Make use of `helm-buffers-sort-transformer` customizable #763
Comments
Matus Goljer notifications@github.com writes:
Because I think this is really handy. Thierry |
Can you explain the reasoning? I don't see any advantage in bringing buffers with short names to the front, they are no more relevant to the search query than any other buffer. Recently used buffers, however, are relevant because they are being used and they should be accessed in the fastest way possible. |
@Fuco1 I think it is because it is used for simple ranking. The shortest candidates have the highest match to the provided Helm patterns. Since @thierryvolpiatto probably we can have an option to disable buffer sorting and leave the order as it is? |
Matus Goljer notifications@github.com writes:
Your more recent buffers are on top at startup, if you search a buffer, You can consider this as Wontfix. Thierry |
I think when one use |
At init helm, buffers are ordered by recency, However after we begin to narrow down to select a candidate, the buffers change to being sorted by length. This is nonsensical because more often then not we are looking for the most recent buffer matching the pattern, not the shortest. See emacs-helm#763 for details.
At init helm, buffers are ordered by recency, However after we begin to narrow down to select a candidate, the buffers change to being sorted by length. This is nonsensical because more often then not we are looking for the most recent buffer matching the pattern, not the shortest. See emacs-helm#763 for details.
At init helm, buffers are ordered by recency, However after we begin to narrow down to select a candidate, the buffers change to being sorted by length. This is nonsensical because more often then not we are looking for the most recent buffer matching the pattern, not the shortest. See emacs-helm#763 for details.
With no narrowing,
helm-buffers-list
lists the buffers in LRU order. However, as soon as I narrow, it sorts them by buffer name length. I consider this an anti-feature: I usually want to jump to the least recently used, not to the shortest buffer matching a pattern.Is there any reason at all to re-sort the candidates?
The text was updated successfully, but these errors were encountered: