Imported from previous forum
[ original email was from John McGuire - jmcguire@pulsion.co.uk ]
I am relatively new to the FIX protocol but have looked over the 4.1 Protocol Specification. Is there a way of extending the protocol? For example if I want to add some other information to a quote message are there any spare fields I could use? Obviously I would have to agree the information held if the spare field with the 3rd party I would be connecting to.
I can understand that you need to define specific fields in a message but I thought there might be a spare text field in each message for customising the message.
Does anyone know if this is possible ?
Thanks
John
[ original email was from Ryan Pierce - rpierce@taltrade.com ]
> I am relatively new to the FIX protocol but have looked over the 4.1 Protocol Specification. Is there a way of extending the protocol? For example if I want to add some other information to a quote message are there any spare fields I could use? Obviously I would have to agree the information held if the spare field with the 3rd party I would be connecting to.
>
> I can understand that you need to define specific fields in a message but I thought there might be a spare text field in each message for customising the message.
>
> Does anyone know if this is possible ?
Field tags from 5000 to 9999 are reserved for this purpose. It makes sense to first check the FIX web site, which maintains a registry of such proprietary field tags, prior to creating any, and to register tags you are using. This helps prevent collision between one firm’s proprietary tags and another.
…
> > Does anyone know if this is possible ?
>
> Field tags from 5000 to 9999 are reserved for this purpose. It makes sense to first check the FIX web site, which maintains a registry of such proprietary field tags, prior to creating any, and to register tags you are using. This helps prevent collision between one firm’s proprietary tags and another.
>
>
This brings me to this question :
Should a fix parser be able to handle user-defined
fields within a fix message without knowing about
the data on the Fix web site (about proprietary field tags)?
The problem occurs when trying to parse a fix message that has data fields. For the standard fields this is no problem since the data field must be preceded by length field. If one is dealing with user-defined fields (for which one does not know the field types) then parsing is not
possible.
Any suggestions ?
> …
> > > Does anyone know if this is possible ?
> >
> > Field tags from 5000 to 9999 are reserved for this purpose. It makes sense to first check the FIX web site, which maintains a registry of such proprietary field tags, prior to creating any, and to register tags you are using. This helps prevent collision between one firm’s proprietary tags and another.
> >
> >
>
>
> This brings me to this question :
> Should a fix parser be able to handle user-defined
> fields within a fix message without knowing about
> the data on the Fix web site (about proprietary field tags)?
>
> The problem occurs when trying to parse a fix message that has data fields. For the standard fields this is no problem since the data field must be preceded by length field. If one is dealing with user-defined fields (for which one does not know the field types) then parsing is not
> possible.
>
> Any suggestions ?
>
I do not think one should be expected to support any user defined fields without prior bi-lateral agreement with a counterparty who needs to provide them.