Andy Parkins [ARCHIVE] on Nostr: 📅 Original date posted:2013-07-23 📝 Original message:On Tuesday 23 July 2013 ...

📅 Original date posted:2013-07-23
📝 Original message:On Tuesday 23 July 2013 10:42:05 Pieter Wuille wrote:
> On Tue, Jul 23, 2013 at 10:30:13AM +0100, Andy Parkins wrote:
> > One additional URL makes this pretty much perfect:
> > GET /rest/block-with-tx/TX-HASH
> >
> > Construction of the transaction-hash-to-block database is something the
> > full client's have to do anyway, so this query is no harder than the
> > others for them to supply; but suddenly makes it possible for an SPV
> > client to trace the providence of any transaction without needing to
> > maintain the entire chain.
>
> There is actually no such index being maintained by default, and doing so
> is an unnecessary burden IMHO (you need to enable -txindex since 0.8 to
> get this). Of course, if enabled, it can be exposed.

Wow. I'm surprised at that. How does a newly received transaction have its
inputs verified then? Multiple linear brute force searches of the block chain
for every new transaction? Or is it that transactions are only recorded if
they were in a block, and just their presence indicates they're valid?


Andy

--
Dr Andy Parkins
andyparkins at gmail.com