. 1960–2005.
    {{Encyclopaedia of Islam, New Edition/sandbox}}
    The Encyclopaedia of Islam, New Edition (12 vols.). Leiden: E. J. Brill. 1960–2005.

There are some examples that compare the live version with the new sandbox version in ~/testcases.

Comments? Opinions? Questions?

Trappist the monk (talk) 23:39, 18 June 2020 (UTC)Reply

Thanks for this work, Trappist. If my understanding is correct, any inline instances of {{sfn|EI2...}} or similar will need to be adjusted. My (possibly incomplete) survey turned up just these: Abbasid Samarra, Al-Abbas ibn Ahmad ibn Tulun, Enderûn, Hypotheses about the identity of Dhu al-Qarnayn, Ibn Zulaq, Itakh, and Sahib al-bab (but I'll check again). But some of them use an explicit |ref= such as (harvid|EI2|2002)—technically they'd still work but could also be modernized. David Brooks (talk) 14:17, 19 June 2020 (UTC)Reply
Adding a more substantive comment, on the Brill online URL:
  • Looking at the source, I think a constructed URL should have a "subscription required" comment, but it doesn't appear in the sandbox. A doi link also doesn't have the comment (you can access without subscription but that just gives me the first half dozen lines).
  • The site specifies a "Cite this page" format that's different from ours. And I'm not clear whether their requirement includes all three lines.
  • Do you know how stable the URL and search behavior of the site is likely to be? I know, hard to see the future, but apparently reliable external references have gone away in the past 15 years. That said, I don't have a better suggestion other than Eternal Vigilance. David Brooks (talk) 14:22, 20 June 2020 (UTC)Reply
Answers
  • this template is a wrapper template around a cs1 template ({{cite encyclopedia}}). cs1 does not support a "subscription required" comment. We should not add such an annotation to this template as it would be inconsistent with adjacent cs1|2 templates. In the sandbox, I specifically disabled the subscription-required icon because the search-results pages are not themselves behind the paywall and each result includes a lock icon indicating that the associated articles are behind the paywall. When editors provide explicit url, presumably directly to the cited article, then the sandbox template includes the subscription required icon. In cs1|2, sources identified with |doi= are normally locked away behind a paywall or registration barrier; cs1|2 does not highlight the norm; for free-to-read sources behind a |doi= identifier, we have |doi-access=free.
  • Many websites offer suggested citations in various formats – sometimes their own, sometimes standardized (apa, cmos, ...); en.wiki articles using cs1|2 (or templates wrapping cs1|2) are not obliged to use any of the offered styles. Were en.wiki articles so obliged, citation consistency would not be achievable.
  • Yeah, Eternal Vigilance.
Trappist the monk (talk) 22:36, 20 June 2020 (UTC)Reply
There having been no objections, I have updated the live version of this template. Articles using {{sfn|EI2...}} mentioned in the above discussion have been tweaked.
Trappist the monk (talk) 22:54, 25 June 2020 (UTC)Reply

And ... reverted by Editor Cplakidas with the edit summary restoring inexplicably removed parameters. I'm pretty sure that the parameter removal was not 'inexplicable' and that the changes should be restored.

