User blog:Inept Wiki User/STOP Notes (22/07/2021)

Series Three Overhaul Project - Loose Notes (22/07/2021) These are some loose notes concerning the Series Three Overhaul Project (STOP) and other recent stuff on the wiki, as of July 22th 2021.

Due to circumstances at present, both on and off the wiki, these notes will not be particularly well sorted for now.

=Opening=
 * Previous note attempt from July 15th aborted and integrated into this post.
 * Transferred raw data (for future use and reference) from auto-sweep/similar projects in past couple of weeks stored on July 15th log. Assembly/processing incomplete. Previous one on July 5th.
 * Desktop Transition Work (04/07/2021) assembly and tasks incomplete.
 * Separate post planned for odd jobs and tasklists. Help would be appreciated with tasks there when it is ready.
 * Recent Index IF event/gacha additions, latest Index manga chapter covering Index V16 Ch3 end & BTL - Ch4 start in mid-month, new Railgun chapter in a few days.
 * Previous STOP Notes batch from July 10th.
 * Edit (27/07/2021): Toaru Kagaku no Mental Out, first mentioned a number of months ago, has finally begun, along with respective edits.

=Assorted Notes=
 * Note: Auto-Sweep project information in section "Recent Auto-Sweep Sub-Projects" below (jump link omitted due to FD's new content jumper).
 * Note: List of stuff to do on Odd Jobs
 * Note: Need to find solution to problem of ending up spending days writing these notes (to detail stuff to others), such that it is nearly time for the next one by the time it is nearly completion. Besides taking up time and effort, it is causing delays for other work, including the transmission of key details to collaborators.

Unsorted:
 * Saw a comment regarding MMO closure a couple of weeks ago. Need to verify truth and if true, see what information needs to be acquired beforehand.
 * On that note, see about the collection of defunct/near-defunct game data
 * Have noted this previously, but so don't have to do much for a while for Featured Articles (switching currently once every two months), build up the 'stock' of readied candidates, so it can then be left for a while.
 * Previously noted, Category:Organizational Redirects created on July 8th for redirects involving site operations.
 * Considerable help will be needed with several large-volume data assembly/processing tasks.

Recent Release Related

 * Railgun, First Year Arc: Assembly continuing on . Repeated chapter splits (likely recombination in volume releases). Given circumstances, images for various unnamed character/abilities for SD202 and . Background used for events.


 * Toaru Kagaku no Mental Out: General assembly to be carried out for the new spinoff.
 * Toaru IF: As ever, data assembly for information relating to Toaru IF is continuing on.
 * Previously, several copies of the table used to list/gather basic information/images about 3 star character releases were created, in order to gather additional pieces of specific information and sort/process them separately while details for the design of the next stage table is being sorted out. Not much progress has been made on the data assembly in these three separated tables.
 * However, things are progressing on the table design front - for this table, several fields are definite (#, Image, Title, Character, Notes/Details) and there are options (Type, Date of Release, Release Type etc) which could be fields for sorting purposes (given this is a sortable table). Given the width required for some of the definite fields, there is a limit to how many optional fields can be included. however several of the definite fields can be used to carry out the effective sorting function that the optional possibilities could (e.g. # could carry out the same effective sort as Date of Release, and depending on what is at the top of the Note-type field, it can carry the sorting pattern of one of the possibilities).
 * There hasn't been much more information assembly concerning the character sidestories and event stories for a while - this information would be useful to have gathered.
 * The other potential project previously discussed with Herald is still on-hold.

GT4-related
GT4-related (/):
 * Numerous sections still being drafted and refined (Personalities among them). Some other additions to pages in need of polishing.
 * Need to avoid heedlessness and haste in doing these.
 * Los Angeles: Given the circumstances and the city's prominent role in the story, this is going to get an article of its own. Information regarding the locations within the city will be gathered on its own Locations section and they will not receive separate articles of their own. There will be no other city/settlement article creations at this time, given the respective circumstances.
 * Helcalia Grocery: Following note regarding reused character design, have assembled a table with the design images of characters with designs derived from reused previous design images of other character on . However, have not yet identified the specific design image for Index Librorum Prohibitorum which was recycled for Helcalia and referred to in that instance.
 * Between the Lines: Had been considered certain specifics regarding the bits regarding these, and then considered that there were other non-standard cases in previous volumes that might need a slight tweak, as well as the usefulness of having data on the subject readily available. Have assigned to the collection of data on non-standard Between the Lines sections, non-standard start-end sections, and split Epilogues (i.e. one bit of the Epilogue, then the Afterword, then a bit afterwards (though not always labelled as such))
 * Melzabeth Grocery/Space Engage: Was planning to make notes on Melzabeth's watch and dwellings. For mobile home, was uncertain about choice between Space Engage or Melzabeth's Trivia.
 * As Ollerus pointed out, Emptiness Marriage is a Transla-Pen translation error and as such, now redirects to Space Engage. It has been noted on Ch3's Trivia Expand, but Nihongo has not been added yet.
 * Elizard: Thinking about making a few notes regarding her drinks collection in Trivia, given NT22 and GT4.
 * Cryptography: Given elements of codes and encryption featured early on in GT4, coupled with Agnese SS and other previous stuff, have assigned to assembling all information in the series connected to the subject of Cryptography.
 * Philosopher's Stone: Am planning to make a large sub-section on Alchemy at a minimum, or make an article for it at maximum, given the role of the stone and the process behind it in GT4, coupled with the circumstances of previous mentions, and Alchemic/Rosicrucian featuring of it in the story.
 * R&C Occultics: Have been planning to add infobox and transfer the initial image to it. I'm working on more stuff for Principles.
 * Mass Driver: The previous categorization before GT4 and its subsequent role in GT4 prompts consideration about technologies formerly AC or non-AC but now found in both.


 * Over-Arc: In light of RCO's destruction in GT4, Ollerus proposed changing the current over-arc's name from R&C Occultics Arc to Anna Sprengel Arc. Another possibility which has been considered, taking into account spoilers, current and potential future events, is the New Age Arc. However, the matter is not particularly urgent (and the name still holds at present), and additional data to be considered will only arise when information emerges for the continuation.

Images

 * Checking required for recent image tweaks on pages and other images on pages between Desktop skins and at different collapse/zoom/side bar settings.
 * Redirect Images: The list of image redirects (redirects leading to images) is up to date, however not sure that the list of redirect images (images on redirects) compiled a number of months back is up to date.
 * Need to verify the state of the list and the state of any operations to verify/update the list.
 * Carry out identification sweeps as necessary and compare results to acquire relevant information.
 * Considering categorizing under new category: Redirect Images - should also consider solutions to the negligence problem that might occur (e.g. similar to people not updating an image's source, categories or other things when they add/remove/change the image).
 * Category:Image Redirects page gives only images on Classic Categories rather than Dynamic Categories - this is on both Oasis and FD.
 * Have been planning to conduct a possible ad-hoc image renaming session (overlapping somewhat with SD15X-SD19X original task). Intend to do it separately from the other major endeavors and had previously been waiting until the TT21-TT23 renewal sweeps were completed so the all-file sweep worked properly (see section "TT21-23: Image DSP-IRI Checks")
 * Thinking it might be an idea to keep a record of the renames and deletions in these cases for reference purposes (either via log extracts or manual recording).
 * See Odd Jobs post when ready for further details on images currently sought or related tasks

User Actions and Personal Feelings
'''Compilation incomplete and unstarted. Unprocessed'''

=Templates=
 * Wish to check how many levels of template within template can be done before malfunctions or other problems arise.
 * Have noted various instances of template documentation (either on the template page or on a separate documentation page), that need to be brought up to scratch (these include ones for D's recent creations).

Minor template ideas which I've toyed with but not properly considered or committed to:
 * A smaller, more discrete or less obtrusive version of the Overhaul tag (possibly for marking non-immediate (but inevitably needing work) sections).
 * A visible marking template for temporary galleries (there are already TrackTags marking most of them, but was thinking about another possibility for future work).
 * A non-full-width version of the Cleanup tag.

