[pdftex] pdftex compliance..

[Tom Kacvinsky]

Hi,

First of all, I am not speaking on behalf of the AMS or NSF.

I have had some experience with FastLane.  For others on this list, I first give
a general rundown of the FastLane submission process:

the PI (principal investigator) or someone working on behalf of the PI uploads
the myriad sections of the proposal.  This includes the biographies of the PI
and co-PIs (co-principal investigators), the budget, the budget justification,
letters of support (some institutes will give matching funds if the PI receives
an NSF grant, these letters state that intent), the actual proposal (what will be
studied, etc...), and so on.

The PI can upload as many times as he/she needs to get the proposal uploaded
properly.  What this usually entails is getting the files uploaded and then
checking to see if the *entire* proposal prints properly.  More on this below.
Also, this is the part Gary mentioned specifially in his message:

    In most instances, pdflatex is useable with FastLane but if you are
    using it you should be very careful to test the printing of your entire
    proposal as it can cause concatenation errors when we attempt to put
    the entire proposal together

The PI has the institute he/she works for actually submit the grant.  This is
just a finalization process.  After this, the PI cannot correct their proposal.
Someone at the NSF has to take care of the problems.

The review process starts by the entire proposal being batch printed for
dissemination to proposal reviewers.  The batch printing starts by concatenaton
of the various proposal sections.  This concatenation also occurs when the PI
sees if he/she can print the entire proposal.

This is where the problems start.  Depending on who you talk to, the problem is
with NSF's concatenation process, or the problem is with the section PDF files
themselves, or the program that made the section PDF files.  In any case, the
concatenation process fails for some PDF files.  The last time I checked, the
reason why pdftex generated PDF files fail to concatenate is because of
different font subsets (in terms of glyphs complements, encodings, etc...) in
different files have the same subset name (for instance, AABBCC+CMR10).  I
currently do not know what the NSF is using to concatenate PDF files (I did at
one time), so I cannot speak as to what the problems are.  In the past, the
problems had to do with using Acrobat 3.0 based technology to concatenate the
PDF files (and to be fair, some of the problems also exist in Acrobat 4.0 and
later technology, and some PDF files are really wacked out).

Anyway, the problems with font subset names was fixed by Thanh over a year
ago, so I suspect you have an out of date version of pdftex.

If you get the latest/greatest version of pdftex, you shouldn't have these
problems.

Tom

On Sat, 4 Aug 2001, K. R. Subramanian wrote:

> This is with regard to doing NSF proposals using pdflatex. Is the following
> from NSF Fastlane Help Desk
> true? I would have thought that using conversion tools to go from dvi-ps-pdf
> would be a little
> less reliable than doing pdflatex..
>
> pdftex Gurus, please respond..
>
>     -- krs
>
> "Walker, Gary" wrote:
>
> > Dr. Subramanian,
> >
> > We are quite well aware of pdflatex.  It is still very much in development,
> > however, and in order to insure that the pdfs uploaded to FastLane are fully
> > compliant with the Adobe PDF specification, we cannot, at this time,
> > recommend its use.  In most instances, pdflatex is useable with FastLane but
> > if you are using it you should be very careful to test the printing of your
> > entire proposal as it can cause concatenation errors when we attempt to put
> > the entire proposal together.  Thank you.
> >
> > Sincerely,
> >
> > Gary E. Walker
> > FastLane Help Desk
> > (800) 673-6188
> > fastlane at nsf.gov
> >
> > > -----Original Message-----
> > > From: K. R. Subramanian [SMTP:krs at cs.uncc.edu]
> > > Sent: Friday, August 03, 2001 3:14 PM
> > > To:   fastlane at nsf.gov
> > > Subject:      update -- using Tex/Latex to produce pdf..
> > >
> > >
> > > These days you can use pdflatex to produce pdf directly from tex/latex
> > > documents - this is
> > > the easiest and most reliable ways to produce good pdf.
> > >
> > > I think website needs to be updated to include this. tex-dvi-ps-pdf is
> > > perhaps more errorprone. Plus,
> > > pdflatex also will now support including image files just as Jpeg..
> > >
> > >     -- krs
> > >
> > >
> > > --
> > > Dr. K.R.Subramanian                           Phone: (301) 402-0042
> > > National Institutes of Health                 FAX:   (301) 402-4080
> > > National Library of Medicine                  Email: krs at cs.uncc.edu
> > > Bldg 38A, Rm. B1N30D                          WWW: www.cs.uncc.edu/~krs
> > > 8600 Rockville Pike, Bethesda, MD 20894
> > >
> > >
>
> --
> Dr. K.R.Subramanian                           Phone: (301) 402-0042
> National Institutes of Health                 FAX:   (301) 402-4080
> National Library of Medicine                  Email: krs at cs.uncc.edu
> Bldg 38A, Rm. B1N30D                          WWW: www.cs.uncc.edu/~krs
> 8600 Rockville Pike, Bethesda, MD 20894
>
>
>
> _______________________________________________
> pdftex mailing list
> pdftex at tug.org
> http://tug.org/mailman/listinfo/pdftex
>