Trappist the monk (talk) 11:11, 28 June 2020 (UTC)Reply

Cplakidas, two requests: 1. Please experiment in the template's sandbox rather than in the live template. 2. Please explain why you have reverted the recent improvements to this template. If possible, link to articles in which this template was not working correctly, and explain the difference between how the template was being rendered and what you expected to see. Thanks. – Jonesey95 (talk) 13:25, 28 June 2020 (UTC)Reply
Trappist the monk and Jonesey95 I was unaware of the discussion here. As the original template editor, I saw some edits that changed the template's behaviour rather strongly, without it being entirely transparent why (I saw this at the article Bajkam, where the citation read something like "Canard, Marius (1960). "Badjkam". In Gibb, H. A. R.; Kramers, J. H.; Lévi-Provençal, E.; Schacht, J.; Lewis, B. & Pellat, Ch. (eds.). The Encyclopaedia of Islam, New Edition, Volume I: A–B. Badjkam, Leiden: E. J. Brill. pp. 866–867". Now the search thing was a cool feature, but why it had to be duplicated after the actual reference, rather than used for the url, I didn't understand at all. I was unaware of Brill having specific editors cited for the entire work, because the editor teams vary enormously with each volume. I also opposed (hence my edit summary) the complete removal of the 'ref=' options, which is something that is potentially necessary when citing several works, or even articles, by the same author. I intended to write to Trappist the monk about this, but didn't get to it in time, as I was AFK. So I am not opposed to a reinstatement of the changes, but insist that the ref options be kept, and suggest that the search function be used in the url parameter. Cheers, Constantine 15:00, 28 June 2020 (UTC)Reply
Here is an EI2 template copied from Bajkam:
The current (Cplakidas's reverted version) template:
The sandbox (updated, Trappist the monk's version) template:
Cplakidas, what differences or problems do you see? – Jonesey95 (talk) 15:33, 28 June 2020 (UTC)Reply
(edit conflict)
I have added the EI2 cite from Bajkam to the ~/testcases. The only difference between the sandbox and the current live version is the ampersand between the last two editor names (see point 7 in my list of changes above).
Now the search thing was a cool feature, but why it had to be duplicated after the actual reference, rather than used for the url, I didn't understand at all. I don't understand what you mean. Can you clarify?
Re: |ref= preset. Because you insist that the ref options be kept, can you show how that form of |ref= preset is superior to the form used by the ~/sandbox version of this template? Real-life examples are more convincing that contrived examples.
Re: search 'function'. The creation of a url that links to the Brill search results page is still present in the ~/sandbox version of the template:
{{Encyclopaedia of Islam, New Edition/sandbox |authorlink=G. R. Hawting |first=G. R. |last=Hawting |pages=623–625 |title=Marwān II |volume=6}}
Hawting, G. R. (1991). "Marwān II". In Bosworth, C. E.; van Donzel, E. & Pellat, Ch. (eds.). The Encyclopaedia of Islam, New Edition, Volume VI: Mahk–Mid. Leiden: E. J. Brill. pp. 623–625. ISBN 978-90-04-08112-3.
cf. the live version:
Hawting, G. R. (1991). "Marwān II". In Bosworth, C. E.; van Donzel, E. & Pellat, Ch. (eds.). The Encyclopaedia of Islam, Second Edition. Volume VI: Mahk–Mid. Leiden: E. J. Brill. pp. 623–625. ISBN 978-90-04-08112-3.
Trappist the monk (talk) 15:37, 28 June 2020 (UTC)Reply
Trappist the monk and Jonesey95 here's what I meant (the problem appears when using the 'article' parameter without specifying a url: Canard, Marius (1960). "Badjkam". In Gibb, H. A. R.; Kramers, J. H.; Lévi-Provençal, E.; Schacht, J.; Lewis, B. & Pellat, Ch. (eds.). The Encyclopaedia of Islam, New Edition, Volume I: A–B. Leiden: E. J. Brill. pp. 866–867. OCLC 495469456. Constantine 16:26, 28 June 2020 (UTC)Reply
Fixed.
Trappist the monk (talk) 16:49, 28 June 2020 (UTC)Reply
Looks good, thanks a lot Trappist the monk. Regarding the 'ref' parameter: First, I am very much in favour of using the 'EI2' abbreviation instead of the editor names. 'EI2' is universally used and known in Islamic and related Near Eastern studies and should be preferred when the article author is not specified, especially since the encyclopaedia is in reality a single, multi-volume work, rather than considering the individual volumes separately. Second, on the manual definition of 'ref', since this remains possible with your version, I am not raising any objections. Constantine 18:10, 29 June 2020 (UTC)Reply
EI2 may very well be universally used and known in Islamic and related Near Eastern studies but, the en.wiki audience is not the Islamic-and-related-Near-Eastern-studies specialist. The en.wiki audience is the general non-specialist reader so it is wholly inappropriate for us to use jargon, like 'EI2', in short-form citations. This template must render citations that are meaningful to the general reader. Specialists reading our articles in the topic area will not be disadvantaged when we don't use 'EI2' in our short-form citations.
So we can update the live version of the template from the sandbox?
Trappist the monk (talk) 21:39, 29 June 2020 (UTC)Reply
I think the general reader is smart enough to follow a link to the full-form reference, but let's agree to disagree on this one. Go ahead. Constantine 22:24, 29 June 2020 (UTC)Reply
Of course they are; but links, tool-tips, and the pale-blue highlighting don't work in all cases (printed and pdf copies of an article etc).
I have resynched the template from the sandbox.
At the start of the template code is this:
<noinclude>{{{volume|}}}</noinclude>
What purpose does that serve?
Trappist the monk (talk) 13:58, 30 June 2020 (UTC)Reply

