View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0003539 | HTML & PERL | Feature Request - Database | public | 2022-11-21 16:17 | 2022-11-28 15:26 |
| Reporter | CDB-Man | Assigned To | |||
| Priority | normal | Severity | minor | Reproducibility | N/A |
| Status | new | Resolution | open | ||
| Summary | 0003539: Add song-anime-rel comments | ||||
| Description | Similar to how char-char-rel allow comments, add comments to song-anime-rel. Example: https://anidb.net/song/108675/anime Songs have various issues that need to be documented, for example: 1. credited vs uncredited 2. timestamp info to help with song crediting 3. different cuts of songs 4. other idiosyncrasies where having the ability to document comments would be helpful Currently, since song-anime-rel do not have a comment field, the comments are being stored in the song entries themselves. The problem with this is that anime-relation-specific data is being stored in the song entry. For example, if an insert song is used across 3 anime in the franchise, then comments entered in the song description for anime 1 will not apply to anime 2 and 3. Further discussion was had in here: https://anidb.net/admin/creq/16411850 | ||||
| Tags | No tags attached. | ||||
|
|
As I asked in https://anidb.net/admin/creq/16411850 and saw no objection, the only data we want to have and maintain in anime-song-rel is broadcast type and timestamps. The first suggestion is: 1 new column to anime-song table: - description (text) In this text field we will fill info by the template, if necessary. This one will be easier and faster to implement, but I think it will create next problems in the future: - it's just unvalidated text field and anybody can write here anything; - even if we will follow the template, it easily could be mistakes, typos, wrong separators; - as ADD creqs process instantly, nobody can notice the errors mentioned above and fix them right away. The second suggestion is: 2 new columns to anime-song table: - broadcast_type (enum) - timestamps (jsonb) (json column should works good there, because we only need to save and read this data without any filtering) timestamps format in the attached file (timestamps is array to support multiply plays of the 1 track in the 1 episode) All data could be validated: - broadcast_type - is final list of options - episode_id / episode_number - should be in this relation episode field; - timestamp - should be time. This one will be harder and slower to implenent, but because of validation, will be easier to maintain. The only question is front view (anime detail, song detail, add/edit forms). Also, as we will have episode in json, this data could be used (shown) on file detail pages for certain episode. timestamp.json (195 bytes)
{
"episode_id_1/episode_number_1": [
"[h]:mm:ss",
"mm:ss",
"m:ss",
...
],
"episode_id_2/episode_number_2": [
"m:ss"
],
...
} |