A number of templates have been noted as having spacing problems post-FD switch (Stub might have had one, but I haven't found any notes made regarding pages on which it was observed, so no way to know if still there, only in certain places, already fixed, or misremembered). For Expand, there was a large gap between the default text and the case-specific text. I've managed to resolve this by removing a line-break (shown below), but further texts might be needed.

to

Character Tabs
One thing that I was recently pondering was whether it would be possible to group the Chronology sub-page links on the Character Tab templates (a combination of top-nav tabs and character infobox, shared between character sub-pages and used to save having to change multiple separate instances of the same thing). I was initially thinking about a tweak to the base template used to produce the top-nav tabs, the Parent Tabs Template, however this seemed too tricky and as though it would cause more trouble than it was worth.

Then another approach came to mind - one which would achieve the desired result but also have another small benefit.
 * Instead of tweaking the Parent Tab Template, the idea is to replace the instance of it in the combined character tab/infobox template with a modified version of the source code used for the Parent Tab Template, tweaked to serve the purpose of the character article and sub-pages.
 * This would allow for modifications in a simpler/specific manner and without causing problems elsewhere.
 * Additionally, it would also allow for the decommissioning of 35 anti-redlink redirects used to prevent the template from producing items in Special:WantedPages (these are because the Parent Tab Template uses PAGENAME in its tab link coding, which results in links on the combined template page itself (e.g. PageName/SubPage on the page itself produces TemplatePage/SubPage link on the combined template page - with specific links to the pages in question, there won't be any redlinks).

For example, Misaka Mikoto's tab template is currently like this:

Which appears like this:

Changing it to something like this:

Would make it appear like this:

Edit (24/07/2021): Proposed edits carried out successfully for the combined character infobox/tab templates that have multiple Chronology subpages. The ones with just one left as they are.

Navboxes
One thing I found in the past few months was a method of separating the links within a navbox's source code into a form which can be used by MassEdit (e.g. for implementation via auto-sweep and other similar purposes). It involves copying the source code to an empty or nearly empty page (such as an unassigned Shared Draft) and then carrying out a series of coordinated auto-sweeps to create line-breaks between the links and the rest of the template code (this will also create gaps between a lot of other parts of the template code, but this won't matter). The main targets will be the square-brackets that are used in the links. Once the line-breaks are placed, the whole thing should be in a form where an external list sorting application can separate the links from all the other stuff. Then one just has to discard the rest, remove the brackets via FindAndReplace, and the list should be ready. Regex alternatives for separating, tagging or direct list compiling not explored yet.

There are a few methods which I have been thinking about but have yet to learn or rediscover; one is a method of excluding links on a navbox template from the WLH (Special:WhatLinksHere) of other pages (besides temporarily disabling or removing the links from the navbox itself and waiting for the cache to clear). Two others, which might be accomplishable through regex applications, are a method of checking to see if there are any Navgation headers missing (i.e. any navboxes or groups of navboxes without a Navigation header above them) and a way of reordering groups of navboxes.

As with a number of other templates, there is the matter of documentation (such as for Damage's recent creations), but that has been noted previously.

With regards to story arc navboxes (two of which were done poorly by A125 and later decommissioned), Damage3245 began making drafts for some a few weeks ago. These are something which shouldn't be rushed and must be done careful (on that note, a tendency for haste and assumption by various people has caused me quite a considerable amount of stress in the past), but for now, as indicated by communications and notes, the drafts are not have anything more than the recent volumes/chapters (novel/manga), episodes and perhaps events for now. Damage has just done the outer frame and is just adding the novels/chapters for now, though with two buttons leading to previous and next arcs (these might need a slight thinning).

The subject of navbox images and cataloguing/categorizing/tracking/adjusting them shall be noted and detailed some other time, but I will note that at the time of writing I'm currently not in favor of fully dispensing with them, as considered by one of my colleagues.

