Commons:Bots/Requests/SamoaBot 4

From Wikimedia Commons, the free media repository
Jump to: navigation, search

SamoaBot 4 (talk · contribs)

Operator: Ricordisamoa (talk · contributions · Number of edits · recent activity · block log · User rights log · uploads · Global account information)

Bot's tasks for which permission is being sought: "condense" multiple {{assessments}}-like templates into single ones (example)

Automatic or manually assisted: automatic (after test run)

Edit type (e.g. Continuous, daily, one time run): continuous

Maximum edit rate (e.g. edits per minute): 8-12 EPM

Bot flag requested: (Y/N): N

Programming language(s): Python, PWB

--Ricordisamoa 04:02, 6 July 2013 (UTC)

Discussion

  • Looks OK for me. --EugeneZelenko (talk) 14:24, 6 July 2013 (UTC)
  • Please can you initiate (or point to) a community discussion involving FPC/QIC/VIC noticeboards to confirm that they want this? --99of9 (talk) 11:00, 11 July 2013 (UTC)
    • Unnecessary clutter/fragmentation of the file description page is no option, not now and not in future. -- Rillke(q?) 08:05, 20 July 2013 (UTC)
@99of9: I'm a bit busy at the moment, could you please initiate (and point to) such community discussion? --Ricordisamoa 17:34, 17 August 2013 (UTC)
@99of9: where should the discussion take place? What are "FPC/QIC/VIC noticeboards"? --Ricordisamoa 03:55, 3 September 2013 (UTC)
Sorry I haven't got around to it either. Here is FPC talk, QIC talk, and VIC talk. Most of the regulars in those communities will have those pages watchlisted. --99of9 (talk) 05:45, 3 September 2013 (UTC)
I have sent them identical messages pointing to this page. --Ricordisamoa 07:55, 3 September 2013 (UTC)
  • Fine for me. — TintoMeches, 08:13, 3 September 2013 (UTC)
  • I endorse the objective of reducing page clutter by having assessment-like templates being shown together in a more consistent manner, and I understand the good intentions from the bot implementer to have a bot do that tedious task. I do have some serious reservations about the {{Assessments}} template itself as I find it adds visual clutter and confusion, especially when used for merging FP, QI and VI into the same template with its current implementation. My concerns are
    1. The visual appearance of the template is for me an eyesore when more than one type of assessment is used, see File:Acrocinus longimanus MHNT femelle.jpg for an example where, the template is used for both FP, QI and VI. There are three frames in different colours and inside the box three logos in different colors and styles. Yak! For the Valued image project we spend a lot of time originally in crafting a nice colour scheme for the frame and background colour, which very nicely matches the logo, on which much time was spend. These colours are a trademark of the individual projects and gives project assessment identity. By putting it all together in one monster templte with frames in frames, as has been done, the project identiity is lost.
    2. The usage for VI is not 'approved' (nor documented) and I would object to using it for VI. For instance the Assessment template does not (as far as I know) support the important scope parameter, which is in the {{Valued image}}, nor am I sure it categorizes the file page correctly as the Valued image template does or correctly supports the subpage parameter (when it deviates from the default).
    3. The implementation for the Assessments template is terribly complicated because it is trying to do way too many things at the same time. As a consequence it is very often broken (see the talk page and archives) for its long and troublesome history. It is only a few users here who can actually edit it and maintain it.
    4. Bots specific to the QI, VI and FP projects will continue to add the old templates. Thus this bot will have to run regularly to merge the templates. It should be checked with the other bot implementers that this merge does not influence any of their bot tasks in unforeseen manners.
    5. Although parameter names for the Assessment templates has improved, I find the 1, 3, etc. values in the template confusing and non-human readable.
  • Due to these concerns, and since there is not community consensus that an all-engulfing template is a better solution than the existing stand-alone templates, I cannot support this template merge as is. And I will revert any such changes made on photos I have uploaded. If there is a community consensus, that we really want to do the merge as suggested, I will respect that (very reluctantly).
  • That said, I also understand that the current situation with the stand-alone templates, placed randomly on the file page and in more or less random order is confusing and improvable. I am just not sure, that merging them into the Assessment monster is the right solution. I would much rather like to see some smart logic by which, when placed side by side in consistent order, the existing templates would render in a neater way, say in in three columns if FP, QI and VI (I do not know if this is technically possible). That would make template maintenance easier (like adding new languages), and it would be easier for the project-specific bots to add them and change them. Something like, when you place a {{Location}} template underneath an {{Information}} template you get a nice consistent view from which it appears when looking at the file page, that it is one integrated table, like here for example. I think a wider community discussion regarding this is needed to get some mockups on the table to understand what is possible in a way, where both a nice visual appearance with project identities are maintained, where the visual clutter is lowered, where it is easy to maintain the template code, and where changes to project-specific bots are minimal.
  • For now I can endorse having the bot order the templates consistently, e.g., first FP, then QI, then VI, under information and location info, but nothing more until a community consensus is achieved.

