Re: alpha dot

[Scott Kostyshak]

On Wed, Feb 15, 2017 at 01:18:41PM +0100, Guillaume Munch wrote:

> I have a series of patches (attached) that make the decoration thickness
> increase with zoom, to improve the appearance on high-dpi. The
> decorations do not look worse than without the added thickness. I have
> not committed it yet because I have not found a way to keep it simple,
> and there are still some issues. But you might like to try them.

The patch works well for me. It is fun to see how the decorations look
zoomed in.

> For \dot in particular it looks like a square rather than a line.

Agreed.

Scott


signature.asc
Description: PGP signature

Re: [patch] Theorem environment: set NextNoIndent to 0

[Scott Kostyshak]

On Wed, Feb 15, 2017 at 09:07:26AM +, Guenter Milde wrote:

> >> >> NextNoIndent was set to 1 at 46611ba1, but the commit message only
> >> >> states "Fixes to the theorem modules." so I'm not sure how else to
> >> >> check.
> 
> Some suggestions:
> 
> * Ask the developer of 46611ba1 for possible reasons of the change.

Richard, how good is your memory? :)

> * Change the document class of your examples and compare output and LyX-GUI

Why? I think the module should provide correct output for the default
(article class). (update: ah, now I see what you mean. Let's discuss
this below)

> * Can indentation/start of a new paragraph be influenced by the new 
>   "new paragraph" special? (the curly arrow popping up every now and then in
>   files converted with lyx2lyx)

I'm not sure.

> No. I am not against the change (just wanted to avoid a situation where
> we solve display in half of the cases but break it in the other half.)

I see what you meant above now. It seems we took different approaches.

My approach: make things work for the default document class.

Your approach: make things work for as many document classes as
possible.

Probably the ideal approach is to maximize a weighted average, where the
weights are how often the document classes are used. Unfortunately, I
don't know how to calculate the weights.

Whatever our decision is we might want to document it (in
Development.lyx or in Customization.lyx?)

> >> Please do commit. I find this very annoying and keep meaning to commit
> >> it myself.
> 
> > OK, I'll wait a couple of days to see if Günter has an objection. He
> > might see something that I missed.
> 
> Please go ahead.

I think the above discussion is important first.

Scott


signature.asc
Description: PGP signature

Re: [LyX/master] ctests: update inversion patterns.

[Scott Kostyshak]

On Wed, Feb 15, 2017 at 09:23:19AM +, Guenter Milde wrote:

> No, I did not even find out where to report...
> So, if you want to take aktion I would be glad.

No thanks :)

> Anyway, the Babel-bug is low priority, as using Xe/LuaTeX with 8-bit
> fonts is a rare cornercase. Also, the simple and obvious workaround is to
> set the input encoding to utf8.

Makes sense.

Scott


signature.asc
Description: PGP signature

Re: ctest fails standard exports of a number of examples and templates files

[Scott Kostyshak]

On Wed, Feb 15, 2017 at 01:26:09PM +, Guenter Milde wrote:

> IMO, it is simpler to add the additional nonstandard dependencies in the
> comments in "unreliableTests" only and reword the start of 4.3.1.4
>   
>   To run all tests, in the build directory simply run the command ctest.
>   A full, up-to-date TeXLive installation is recommended to run the
>   tests. Otherwise, some tests will fail. Tests with additional
>   requirements are labeled “unreliable:nonstandard”. 
>   
> to something like
> 
>   A full, up-to-date TeXLive installation is required to run the
>   tests. Otherwise, some tests will fail. 
>   Tests with the label “unreliable:nonstandard” have additional
>   requirements (see development/autotests/unreliableTests for details).
>   To exclude these tests, run `ctest -LE nonstandard`.
>   To run all tests, in the build directory simply run the command ctest.

That looks good.

> This would safe us some work updating requirements and prevent
> desynchronisation of the requirement lists.

Good point. I wonder if an outdated list would be better than no list.

Scott


signature.asc
Description: PGP signature

Re: lyx errors with lualatex and non-tex fonts

[Kornel Benko]

Am Mittwoch, 15. Februar 2017 um 18:19:55, schrieb Jürgen Spitzmüller 

> Am Mittwoch, den 15.02.2017, 17:43 +0100 schrieb Kornel Benko:
> > Here they fail both.
> 
> FWIW both XeTeX with and without TeX Fonts works without problems here.
> This is up to date TeXLive 2016.
> 
> *File List*

I have the same. But I found a reason:

Besides texlive from CTAN I also had installed font awesome as a ubuntu-package.
Somehow xelatex takes this font, while lualatex is OK with the one from ctan.
De-installing this package, lualatex still compiles the file while xelatex is 
complaining:

!
! fontspec error: "font-not-found"
! 
! The font "FontAwesome" cannot be found.
! 
! See the fontspec documentation for further information.
! 
! For immediate help type H .
!...  

Making symbolic link of `kpsewhich --var-value 
TEXMFSYSVAR`/fonts/conf/texlive-fontconfig.conf to ~/.fonts.conf
and rerunning 'fc-cache -fv' cured it for me.

Now it compiles.