Why is "(search results)" visible in the citation? That does not seem appropriate to me. (1) It isn't part of the article title and (2) it almost seems to imply uncertainty about the actual article, as if we aren't sure which one we're citing. Srnec (talk) 23:44, 19 July 2020 (UTC)Reply

Pretty sure that I gave a reason for that in item 4 of my initial post in this topic. If you are seeing the ' (search results)' annotation when the template has a value assigned to |url= and / or |doi=, please show me where that is because a link to an actual article should not show the annotation.
Trappist the monk (talk) 23:52, 19 July 2020 (UTC)Reply
The reason you gave was "because under those conditions, this template creates a url to Brill's search engine". I don't think that's a good reason for the reasons I gave. We aren't citing search results. Better to have the search results link tacked on at the end or else omit it altogether than to put "(searh results)" in the title. It even appears between the quotation marks. I'm not talking about cases where the url or doi has been assigned. Srnec (talk) 03:55, 20 July 2020 (UTC)Reply
Without |url= or |doi= and with an article title, the template creates a link to Brill's search page using the article title as a search term. Others in this discussion apparently think it a worthwhile tool. I think that such links should be annotated because it is unprofessional of us to make our readers hunt through a listing of search results to get to the article that the editor consulted.
We might lift the annotation from the article title and put it in |type=:
|type={{#if:{{{url|{{{doi|}}}}}}||{{#if:{{{article|}}}{{{title|}}}|search results}}}}
This is less than optimal as a warning for editors because |type= doesn't attach to the article title so will likely to go unnoticed. Here is a {{cite encyclopedia}} mock-up of how that would render:
Canard, Marius (1960). "Badjkam". In Gibb, H. A. R.; Kramers, J. H.; Lévi-Provençal, E.; Schacht, J.; Lewis, B.; Pellat, Ch. (eds.). The Encyclopaedia of Islam, New Edition, Volume I: A–B (search results). pp. 866–867. OCLC 495469456.
Trappist the monk (talk) 11:35, 20 July 2020 (UTC)Reply

"Search result"

edit

Hi Trappist the monk, Sorry I reverted the latest change because it adds "Search results" to the title, which I don't think is appropriate. For example, when I cite like this:

What I mean to do is citing the "Ka'ba" entry in the fourth volume of EoI2, and not citing a search result for Ka'ba in the online website. But this version (that I reverted) rendered the title as "Ka'ba (search results"). HaEr48 (talk) 22:39, 1 August 2020 (UTC)Reply

