[Bitcoin-development] Accepting broken QRcodes

[Bitcoin-development] Accepting broken QRcodes

[Mike Hearn]

Hi bitcoin-development,

blockchain.info generates non-BIP-compliant URIs in its QRcodes, as
does its iPhone app. They are of the form bitcoin://address not
bitcoin:address.

I asked Ben to fix this (social networks don't parse QRcodes after
all), but after explaining that social networks don't parse URLs
without :// in them, he stopped responding to my emails. So I've gone
ahead and added support for reading these types of URLs to bitcoinj,
in the interests of "just works" interoperability.

This mail is just a heads up in case anyone else wants to do the same
thing. Hopefully at some point, Ben will stop generating such QRcodes
and we can remove these hacks and get back to BIP compliance.

Re: [Bitcoin-development] Accepting broken QRcodes

[Andreas Schildbach]

> I asked Ben to fix this (social networks don't parse QRcodes after
> all), but after explaining that social networks don't parse URLs
> without :// in them, he stopped responding to my emails. So I've gone
> ahead and added support for reading these types of URLs to bitcoinj,
> in the interests of "just works" interoperability.
> 
> This mail is just a heads up in case anyone else wants to do the same
> thing. Hopefully at some point, Ben will stop generating such QRcodes
> and we can remove these hacks and get back to BIP compliance.

The problem with this "accept everything even if broken" approach is
that people will probably never fix the broken stuff. So we likely end
up with a fragmented de-facto standard.

That does not mean I am totally against accepting broken URLs, but there
should be at least a promise that they will be fixed at the source.

Re: [Bitcoin-development] Accepting broken QRcodes

[Gary Rowe]

[-- Attachment #1: Type: text/plain, Size: 1735 bytes --]

Is it worth having a few more people email Ben to ask him politely to fall
into line with the BIP? No point encouraging broken windows by not speaking
out.

On 16 July 2012 09:16, Andreas Schildbach <andreas@schildbach.de> wrote:

> > I asked Ben to fix this (social networks don't parse QRcodes after
> > all), but after explaining that social networks don't parse URLs
> > without :// in them, he stopped responding to my emails. So I've gone
> > ahead and added support for reading these types of URLs to bitcoinj,
> > in the interests of "just works" interoperability.
> >
> > This mail is just a heads up in case anyone else wants to do the same
> > thing. Hopefully at some point, Ben will stop generating such QRcodes
> > and we can remove these hacks and get back to BIP compliance.
>
> The problem with this "accept everything even if broken" approach is
> that people will probably never fix the broken stuff. So we likely end
> up with a fragmented de-facto standard.
>
> That does not mean I am totally against accepting broken URLs, but there
> should be at least a promise that they will be fixed at the source.
>
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Bitcoin-development mailing list
> Bitcoin-development@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/bitcoin-development
>

[-- Attachment #2: Type: text/html, Size: 2467 bytes --]

Re: [Bitcoin-development] Accepting broken QRcodes

[Wladimir]

[-- Attachment #1: Type: text/plain, Size: 2574 bytes --]

But is he the only one using the broken URLs? It was my impression that
they were widespread already.

Wladimir

On Mon, Jul 16, 2012 at 10:52 AM, Gary Rowe <g.rowe@froot.co.uk> wrote:

