Delete legacy blob component#50915
Conversation
Bundle size report 📦
Look at the Statoscope report for a full comparison between the commits 6a4bf80 and ff90add or learn more. Open explanation
|
This reverts commit 34f1b98.
96b1e20 to
6e5f46f
Compare
|
Codenotify: Notifying subscribers in CODENOTIFY files for diff ff90add...6a4bf80.
|
|
Awesome! Does this mean we can also get rid of the fact that the first of the two For this first query we only need the |
e31bd4d to
a8efd30
Compare
I guess we can. Setting the default value |
|
We can even remove the |
|
Oops sorry (missed the comment box...) |
|
I think we will still see code load noticeably faster with I did a quick test loading this file:
That files pretty big but I still see a 50ms-100ms improvement when loading medium sized files. For small files the difference is barely noticeable. I think we should keep it. Btw, |
| // We only want to include HTML data if explicitly requested. We always | ||
| // include LSIF because this is used for languages that are configured | ||
| // to be processed with tree sitter (and is used when explicitly | ||
| // requested via JSON_SCIP). | ||
| const html = [HighlightResponseFormat.HTML_PLAINTEXT, HighlightResponseFormat.HTML_HIGHLIGHT].includes(format) |
There was a problem hiding this comment.
I think we'll get a perf boost from this PR. Looks like we were fetching html for plaintext responses when we didn't actually need to, nice!


Removes legacy (table-based) blob component.
CodeMirror has been a default blob viewer since Sourcegraph
4.5.0. Table-based legacy blob viewer was left as an opt-out option in case we encounter major issues with the CodeMirror blob viewer. Luckily no significant issues were reported so we can remove the legacy (table-based) blob view.Test plan