Shared Draft Header
As noted on the Desktop Transition Work posts, there are a number of templates which were affected by the transition to FandomDesktop - in this case, their background/text settings caused a problem (mainly visibility) when switching between Light and Dark Modes. OH&S has fixed the infobox templates and is looking to fix other noted ones (and we should probably check to see if there are any missing from the list), but there is one case in particular which I'm going to have to work on.

When I first devised and created the Shared Drafts, I did not create a header template to share between them (there were a number of reasons if I recall correctly - there was a conversion switch that was being encouraged at the time and functions would be disrupted if someone went through with it for the header (though Parent Tab Template wasn't forced through), there weren't so many the Shared Drafts originally and they could be modified individually without much hassle, there would be complications in implementing the return links which were intended to assist the original intentions for joint editing (the template being placed on a sandbox, edited via the edit link and then returning to the sandbox via the return link - rather than just visiting and editing the template itself).

However, with circumstances having changed considerably since then, I'm considering making a Shared Draft Header and implementing it between the Shared Drafts, in order to support maintenance and monitoring (making sure the return links are up-to-date etc - though an improved WLH plan might be needed), make modifying the shared headers easier, and to solve the night-mode display problem. There are a few problems concerning bulk implementation and the return links, but I think a few carefully planned and coordinated auto-sweeps with MassEdit will be able to accomplish them. Countermeasures can also be taken against an inappropriate shift.

Side Note: Some of the Shared Drafts have other Shared Drafts within them (e.g. the GTX tasklist with the GTX plans)(but don't have an SD within a SD within SD) - that is partly why I intended to check how many levels of template within template were possible/advisable before a malfunction or problems occur.

Side Note: Potential coding and sweeping problems

Shared Drafts and TrackTags

 * Number of these has greatly expanded due to various circumstances, including:
 * Need for them when none are available (and breathing room).
 * Simultaneous large-scale and/or long-term projects tying up existing SDs and TTs (and sometimes getting into a muddle).
 * Suggestion to reign them in and consolidate them to a more manageable amount.
 * Note: Opinions regarding the right amount might differ.
 * The above is NOT to be carried out at present. Much more work and preparation is needed before then, and circumstances are still problematic.
 * Cleaning and sorting out is needed.
 * Necessary to retain and preserve certain SD/TT Histories for future reference and work purposes (several of them are currently linked with History quick-view template for those purposes) - possibility of using the Merge trick to transfer them to another SD/TT, blog post entry or sandbox page. Straight deletion without heeding this is NOT an option.

Qref Project
Compilation incomplete
 * Planning to put notes for templates relating to OH&S's quick reference project (planning/developing Qref/Qref2 and ref guide) on a separate section for that project.

=Recent Auto-Sweep Sub-Projects=
 * Much of the raw data from these type of projects is recorded for future reference on STOP Notes or similar posts. For the most recent ones, the data was recorded on two separate log posts, dated July 5th (this particular one also has a certain amount of offload (with SD history links) and overflow holding) and July 15th.
 * Intend to extract anomaly data from the raw data log and perform supplementary sweeps.
 * Series of reference-related sweeps were conducted between June and July to locate and switch remaining old-style references (name/content) to the current format (including Chap#P, Chap#X, XMajutsuNovelX, "Chapter 0X_", ChXBTL, V0Ch0Pt0, MangaCh etc). Large amount done but currently suspended due to the matter of OH&S's recent work.
 * Sweep conducted in the course of locating missed uncategorized redirects during the assembly of the most recent sweep aid lists (July 5th).

Gallery HideAddButton Sweeps
One of the main sweep-related sub-projects which was carried out in the middle of July (sweeps executed on July 14th-15th after testing). It involved the addition of hideaddbutton parameters to numerous galleries across the wiki, which due to the volume had to be done by bot (avoiding potential RecentChanges clogs) and in bulk.

