Missing character

[Wolfgang Engelmann]

Am 01.09.22 um 16:03 schrieb Dr Eberhard Lisse:
> C.O.M.P.I.I.L.E the TeX file with lualatex and note where it stops (on
> or near the error line ).
> 
> Also what does googling for the error message show you?
> 
> Or just for "U+0098"?
> 
> 
> el
> 
> 
> On 31/08/2022 17:01, Wolfgang Engelmann wrote:
>>
>>
>> Am 31.08.22 um 15:20 schrieb Dr Eberhard Lisse:
>>> compile the file with lualatex and not the line the error is reported
>>> on.
>>>
>>> el
> [...]
>> Thanks, El,
>>
>> neither checking the lyx file nor the luatex exported file with bless
>> and decimal character found the 0098
>>
>> Wolfgang

El,
I do get Log-outputs of the luatex run such as that one (between #### 
and ####:
####
[77<./export_CircardianRyth-Mensch-Sig_Puls-65p0-5p0-___a0b8ea4063538c2239436b1d
41794c6fb968832f6bd.pdf>] 
[78<./export_VanPol_Puls-SB-PRC_sb_V4_Dia_pdf_0ffff72_
__c7e7c78a8954a038bd2ba6f8256318f8302a367cea5.pdf>] [79
Missing character: There is no ˜ (U+0098) in font 
[LibertinusSerif-Italic.otf]:
mode=node;script=latn;language=dflt;+tlig;+tnum;+lnum;!
.
<argument> ...not:N \tex_shipout:D \box_use:N \l_shipout_box
 
\__shipout_drop_firstpage_...
l.2612
####
and going to the line (see >> below) where the U+0098 character should be

A smaller nonlinearity (with a VP oscillator $\epsilon=0.2$) the
maximal amplitude is reached earlier after the pulse. The transient
time is also smaller (see Fig. \ref{fig:Phasenverschiebung nach Puls-1}).}
\end{figure}

 >>this is line 2612 >> \label{scilab script-1}To avoid changing the 
parameters for each
pulse time and the following simulations sequentially, one can use
a Scilab script, which carries this out (see the algorithm 
\ref{alg:Skript-zur PRC-Erzeugung}).

As to your Google proposal:
The Internet tells me, e.g.:
'START OF STRING' (U+0098)
and:
Character U+0098 (152) is an invisible control code that is almost never 
used. The trick is that "ISO-8859-1" has a different meaning to Java and 
web browsers. For Java it really is the ISO-8859-1 standard, which maps 
exactly to the first 256 code points of Unicode. That includes a range 
of little-used C1 control characters at 128-159.

What I tried was to put the surrounding of the error (was a figure + 
legend) in a lyx note, run it again and check the tex file again. The 
0098 complain is now at another place, which had not been given before 
as an error.

Paul recommended to run a bisection search, but its the same problem as 
described above.

I was hoping to localize the 0099 character by using the editor 'best', 
which allows to search for decimal codes, but that did not show it at 
all. Perhaps somebody knows a better editor to check for those characters.

FAZIT:
I gave up, since it is a warning only and it takes me too much time (and 
time of those trying to help me)
Thanks anyway!

Wolfgang