Wikipedia

Template talk:Infobox NFL biography/Archive 13

< Template talk:Infobox NFL biography
Latest comment: 8 years ago by Bagumba in topic Cleanup of Infobox NFL coach
Archive 10Archive 11Archive 12Archive 13Archive 14Archive 15→Archive 18

CFL stats

Latest comment: 8 years ago6 comments3 people in discussion

cfl.ca changed their website and the way their url formats are for player bios so this players with this infobox are linking to "PAGE NOT FOUND" Joeykai (talk) 23:44, 9 December 2015 (UTC)

The problem seems to be that this sort of data is no longer hosted on CFL.ca at all. Rather, it looks like this sort of player statistical information is now being hosted by STATS LLC on a page like this one: Jabari Arthur stats. The larger problem behind this change is that the player identifiers on the new site are completely different than those used on the old site, so in addition to changing the URL in the template the fix for this would also entail changing the ID on every single article that uses this parameter. I guess this points up one major reason why some have argued that these sorts of external links may not belong in Infoboxes to begin with. Maybe it's time that we had that discussion on the larger topic before we invest the time in fixing these links. — DeeJayK (talk) 14:54, 10 December 2015 (UTC)
DJK, I suggest you ping the WP:CFL guys on their talk page to see what they are doing about this. Also, take a look at Templete:Infobox gridiron football player, and see if anyone has already updated the player page coding. Hopefully, the CFL is moving toward permanent player pages (like NFL.com); in the past, they have deleted player pages for retired or otherwise inactive players. Dirtlawyer1 (talk) 20:31, 10 December 2015 (UTC)
I've raised this issue over on WP:CFL. We'll see if the CFL folks have any opinions to share. — DeeJayK (talk) 17:43, 15 December 2015 (UTC)
UPDATE: I've removed (commented out) the external links to cfl.ca in {{Infobox NFL player}} until this issue can be addressed. I've also added a tracking category to see how many pages are currently using this parameter: Category:Pages using infobox NFL player with cfl parameter. — DeeJayK (talk) 22:00, 22 December 2015 (UTC)
We do what we can. CFL.com/CFL.ca has repeatedly changed the links for their active player profiles over the last several years, and they routinely delete the webpages for former players. Unless the WP:CFL guys can provide an easy solution, I think setting this challenge aside for the time being is the best use of your valuable editing time. Dirtlawyer1 (talk) 18:08, 26 December 2015 (UTC)

Cleanup of Infobox NFL coach

Latest comment: 8 years ago27 comments4 people in discussion

I'm noticing that the existing Infoboxes (player and coach) contain lots of alternate parameters. For example the parameter "current_team" can also be used as "currentteam". While I can understand how these sorts of duplications might make the templates slightly more forgiving to the user, they also add quite a bit of complexity and bulk to the syntax of the template code. The problem is exacerbated when we're attempting to merge two templates which both contain lots of this sort of duplication. It seems the most common use case for employing these templates is going to be either copying and pasting the infobox from an existing article or from the template doc page, either of which approach makes these sorts of duplicate parameters less necessary.

It seems their only value is to allow lazy editors (who don't want to take the time look up the correct parameter) to sort of guess at what the parameter should be or type the parameter with the incorrect case or spacing. As such, I really don't think that these sorts of alternate spellings provide enough value to be worth the maintenance hassle.

I'm undertaking an effort to identify articles which currently use any of these alternate or deprecated parameters starting with the articles that use {{Infobox NFL coach}}. Once I've eradicated these alternate spellings from existing articles, I intend to excise them from the infoboxes. Comments? — DeeJayK (talk) 14:13, 26 November 2015 (UTC)