--Slaunger (talk) 19:12, 3 September 2013 (UTC)

  • I used to be sceptical about merging QI and VI into the Assessments template as well. I think condensation into a single template actually reduces clutter and presents a single compact box to the user. I think this is a usability improvement. I have no reservations about adding this functionality to the QICbot (and the VICbot). I've been playing with the excellent mwparserfromhell python module lately and using that module the task of adapting the bot should be super-easy. --Dschwen (talk) 21:09, 3 September 2013 (UTC)
    • I just had a look at mwparserfromhell. This looks like a most useful python module! So much easier to use than (arcane) regexes. --Slaunger (talk) 19:21, 4 September 2013 (UTC)
  • Furthermore if the current implementation of the template is considered an eyesore ;-) I would invest a little time in improving the template layout. The consolidation in wikitext space should be independent of that effort. Actually, the {{Assessments}} template could even be adapted to spit out an old style separate QI/VI box (although I think this would be counterproductive). --Dschwen (talk) 21:12, 3 September 2013 (UTC)
    • But it has to support the needed parameters before consolidating in wikitext such that no information is lost in the transition, and currently I do not think the template does - at least it is not documented. If consolidated, I agree it should not spit out separate QI/VI/FP frames exactly as the separate templates does now, but I think some middleground in the layout needs to be found, otherwise the increase in uglyness, outweighs the decrease in clutter IMO.
    • So you are not concerned about the continued maintenance of this monster template? I would much rather solve the issue with composition (make the overall template call the (possibly adapted) individual templates) instead of the current very complicated logic. Will be much easier to maintain. It is fine if you are willing to adapt two of the bots of course. Greatly appreciated. --Slaunger (talk) 21:23, 3 September 2013 (UTC)
      • This template could be made into a well documented easy to maintain LUA module :-). Any takers? Which template parameters are you worried about preserving? --Dschwen (talk) 22:49, 3 September 2013 (UTC) P.S.: Oh, while {{Quality image}} has no parameters the {{Valued image}} template does take a few. Yep, those should be conserved somehow. --Dschwen (talk) 22:51, 3 September 2013 (UTC)
        • I have never tried coding in Lua, but I had a look at it as used for Mediawiki templates. I looks like it has a low entry barrier. It reminds me very much of Python, which I am proficient in. So as a matter of fact I am tempted to 'take' that task of making a well-documented, easy to maintain and easy to test Lua template, with sensible and logical parameter names and values Face-smile.svg. Considering how much I have been bitching about the assessments template I ought to take responsibility for improving Clin. However, it will take some time, as I need to understand a couple of things, like how to do internationalization in the best way, how to best structure the code, how to make sandbox mockup templates for testing and evaluation of different possibilities such that extensions with new languages and support for FP programs on different wikis can be done most easily. I need to understand what the possibilities are for making things look visually consistent, with minimum clutter and 718smiley.svg Awesome!

. And most importantly, I will need to get consensus from the community about how this shall work, what shall be the scope of the template, by actually taking into account other users expectations and desires, unlike... I estimate this process will take the rest of the year 2013 given the limited spare time I have for this and the rigour with which I will approach this. If putting this bot on hold until then is acceptable, I will be happy to do that. --Slaunger (talk) 19:11, 4 September 2013 (UTC)

          • Alas User:Slaunger is on wikibreak. Would anyone else like to have a go at this? Otherwise I'm afraid we are not ready for this bot task. --99of9 (talk) 10:26, 13 January 2014 (UTC)
  • Interesting discussion, even if a bit too technical for me. As a "basic" user, I think that current templates are not so bad, but it looks sometimes very "random" on the file page (my choice should be: description, assessment templates, before license). I think that awarded pictures are not visible enough in "Commons". And I notice that if FP or VI pictures can be "delisted", it is not the case for QI (QI for one day, QI for ever), this should not been forgotten in case of a "merging" is decided (a template for "former FP" exists too...). No further opinion for the moment. Only my three cents.--Jebulon (talk) 13:53, 6 September 2013 (UTC)
  • @Ricordisamoa, @Slaunger, @99of9, @Dschwen: What's the state of this? Are we waiting for anything in particular, or can we push this request forwards (or else close as stale)? odder (talk) 19:33, 2 April 2014 (UTC)
  • Pictogram voting info.svg Bureaucrat note: I'm closing this request as stale given there have been no updates on it for more than two months, and no updates even though I pinged several people a week ago. odder (talk) 13:59, 9 April 2014 (UTC)