Auto-sweep work had been previously carried out starting in April and going into May, in order to catalogue the different gallery variations (in terms of parameter settings) present on the wiki, a foundation task towards further operations. The initial identification involved tagging all instances of the first part of the standard gallery opening tag ( <gallery ), then performing a sweep to switch the tags for the ones without parameters (if I recall correctly, TT44 for the parameter ones and TT50 for the non-parameter ones (infobox cases exempted)). Successive coordinated sweeps were conducted to identify the locations and parameters for the galleries (mainly involving taking one set of parameters from a sample, performing a switch sweep on the tagged items (switching them to a tag assigned for done items), recording the results of the sweep (RecentChanges/MassEdit), and then repeating the cycle until all items were done) - a complicated and slow process, somewhat hampered by some others' actions at times. The results were recorded on SD208, several close SDs, and several STOP Notes and similar log posts from April and May.

Adding the hideaddbutton parameter to the non-parameter cases was fairly simple, but for the ones with parameters, a coventional text-based MassEdit auto-sweep was not feasible due to the considerable variation, increased slightly further by actions since the ID sweeps. Hence, regex would have to be used instead.

My attempts to learn regex earlier this month were hampered by workload and other circumstances, but thankfully OH&S helped me by devising a pair of regex formulae which would carry out the functions of sorting and addition the target gallery parameter. I had made an oversight when describing the circumstances (hideaddbutton=true rather than hideaddbutton="true"), but a minor adjustment to the sweep sequence before the start prevented any problems.

First, an all-page sweep would remove the quotation marks from hideaddbutton="true" would be removed, then the re-arrangement regex sweep would be conducted to move all instances of hideaddbutton=true to the end of the gallery opening tags. Next, a sweep would be conducted to add hideaddbutton=true to the end of all gallery opening tags which lack them (excluding the infobox galleries). Then the quotes would be re-added, the end result being that all of the targets should have hideaddbutton="true" at the end of their gallery opening tags.

A variation of the reordering regex sequence could be used in future sweeps to reorganize the other gallery parameters on the various galleries across the wiki to get them all in a specific order.

TT21-23: Image DSP-IRI Checks
During the Image Information Survey, TrackTags 21, 22 and 23 were assigned to locate and track cases where an image's Description, Source and Purposes have IRIs (ImageReqInfo) in place. In the time since then, there have been cases where new images have been uploaded and IRIs have been used for those image's DSPs, but the respective tags not placed. There have also been instances where the tags were added in cases where they shouldn't have been or left in where they should have been removed.

The objective of these check sweeps were to add the tags in cases where they should be but weren't. While there was also the matter of removing them from where they shouldn't be, the priority was to get the newer cases logged first - as though there was the likelihood of ad-hoc renaming of certain images, which would render the AllFile list out-of-date and necessitate adjustments (extra hassle) in order to carry out addition sweeps properly without missing stuff out. The addition/check sweeps were carried out on July 15th-16th and the results recorded on the data log.

Side Note: A removal sweep from non-applicable cases could be done by adjusting the proper case tags slightly (e.g. {TT21} to {TT21|X}) and then executing a sweep to tweak or remove the unadjusted cases, so as to identify and verify the nature of the non-applicable cases, before switching the applicable ones back. Regex might also be a possibility.

Side Note: Several of the cases of the above are connected to the improper integration of unprocessed content preceded by TT30. TT27-TT29-related cases as well.

TT51/54/60: Mid-Paragraph Images and Others
This sub-project is connected to a problem which occurred following UCP transition at the end of last year. Before the UCP transition, thumbnails and slideshows within a body of text had that text wrap around them. After the transition, they create a large gap between the points where they are placed instead.

In order to identify these cases and deal with them as appropriate to the circumstances of the respective case, a series of all-page sweeps were conducted in Spring, based on an idea of something which would be common to many of them but not to many others
 * The idea was this - if a thumbnail image is present within a paragraph, there is a considerable chance that one of the two sets of square-brackets at the start and the end of the thumbnail will be in direct contact with the words next to it (Note: At this time, regex was not being considered).
 * In the case of closing brackets, such sweeps would also pick up cases where thumbnails at the top of a paragraph are in direct contact with the first word of the paragraph and need a line-break.
 * The sweeps were conducted sequentially, Aa-Zz, tagging with TT51. They did not cover direct contact with the opening brackets, or non-A-Z characters being the ones in contact with the closing brackets.
 * It was realized after the start, that the cases of pluralization of links (e.g. demons ) would also be caught by the sweep, however the sweeps were continued as information on the subjects of this oversight would be useful in other ways.