> Jürgen

Kornel

signature.asc
Description: This is a digitally signed message part.

Re: lyx errors with lualatex and non-tex fonts

[Jürgen Spitzmüller]

Am Mittwoch, den 15.02.2017, 17:43 +0100 schrieb Kornel Benko:
> Here they fail both.

FWIW both XeTeX with and without TeX Fonts works without problems here.
This is up to date TeXLive 2016.

*File List*
moderncv.cls2015/07/28 v2.0.0 modern curriculum vitae and letter
document class
  size12.clo2014/09/29 v1.4h Standard LaTeX file (size option)
etoolbox.sty2017/01/02 v2.4 e-TeX tools for LaTeX (JAW)
  ifthen.sty2014/09/29 v1.1c Standard LaTeX ifthen package (DPC)
  xcolor.sty2016/05/11 v2.12 LaTeX color extensions (UK)
   color.cfg2016/01/02 v1.6 sample color configuration
   xetex.def2016/07/11 v4.10 LaTeX color/graphics driver for XeTeX 
(L3/RRM/JK)
colortbl.sty2012/02/13 v1.0a Color table columns (DPC)
   array.sty2016/10/06 v2.4d Tabular extension package (FMi)
 ifxetex.sty2010/09/12 v0.6 Provides ifxetex conditional
ifluatex.sty2016/05/16 v1.4 Provides the ifluatex switch (HO)
   t1enc.def2017/01/24 v2.0c Standard LaTeX file
   t1lmr.fd2009/10/30 v1.6 Font defs for Latin Modern
 lmodern.sty2009/10/30 v1.6 Latin Modern Fonts
 url.sty2013/09/16  ver 3.4  Verb mode for urls, etc.
graphicx.sty2014/10/28 v1.0g Enhanced LaTeX Graphics (DPC,SPQR)
  keyval.sty2014/10/28 v1.15 key=value parser (DPC)
graphics.sty2016/10/09 v1.0u Standard LaTeX Graphics (DPC,SPQR)
trig.sty2016/01/03 v1.10 sin cos tan (DPC)
graphics.cfg2016/06/04 v1.11 sample graphics configuration
fancyhdr.sty2016/09/06 3.8 Extensive control of page headers and 
footers
tweaklist.sty
calc.sty2014/10/28 v4.3 Infix arithmetic (KKT,FJ)
  xparse.sty2017/02/10 v6878 L3 Experimental document command 
parser
   expl3.sty2017/02/10 v6878 L3 programming layer (loader) 
expl3-code.tex2017/02/10 v6878 L3 programming layer 
l3xdvipdfmx.def
microtype.sty2016/05/14 v2.6a Micro-typographical refinements (RS)
microtype-xetex.def2016/05/14 v2.6a Definitions specific to xetex 
(RS)
microtype.cfg2016/05/14 v2.6a microtype main configuration file 
(RS)
moderncvcollection.sty2015/07/28 v2.0.0 moderncv collections
moderncvcompatibility.sty2015/07/28 v2.0.0 modern curriculum vitae 
and letter compatibility patches
fontspec.sty2017/02/12 v2.6 Font selection for XeLaTeX and LuaLaTeX
fontspec-xetex.sty2017/02/12 v2.6 Font selection for XeLaTeX and 
LuaLaTeX
   tuenc.def2017/01/24 v2.0c Standard LaTeX file
fontspec.cfg
geometry.sty2010/09/12 v5.6 Page Geometry
   ifpdf.sty2016/05/14 v3.1 Provides the ifpdf switch
  ifvtex.sty2016/05/16 v1.6 Detect VTeX and its facilities (HO)
moderncvstylecasual.sty2015/07/28 v2.0.0 modern curriculum vitae 
and letter style scheme: casual
moderncviconsawesome.sty2015/07/28 v2.0.0 modern curriculum vitae 
and letter icons: awesome
fontawesome.sty2016/05/15 v4.6.3.1 font awesome icons
fontawesomesymbols-generic.tex
fontawesomesymbols-xeluatex.tex
moderncvheadii.sty2015/07/28 v2.0.0 modern curriculum vitae and 
letter header variant: 2
moderncvbodyi.sty2015/07/28 v2.0.0 modern curriculum vitae and 
letter body variant: 1
moderncvfooti.sty2015/07/28 v2.0.0 modern curriculum vitae and 
letter footer variant: 1
moderncvcolorblue.sty2015/07/28 v2.0.0 modern curriculum vitae and 
letter color scheme: blue
xunicode.sty2011/09/09 v0.981 provides access to latin accents and 
many other characters in Unicode lower plane
polyglossia.sty2015/03/25 v1.42.4 Alternative to Babel for XeLaTeX 
and LuaLaTeX
makecmds.sty2009/09/03 v1.0a extra command making commands
 xkeyval.sty2014/12/03 v2.7a package option processing (HA)
 xkeyval.tex2014/12/03 v2.7a key=value parser (HA)
gloss-english.ldfpolyglossia: module for english
hyperref.sty2016/06/24 v6.83q Hypertext links for LaTeX
hobsub-hyperref.sty2016/05/16 v1.14 Bundle oberdiek, subset 
hyperref (HO)
hobsub-generic.sty2016/05/16 v1.14 Bundle oberdiek, subset generic 
(HO)
  hobsub.sty2016/05/16 v1.14 Construct package bundles (HO)
infwarerr.sty2016/05/16 v1.4 Providing info/warning/error messages 
(HO)
 ltxcmds.sty2016/05/16 v1.23 LaTeX kernel commands for general use 
(HO)
 intcalc.sty2016/05/16 v1.2 Expandable calculations with integers 
(HO)
etexcmds.sty2016/05/16 v1.6 Avoid name clashes with e-TeX commands 
(HO)
kvsetkeys.sty2016/05/16 v1.17 Key value parser (HO)
kvdefinekeys.sty2016/05/16 v1.4 Define keys (HO)
pdftexcmds.sty2016/05/21 v0.22 Utility functions of pdfTeX for 
LuaTeX (HO)
pdfescape.sty2016/05/16 v1.14 Implements pdfTeX's escape features 
(HO)
bigintcalc.sty