@Deejayk: Yes, the alternate parameter names need to be replaced. For certain parameters, there are multiple alternate parameter names -- what Bagumba called "legacy parameter names". Both of these templates are the result of previous merges, but especially Infobox NFL player, which includes a half dozen or more merges with older templates. When this process is done, each parameter should have a single simple name. I spent a couple weeks last summer compiling a list of alternate "legacy" parameters with the plan that BU Rob13 would run a bot to eliminate/replace all "legacy" parameter names. I suggest that you ping BU Rob13, and see when he will be finished with his final exams and what his availability is. Dirtlawyer1 (talk) 21:19, 26 November 2015 (UTC)
"only value is to allow lazy editors ...": I know of at least one project were someone thought underscores made the params more readable, so they universally changed them. Maybe the same happened here. As long as any information that is presented today remains available after a merge (and after any other massaging needed), I don't have any objection. Presumably we would not make any presentation changes in this iteration unless a merge conflict requires it be addressed. Any other improvements should be discussed after the merge.—Bagumba (talk) 05:55, 27 November 2015 (UTC)
Have we decided on a final list of parameters for the merged template? If so, I am willing to go through the Coach template uses and replace any alternate names with the base names (removing the need for multiple alt names). I know Player also uses alt names, but it looks like only 1-2 (and a few alt names is fine). My apologies for being absent on this, I was busy with other things (though it looks like things are proceeding well). Primefac (talk) 16:37, 28 November 2015 (UTC)
@Bagumba: point taken; my "lazy editors" comment was a gross oversimplication. I admit I don't have any idea WHY these duplicate spellings were introduced, and your conjecture seems plausible. My thinking is that, regardless of how or why these duplicates came about, having multiple parameters representing the same thing doesn't really serve a useful purpose. As such, ,minimizing or eliminating these duplications will allow us to simplify these templates and the process of merging them. — DeeJayK (talk) 15:55, 30 November 2015 (UTC)
@Deejayk: "I don't have any idea WHY these duplicate spellings were introduced": as I noted above, most of these alternate parameter names were introduced as a result of the multiple merges of at least six different infobox templates into the present Infobox NFL player. Others were a result of cross-overs from related templates, such Infobox college coach, which shares a common lineage with Infobox NFL coach. Dirtlawyer1 (talk) 16:18, 30 November 2015 (UTC)
@Primefac: I don't necessarily have a preference as to which of the parameters we settle on (although, see Bagumba's comment above). I guess for simplicity my thinking is that we would choose whichever variant is most prevalent in usage. To that point, does anyone understand how "Check for unknown parameters" section of the {{Infobox NFL coach}} code works? I see that it populates a maintenance category Category:Pages using infobox NFL coach with deprecated syntax, but I don't understand how it determines which parameters are deprecated as it seems to be given a list of all the parameters. Since I don't grok the wizardry behind this piece of code, I've added a bit of code to the coach Infobox template which does a similar thing, but only for a single parameter at a time; this can be used to determine which pages (if any) are using a particular parameter. Although it will be a bit time consuming to go through every parameter, that's my best idea at this point in time. Edit: I've figured out that this is using MediaWiki:Check for unknown parameters (edit | talk | history | links | watch | logs); I'm working on determining how this module can be harnessed to our ends. — DeeJayK (talk) 15:55, 30 November 2015 (UTC)
Likewise, I only care at the margins what the parameter names are; only that the resulting parameter names should be logical, relatively intuitive, and consistent throughout. At the completion of this process the parameter names should be consistent, and no alternative "legacy" parameters should exist. Likewise, data entry, data formatting, and parameter architecture should also be consistent throughout. The 340 articles that presently use Template:Infobox NFL coach should not present team tenures in a manner different from the majority of coach articles that already use Template:Infobox NFL player. The purpose of any well executed template merge is to simplify, make consistent, and conform presentation of data to a common standard. If we are simply engrafting even more legacy parameters -- and even worse, importing parameter data in a completely different format from the established standards of Infobox NFL player -- then we would be better served by manually replacing the remaining 340 instances of Infobox NFL coach, rather than further complicating the existing template in order to accomplish a "merge". Dirtlawyer1 (talk) 16:13, 30 November 2015 (UTC)
