Terry,
Thanks for your review and comments.
1. E1.2.1.3. New Fields, GlobalParametersIFD
We do not believe there is much ambiguity as to which fields a reader needs
to look for within the GlobalParametersIFD.
First, the GlobalParametersIFD definition further restricts the fields by
stating that the legal set is listed within TIFF-FX or a TIFF-FX extension.
"Fields in the GlobalParametersIFD MAY be TIFF fields that are legal
in any IFD. Only a limited number of fields are designated as
belonging inside the GPIFD, and they are specifically mentioned by
RFC XXXX or an extension when they do."
Second, the "SHOULD" recommendation is considered very strong with potential
interoperability consequences if ignored. It does, however, leave the door
open for other actions should a compelling situation arise. The SharedData
field is one example of a field that is required for proper interpretation
of a data stream, yet it is located within the GlobalParametersIFD. If the
TIFF-FX GlobalParametersIFD constraint were MUST rather than SHOULD, the
TIFF-FX extension would not have been able to include SharedData.
2. TIFF-FX Section 8.2.1 NewSubFileType fields
The NewSubFileType field definition appears to be clear enough, as each
SubIFD is NOT a page, so Bit 1 doesn't make sense on a SubIFD. A reader
could ignore that bit if it appeared on a SubIFD, without changing the
meaning of anything. As you have pointed out, the Section 8.4 rules do
clarify this further.
3. E1.4.5 Rules and Requirements for Images
We agree with your clarification recommendations. The 5th sentence of Rule 8
has being change to read:
"When the image or coding parameters change between stripes in the
foreground/Background/secondary mask layers these stripes SHALL
be stored in separate single-stripped IFDs."
4. E1.4.1 typos
Done.
Regards,
Lloyd
-----Original Message-----
From: Terry Stewart [mailto:terryandlori@earthlink.net]
Sent: Monday, July 30, 2001 1:00 AM
To: ietf-fax@imc.org
Subject: Suggestions for draft-ietf-fax-tiff-fx-Extension-1-02.txt
Hi All,
Below are suggestions for draft-ietf-fax-tiff-fx-Extension-1-02.txt.
Terry Stewart
============================================================================
====
E1.2.1.3. New Fields, GlobalParametersIFD Field:
The current wording regarding the GlobalParametersIFD contains the sentence:
"Required baseline fields should not be located in the
GlobalParametersIFD, but should be in each image IFD."
The "should" in this sentence appears to be a suggestion only and a TIFF
writer could place required baseline fields in the GlobalParametersIFD
and still be considered a valid TIFF-FX file. This means that TIFF-FX
readers would need to handle the cases where TIFF-FX writers decided to
ignore the suggestion and place required fields into the
GlobalParametersIFD.
It is ! my! understanding that the GlobalParametersIFD should never contain
fields that would directly affect the rendered image (e.g. Photometric
Interpretation, ImageWidth). Therefore, it would be better to enforce
this by declaring that the GlobalParametersIFD must not contain required
baseline fields.
"Required baseline fields MUST NOT be located in the
GlobalParametersIFD, but should be in each image IFD."
This would remove ambiguity about the intended use of the
GlobalParametersIFD
and reduce the complexity of TIFF readers. Otherwise a tolerant TIFF reader
would have to treat the GlobalParametersIFD as a type defaulting mechanism
for
required baseline TIFF fields since the GlobalParametersIFD would need to
be checked when IFDs are missing required baseline fields.
============================================================================
====
The definit! io! n for NewSubFileType in Section 8.2.1 is:
NewSubFileType(254) = 16, 18. LONG
For Profile M, the NewSubFileType field has two bits that are
required. Bit 1 indicates a single page of a multi-page document
and must be set for the Primary IFD; Bit 4 indicates the MRC imaging
model as described in ITU-T Recommendation T.44 [T.44], and must be
set for Primary IFDs and all MRC-specific child IFDs.
This definition allows Bit 1 to be on for MRC-specific child IFDs. I believe
the description in Section 8.4 "Rules and Requirements for Images" cla! ri!
fies
that the child IFDs must not set bit 1 on, but it would be clearer if
the definition of NewSubFileType mentioned this as well:
NewSubFileType(254) = 16, 18. LONG
For Profile M, the NewSubFileType field has two bits that are
required. Bit 1 indicates a single page of a multi-page document
and must be set for the Primary IFD and MUST NOT be set for
MRC-specific child IFDs; Bit 4 indicates the MRC imaging
model as described in ITU-T Recommendation T.44 [T.44], and must be
set for Primary IFDs and all MRC-specific chil! d ! IFDs.
============================================================================
====
E1.4.5. Rules and Requirements for Images
The current description of multi-stripped IFDs for non-Primary IFDs is
currently described as follows:
8. In MRC Internet Fax, each layer is transmitted as a sequence of
:
single Primary IFD as a series of corresponding strips. Without
constraint on coding parameter changes, all stripes of the
foreground/Background/secondary mask layers SHALL be stored in
separate single-stripped IFDs. These IFDs are referenced by the
:
10. When there are no image and coding parameter changes be! tw! een
stripes, a single multi-stripped IFD (i.e. each strip
corresponding to a stripe) MAY be used to represent a multi-
striped layer that! is not a Primary Mask layer.
The phrase "Without constraint on coding parameters changes" in item 8
wasn't very clear to me. Recommend clarifying this area of 8 with the
following:
8. In MRC Internet Fax, each layer is transmitted as a sequence of
:
single Primary IFD as a series of corresponding strips. When
the image or coding parameters change between stripes in the
foreground/Background/secondary mask layers these stripes SHALL
be stored in separate single-stripped IFDs. These IFDs are
! &n! bsp; referenced by the ...
============================================================================
====
Typos:
E1.4.1 Introduction, last paragraph:
"in-dept" should be "in-depth"
============================================================================
====
This archive was generated by hypermail 2b29 : Tue Jul 31 2001 - 17:19:42 EDT