Re: lyx errors with lualatex and non-tex fonts

[Kornel Benko]

Am Mittwoch, 15. Februar 2017 um 13:59:02, schrieb Guenter Milde 

> On 2017-02-15, Kornel Benko wrote:
> > Am Mittwoch, 15. Februar 2017 um 08:37:46, schrieb Jürgen Spitzmüller 
> > 
> >> Am Dienstag, den 14.02.2017, 20:51 +0100 schrieb Kornel Benko:
>
> >> > Apart from this there are still 72 messages with
> >> > ! LaTeX Error: Command \textquotedbl unavailable in encoding
> >> > TU.
> >> > in
> >> > export/examples/modernCV_dvi3_texF
> >> > export/examples/es/modernCV_pdf4_texF
> >> > export/examples/es/modernCV_pdf5_texF
>
> >> Should be fixed now.
>
> > It is.
>
> >> Jürgen
>
> > Still failing for instance export/examples/modernCV_pdf4_texF
> > Missing character: There is no  in font FontAwesome/OT!
>
> It works here. Can you have a look at the output, maybe you can spot a
> missing character. FontAwesome is used for the icons like seen in the
> bottomlines:
>
> Paul Testman
> Teststreet 17 -- 0 Nicecity -- Switzerland
>  +43(0)999 888 •  +43(0)999  •  +43(0)999 
>  test...@testman.xx •  www.johndoe.com •  ptestman
>  ptestman •  ptestman • www.lyx.org
> 6/6
>
> Also I wonder how modernCV_pdf4_texF can fail when
> modernCV_pdf4_systemF compiles -- both use the same Unicode fonts!

Here they fail both. The above was only an example. (I did not say that 
modernCV_pdf4_systemF compiles)
OTOH, lualatex uses the same font and output is OK.

> Maybe updating texlive-fonts-extra (or whereever you get your FontAwesome
> from) or the modernCV document class helps.

It is directly from CTAN here.
Comparing with correct output (e.g. lualatex) it is the glyph for mobile.
(Wrong output attached)
Looking into log:
(/usr9/local/texlive/2016/texmf-dist/tex/latex/fontawesome/fontawesome.sty
Package: fontawesome 2016/05/15 v4.6.3.1 font awesome icons

(/usr9/local/texlive/2016/texmf-dist/tex/latex/fontawesome/fontawesomesymbols-g
eneric.tex) (/usr9/local/texlive/2016/texmf-dist/tex/latex/fontspec/fontspec.sty
Package: fontspec 2017/02/12 v2.6 Font selection for XeLaTeX and LuaLaTeX
(/usr9/local/texlive/2016/texmf-dist/tex/latex/fontspec/fontspec-xetex.sty
Package: fontspec-xetex 2017/02/12 v2.6 Font selection for XeLaTeX and LuaLaTeX
...
But I have also many messages like:
Package microtype Info: Character `A' is missing
(microtype) in font `\TU/FontAwesome(0)/m/n/6'.
(microtype) Ignoring protrusion settings for this character.
Package microtype Info: Character `F' is missing
(microtype) in font `\TU/FontAwesome(0)/m/n/6'.
(microtype) Ignoring protrusion settings for this character.

> Otherwise, we can add
>
>export/examples/es/modernCV_pdf4_texF
>export/examples/es/modernCV_pdf5_texF
>
> to unreliableTests.
>
> > Now only 5 failed tests remain.
>
> Which?

1771 - export/doc/ar/Intro_pdf4_systemF (Failed)
3492 - export/examples/modernCV_pdf4_texF (Failed)
3493 - export/examples/modernCV_pdf4_systemF (Failed)
4649 - export/examples/es/modernCV_pdf4_texF (Failed)
4650 - export/examples/es/modernCV_pdf4_systemF (Failed)

> Günter

Kornel

signature.asc
Description: This is a digitally signed message part.

