bitcoin:12345 *is* a "real" URI. It's just not an absolute, hierarchical URI (a.k.a. a URL); rather, it's an opaque URI.
And your suggestion is actually in violation of the URI spec, since "blockhash", "txid", "block", and "address" are not host names. More correct would be: blockchain:?blockhash=00000000000000001003e880d500968d51157f210c632e08a652af3576600198 blockchain:?txid=3b95a766d7a99b87188d6875c8484cb2b310b78459b7816d4dfc3f0f7e04281a blockchain:?block=189000 blockchain:?address=1RicMooMWxqKczuRCa5D2dnJaUEn9ZJyn You should read the URI syntax RFC: https://tools.ietf.org/html/rfc3986 On Saturday, 29 August 2015, at 12:31 pm, Richard Moore via bitcoin-dev wrote: > I like the idea of having a standard for this, that all explorers (and even > core, eventually) would understand. > > I would recommend 2 changes though. First, using a real URI scheme, > blockchain:// so that we can just use normal URL parsing libraries. The > bitcoin: thing leads to additional code to mutate it into a proper URI before > passing it to URL parsing. And I think it would be fine to include the type > looking up. For example: > > blockchain://blockhash/00000000000000001003e880d500968d51157f210c632e08a652af3576600198 > blockchain://txid/3b95a766d7a99b87188d6875c8484cb2b310b78459b7816d4dfc3f0f7e04281a > blockchain://block/189000 > blockchain://address/1RicMooMWxqKczuRCa5D2dnJaUEn9ZJyn > > I think this would help the URI be more human understandable as well as give > the explorers the ability to optimize a bit what they are looking for when > hitting various databases. > > A possible future path could also include blockchain://tx/123000/4 for block > height, tx index... Another possibility could be blockchain://version which > would return a list of supported paths, version of the BIP supported, etc. > > The BIP should also specify little endian searching. I'm not sure, but would > it also make sense for this BIP to include what the return results should > look like? Maybe another, related BIP. > > > On Aug 29, 2015, at 7:48 AM, Marco Pontello via bitcoin-dev > > <bitcoin-dev@lists.linuxfoundation.org> wrote: > > > > Hi! > > My first post here, hope I'm following the right conventions. > > I had this humble idea for a while, so I thought to go ahead and propose > > it. > > > > BIP: XX > > Title: URI scheme for Blockchain exploration > > Author: Marco Pontello > > Status: Draft > > Type: Standards Track > > Created: 29 August 2015 > > > > Abstract > > ======== > > This BIP propose a simple URI scheme for looking up blocks, transactions, > > addresses on a Blockchain explorer. > > > > Motivation > > ========== > > The purpose of this URI scheme is to enable users to handle all the > > requests for details about blocks, transactions, etc. with their preferred > > tool (being that a web service or a local application). > > > > Currently a Bitcoin client usually point to an arbitrary blockchain > > explorer when the user look for the details of a transaction (es. Bitcoin > > Wallet use BitEasy, Mycelium or Electrum use Blockchain.info, etc.). > > Other times resorting to cut&paste is needed. > > The same happens with posts and messages that reference some particular > > txs or blocks, if they provide links at all. > > > > Specification > > ============= > > The URI follow this simple form: > > > > blockchain: <hash/string> > > > > Examples: > > > > blockchain:00000000000000001003e880d500968d51157f210c632e08a652af3576600198 > > blockchain:001949 > > blockchain:3b95a766d7a99b87188d6875c8484cb2b310b78459b7816d4dfc3f0f7e04281a > > > > Rationale > > ========= > > I thought about using some more complex scheme, or adding qualifiers to > > distinguish blocks from txs, but in the end I think that keeping it simple > > should be practical enough. Blockchain explorers can apply the same > > disambiguation rules they are already using to process the usual search > > box. > > > > From the point of view of a wallet developer (or other tool that need to > > show any kind of Blockchain references), using this scheme mean that he > > can simply make it a blockchain: link and be done with it, without having > > to worry about any specific Blockchain explorer or provide a means for the > > user to select one. > > > > Blockchain explorers in turn will simply offer to handle the blockchain: > > URI, the first time the user visit their website, or launch/install the > > application, or even set themselves if there isn't already one. > > > > Users get the convenience of using always their preferred explorer, which > > can be especially handy on mobile devices, where juggling with cut&paste > > is far from ideal. _______________________________________________ bitcoin-dev mailing list bitcoin-dev@lists.linuxfoundation.org https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