The courtesy link to the online search result is not my doing. That functionality was added to the template with this edit. The ' (search results)' annotation is my doing. I think that it should be retained because linked titles should link to the cited source, not to the search results du jour. We should not be making readers hunt-about in search results for the source that you consulted for our en.wiki article.
If what you mean to do is [cite] the "Ka'ba" entry in the fourth volume of EoI2, and not [cite] a search result for Ka'ba in the online website, which I understand to mean 'cite the print version' without a linked article title, the template before your revert supported that as this ~/sandbox version of your example citation shows:
{{Encyclopaedia of Islam, New Edition/sandbox|volume=4|title=Kaʿba |pages=317–322|first=Arent Jan|last=Wensinck |url=unset |url-access=unset}}
Wensinck, Arent Jan (1978). [unset "Kaʿba"]. In van Donzel, E.; Lewis, B.; Pellat, Ch. & Bosworth, C. E. (eds.). The Encyclopaedia of Islam, New Edition, Volume IV: Iran–Kha. Leiden: E. J. Brill. pp. 317–322. OCLC 758278456. {{cite encyclopedia}}: Check |url= value (help)
And yes, the unset keyword is mentioned in the template documentation. I have since tweaked the ~/sandbox so that |url=unset also unsets |url-access=.
I think that the template should be restored from its sandbox so that the other changes mentioned at Template talk:Encyclopaedia of Islam, New Edition § proposed changes to this template are also restored.
Trappist the monk (talk) 12:07, 2 August 2020 (UTC)Reply
I think the (search result) in the citation title is so jarring that I do not think it should show up by default. I suspect most editors who do not set the url do not really want to cite to a search result, so setting it as the citation title seems wrong. I'll have to go through all the articles I wrote that cite EoI and add |url=unset and |url-access=unset. Not sure I agree with setting the search result URL by default too, but it is less jarring. HaEr48 (talk) 15:39, 3 August 2020 (UTC)Reply
I suspect most editors who do not set the url do not really want to cite to a search result. I agree. I suspect that no editor wants to cite the search results page and they may not realize that the template is making it appear that they are without the template also explicitly states that 'this-title-links-to-a-search-result-page'. If editors don't know that something is amiss, that something will never, ever, be fixed. When the (search result) in the citation title is ... jarring enough to get editor attention, that is a good thing because the link can be fixed.
I think that the template should be restored from its sandbox.
Trappist the monk (talk) 14:15, 4 August 2020 (UTC)Reply
Trappist the monk, maybe it's a good thing to notice that if you're actively editing the reference section, but most of the usage is on already-developed article. The editors won't notice it because they're not editing the reference anymore, but readers will see the jarring part which looks unprofessional. Plus, even if the editors notice, most people are not technical enough to be able to read the implementation of this template and find out how "search results" gets magically added to the title, or find out how to fix it. The solution you mentioned, setting the url parameter to "unset" is not obvious at all unless you happen to know that a priori. HaEr48 (talk) 23:29, 4 August 2020 (UTC)Reply
Asking Cplakidas and Srnec who have commented above to also weigh in. HaEr48 (talk) 23:30, 4 August 2020 (UTC)Reply
I am conflicted, TBH. On the one hand, I get HaEr48's objection that we cite an article, not a search results page, on the other, I do see the potential utility of the search feature. Since the access to the actual articles is restricted, the utility of the search results is limited by default, so the only real uses of this feature are a) as a verification mechanism (i.e. the article exists) and b) through the 'search result' annotation as a hint to redress the problem by providing the actual article url (which can be easily retrieved from the search results). In other words, the question is IMO not about whether to include '(search result)' or not, but whether the search feature itself is necessary/useful. If we provide the search feature, then we are IMO obliged to clarify it with the '(search result)' annotation for truthfulness' sake. Constantine 10:28, 5 August 2020 (UTC)Reply
If the utility is limited, and adding it by default causes counter-intuitive/jarring behaviour, I suggest that we don't provide the search result by default, and provide a parameter for editors who want to turn it on. HaEr48 (talk) 14:10, 5 August 2020 (UTC)Reply
That's a suggestion I can get behind, simply change the default setting to |url=unset. Constantine 14:32, 5 August 2020 (UTC)Reply
You noticed it, didn't you? Reverting the template is one way to 'fix' the rendering though certainly not a good way.
The citation Wensinck 1978 at Qibla, for example, still links to Brill's search results page so any reader clicking that citation's title link expecting to go to an article titled Kaʿba will be surprised because they do not land at Kaʿba. When the title is Kaʿba (search results), the reader isn't surprised when they land on the search results page because they know exactly where that link takes them. You are 'jarred' because you know that this citation did not say that it linked to a search results page when you wrote it. The addition of the convenience link to the search results page did not jar you because many upon many citation titles are linked – you are so accustomed to seeing linked titles that you did not notice that the citation had a courtesy link or that the link pointed to someplace that you had not intended. You did notice when the template told you that the linked title target is a search results page. That you noticed was, along with not surprising the reader, an important reason for the annotation. With the template no-longer rendering the annotation, without careful scrutiny, editors are much less likely to notice that the EI2 templates link to a search results page when |title= has a value. And, of course, we're back to confounding reader expectation when they end up at some place that they did not expect. That, is unprofessional.
The good way to fix Wensinck 1978 is to restore the template from its sandbox and then do one of these
  1. add |url=unset
  2. add one or both of:
    1. |url=https://referenceworks.brillonline.com/entries/encyclopaedia-of-islam-2/kaba-COM_0401?s.num=0&s.f.s2_parent=s.f.book.encyclopaedia-of-islam-2&s.q=Ka%CA%BFba
    2. |doi=10.1163/1573-3912_islam_COM_0401
Trappist the monk (talk) 18:44, 5 August 2020 (UTC)Reply
Get rid of the feature. Most readers won't have access to the correct page anyway. Adding "(search results)" is being honest about the link but lying about the citation. If we can't be honest all around, best to drop the automatic search results link. Srnec (talk) 20:50, 5 August 2020 (UTC)Reply