Re: lyx errors with lualatex and non-tex fonts

[Guenter Milde]

On 2017-02-15, Kornel Benko wrote:
> Am Mittwoch, 15. Februar 2017 um 08:37:46, schrieb Jürgen Spitzmüller 
> 
>> Am Dienstag, den 14.02.2017, 20:51 +0100 schrieb Kornel Benko:

>> > Apart from this there are still 72 messages with
>> > ! LaTeX Error: Command \textquotedbl unavailable in encoding
>> > TU.
>> > in
>> > export/examples/modernCV_dvi3_texF
>> > export/examples/es/modernCV_pdf4_texF
>> > export/examples/es/modernCV_pdf5_texF

>> Should be fixed now.

> It is.

>> Jürgen

> Still failing for instance export/examples/modernCV_pdf4_texF
>   Missing character: There is no  in font FontAwesome/OT!

It works here. Can you have a look at the output, maybe you can spot a
missing character. FontAwesome is used for the icons like seen in the
bottomlines:

Paul Testman
Teststreet 17 -- 0 Nicecity -- Switzerland
 +43(0)999 888 •  +43(0)999  •  +43(0)999 
 test...@testman.xx •  www.johndoe.com •  ptestman
 ptestman •  ptestman • www.lyx.org
6/6

Also I wonder how modernCV_pdf4_texF can fail when
modernCV_pdf4_systemF compiles -- both use the same Unicode fonts!

Maybe updating texlive-fonts-extra (or whereever you get your FontAwesome
from) or the modernCV document class helps. 

Otherwise, we can add

   export/examples/es/modernCV_pdf4_texF
   export/examples/es/modernCV_pdf5_texF

to unreliableTests.

> Now only 5 failed tests remain.

Which?

Günter

Re: ctest fails standard exports of a number of examples and templates files

[Guenter Milde]

Dear Jean-Pierre, Scott, and Kornel, dear others,

On 2017-02-15, Scott Kostyshak wrote:
> On Tue, Feb 14, 2017 at 04:31:22PM +0100, Jean-Pierre Chrétien wrote:
>> Hello,

>> I discovered that quite a number of examples and templates files
>> depend on resources which do not come with TeXLive 2016. This is
>> normal with external applications (R, lilypond, etc.) but there are a
>> number of LaTeX packages from editors which are not in the appropriate
>> collection of TeXLIve 2016.

>> I made an addition to Development.lyx to record what is needed to avoid
>> ctest failures on files which are labeled as standard exports, see
>> attachment.

> Thanks for taking a look at the tests! And I agree, it's good to specify
> all the dependencies. I have some comments below.

> I think there are tests that should be labeled "nonstandard" but
> currently are not.

So do I. 
All tests that require more than the agreed "standard requirements"
should get the label "unreliable:nonstandard".

> Once these are corrected, we would then change the
> following sentence you added:

>   With TeXLive 2016, here are the resources which should be installed to
>   avoid standard tests failure:

> to something like:

>   Here are the resources which should be installed to
>   avoid failure of tests with the "unreliable:nonstandard" label:

> Do you agree?

> For Development.lyx, the section
> now reads:

> To run all tests, in the build directory simply run the command ctest. A
> full, up-to-date TeXLive installation is recommended to run the tests.
> Otherwise, some tests will fail. Tests with additional requirements are
> labeled "unreliable:nonstandard" [footnote: With an up-to-date Texlive 2016
> installation, there are standard tests which will fail because of
> missing packages or external applications, see below for a list of those
> which should be installed to avoid these failures.]

The footnote would no longer be required (i.e. replaced by adding the tests
from the list below to the "unreliableTests" filter file.

> I think this could be made more concise. Also, we need to be sure not to
> confuse two things: (1) a *full* (I think we should mention this,
> right?) up-to-date TeXLive installation is assumed for tests without the
> "unreliable:nonstandard" label; and (2) for tests with the
> "unreliable:nonstandard" label, there are requirements in addition to a
> full TeXLive installation.

> Instead of "see below", I would prefer a reference to a label.

> Currently the list of dependencies you added is under the section
> "4.3.1.7 What action should you take if a test fails?".
> I think it would make more sense to add a section before e.g.
> "Configuring the tests" that is dedicated to "Test dependencies".
> What do you think?

IMO, it is simpler to add the additional nonstandard dependencies in the
comments in "unreliableTests" only and reword the start of 4.3.1.4
  
  To run all tests, in the build directory simply run the command ctest.
  A full, up-to-date TeXLive installation is recommended to run the
  tests. Otherwise, some tests will fail. Tests with additional
  requirements are labeled “unreliable:nonstandard”. 
  
to something like

  A full, up-to-date TeXLive installation is required to run the
  tests. Otherwise, some tests will fail. 
  Tests with the label “unreliable:nonstandard” have additional
  requirements (see development/autotests/unreliableTests for details).
  To exclude these tests, run `ctest -LE nonstandard`.
  To run all tests, in the build directory simply run the command ctest.

This would safe us some work updating requirements and prevent
desynchronisation of the requirement lists.

