Template talk:This is a redirect/Archive 2

Archive 1

Archive 2

subst

@Primefac: Your changes looks nice :-) I found one case where it doesn't work. An example from the /doc [1]:

{{This is a redirect
  |e0=See '''{{-r|(Other redirect title)}}''' for printworthy redirect.
 |from airport code
  |e1=* ''This airport code has been discontinued.''
 |from ambiguous page
  |e2=* '''Note:'''  ''The ambiguity is easily disambiguated.''
 |from London bus route
  |e3=* (This bus route is to and from Maidenhead.)
 |from Unicode
 |unprintworthy
  |e5=* ''Up to '''eight''' of these '''{{para|e#}}''' parameters may be used – the '''{{para|e0}}''' as a '''TOP note''' as well as one {{para|e#}} parameter for each different rcat.''
}}

"subst:" gives:

{{redirect category shell|{{R from airport code
  }}
* ''This airport code has been discontinued.''{{R from ambiguous page
  }}
* '''Note:'''  ''The ambiguity is easily disambiguated.''{{R from London bus route
  }}
* (This bus route is to and from Maidenhead.){{R from Unicode
 }}{{#if:
   |{{#if:
    |{{R unprintworthy
  |{{{p5}}}|{{{n5}}}}}
    |{{R unprintworthy
  |{{{p5}}}}}
   }}
   |{{R unprintworthy
  }}
  }}
* ''Up to '''eight''' of these '''{{para|e#}}''' parameters may be used – the '''{{para|e0}}''' as a '''TOP note''' as well as one {{para|e#}} parameter for each different rcat.''|h=See '''{{-r|(Other redirect title)}}''' for printworthy redirect.}}

It produces the right visual output, but my impression was that a bot should subst them all. Christian75 (talk) 11:17, 17 April 2017 (UTC)

Christian75, that's very odd. As far as I'm aware, when you pass a parameter to a template it ignores whitespace (e.g. putting every param on a new line in an {{infobox}}), so I don't know why it would re-insert it here. I'll look into it.

I know the e, n, and p parameters are deprecated, so I've added some tracking cats to see what's actually being used. It might be that we can just cut a lot of the code and make it less likely to break when substed. Thanks for finding this! Primefac (talk) 12:40, 17 April 2017 (UTC)

It depends on the type of parameter. For named parameters, and also for explicitly numbered parameters, leading and trailing whitespace is stripped from parameter values. However, for positional parameters, all whitespace is preserved. Consider the two lines

 |from airport code
  |e1=* ''This airport code has been discontinued.''

- here there is one positional parameter and one named parameter, and the newline after the first instance of the word "code" is significant. If however they had been written as

 |1=from airport code
  |e1=* ''This airport code has been discontinued.''

the only difference is the addition of two characters 1= but we now have one numbered and one named parameter, and the newlines are all stripped. When doing this it's important to explicitly number all of them, otherwise you may find unexpected effects. --Redrose64 🌹 (talk) 20:45, 17 April 2017 (UTC)

Of course, the other issue we have to consider is that (technically speaking) {{rcat shell}} doesn't actually accept random comments made in the code, so even when substing something inline like -ihah you end up with some odd-looking code. I think we have to ask ourselves whether the extra note is proper to include.

{{redirect category shell|{{R move}}{{R hatnote}}
* '''Note:'''  ''See '''[[Instituto Hondureño de Antropología e Historia]]''' for hatnote mention.''{{R related}}{{R sect}}{{R unprintworthy}}}}

vs

{{redirect category shell|{{R move}}{{R hatnote}}
* '''Note:'''  ''See '''[[Instituto Hondureño de Antropología e Historia]]''' for hatnote mention.''
{{R related}}{{R sect}}{{R unprintworthy}}}}

vs

{{redirect category shell|
{{R move}}
{{R hatnote}}
* '''Note:'''  ''See '''[[Instituto Hondureño de Antropología e Historia]]''' for hatnote mention.''
{{R related}}
{{R sect}}
{{R unprintworthy}}
}}

I'm not really keen on any of those options, but if we're trying to keep those random notes then we're going to have issues with formatting (regardless of whitespace). Primefac (talk) 15:09, 18 April 2017 (UTC)

If that third option is desired, it might require a bot run for formatting. Primefac (talk) 15:22, 18 April 2017 (UTC)

• One more ping to see if anyone's keen on commenting, otherwise I'll just remove the reliance on the eX parameter and only subst the actual templates. Primefac (talk) 18:57, 26 April 2017 (UTC)

Keeping this template

It could be really nive if this template was not deleted after all the subst'ing. and kept as a subst only template. Does that require a new TfD? Christian75 (talk) 13:42, 4 May 2017 (UTC)

Christian75, my original intention (once all extant uses are substed) was to simply redirect this to the rcat shell. I don't think changing that would necessarily require a new TFD, but it would definitely require some sort of consensus. Primefac (talk) 14:21, 4 May 2017 (UTC)

Thanks, I hoped someone would see this - but I havn't pinged anyone and do not think I will do it. Christian75 (talk) 14:01, 16 May 2017 (UTC)

It is a good idea to keep as a subst: only, because it can be very fast. However subst will only replace the cabalistic symbols with marginally less cabalistic ones. Better for HPB to make intelligent replacements. All the best: Rich Farmbrough, 09:53, 28 May 2017 (UTC).

BRD

User:Primefac boldly added made this auto-subst only. I reverted him. He re-reverted. We had I thought, agreed to a compromise where I would stop doing good replacements, and he would stop doing bad replacements. Therefore I restored the status-quo ante. He has reverted me again - and 6000 more items have been subste'd.

I am restoring the status quo once again. We can now discuss.

Discussion

Further to the request at Bot requests I have invested a significant amount of time in producing a method of replacing this template, in the most useful way I can. Conversely I understand Primefac has invested effort in making the template substable. Nonetheless the substing method is comparatively poor and sometimes wrong.

For this reason I propose that the template not be further substed, at least with the current set-up, but instead we suspend work on this until the BRFA is approved, and do the thing that actually gives better results.

All the best: Rich Farmbrough, 10:04, 28 May 2017 (UTC).

Rich Farmbrough, I ran the autosubst past a few users, double-checked everything, and even cleaned up a few messes I made along the way. Yet all I have heard from you is complaining and not a single reason why the autosubst is sub-par. The only thing I can see that your bot run does that autosubst doesn't is add white space. I don't think that's enough to sideline the substitution of the template. Primefac (talk) 13:06, 28 May 2017 (UTC)

• It deals with {{{e1}}}.... etc.
• It leaves those that have parameters outside the e, p, n series for manual processing.
• It replaces parameters with carefully chose templates rather than blindly prefixing "R ".
• It doesn't leave red-linked R cats.
• It follows the layout on the doc page.
• It removes redundant colons.

Is that not enough?

All the best: Rich Farmbrough, 14:41, 28 May 2017 (UTC).

It's great. Some of them are minor points (the layout given by the autosubst is listed as an acceptable arrangement on the /doc), but overall I can see the benefits. While I am partially culpable, it would have been nice to have had this conversation two weeks ago and saved all the drama. Primefac (talk) 14:50, 28 May 2017 (UTC)

Good! Just to clarify on layout, in the deep dark days of yore, redirects could only have text on the fist line, and there were some of these that used the template.

All the best: Rich Farmbrough, 15:56, 28 May 2017 (UTC).