We might have to do that anyway, if there are that many legacy parameters and varying coding (specifically the "past teams" and "highlights" sections). Primefac (talk) 16:21, 30 November 2015 (UTC)
I agree with Dirtlawyer1 description of the task. Looking at where we are with the process, I still feel like a merge can be accomplished. — DeeJayK (talk) 16:24, 30 November 2015 (UTC)
The merge needs to happen for a lot of reasons, guys, not least of which is most NFL coach bios already use Infobox NFL player. The question is: does it make more sense to execute the merge using a bot or AWB, whether doing it properly requires manually editing, or it can be most efficiently accomplished with a combination of manual editing and automation. In order, I suggest we focus on (1) which parameters will be included in the merged template, (2) what the merged template should look like (including any shift of parameters from one section to another), (3) determining what the single conformed parameter names for each surviving parameter should be, and (4) only then deciding how best to execute the merge. I assisted User:Eagles247 with the last major round of merges into Infobox NFL player in 2010, involving four or five times as many articles as the 340 under discussion here, and I can tell you a little pre-planning goes a long way and the need for follow-through is absolute. I still encounter older articles that were the subject of that prior merge whose infoboxes do not display properly, either because of legacy parameter names or because the formatting of data was not correct. A lot of those problems were the result of relying on third-party bot operators for automated merges/conversions. Given the relatively small number involved here (340), every affected article should be reviewed after the merge to make sure there are no problems. Dirtlawyer1 (talk) 17:41, 30 November 2015 (UTC)
@Dirtlawyer1: "... we would be better served by manually replacing the remaining 340 instances of Infobox NFL coach} ...": I don't really care if that happens or not. I'd only caution that this merge has already been mired for 3 year at Wikipedia:Templates_for_discussion/Holding_cell. Since a manual merge hadn't happened by now, that should be a point of consideration in deciding whether to halt or continue the recent momentum of a programmatic solution—Bagumba (talk) 20:41, 30 November 2015 (UTC)
@Bagumba: See my comment above. Will you be available to do the manual editing to clean up the mess created by adding all of the conflicting and redundant parameters of Infobox NFL coach to Infobox NFL coach? The rational way to handle this is to decide what parameters are to be kept, what the finished product will look like, and then determine the most efficient way to implement it. Slapping together two templates, with different architecture and different data formatting, for the sake of saying a merge has been done is just plain goofy. Having a plan before proceeding is common sense, even if you have never implemented a merge of this nature before. Let's not waste our valuable (and limited) volunteer time by not having a plan, and not doing this in the most time-efficient manner. Dirtlawyer1 (talk) 21:23, 30 November 2015 (UTC)
"... for the sake of saying a merge has been done is just plain goofy" It's overly simplistic to discount the reduced maintenance of the majority of the code that is common. I'm probably in the minority of people that don't have a huge problem with deprecated parameters that are removed from documentation, but continue to work for backwards compatibility. It's done in industry all the time to reduce the risk of things breaking, and has been working as such in the existing templates. However, if enough people are bugged that "old" parameters still lurk, then so be it. I might volunteer to brute-force merge a few if that is the direction that is decided, but I'm typically not much of a gnome.—Bagumba (talk) 21:44, 30 November 2015 (UTC)
Common parameter coding with AWB is easy; common data formatting, however, requires manual editing. The plan would be to use AWB to replace the parameter names in the 340 coach articles. That can be done in a matter of minutes once we settle on the final parameter names (we can get BU Rob13 to clean up the legacy parameter names in the current 16,000 transclusions of Infobox NFL player when he's done with exams). The manual work -- the real pain-in-the-ass manual work -- involves cleaning up the parameter data for team tenures, championships, awards and other highlights and putting it into the standard format for Infobox NFL player. There are 340 articles that need to have their infobox "merged" in this exercise; if we split that manual editing among 7 to 10 knowledgeable editors we can be done in about a week. It's a bit more than typical "gnome" work because it requires a working knowledge of NFL/AFL history, championships, awards and honors. I've already pinged DeeJayK and Primefac to see if they're willing, and I'm pretty sure Wikioriginal-9 will take a slice of the work. If we can recruit another 3 to 5 editors to help, this becomes a very manageable exercise -- nothing of the scale of the merge work Eagles oversaw 5 years ago. Dirtlawyer1 (talk) 22:26, 30 November 2015 (UTC)

