Hi Steen, Much truth in what you say below. And for me, the best of all worlds is to use ASN.1 definitions for the schema underlying the encodings required for specifications. As you note, ASN.1 is a mature, reliable schema and unlike TRex, RELAX, XSD, etc. does not restrict me to a single verbose encoding.
ASN.1 allows me to rigorously define one set of abstract values from which I can apply BER, PER or XML Encoding Rules (XER) as needed. For volume burdened or resource challenged environments, I can use more than one set of rules, sending compact PER over the wire and converting the transferred values to XER for local use. XER allows you to tap into the best of "XML", XML 1.0 markup, and avoid the flux and churn of the many and every growing complex "arms and legs" in the XML space. This allows you to leverage the browser when needed. And the new dotted format for object identifier values, <OID> 1.2.3.4 </OID>, allows you to uniquely identify objects such as common security algorithms and types, while avoiding the file path issues you describe below. Phil Steen Oluf Karlsen wrote: > > Hello Salim > > Just a quick remark on your question - > > going from XML to ASN.1 and vice versa > > After studying the effects of going from a message set > specified in the header file format of ANSI C > to an ASN.1 specified set versus one specified in XML > I made the personal conclusion that > > ASN.1 is a mature and consistent language to specify > message syntaxes. > > XML-schema is still underway. While the actual instances > of messages in XML are straightforward their formal > specification is not. Efforts have been made to unify > the various xml-schema variants - but the results are > not promising - and worse - the resulting protocol specs > are bulky, badly structured and not logic. Particularly I > fear the effects of having file paths in the xml-schema > headers to inherited XML defnitions. Mingling with file paths > and trying to put them in a standard format is - to me - > deemed to create trouble. > > My personal conclusion is that I would prefer (given an > open standard is asked for with origins that speak to > internet technology lovers ) messages specified in XML > (given that bandwidth is not an issue) - but I would > suggest the message protocol to be specified in ASN.1. > > Maybe a subset of ASN.1. All the "depreciated" parts of the > language makes it difficult for newcomers to master the > abstract language fast. > > Comments wellcome ! > > Steen Oluf Karlsen > S�holtvej 6 > Vester Vandet > DK-7700 Thisted > Danmark > Tel +45 97 97 72 72 > email [EMAIL PROTECTED] > > -----Oprindelig meddelelse----- > Fra: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]P� vegne af salai > sivamal.G > Sendt: 2. januar 2002 10:34 > Til: [EMAIL PROTECTED] > Emne: Re: [ASN.1] xml > > try www.cryptix.org > > if donot get let me know. > > -Salai > www.icope.com > Bangalore. India. > > ----- Original Message ----- > From: "Salim Mounir AlAoui" <[EMAIL PROTECTED]> > To: <[EMAIL PROTECTED]> > Sent: Friday, December 21, 2001 10:43 PM > Subject: [ASN.1] xml > > > > > hello, > > > > does anyone have some pointers for me, i am trying to move from xml to > > asn.1 and vice versa, is there a compiler that does that on the market or > > as a freeware? > > > > thanks in advance. > > > > > > > > > > > > > > > > > > > > -------------------------------------------------------------------------- > > Salim Mounir Alaoui [EMAIL PROTECTED] > > Computer Science Dept. [EMAIL PROTECTED] > > Research Assistant. [EMAIL PROTECTED] > > Florida Institute of Technology > > Melbourne, Florida > > Voice: (407) 537-8025. > > -------------------------------------------------------------------------- > > > > > >