Günter

Re: alpha dot

[Guillaume Munch]

Le 15/02/2017 à 13:46, Guenter Milde a écrit :

On 2017-02-15, Guillaume Munch wrote:


[-- Type: text/plain, Encoding: 8bit --]



Le 15/02/2017 à 05:02, Scott Kostyshak a écrit :

On Wed, Jan 04, 2017 at 09:00:44AM +0100, Wolfgang Engelmann wrote:

On 03.01.2017 21:32, Scott Kostyshak wrote:

On Mon, Jan 02, 2017 at 11:25:07AM +0100, Wolfgang Engelmann wrote:



How could I get a dot on an alpha by using the math function of
lyx? Or do I have to use a tex insert:



$\dot \alpha$


Instead of a tex insert (ERT), use a mathed inset (Ctrl-M, say) and
insert \dot \alpha there. This is converted to the right symbol(s) in
both, LyX and output document.


Click on the "frame decorations" icon in the math toolbar. See the
screenshot. Does that work for you?


This gives the same result.


In the LyX math preview (i.e. without instant preview), the "dot" looks
more like a small line. Do you agree or does it look correct to you?



Yes, Scott, in the instant preview it does look like a short line, but its
correct in the pdf output.



I tried to fix it, but it seems that in this part of the code we can't
really draw circles; we can only trick the user into thinking its a
little circle by drawing a short line. We could draw a polygon with many
sides (I guess this is how all circles are drawn on screens?). Is there
a program that outputs coordinates in the format required by the code in
src/mathed/MathSupport.cpp ?



Another interesting issue is that the math decorations do not zoom with
the font; (because)? if they did, our approximations would look pretty
rough I guess. I'm actually amazed with how well most of them work. I
never realized a problem before.



I have a series of patches (attached) that make the decoration thickness
increase with zoom, to improve the appearance on high-dpi. The
decorations do not look worse than without the added thickness. I have
not committed it yet because I have not found a way to keep it simple,
and there are still some issues. But you might like to try them.



For \dot in particular it looks like a square rather than a line.



Did you think about using Unicode and, e.g., the STIX fonts for mathed?
I know this is a major change but it should solve issues like scaling and
curved lines.


Harfbuzz has been providing access to the the math table of opentype
fonts since very recently, and even computes the glyph assemblies for
decorations. So in theory it is possible to implement variable-size
decorations with opentype fonts. See
http://www.lyx.org/trac/ticket/9014#comment:7. But that's a big project.



For the "frame decorators", one could explore the use of combining
characters - maybe there are issues with placement but α̇ or 훼̇ looks better
in mathed than the current "square dot".


Indeed, for standard-sized decorations, it would be simpler to implement.

Re: CopyrightYear is the correct command name in acmsiggraph-0.92.layout

[Guenter Milde]

On 2017-02-15, Jean-Pierre Chrétien wrote:
> Le 14/02/2017 à 17:20, Jürgen Spitzmüller a écrit :
>> Am Dienstag, den 14.02.2017, 16:16 +0100 schrieb Jean-Pierre Chrétien:
>>> diff --git a/lib/layouts/acmsiggraph-0-92.layout
>>> b/lib/layouts/acmsiggraph-0-92.layout
>>> index 5732ecc..30a35df 100644
>>> --- a/lib/layouts/acmsiggraph-0-92.layout
>>> +++ b/lib/layouts/acmsiggraph-0-92.layout
>>> @@ -108,7 +108,7 @@ End

>>>   Style "Copyright_year"
>>>  CopyStyle   "TOG online ID"
>>> -   LatexName   copyrightyear
>>> +   LatexName   CopyrightYear
>>>  LabelString "Year of copyright:"

>> Are you sure this command was not valid in previous versions? The
>> changelog of 0.93 says: "Corrected \CopyrightYear command". Perhaps you
>> should double-check with version 0.92.

> Sure, it was copyrightyear in earlier versions.


>> BTW it seems like the class is deprecated anyway in favour of
>> acmart.cls. See

>> http://www.siggraph.org/learn/instructions-authors

> You're right, and acmart.cls is in TeXLive2016.
> So what? Should we remove the acmsiggraph template and layout once a template 
> and layout are available for acmart? I can look at this when I find some time 
> to 
> do so.

> In the meantime, the asmsiggraph template fails to compile without my patch. 
> Shoud it be applied (this is for stable)?

If it was copyrightyear in earlier versions, it will fail with your patch
for people using the earlier versions.
I propose to add compatibility code in the preamble, e.g.

   \providecommand*{\CopyrightYear}{\copyrightyear}
   
This way, we don't need one more layout for a deprecated document class'
version.

In addition, the template and layout should get a comment that 

>>  the class is deprecated anyway in favour of
>> acmart.cls. See http://www.siggraph.org/learn/instructions-authors

Günter

Re: alpha dot

[Guenter Milde]

On 2017-02-15, Guillaume Munch wrote:

> [-- Type: text/plain, Encoding: 8bit --]

> Le 15/02/2017 à 05:02, Scott Kostyshak a écrit :
>> On Wed, Jan 04, 2017 at 09:00:44AM +0100, Wolfgang Engelmann wrote:
>>> On 03.01.2017 21:32, Scott Kostyshak wrote:
 On Mon, Jan 02, 2017 at 11:25:07AM +0100, Wolfgang Engelmann wrote:

> How could I get a dot on an alpha by using the math function of
> lyx? Or do I have to use a tex insert:

> $\dot \alpha$

Instead of a tex insert (ERT), use a mathed inset (Ctrl-M, say) and
insert \dot \alpha there. This is converted to the right symbol(s) in
both, LyX and output document.