I'm (finally) finishing up my cleanup/simplification of the existing {{Infobox NFL coach}}. I've whittled the duplicate parameters down to the minimum. Here is a list of the parameters I've managed to remove (bear in mind template parameters are case sensitive):

  1. currentteam (current_team should be used instead)
  2. Image (image should be used)
  3. Height_ft (height_ft should be used)
  4. Height_in (height_in should be used)
  5. Caption (caption should be used)
  6. weight_lb (weight_lbs should be used)
  7. weight_lb (weight_lbs should be used)
  8. Weight_lb (weight_lbs should be used)
  9. Weight_lbs (weight_lbs should be used)
  10. Position (position should be used)
  11. College (college should be used)
  12. HighSchool (high_school should be used)
  13. high school (high_school should be used)
  14. Jersey (jersey should be used)
  15. DraftedYear (drafted_year should be used)
  16. DraftedRound (drafted_round should be used)
  17. AFLDraftedYear (AFL_drafted_year should be used)
  18. AFLDraftedRound (AFL_drafted_round should be used)
  19. Awards (awards should be used)
  20. Honors (honors should be used)
  21. Retired #s (retired_numbers should be used)
  22. Records (records should be used)
  23. record (overall_record should be used)
  24. Record (overall_record should be used)
  25. RegularRecord (regular_record should be used)
  26. PlayoffRecord (playoff_record should be used)
  27. Super_Bowls (SuperBowls should be used)
  28. Championships (championships should be used)
  29. years (player_years1 or coach_years1 should be used)
  30. teams (player_team1 or coach_team1 should be used)
  31. player_years (player_years1, player_years2, etc. should be used)
  32. player_teams (player_team1, player_team2, etc. should be used)
  33. coaching_years (coach_years1, coach_years2, etc. should be used)
  34. coachingyears (coach_years1, coach_years2, etc. should be used)
  35. coach_years (coach_years1, coach_years2, etc. should be used)
  36. coaching_teams (coach_team1, coach_team2, etc. should be used)
  37. coachingteams (coach_team1, coach_team2, etc. should be used)
  38. coach_teams (coach_team1, coach_team2, etc. should be used)
  39. debutyear (no direct replacement, but this was not implemented in the code and was not used by any page)
  40. debut_year (no direct replacement, but this was not implemented in the code and was not used by any page)
  41. debutteam (no direct replacement, but this was not implemented in the code and was not used by any page)
  42. debut_team (no direct replacement, but this was not implemented in the code and was not used by any page)
  43. finalyear (no direct replacement, but this was not implemented in the code and was not used by any page)
  44. final_year (no direct replacement, but this was not implemented in the code and was not used by any page)
  45. finalteam (no direct replacement, but this was not implemented in the code and was not used by any page)
  46. final_team (no direct replacement, but this was not implemented in the code and was not used by any page)
  47. player_positions (player_position should be used)
  48. DatabaseFootball (NFL or PFR should be used instead)
  49. DatabaseFootballCoach (PFRCoach should be used instead)
  50. ESPN (NFL or PFR should be used instead)
  51. CBS (NFL or PFR should be used instead)
  52. CFHOFYear (no direct replacement, but this was not implemented in the code and was not used by any page)

Through the course of this effort I did quite a bit of manual cleanup of the articles that use this template (to the point that I began to wonder whether it would have made more sense to just to manually migrate them all to use {{Infobox NFL player}}). I also added two parameters: SRCollCoach provides an external link for college coaching records and CFLCoach provides an external link for CFL coaching records. I realize there has been discussion as to whether these types of external links are appropriate in infoboxes, but I figured that as long as we had some, why not make it more complete. If someone wants to initiate a discussion to remove all of the external links, I would be happy to participate.

Next steps are to update the {{Infobox NFL biography/sandbox}} with the changes that I've made to streamline {{Infobox NFL Coach}}. However, before I begin that, I'd like to solicit feedback on the changes I've made to this point. — DeeJayK (talk) 21:10, 9 December 2015 (UTC)

Thanks for the thorough update! You mentioned wondering if a manual update made more sense. Am I to presume that you determined that a merge is the best direction to go?—Bagumba (talk) 05:33, 10 December 2015 (UTC)
I'm still not sure I've really determined whether a merge remains the best way forward. Either way we go, the work I've done in simplifying this template and cleaning up the pages that use it should make the migration process simpler once we determine how that will look. — DeeJayK (talk) 15:21, 10 December 2015 (UTC)

FYI, I've already recruited several editors to help with the manual clean-up of the merged coach infobox, including myself, CrashUnderride, Jweiss11, Primefac, Wikioriginal-9, Yankees10 -- plus Bagumba and DeeJayK. That makes eight so far, and we can recruit several more if needed. The coding and data entry for the "merged" coach infoboxes should match that of existing players and coaches that already use Infobox NFL player. There should be no legacy coding and no legacy formatted data when we're done. One template to rule them all. The biggest problem, of course, if the inverted teams and tenures data for the existing uses of Infobox NFL coach, which will need to be put in the standard format:

The data formatting for coaches parallels that for players:

Easy-peasy, but the manual work should be performed by editors familiar with NFL/AFL history (so as to catch various errors, etc.), and fully briefed on the merge and surviving parameter names, etc.. Dirtlawyer1 (talk) 20:50, 10 December 2015 (UTC)

