Welcome, Guest
Username: Password: Remember me
  • Page:
  • 1

TOPIC: Comments to Annex E

Comments to Annex E 20 Sep 2001 09:19 #5945

Hi all,

i have some problems with Annex E.1 (ASN.1), that I think need to be resolved/clarified.

1. The note in subclause E.1.2 is unclear.
a) Are subtypes accessible or not? Implicitly they are but refining the note could avoid possible misunderstanding.

b) Why value sets are not accessible? I can not see (at the moment) a justification for it as TTCN-3 also supports list of values. Note that subtypes and value sets are semantically equal, differs syntactically only. So if subtypes are accessible (see item a), value sets also should be.
If defined in ASN.1:
SampleMessage ::= SEQUENCE {
field1 Field1,
<further components>
}
Field1 ::= INTEGER (0..15)
DefinedValuesForField1 Field1 ::= {0 | 1}

Why this definition is forbidden?:
template SampleMessage myMessage :=
{
field1 := DefinedValuesForField1,
<further components>
}

2. Subclause E.1.3:
This subclause is too restrictive. Implicitly it does not allow to use parameterized ASN.1 objects in TTCN message definitions. Consequently templates can be based on either fully ASN.1-defined or fully TTCN-defined definitions. Personally I think that TTCN-3 can not compete with ASN.1 but should provide full support instead. Taking into account the significance and widespread of ASN.1, such severe restrictions may become a drawback for TTCN-3.

Also not clear the general classification of value sets as uniquely ASN.1 objects as TTCN-3 also supports list of values at list for strings and integers.

3. Subclause E.1.4
The purpose of the subclause is unclear. It contains no requirement and claiming that ASN.1 messages are only SEQUENCEs or SETs is not true: e.g. in NBAP, RANAP and many other protocols it is a CHOICE but there is no restriction in ASN.1 on this respect (e.g. in principle why a protocol could not have a single message which is an EMBEDDED-PDV or just an INTEGER?).

4. "UnalignedPER-97" and "AlignedPER-97" are attributes for basic or canonical PER encoding?

Best Regards, György

============================================
dr. György RÉTHY
Ericsson Communications Systems Hungary Lim.
Conformance Center
tel.: +36 1 437-7006; fax: +36 1 437-7767
mobile: +36 30 297-7862
e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
web: www.r.eth.ericsson.se/~ethgry
============================================
The administrator has disabled public write access.
  • Page:
  • 1

FacebookTwitterGoogle BookmarksRedditNewsvineTechnoratiLinkedin