Template talk:GAList2
Broken
[edit]This template seems to be broken. Look at any handful of pages it's used on and you will see much exposed raw wikicode. Not sure how to fix this. - dcljr (talk) 04:44, 31 March 2019 (UTC)
- Looks like the template just needs to be subst'd (which is required, according to the docs) to fix most of the problems. Compare one instance before and after subst'ing. Still one small snippet of exposed wikicode (after the word "plagiarism:"), which seems to be caused by a missing parameter (added to the template after it was used on the page in question). I guess "someone" needs to substitute this template on all pages where it's being transcluded (which causes said pages to drop off the list). That's 1,096 pages at the moment. - dcljr (talk) 10:47, 11 April 2019 (UTC)
- Ugh. Because the template has changed over the years, subst'ing it today would result in (possibly) "ahistorical" text being substituted into any given page — just as "ahistorical" text is currently being transcluded into many of the pages. It looks like the only way this could be done "correctly" is to:
- note the date the template was added to a given "target page";
- edit that version of the template and copy its wikicode into a temporary template (or user subpage?) created for this purpose;
- replace the name of this template in the target page with the temporary name (with "subst:") and save;
- fix any remaining formatting problems.
- Obviously, no one is going to do this for 1000+ pages [frown] — although the task would be somewhat streamlined if the list of target pages could be sorted by which version of the template was intended to be used. I guess I could start to do that. Maybe. Someday. - dcljr (talk) 12:24, 11 April 2019 (UTC)
- I've fixed the exposed wikicode issue. The "ahistorical" issue is more difficult to fix, though. Olivaw-Daneel (talk) 17:39, 23 February 2023 (UTC)
- Thank you for addressing this issue. I had forgotten about it. I'll have to do more research to figure out / remember how serious the "ahistorical text" problem is. In the meantime, I'm pinging the most frequent editor of this template, @SilkTork, to see what they think. - dcljr (talk) 03:09, 24 February 2023 (UTC)
- Thanks for the ping dcljr. My contributions to the template have all been text based rather than code, and my last edit was 10 years ago, so I'm not sure how much help I can be with this issue. However, if you have a specific question, I'll attempt to answer it. SilkTork (talk) 10:45, 24 February 2023 (UTC)
- This template should be substituted, but it is being transcluded on many pages. Since the output of the template has changed from time to time over the years, the current output being displayed on those pages (and the output that would be "frozen" in place if these template calls were subst'd) doesn't necessarily match what the users of the template intended to say "yes" or "no" to at the time they used the template (which is the precise reason the template's documentation says it should be substituted). For reference, the handful of instances I checked ("at random") were all set up at least 12 years ago. Theoretically, this problem could be fixed by doing something like the sequence of steps outlined above in my comment starting with the word "Ugh." Do you have an opinion as to whether this is worth trying to do? Or should we just subst the templates and be done with it? Or just ignore the problem altogether? - dcljr (talk) 03:51, 25 February 2023 (UTC)
- Thanks for the ping dcljr. My contributions to the template have all been text based rather than code, and my last edit was 10 years ago, so I'm not sure how much help I can be with this issue. However, if you have a specific question, I'll attempt to answer it. SilkTork (talk) 10:45, 24 February 2023 (UTC)
- Thank you for addressing this issue. I had forgotten about it. I'll have to do more research to figure out / remember how serious the "ahistorical text" problem is. In the meantime, I'm pinging the most frequent editor of this template, @SilkTork, to see what they think. - dcljr (talk) 03:09, 24 February 2023 (UTC)
- I've fixed the exposed wikicode issue. The "ahistorical" issue is more difficult to fix, though. Olivaw-Daneel (talk) 17:39, 23 February 2023 (UTC)
- Ugh. Because the template has changed over the years, subst'ing it today would result in (possibly) "ahistorical" text being substituted into any given page — just as "ahistorical" text is currently being transcluded into many of the pages. It looks like the only way this could be done "correctly" is to:
What you are saying, dcljr, is that the records of past GA reviews are being altered each time the template text is updated, such that the records may be misleading. But that you can correct this by doing a sequence of steps, but you're not entirely sure it would work. Is it possible for you to do some form of test on a test wiki before doing the changes? I think if the misinformation can be corrected, that should be done. But I think we'd need some confirmation that the fix will improve matters rather than make them worse. Once your fix has been applied, you can then make changes to this template which will prevent future problems? SilkTork (talk) 10:18, 27 February 2023 (UTC)
- Yes to your first statement. As for the second, the issue is not so much that it might not work, but that it might not be worth doing. The only reason I brought you into this is, I thought you (as the user who seems to have made the most individual edits to this template) might have some insight as to whether the text of individual items in the template's output have changed enough over the years that it's worth going to the trouble of subst'ing the text of past states of the template rather than simply subst'ing the current output of the template. It looks like I was wrong about that (i.e., you don't have any special insight). [grin] In which case, I'll just have to do more research… - dcljr (talk) 03:17, 28 February 2023 (UTC)
- Is there something I can do to help out with your research? SilkTork (talk) 10:59, 28 February 2023 (UTC)
Missing parameters
[edit]I just fixed the (relatively minor) problem mentioned in my earlier comment, where raw wikicode was being displayed (instead of an icon) when a missing or empty "criterion parameter" was passed to the {{GAList/check}} template. I used "neu" as the default value passed to {{GAList/check}} because that seems to be the intended default value in the implementation of that template (unless I am misunderstanding the situation there). A possible alternative would be to use "dunno" (or "?") as the default in calls to that template from this one. - dcljr (talk) 12:41, 11 April 2019 (UTC)