This distinction between the way playing/coaching histories are displayed seems to me to be the biggest unaddressed hurdle remaining for this merge effort. As someone who values the consistency and simplicity of reorganization that comes with highly structured data, I strongly prefer the way these histories are handled by {{Infobox NFL coach}} (with the playeryears1, playerteam1, playeryears2, playerteam2, coach_years1, coach_team1, etc., parameters) over the unstructured data in {{Infobox NFL player}} (with "pastteams", "pastcoaching", etc.). In fact, if I were designing this template from scratch I would consider extending the structure even further to capture coaching position (i.e. coach_pos1 = defensive backs, coach_pos2 = defensive coordinator, etc.). However, the existing player template has a MUCH larger installed base, and I can't think of any sort of automated/bot method of converting the unstructured data in the existing "pastteams", "pastcoaching", "pastadmin", etc. to a more structured form like the coach template uses. However, since the player template has a much larger installed base than the coach template (16K+ to 300K articles) the effort it would take to implement the structured format in the player articles is daunting. As such, I think our best way forward with this merge effort is to retain the playing/coaching history structures existing in both templates. My hope would be that the structured format would replace the unstructured format over time. — DeeJayK (talk) 18:15, 18 December 2015 (UTC)
I completely agree, both in that the structured format is better, and that hopefully things will swing that way in the future. Certainly from what I've seen, templates of this nature should encourage team1/team2/etc format instead of a generic "team" param. My thoughts would be (once we get everything sorted and merged) to make the doc show the 1/2/3 params and essentially depreciate the generic catch-all fill-in-the-param params, slowly replacing them over time. Primefac (talk) 06:01, 22 December 2015 (UTC)
I'm think that the approach Primefac outlines above would be the best way forward. I'm planning on updating the sandbox with that in mind. If anyone feels differently, please state your case. Thanks. — DeeJayK (talk) 16:43, 22 December 2015 (UTC)
I just saw this, and I could not disagree more strongly. Over 16,000 existing articles are formatted using the "list" format for player and coach tenures, and that includes the majority of NFL coach articles that already use the existing NFL player infobox. There is no benefit in restructuring the coaching tenure data for each individual coaching tenure into three or more separate datapoints; even less so when the majority of NFL coach infoboxes are already formatted in a different manner -- a manner, I might add, that is consistent with the way all 16,000 player histories are already formatted. So, no, adopting the minority practice this is not the best way forward. The goal is a single, uniform format for all NFL infoboxes, not another 6 or 7 years of multiple minority formats that lead to confusion, edit-warring and wasted editor time. Dirtlawyer1 (talk) 17:39, 26 December 2015 (UTC)
IIRC, the itemized team/year params were in line with WP:ACCESS to discourage the use of <br> to format separate stints. If nothing else, the old format should at the very least be deprecated in documentation (if it hasn't been already).—Bagumba (talk) 16:20, 27 December 2015 (UTC)
Uh, no. There is nothing in WP:ACCESS that requires or prefers multiple parameters for tenures. You are recalling recent changes to Infobox college coach, which did not previously employ plainlist formatting, but instead used a single field for all coach tenures (and other single fields for player tenures and awards), with each individual tenure separated with a <br> coded hard return. The plainlist format used by Infobox NFL player does not require coded hard returns, only an asterisk-coded bullet point at the beginning of each tenure and an uncoded hard return at the end of each. In fact, we just converted virtually every navbox on Wikipedia to the bullet-point hlist format over the last two years, and both plainlist and hlist are WP:ACCESS compliant per WP:HLIST and WP:PLIST. Moreover, we absolutely should not start reversing 8.5 years of formatting the year-span tenures of 16,000 infoboxes to accommodate a relative handful of infoboxes in which the year-spans of the coaching tenures precede the team names (which is what requires the multiple parameters in order to achieve proper alignment and internal spacing for year spans of differing character lengths, without using the coded hard returns). Again, this is why these things need to be discussed with editors who know and understand the history of the template. If you want to reverse 8.5 years of dozens of editors moving toward a consistent and logical format for tenures for our American football player infoboxes, we can just undo this merge right now, Bags. We're not going to format tenures for all players and most coaches one way, and some coaches another way -- thus defeating the original logic of the merge to simplify and conform. Dirtlawyer1 (talk) 18:14, 27 December 2015 (UTC)
You had me at "plainlist". If <br> is not a concern, then WP:ACCESS is not an issue.—Bagumba (talk) 22:49, 28 December 2015 (UTC)
@DL: "this is why these things need to be discussed with editors who know and understand the history of the template": And I believe this is being discussed here, no?—Bagumba (talk) 22:49, 28 December 2015 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Template_talk:Infobox_NFL_biography/Archive_13&oldid=856490140"