> Is it worth having a few more people email Ben to ask him politely to fall
> into line with the BIP? No point encouraging broken windows by not speaking
> out.
>
>
> On 16 July 2012 09:16, Andreas Schildbach <andreas@schildbach.de> wrote:
>
>> > I asked Ben to fix this (social networks don't parse QRcodes after
>> > all), but after explaining that social networks don't parse URLs
>> > without :// in them, he stopped responding to my emails. So I've gone
>> > ahead and added support for reading these types of URLs to bitcoinj,
>> > in the interests of "just works" interoperability.
>> >
>> > This mail is just a heads up in case anyone else wants to do the same
>> > thing. Hopefully at some point, Ben will stop generating such QRcodes
>> > and we can remove these hacks and get back to BIP compliance.
>>
>> The problem with this "accept everything even if broken" approach is
>> that people will probably never fix the broken stuff. So we likely end
>> up with a fragmented de-facto standard.
>>
>> That does not mean I am totally against accepting broken URLs, but there
>> should be at least a promise that they will be fixed at the source.
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Live Security Virtual Conference
>> Exclusive live event will cover all the ways today's security and
>> threat landscape has changed and how IT managers can respond. Discussions
>> will include endpoint security, mobile security and the latest in malware
>> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
>> _______________________________________________
>> Bitcoin-development mailing list
>> Bitcoin-development@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/bitcoin-development
>>
>
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Bitcoin-development mailing list
> Bitcoin-development@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/bitcoin-development
>
>

[-- Attachment #2: Type: text/html, Size: 3840 bytes --]

Re: [Bitcoin-development] Accepting broken QRcodes

[Gary Rowe]

[-- Attachment #1: Type: text/plain, Size: 3396 bytes --]

I'm sure that there are many but my Google Search-Fu is not strong enough
to build a query to identify how widespread they are.

Maybe once we have sufficient evidence to support the suspicion we should
post to the main developer forum asking for a cleanup. After all, a Bitcoin
URI starting bitcoin://<address> doesn't actually make much sense because
there is no hierarchy in Bitcoin - it's flat with only an address being a
mandatory element.

I don't want to be all anal about this, but looking at RFC 3986 #10 (
http://tools.ietf.org/html/rfc3986#page-10) it's pretty clear that
introducing a false hierarchy is breaking the specification since it
presumes the existence of a relative URI.

On 16 July 2012 10:02, Wladimir <laanwj@gmail.com> wrote:

> But is he the only one using the broken URLs? It was my impression that
> they were widespread already.
>
> Wladimir
>
>
> On Mon, Jul 16, 2012 at 10:52 AM, Gary Rowe <g.rowe@froot.co.uk> wrote:
>
>> Is it worth having a few more people email Ben to ask him politely to
>> fall into line with the BIP? No point encouraging broken windows by not
>> speaking out.
>>
>>
>> On 16 July 2012 09:16, Andreas Schildbach <andreas@schildbach.de> wrote:
>>
>>> > I asked Ben to fix this (social networks don't parse QRcodes after
>>> > all), but after explaining that social networks don't parse URLs
>>> > without :// in them, he stopped responding to my emails. So I've gone
>>> > ahead and added support for reading these types of URLs to bitcoinj,
>>> > in the interests of "just works" interoperability.
>>> >
>>> > This mail is just a heads up in case anyone else wants to do the same
>>> > thing. Hopefully at some point, Ben will stop generating such QRcodes
>>> > and we can remove these hacks and get back to BIP compliance.
>>>
>>> The problem with this "accept everything even if broken" approach is
>>> that people will probably never fix the broken stuff. So we likely end
>>> up with a fragmented de-facto standard.
>>>
>>> That does not mean I am totally against accepting broken URLs, but there
>>> should be at least a promise that they will be fixed at the source.
>>>
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Live Security Virtual Conference
>>> Exclusive live event will cover all the ways today's security and
>>> threat landscape has changed and how IT managers can respond. Discussions
>>> will include endpoint security, mobile security and the latest in malware
>>> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
>>> _______________________________________________
>>> Bitcoin-development mailing list
>>> Bitcoin-development@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/bitcoin-development
>>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Live Security Virtual Conference
>> Exclusive live event will cover all the ways today's security and
>> threat landscape has changed and how IT managers can respond. Discussions
>> will include endpoint security, mobile security and the latest in malware
>> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
>> _______________________________________________
>> Bitcoin-development mailing list
>> Bitcoin-development@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/bitcoin-development
>>
>>
>

[-- Attachment #2: Type: text/html, Size: 5119 bytes --]