After the completion of these sweeps following this approach, the next stage was to separate the thumbnails from the pluralized/otherwise links. One idea which would separate some of them was that a full stop before the closing square brackets would occur in an image caption but not in a link. These cases were separated by a sweep to switch them to TT54.

There was a large gap while other things were worked on and a method of automatically separating them rather than a manual one was sought. The possibility of using regex via MassEdit was eventually considered and explored. My attempts at learning in early July was hampered by the volume of other tasks and problems on the go, but thankfully OH&S, more experienced in the use of regex, found a regex formula which would carry out the desired function.

The regex sweeps were carried out mid-July (around the 16th) following tests to confirm that they worked without issue, however at the time of initiation, the TT54 cases had been forgotten and only remembered when errors cropped up during the sweeps and checking of the results. This necessitated a few extra sweeps to correct the errors, but the objective was accomplished successfully (with a few manual adjustments). The thumbnails are now tagged under TT60, the links under TT51, and TT54 is now de-assigned, all of the TT54 items being switched to TT60. Sweep data from both MassEdit's log and RecentChanges has been recorded, which provides additional information (providing support for future work), both directly and indirectly.

The next stages are to deal with the tagged thumbnails as appropriate to their situation (something which I intend to do carefully), and to identify the cases of links which were tagged with TT51 - the latter can be done through successive specific removal sweeps but this would take a fair bit of time and expenditure, similar to the gallery parameter log of a few months ago. A few other minor manual adjustments will properly be needed.

TT53: Paragraph Ends without References
A recent auto-sweep sub-project was carried out during the middle part of the month (tagging active between July 13th-22nd), started partly due to irritation with repeated instances being seen during recent checks and edits, partly to have the bot be useful and make some progress while I was occupied with other stuff, and partly to obtain data to support future work.

The objective was to locate various instances in paragraphs where references are not present, in this particular case - at the end of the paragraph (there were various other types of cases, but given the circumstances, I thought "just this one type for now" ). One might think that a sweep for these targets could use Cite rather than a TrackTag, but there are many instances which would be tagged by the sweep for which Cite would not be appropriate, leading to further problems, and the objective this time was to find instances and catalogue their nature for future work.

The idea was to tag instances of paragraphs without references at their end (looking for a full stop followed by two line breaks - done specifically to avoid various non-applicable things), and then, through coordinated removal sweeps, ascertain their location and nature (as there are various cases where the target criteria might apply besides a case where people have neglected to put proper references (e,g, real-life/historical information of real-life subjects, chapter opening paragraphs, and certain non-reference-able sentences)).

The sweeps were carried out, with the initial tagging done followed by coordinated removal sweeps for instances immediately ahead of Synopsis (e.g. chapter opening paragraphs), Etymology (opening paragraphs) and Appearance (opening paragraphs and Etymology) section headers. However, given the circumstances (including time constraints, tediousness, and the possibility of a regex alternative (either sentence targeting or finding a way to find to isolate a search to a specific section)), I decided to abort the current plan partway through and de-tagged the instances in a reduced series of sweeps (one which still provided data for future work - ahead of Level 4 headers, then Level 3 then Level 2, and then all the others (which would not be at a section head)). Results and notes were taken during the course of the aborted operation, providing data for future work.

The non-section end abort sweep edits are still high up in the bot edit list in RecentChanges at time of writing, and I've been considering minimizing bot edits for a little bit in order to have them stay around for quick access.

=Other=

=Links=
 * User blog:Inept Wiki User/Desktop Transition Work (04/07/2021)
 * User blog:Inept Wiki User/Raw Data Log (15/07/2021)
 * User blog:Inept Wiki User/Odd Jobs (22/07/2021)

Previous Links

 * User blog:Inept Wiki User/STOP Notes (10/07/2021)‎