Restart

  • Pictogram voting info.svg Bureaucrat note: I just heard back from @Ricordisamoa, so let's re-open this request and see where it'll get us. odder (talk) 18:51, 14 June 2014 (UTC)
    • Hello? odder (talk) 21:50, 25 June 2014 (UTC)

The bot could run on FPs that are also QIs, skipping VIs until {{Assessments}} supports them properly. Does anyone oppose simple mergers like that? --Ricordisamoa 00:09, 1 July 2014 (UTC)

That seems reasonable to me. --99of9 (talk) 01:43, 1 July 2014 (UTC)
Good; but there is an ongoing discussion on how to handle sets which may affect some previous cases too. Further, a change may be needed for {{Assessments}}. Jee 02:25, 1 July 2014 (UTC)

And what about the placement of {{Assessments}} within the page? If desired, the bot can adjust its position according to some policy/recommendation. --Ricordisamoa 15:22, 17 July 2014 (UTC)

I prefer it below {{Information}}. Some people prefer/move it above {{Information}}. I think it should be discouraged. Jee 15:55, 17 July 2014 (UTC)

If a policy about the placement of the Assessments template doesn't exist, I can't enforce it. But I made a short test-run on file description pages that included {{Assessments}} and {{Quality image}} but not {{Valued image}}, according to catscan2. There are about 1850 of them left. Please keep in mind that the script is still in development and far from ready for 'production' use. --Ricordisamoa 15:39, 11 August 2014 (UTC)

  • @Jkadavoor: Have there been any updates to the discussion that you linked? I had a quick look at it, and it seems to have ended on 30 June, without any further updates or consensus. @Ricordisamoa: Have you managed to finish your work on the script to make it ready for use? If not, how far into the future are we talking? :-) Thank you for your time! odder (talk) 11:09, 28 October 2014 (UTC)
  • odder, sets are banned now; and no progress in re-enabling them. But we can conclude that there is no plan to change the way in handling existing sets. Jee 11:53, 28 October 2014 (UTC)
  • odder, the script is getting better and better, but since it touches templates I'm in no way going to run it automatically. And we have file metadata access from Lua now! It shouldn't be hard to add a 'wallpaper'-like check to Module:Assessments, make the template use the module, deprecate the 'wallpaper' parameter and remove it while merging {{Assessments}} with {{QualityImage}} and such ;-) --Ricordisamoa 10:03, 11 December 2014 (UTC)
  • Pictogram voting comment.svg Comment -- PLEASE stop the bot and read the previous discussions on this matter before forcing the changes into the image files: in here and also in archives 1 and 2 of the same page. As Slauger, I don't acknowledge the legitimacy of those changes and will continue to revert them. Alvesgaspar (talk) 17:08, 11 December 2014 (UTC)
  • Pictogram voting comment.svg Comment First of all, I would like to excuse for having been absent in this discussion. I am sorry, but I still have concerns about the merger of all this information in one template and especially the manner in which they are presented. I expressed it previously in September 2013 above, and I feel like we are pursuing the wrong technical solution.
  1. I agree on the objective of cleaning up file page real estate, but there are IMO other, more maintainable and 'prettier' ways to to that
  2. Question: I am not good at wiki-markup, but could an alternative solution be to tweak the current implementation of the FP, QI and VI templates, such that if they are placed in order
    {{FP...}}{{QI...}}{{VI...}}
    
    they would render as a table with three columns, the first column with the FP frame and background colour, the second one with the green QI frame and background and the third with the golden frame and background colors of the VI? If so, we would get the saving in file page real estate whilst maintaining the color trademarks/identities for the three individual assessment projects. In that case the main job of Samoabot would be to simply order the templates consistently. It would have the benefit of simplifying template maintenance also, as well as bot scripts. -- Slaunger (talk) 18:21, 15 December 2014 (UTC)
  3. I still do not like the incomprehensible arguments to the Assessment templates with giving weird numbers as arguments, which has no clear mneomotecnic meaning, which is a legacy of the old terrible temple. (I am not fluent in Lua, but the new Module looks much more maintainable and better structured). -- Slaunger (talk) 18:21, 15 December 2014 (UTC)
  • @Slaunger: '1' is a standard and cross-language value to mean the boolean 'true'. In template code it is much easier to check for named parameters with a given name than for positional ones with a given value. In Lua the latter is fairly easy, so Module:Assessments has an option to allow things like {{Assessments|featured|quality}}, but I haven't enabled it yet. Once the module has been deployed, we may choose that form as the preferred one and make the existing FPCBot and QICbot use it. Regarding the UI design, I am happy that you're proposing an alternative, but it doesn't appear to cover parameters like 'enwiki' and I feel it is less machine-readable and not properly suitable for mobile devices. --Ricordisamoa 20:22, 26 December 2014 (UTC)

As there is not progress for a reasonable amount of time, I'm closing this as stale / not approved. --Krd 17:44, 5 June 2015 (UTC)