tesshucom
commented
1 week ago Now that we've finally caught up, let's think about it.
Subsonic's Year classification was by decade. No doubt there would be no other default. More as a matter of common sense than a music software specification. We often use terms like 80's and 90's to distinguish musical generations!
That probably won't be enough for large libraries. Let's consider some derived features.
- Change the interval to 5 years or even annually, instead of 10 years.
- It's the most powerful and unquestionable way we can show the songs included in the year. We can narrow the number of nodes by narrowing the year range.
- Divide by the number of songs. In other words, paging. Every 100 songs, every 200 songs, every 300 songs...
- You can extract songs from chronologically arranged songs by their approximate position based on the number and position of the songs. Of course, they are arranged by year, but songs from different adjacent years may be mixed on the same page. If you want to extract a specific number of songs to listen to, or even just see the percentages by year, it's easy.
- Frequency distribution
- Automatic classification based on the number of songs in the entire library
When talking about topics like this, we tend to get caught up in a narrow discussion of how customizable the feature is, anyway. Perhaps providing too many detailed options would overwhelm most users, in a bad way.
While users often have a vague idea of the era of the songs in their library, they often don't know the exact era. There is a contradiction because we are trying to customize something ambiguous. Customization would be impossible, or you would have to go through a lot of trial and error. This would be putting the cart before the horse, and would be considered an unhelpful functional design that takes away from the software's purpose.
So, here are the essential menus to provide to users:
- Variable interval view. 10 years (default)/5 years/1 year
- Frequency distribution
Please note that this can be used with very little effort on your part or even a deep understanding of the library.
Next, we provide a simple customization menu.
- Paging
Note that this only works if you have some idea of the volume of your data.
Prerequisites: #2349
Overvew
Year will be added to UPnP.
Details
Of course, Year, which is similar to Subsonic, will also be ported, but the existing version is difficult to use. (Although its operating speed is the fastest) Well, I have some ideas. Modify the year class width to be configurable. Every 5 years, every 10 years, etc.
Or, completely different algorithm. For example, you can decide on 8 layers in advance, distribute the songs somewhat evenly, and think of the year range as just the result.
for example, [1970-1978] [1979] [1980] [1981-1983] [1984-1987] [1987-1994] [1995-2010] [2011-2024]
Ha, the results will vary depending on the songs the user has. The topic will be something like the basics of statistics. In other words, there is no right answer. There's no right answer, but paradoxically, I can intuitively feel that it's best to keep the number of settings to a minimum.
My hypothesis is that by creating a few realistic Presets, we can save a lot of the effort of maintaining smart lists that are not smart. Of course, in order to have complete control, functions that can be specified for all are considered, but the cost and performance will be extremely poor. (Actually, most of the software on that route is full of bugs or lacks resource estimation.) If I were a user, I wouldn't want to use software that forces me to keep managing smart lists. (It is better not to consume the user's human resources either. That is the mission of the software.)