From: Karthik Nayak <karthik.188@gmail.com>
Date: Mon, 03 Nov 2025 13:39:08 GMT
Subject: Re: [PATCH v3 00/13] Carve out loose object source
Message-ID: <CAOLa=ZToV8KvgG44Yn5QniTNYwKo2d1QS9ngfazQ63uOWrc60A@mail.gmail.com>
In-Reply-To: <20251103-b4-pks-odb-loose-backend-v3-0-6a61ea977393@pks.im>

Patrick Steinhardt <ps@pks.im> writes:

[snip]

> Range-diff versus v2:
>
>  1:  330b7c17e6a =  1:  ed548a7ee4e odb: fix subtle logic to check whether an alternate is usable
>  2:  68b2e736d20 =  2:  59d5548ddfa odb: introduce `odb_source_new()`
>  3:  b6f9a3f6d26 =  3:  a417bd30153 odb: adjust naming to free object sources
>  4:  fe38c58ab20 =  4:  add2b7d112f object-file: move `fetch_if_missing`
>  5:  5fed6e7c429 !  5:  5245ba0c6a5 object-file: introduce `struct odb_loose_source`
>     @@ Metadata
>      Author: Patrick Steinhardt <ps@pks.im>
>
>       ## Commit message ##
>     -    object-file: introduce `struct odb_loose_source`
>     +    object-file: introduce `struct odb_source_loose`
>
>          Currently, all state that relates to loose objects is held directly by
>     -    the `struct odb_source`. Introduce a new `struct odb_loose_source` to
>     +    the `struct odb_source`. Introduce a new `struct odb_source_loose` to
>          hold the state instead so that it is entirely self-contained.
>
>          This structure will eventually morph into the backend for accessing
>     @@ object-file.c: void object_file_transaction_commit(struct odb_transaction *trans
>       	free(transaction);
>       }
>      +
>     -+struct odb_loose_source *odb_loose_source_new(struct odb_source *source)
>     ++struct odb_source_loose *odb_source_loose_new(struct odb_source *source)
>      +{
>     -+	struct odb_loose_source *loose;
>     ++	struct odb_source_loose *loose;

This is one of the points that Junio mentioned and I think the new
naming schemes flows more naturally into. Nice.

The range-diff looks good to me. Thanks.