Re: AquaTerm (all solved)

--===============8182547607173195806==
Content-Type: multipart/alternative; boundary="000000000000b71a24062ca3138b"

--000000000000b71a24062ca3138b
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hi Ethan,

The new configure script works correctly (at least on my ARM-based Mac
running masOS 15.2 Sequioa); thank you.

As for "-g": As long as it's a deliberate choice, I have no problem with
it. (I thought this flag could interfere with optimization, but I realize
now that this is not true; thus, the only difference is the size of the
executable.)

Erik

On Sun, Jan 26, 2025 at 3:03=E2=80=AFPM Ethan A Merritt <[email protected]> wr=
ote:

> On Sunday, 26 January 2025 08:59:41 PST Erik Luijten wrote:
> > Dear all,
> >
> > I apologize for the clutter on this mailing list, but to save others
> work,
> > I wanted to let you know that all AquaTerm issues listed in my email fr=
om
> > yesterday have been resolved:
> >
> > 1. Thanks to the MacPorts project, I figured out what patches are neede=
d
> to
> > compile AquaTerm 1.1.1 on a current version of macOS.
> >
> > 2. Moreover, Ethan, I found out what needs to be changed in the Gnuplot
> > configure script. Line 10668 needs to read:
> >
> > CFLAGS=3D"$CFLAGS -ObjC"; LDFLAGS=3D"$LDFLAGS -framework Foundation
> -framework
> > AquaTerm -F/Library/Frameworks"
> >
> > (the missing -F flag causes subsequent tests to go wrong)
>
> This is getting deep into the autoconf/packaging weeds.
>
> <begin weedy section>
> The ./configure file is not a primary source;  it is produced by running
> various
> autoconf tools  (aclocal autoheader automake m4 ...) using the master
> template
> file configure.ac and the scripts in the subdirectory .../m4/
> If you are starting from the files in the git repository (i.e. not from a
> packaged tarball) this is done by running the script named "prepare"
> which does all this magic and produces the ./configure script.
>
> I didn't write that prepare script and have only a superficial
> understanding of
> what all it is doing.  I can see that there is script in the m4 directory
> .../m4/apple.m4 that is supposed to set the appropriate flags for buildin=
g
> on a Mac.
> <end weedy section>
>
> I can tweak it, but I'll be working blind since I don't have a Mac to tes=
t
> it on.
> Please try the attached configure script (for 6.0.2) generated after
> modifying
> apple.m4
>
> >
> > 3. I need to create an ARM package for AquaTerm and then I will provide
> new
> > packages on my Wiki page. Please allow a few days for me to finish this=
.
> >
> > Lastly, on a side note: I noticed that gnuplot on my machine always
> builds
> > with "-g -O2"; is this default on purpose? ("-g" seems undesirable for
> > production builds)
>
> Why undesirable?
> The advantage is that it allows people to provide more information in a
> bug report.
> As in "here's a backtrace I got for the segfault I'm reporting".
> Is there a disadvantage?
>
>     Ethan

--000000000000b71a24062ca3138b
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hi Ethan,<div><br></div><div>The new configure script work=
s correctly (at least on my ARM-based Mac running masOS 15.2 Sequioa); than=
k you.=C2=A0<br><br>As for &quot;-g&quot;: As long as it&#39;s a deliberate=
 choice, I have no problem with it. (I thought this flag could interfere wi=
th optimization, but I realize now that this is not true; thus, the only di=
fference is the size of the executable.)<br><br>Erik</div></div><br><div cl=
ass=3D"gmail_quote gmail_quote_container"><div dir=3D"ltr" class=3D"gmail_a=
ttr">On Sun, Jan 26, 2025 at 3:03=E2=80=AFPM Ethan A Merritt &lt;<a href=3D=
"mailto:[email protected]">[email protected]</a>&gt; wrote:<br></div><blockquote =
class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px sol=
id rgb(204,204,204);padding-left:1ex">On Sunday, 26 January 2025 08:59:41 P=
ST Erik Luijten wrote:<br>
&gt; Dear all,<br>
&gt; <br>
&gt; I apologize for the clutter on this mailing list, but to save others w=
ork,<br>
&gt; I wanted to let you know that all AquaTerm issues listed in my email f=
rom<br>
&gt; yesterday have been resolved:<br>
&gt; <br>
&gt; 1. Thanks to the MacPorts project, I figured out what patches are need=
ed to<br>
&gt; compile AquaTerm 1.1.1 on a current version of macOS.<br>
&gt; <br>
&gt; 2. Moreover, Ethan, I found out what needs to be changed in the Gnuplo=
t<br>
&gt; configure script. Line 10668 needs to read:<br>
&gt; <br>
&gt; CFLAGS=3D&quot;$CFLAGS -ObjC&quot;; LDFLAGS=3D&quot;$LDFLAGS -framewor=
k Foundation -framework<br>
&gt; AquaTerm -F/Library/Frameworks&quot;<br>
&gt; <br>
&gt; (the missing -F flag causes subsequent tests to go wrong)<br>
<br>
This is getting deep into the autoconf/packaging weeds.<br>
<br>
&lt;begin weedy section&gt;<br>
The ./configure file is not a primary source;=C2=A0 it is produced by runni=
ng various<br>
autoconf tools=C2=A0 (aclocal autoheader automake m4 ...) using the master =
template<br>
file <a href=3D"http://configure.ac" rel=3D"noreferrer" target=3D"_blank">c=
onfigure.ac</a> and the scripts in the subdirectory .../m4/<br>
If you are starting from the files in the git repository (i.e. not from a<b=
r>
packaged tarball) this is done by running the script named &quot;prepare&qu=
ot;<br>
which does all this magic and produces the ./configure script.<br>
<br>
I didn&#39;t write that prepare script and have only a superficial understa=
nding of<br>
what all it is doing.=C2=A0 I can see that there is script in the m4 direct=
ory<br>
.../m4/apple.m4 that is supposed to set the appropriate flags for building =
on a Mac.<br>
&lt;end weedy section&gt;<br>
<br>
I can tweak it, but I&#39;ll be working blind since I don&#39;t have a Mac =
to test it on.<br>
Please try the attached configure script (for 6.0.2) generated after modify=
ing<br>
apple.m4<br>
<br>
&gt; <br>
&gt; 3. I need to create an ARM package for AquaTerm and then I will provid=
e new<br>
&gt; packages on my Wiki page. Please allow a few days for me to finish thi=
s.<br>
&gt; <br>
&gt; Lastly, on a side note: I noticed that gnuplot on my machine always bu=
ilds<br>
&gt; with &quot;-g -O2&quot;; is this default on purpose? (&quot;-g&quot; s=
eems undesirable for<br>
&gt; production builds)<br>
<br>
Why undesirable?<br>
The advantage is that it allows people to provide more information in a bug=
 report.<br>
As in &quot;here&#39;s a backtrace I got for the segfault I&#39;m reporting=
&quot;.<br>
Is there a disadvantage?<br>
<br>
=C2=A0 =C2=A0 Ethan</blockquote></div>

--000000000000b71a24062ca3138b--


--===============8182547607173195806==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline


--===============8182547607173195806==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
gnuplot-beta mailing list
[email protected]
Membership management via: https://lists.sourceforge.net/lists/listinfo/gnuplot-beta

--===============8182547607173195806==--