 Click on the "frame decorations" icon in the math toolbar. See the
 screenshot. Does that work for you?

This gives the same result.

 In the LyX math preview (i.e. without instant preview), the "dot" looks
 more like a small line. Do you agree or does it look correct to you?

>>> Yes, Scott, in the instant preview it does look like a short line, but its
>>> correct in the pdf output.

>> I tried to fix it, but it seems that in this part of the code we can't
>> really draw circles; we can only trick the user into thinking its a
>> little circle by drawing a short line. We could draw a polygon with many
>> sides (I guess this is how all circles are drawn on screens?). Is there
>> a program that outputs coordinates in the format required by the code in
>> src/mathed/MathSupport.cpp ?

>> Another interesting issue is that the math decorations do not zoom with
>> the font; (because)? if they did, our approximations would look pretty
>> rough I guess. I'm actually amazed with how well most of them work. I
>> never realized a problem before.

> I have a series of patches (attached) that make the decoration thickness
> increase with zoom, to improve the appearance on high-dpi. The
> decorations do not look worse than without the added thickness. I have
> not committed it yet because I have not found a way to keep it simple,
> and there are still some issues. But you might like to try them.

> For \dot in particular it looks like a square rather than a line.


Did you think about using Unicode and, e.g., the STIX fonts for mathed?
I know this is a major change but it should solve issues like scaling and
curved lines.

For the "frame decorators", one could explore the use of combining
characters - maybe there are issues with placement but α̇ or 훼̇ looks better
in mathed than the current "square dot".


Günter

Re: alpha dot

[Guillaume Munch]

Le 15/02/2017 à 05:02, Scott Kostyshak a écrit :

On Wed, Jan 04, 2017 at 09:00:44AM +0100, Wolfgang Engelmann wrote:



On 03.01.2017 21:32, Scott Kostyshak wrote:

On Mon, Jan 02, 2017 at 11:25:07AM +0100, Wolfgang Engelmann wrote:

How could I get a dot on an alpha by using the math function of lyx? Or do I
have to use a tex insert:

$\dot \alpha$

Click on the "frame decorations" icon in the math toolbar. See the
screenshot. Does that work for you?

In the LyX math preview (i.e. without instant preview), the "dot" looks
more like a small line. Do you agree or does it look correct to you?

Scott

Yes, Scott, in the instant preview it does look like a short line, but its
correct in the pdf output.


I tried to fix it, but it seems that in this part of the code we can't
really draw circles; we can only trick the user into thinking its a
little circle by drawing a short line. We could draw a polygon with many
sides (I guess this is how all circles are drawn on screens?). Is there
a program that outputs coordinates in the format required by the code in
src/mathed/MathSupport.cpp ?

Another interesting issue is that the math decorations do not zoom with
the font; (because)? if they did, our approximations would look pretty
rough I guess. I'm actually amazed with how well most of them work. I
never realized a problem before.



I have a series of patches (attached) that make the decoration thickness
increase with zoom, to improve the appearance on high-dpi. The
decorations do not look worse than without the added thickness. I have
not committed it yet because I have not found a way to keep it simple,
and there are still some issues. But you might like to try them.

For \dot in particular it looks like a square rather than a line.

Guillaume
>From 45d8c5edc2c8e26065468c691f90a1d50ddc20a9 Mon Sep 17 00:00:00 2001
From: Guillaume Munch 
Date: Mon, 28 Nov 2016 03:27:30 +0100
Subject: [PATCH 1/3] Increase thickness of math delimiter with zoom

---
 src/mathed/InsetMathAMSArray.cpp   | 15 ++-
 src/mathed/InsetMathBig.cpp| 19 +--
 src/mathed/InsetMathCases.cpp  |  4 +++-
 src/mathed/InsetMathDecoration.cpp | 15 +--
 src/mathed/InsetMathDelim.cpp  | 25 ++---
 src/mathed/InsetMathDots.cpp   | 13 +
 src/mathed/InsetMathFrac.cpp   | 11 +++
 src/mathed/InsetMathXArrow.cpp | 19 +--
 src/mathed/MathSupport.cpp | 26 +++---
 src/mathed/MathSupport.h   | 10 +-
 10 files changed, 110 insertions(+), 47 deletions(-)

diff --git a/src/mathed/InsetMathAMSArray.cpp b/src/mathed/InsetMathAMSArray.cpp
index 9a9a349..df022b6 100644
--- a/src/mathed/InsetMathAMSArray.cpp
+++ b/src/mathed/InsetMathAMSArray.cpp
@@ -88,19 +88,24 @@ void InsetMathAMSArray::metrics(MetricsInfo & mi, Dimension & dim) const
 	Changer dummy2 = mi.base.changeEnsureMath();
 	Changer dummy = mi.base.changeArray();
 	InsetMathGrid::metrics(mi, dim);
+	mathed_deco_metrics(mi.base, dim);
 }
 
