💬 Crypt-iQ commented on pull request "net: check for empty header before calling FillBlock":
(https://github.com/bitcoin/bitcoin/pull/33296#discussion_r2319795093)
Yup good point, will do.

💬 achow101 commented on pull request "wallet: Identify transactions spending 0-value outputs, and add tests for anchor outputs in a wallet":
(https://github.com/bitcoin/bitcoin/pull/33268#discussion_r2319820571)
Done as suggested

💬 achow101 commented on pull request "wallet: Identify transactions spending 0-value outputs, and add tests for anchor outputs in a wallet":
(https://github.com/bitcoin/bitcoin/pull/33268#discussion_r2319820786)
Moved

💬 achow101 commented on pull request "wallet: Identify transactions spending 0-value outputs, and add tests for anchor outputs in a wallet":
(https://github.com/bitcoin/bitcoin/pull/33268#discussion_r2319821210)
Done

💬 achow101 commented on pull request "wallet: Identify transactions spending 0-value outputs, and add tests for anchor outputs in a wallet":
(https://github.com/bitcoin/bitcoin/pull/33268#discussion_r2319821459)
Done

💬 achow101 commented on pull request "wallet: Identify transactions spending 0-value outputs, and add tests for anchor outputs in a wallet":
(https://github.com/bitcoin/bitcoin/pull/33268#discussion_r2319823101)
Done

💬 glozow commented on pull request "Revert compact block cache inefficiencies":
(https://github.com/bitcoin/bitcoin/pull/33253#issuecomment-3250303444)
I suppose it depends on how much of compact block relay latency these changes account for?

💬 romanz commented on pull request "index: store per-block transaction locations for efficient lookups":
(https://github.com/bitcoin/bitcoin/pull/32541#issuecomment-3250323045)
> Are you seeing greater improvements on other setups?

I think that the performance gain will increase when fetching transactions near the end of the block.

For example, when fetching the 5000th transaction of [block #90005](https://mempool.space/block/000000000000000000017bfd05b5fa367a424c4a565a4baf7950d9e8605df8ec) using `ab -k -c 1 -n 100000`[^1], enabled `locationsindex` improves the performance ~19x (2.574ms → 0.136ms).

## `-locationsindex=1`

```
Document Path: /rest/
...

📝 jmoik opened a pull request: "cmake: Inherit WERROR setting for secp256k1 build"
(https://github.com/bitcoin/bitcoin/pull/33297)
This change prevents unexpected CLI output when compiling with -DWERROR=ON (treat warnings as errors). This setting is now properly propagated to the secp256k1 subproject build. Previously, secp256k1 would be built without the -Werror flag even when the main project was configured to treat warnings as errors, creating inconsistent build behavior.
https://github.com/bitcoin/bitcoin/issues/33284

💬 sipa commented on pull request "Add functional test for IPC interface":
(https://github.com/bitcoin/bitcoin/pull/33201#discussion_r2319855566)
Yeah, the import is needed at the top level, I believe.

💬 sipa commented on pull request "Add functional test for IPC interface":
(https://github.com/bitcoin/bitcoin/pull/33201#discussion_r2319857824)
Thanks, included.

💬 sipa commented on pull request "Add functional test for IPC interface":
(https://github.com/bitcoin/bitcoin/pull/33201#discussion_r2319855049)
Done.

💬 sipa commented on pull request "Add functional test for IPC interface":
(https://github.com/bitcoin/bitcoin/pull/33201#discussion_r2319858473)
Fixed.

💬 sipa commented on pull request "Add functional test for IPC interface":
(https://github.com/bitcoin/bitcoin/pull/33201#discussion_r2319856069)
Leaving as-is.

💬 sipa commented on pull request "Add functional test for IPC interface":
(https://github.com/bitcoin/bitcoin/pull/33201#discussion_r2319857541)
Done.

💬 sipa commented on pull request "Add functional test for IPC interface":
(https://github.com/bitcoin/bitcoin/pull/33201#discussion_r2319857383)
Fixed.

💬 sipa commented on pull request "Add functional test for IPC interface":
(https://github.com/bitcoin/bitcoin/pull/33201#discussion_r2319857246)
Fixed.

💬 sipa commented on pull request "Add functional test for IPC interface":
(https://github.com/bitcoin/bitcoin/pull/33201#discussion_r2319857067)
I've removed it; let's see where CI breaks.

💬 romanz commented on pull request "index: store per-block transaction locations for efficient lookups":
(https://github.com/bitcoin/bitcoin/pull/32541#issuecomment-3250359798)
Ran the same test using `hey` 0.1.4 - enabling `locationsindex` improves the performance ~15.8x (2.963ms → 0.187ms)

## `-locationsindex=1`
```
hey -c 1 -n 10000 http://localhost:8332/rest/txfromblock/$BLOCKHASH-5000.bin

Summary:
Total: 1.8766 secs
Slowest: 0.0686 secs
Fastest: 0.0001 secs
Average: 0.0002 secs
Requests/sec: 5328.8354

Total data: 2340000 bytes
Size/request: 234 bytes

Response time histogram:
0.000 [1] |
0.007 [9998] |■■■■■■■■■■■■■■■■■■■■■
...

💬 romanz commented on pull request "index: store per-block transaction locations for efficient lookups":
(https://github.com/bitcoin/bitcoin/pull/32541#discussion_r2319877158)
Yes - but if it's OK, I would prefer to do it in a separate PR.