 
 void InsetMathAMSArray::draw(PainterInfo & pi, int x, int y) const
 {
 	Changer dummy2 = pi.base.changeEnsureMath();
+	{
+		Changer dummy = pi.base.changeArray();
+		InsetMathGrid::draw(pi, x, y);
+	}
 	Dimension const dim = dimension(*pi.base.bv);
 	int const yy = y - dim.ascent();
-	// Drawing the deco after changeStyle does not work
-	mathed_draw_deco(pi, x + 1, yy, 5, dim.height(), from_ascii(name_left()));
-	mathed_draw_deco(pi, x + dim.width() - 8, yy, 5, dim.height(), from_ascii(name_right()));
-	Changer dummy = pi.base.changeArray();
-	InsetMathGrid::draw(pi, x, y);
+	int const t = mathed_deco_thickness(pi.base);
+	mathed_draw_deco(pi, x + 1 - t/2, yy,
+	 5, dim.height(), from_ascii(name_left()));
+	mathed_draw_deco(pi, x + dim.width() - t/2 - 8, yy,
+	 5, dim.height(), from_ascii(name_right()));
 }
 
 
diff --git a/src/mathed/InsetMathBig.cpp b/src/mathed/InsetMathBig.cpp
index 889f8d4..e5848aa 100644
--- a/src/mathed/InsetMathBig.cpp
+++ b/src/mathed/InsetMathBig.cpp
@@ -24,6 +24,10 @@
 #include "support/docstream.h"
 #include "support/lstrings.h"
 
+#include 
+
+using namespace std;
+
 
 namespace lyx {
 
@@ -90,10 +94,12 @@ void InsetMathBig::metrics(MetricsInfo & mi, Dimension & dim) const
 {
 	Changer dummy = mi.base.changeEnsureMath();
 	double const h = theFontMetrics(mi.base.font).ascent('I');
-	double const f = increase();
-	dim.wid = 6;
-	dim.asc = int(h + f * h);
-	dim.des = int(f * h);
+	double const height = h * (1 + 2 * increase());
+	int const axis = axis_height(mi.base);
+	dim.wid = max(6, mathed_mu(mi.base.font, 6.0));
+	dim.asc = int(height/2 + axis);
+	dim.des = int(height/2 - axis);
+	mathed_deco_metrics(mi.base, dim);
 }
 
 
@@ -115,8 +121,9 @@ void InsetMathBig::draw(PainterInfo & pi, int x, int y) const
 {
 	Changer dummy = pi.base.changeEnsureMath();
 	Dimension const dim = dimension(*pi.base.bv);
-	mathed_draw_deco(pi, x + 1, y - 

Re: CopyrightYear is the correct command name in acmsiggraph-0.92.layout

[Jean-Pierre Chrétien]

Le 14/02/2017 à 17:20, Jürgen Spitzmüller a écrit :

Am Dienstag, den 14.02.2017, 16:16 +0100 schrieb Jean-Pierre Chrétien:

diff --git a/lib/layouts/acmsiggraph-0-92.layout
b/lib/layouts/acmsiggraph-0-92.layout
index 5732ecc..30a35df 100644
--- a/lib/layouts/acmsiggraph-0-92.layout
+++ b/lib/layouts/acmsiggraph-0-92.layout
@@ -108,7 +108,7 @@ End

  Style "Copyright_year"
 CopyStyle   "TOG online ID"
-   LatexName   copyrightyear
+   LatexName   CopyrightYear
 LabelString "Year of copyright:"


Are you sure this command was not valid in previous versions? The
changelog of 0.93 says: "Corrected \CopyrightYear command". Perhaps you
should double-check with version 0.92.


Sure, it was copyrightyear in earlier versions.



BTW it seems like the class is deprecated anyway in favour of
acmart.cls. See

http://www.siggraph.org/learn/instructions-authors


You're right, and acmart.cls is in TeXLive2016.
So what? Should we remove the acmsiggraph template and layout once a template 
and layout are available for acmart? I can look at this when I find some time to 
do so.


In the meantime, the asmsiggraph template fails to compile without my patch. 
Shoud it be applied (this is for stable)?


--
Jean-Pierre

Re: lyx errors with lualatex and non-tex fonts

[Kornel Benko]

Am Mittwoch, 15. Februar 2017 um 09:39:57, schrieb Guenter Milde 

> On 2017-02-14, Kornel Benko wrote:
> 
> >> BTW: was does the Sublabel: RESET mean?
> 
> > RESET means 'Reset the set of active sublabels'. Ergo from now on no
> > sublabel is defined.
> 
> Thank you for the explanation. So, instead of sublabel RESET, we can also
> move the patterns to the top of the file.

Yes, same effect.

> However, I'd prefer if every clause in invertedTests were sorted under some
> sublabel - either one of the existing ones (including TODO for cases where
> the reason is not yet known) or a new one like "unreliable-fail").

OK.
 Note 1: You can assign more than one label at the same 'Sublabel:'-line.
 Note 2: Each 'Sublabel:'-line implies RESET as the first entry.

Also note that there are hardcoded labels. They are not affected by RESET. They 
are assigned because of
 1.) path of the lyx-file (e.g. lib/examples --> label examples, lib/doc --> 
manuals, ...)
 2.) file in which the test is determined (e.g. unreliableTests --> label 
unreliable, ...)
 3.) kind of a test (e.g. lyx2lyx, load)

> Günter

Kornel

signature.asc
Description: This is a digitally signed message part.

Re: lyx errors with lualatex and non-tex fonts

[Kornel Benko]

Am Mittwoch, 15. Februar 2017 um 08:37:46, schrieb Jürgen Spitzmüller 

> Am Dienstag, den 14.02.2017, 20:51 +0100 schrieb Kornel Benko:
> > Apart from this there are still 72 messages with
> > ! LaTeX Error: Command \textquotedbl unavailable in encoding
> > TU.
> > in
> > export/examples/modernCV_dvi3_texF
> > export/examples/es/modernCV_pdf4_texF
> > export/examples/es/modernCV_pdf5_texF
> 
> Should be fixed now.

It is.

> Jürgen

Still failing for instance export/examples/modernCV_pdf4_texF
Missing character: There is no  in font FontAwesome/OT!

Now only 5 failed tests remain.

Kornel


signature.asc
Description: This is a digitally signed message part.

Re: lyx errors with lualatex and non-tex fonts

[Guenter Milde]

On 2017-02-14, Kornel Benko wrote:

>> BTW: was does the Sublabel: RESET mean?

> RESET means 'Reset the set of active sublabels'. Ergo from now on no
> sublabel is defined.

Thank you for the explanation. So, instead of sublabel RESET, we can also
move the patterns to the top of the file.

However, I'd prefer if every clause in invertedTests were sorted under some
sublabel - either one of the existing ones (including TODO for cases where
the reason is not yet known) or a new one like "unreliable-fail").

Günter

Re: [LyX/master] ctests: update inversion patterns.

[Guenter Milde]

On 2017-02-15, Scott Kostyshak wrote:
> On Tue, Feb 14, 2017 at 10:34:25PM +0100, Günter Milde wrote:
>> commit b863fb0723104bf702e4b4d8b967a7ce0c2255b5
>> Author: Günter Milde 
>> Date:   Tue Feb 14 22:31:12 2017 +0100

>> ctests: update inversion patterns.

>> The new TeXLive uses font encoding TU for Unicode fonts with Xe- and 
>> LuaTeX.
>> The command \textquotedbl for straight quotes is no longer supported,
>> \textipa no longer supported with LuaTeX.

Missing \textquotedbl is clearly a bug as this is the standard way to get
straight double quotes with fonts supporting the "TeX ligatures".

Missing \textipa could be intentional -- maybe this is "outsourced" to a
Luatex package (for XeTeX, \textipa support is in xunicode.sty). 


>> Problems with Spanish Babel and Xe/LuaTeX with 8-bit fonts lead to new 
>> errors
>> in some cases.

> Did you by chance file a bug report? If so, we should link to it in the
> invertedTests comments.

No, I did not even find out where to report...
So, if you want to take aktion I would be glad.

Anyway, the Babel-bug is low priority, as using Xe/LuaTeX with 8-bit
fonts is a rare cornercase. Also, the simple and obvious workaround is to
set the input encoding to utf8.

We could simply *ignore* all Russian, Ukrainean, and Spanish documents with
.*_texF because they produce either wrong output or fail.

In the long run, LyX should use utf8 as default inputenc for these languages.



Thanks,

Günter

Re: [patch] Theorem environment: set NextNoIndent to 0

[Guenter Milde]

On 2017-02-15, Scott Kostyshak wrote:
> On Tue, Feb 14, 2017 at 10:51:44PM -0500, Richard Heck wrote:
>> On 02/14/2017 10:12 PM, Scott Kostyshak wrote:
>> > On Mon, Nov 14, 2016 at 04:36:12PM -0500, Scott Kostyshak wrote:
>> >> On Sun, Nov 13, 2016 at 07:18:43PM +, Guenter Milde wrote:
>> >>> On 2016-11-11, Scott Kostyshak wrote:
>> >>>
>>  The attached patch fixes the display of the attached .lyx file. When
>>  compiled, the .lyx file indents the first line of the environment
>>  following the Theorem environment. But LyX shows the line as not
>>  indented.
>>  There might be similar fixes needed, such as in elsart.layout

...

>> >> NextNoIndent was set to 1 at 46611ba1, but the commit message only
>> >> states "Fixes to the theorem modules." so I'm not sure how else to
>> >> check.

Some suggestions:

* Ask the developer of 46611ba1 for possible reasons of the change.

* Change the document class of your examples and compare output and LyX-GUI

* Can indentation/start of a new paragraph be influenced by the new 
  "new paragraph" special? (the curly arrow popping up every now and then in
  files converted with lyx2lyx)
  

>> > Günter (or anyone else), do you object to me committing the patch?

No. I am not against the change (just wanted to avoid a situation where
we solve display in half of the cases but break it in the other half.)


>> Please do commit. I find this very annoying and keep meaning to commit
>> it myself.

> OK, I'll wait a couple of days to see if Günter has an objection. He
> might see something that I missed.

Please go ahead.

Thanks,

Günter