Re: LyX crashes when using Arabic versions 2.4.3+

2025-07-29 Thread Jean-Marc Lasgouttes
Le 29/07/2025 à 00:13, daky dakyion a écrit : *The problem:* When creating a new file and changing the language of the document to Arabic. After typing the first letter in Arabic and sometimes even in English, the app crashes instantly with SIGSEGV error. This happened on two different PCs, one

Re: LyX crashes when using Arabic versions 2.4.3+

2025-07-29 Thread Jean-Marc Lasgouttes
Le 29/07/2025 à 00:13, daky dakyion a écrit : *The problem:* When creating a new file and changing the language of the document to Arabic. After typing the first letter in Arabic and sometimes even in English, the app crashes instantly with SIGSEGV error. This happened on two different PCs, one

Re: Backtick Display Problem

2025-07-28 Thread Jean-Marc Lasgouttes
Le 28/07/2025 à 17:11, Richard Kimberly Heck a écrit : When using the Droid Serif font for display, I'm seeing a problem displaying the backtick (used as left single quote), in stable only. It gets drawn over the following letter, as if it were a grave accent. The LaTeX code is correct, however

Re: [LyX/master] Add plimsoll symbol, for Bernt

2025-07-26 Thread Jean-Marc Lasgouttes
Le 26 juillet 2025 19:01:05 GMT+02:00, Richard Kimberly Heck a écrit : >It seems unfortunate that I had to mess with the CPP code to add this symbol. >That means an ordinary user couldn't have done so. Without that, LyX did not >try to load the package. Couldn't we let LyX load unknown packages

Re: The features/breakinsets branch

2025-07-23 Thread Jean-Marc Lasgouttes
Le 23/07/2025 à 10:49, Pavel Sanda a écrit : On Tue, Jul 22, 2025 at 06:11:00PM +0200, Jean-Marc Lasgouttes wrote: Some of the not working things above are easy to fix, some are not. And there are all the bugs I have not identified yet. If you are interested in bug reports, I see painting

Re: [LyX/master] Fix math-mode control of input method on Linux and Windows

2025-07-23 Thread Jean-Marc Lasgouttes
Le 23/07/2025 à 13:59, Yokota K. a écrit : 2025/07/23 16:29、Jean-Marc Lasgouttes のメール: Le 23 juillet 2025 03:28:12 GMT+02:00, Koji Yokota a écrit : This part of code prohibits only CJK from entering their characters in the math mode, so it is a minimal list. I’m thinking of taking a

Re: The features/breakinsets branch

2025-07-23 Thread Jean-Marc Lasgouttes
Le 23/07/2025 à 11:15, Jürgen Spitzmüller a écrit : Am Mittwoch, dem 23.07.2025 um 10:49 +0200 schrieb Pavel Sanda: Towards the end of line the end of inset is clearly wrongly computed and text within inset and behind gets mixed. This is something that I did not see in my example (also the no

Re: [LyX/master] Fix math-mode control of input method on Linux and Windows

2025-07-23 Thread Jean-Marc Lasgouttes
Le 23 juillet 2025 03:28:12 GMT+02:00, Koji Yokota a écrit : >This part of code prohibits only CJK from entering their characters in the >math mode, so it is a minimal list. I’m thinking of taking a strategy to add >another language if it is found to be appropriate to do so. Does it look good >

Re: The features/breakinsets branch

2025-07-22 Thread Jean-Marc Lasgouttes
Le 22/07/2025 à 23:41, Scott Kostyshak a écrit : On Tue, Jul 22, 2025 at 06:11:00PM +0200, Jean-Marc Lasgouttes wrote: Comments welcome. Don't get your expectations too high, this is far from usable in real life. Thanks for your work on this, JMarc! I pulled in the branch and compile

The features/breakinsets branch

2025-07-22 Thread Jean-Marc Lasgouttes
this a bit today and tomorrow, and then stop for a month. Comments welcome. Don't get your expectations too high, this is far from usable in real life. JMarc From c04d9691efdfa46cf5d066553611d7cdfcd8d00f Mon Sep 17 00:00:00 2001 From: Jean-Marc Lasgouttes Date: Sun, 13 Jul 2025 23:

Re: [LyX/master] Fix math-mode control of input method on Linux and Windows

2025-07-22 Thread Jean-Marc Lasgouttes
Le 22/07/2025 à 12:48, Koji Yokota a écrit : commit 5f290476f809e75265e9a61a5bbcf16438ee8bd0 Author: Koji Yokota Date: Tue Jul 22 19:44:02 2025 +0900 Fix math-mode control of input method on Linux and Windows Since QInputMethod::locale() doesn't work on these systems, use lang

Re: clang compilation with options -Winconsistent-missing-destructor-override and -Wsuggest-destructor-override

2025-07-17 Thread Jean-Marc Lasgouttes
Le 16/07/2025 à 13:53, Kornel Benko a écrit : There are many suspicious places requiring 'override' attribute. The attached takes care of it. Does it harm? I read this https://github.com/isocpp/CppCoreGuidelines/issues/721 and I still do not know what I think about it :( JMarc -- lyx-devel m

Re: 5 warnings from -Woverloaded-virtual (Koji?)

2025-07-14 Thread Jean-Marc Lasgouttes
Le 14/07/2025 à 14:41, Yokota K. a écrit : Along with the 3/ plan, I hope the attached patch suppresses the warnings, but I’m not sure how much of resemblance of functions leads to a warning (it still uses overloaded text() functions). Looks good to me. What is not good is when you overload a

Re: [LyX/master] Avoid clang warning

2025-07-14 Thread Jean-Marc Lasgouttes
Le 14/07/2025 à 13:46, Kornel Benko a écrit : BTW, what about warnings we get when compiling with -fsanitize=address? Mainly -Wmaybe-uninitialized. I was planning to do useful summer work this week, but if you prefer less warnings to breakable insets, we can do that ;) JMarc -- lyx-devel ma

Re: [LyX/master] Avoid clang warning

2025-07-14 Thread Jean-Marc Lasgouttes
Le 14/07/2025 à 13:27, Scott Kostyshak a écrit : Current master compiles fine for me, which means you didn't introduce any new warnings for the categories that we've checked off, which is a good amount. That is, I use -Werror and -Weverything-we-fixed-before. Note that I have obviously not chec

Re: [LyX/master] Avoid clang warning

2025-07-14 Thread Jean-Marc Lasgouttes
Le 14/07/2025 à 12:43, Kornel Benko a écrit : Am Mon, 14 Jul 2025 10:20:59 + schrieb Jean-Marc Lasgouttes : commit c6aee1c65531578d6128e06839dde168ede1d988 Author: Jean-Marc Lasgouttes Date: Mon Jul 14 12:19:47 2025 +0200 Avoid clang warning Still 1 remaining: /usr2/src/lyx/lyx

Re: Worth fixing warnings about deprecated copy constructors and copy assignment operators?

2025-07-13 Thread Jean-Marc Lasgouttes
Le 12/07/2025 à 09:53, Jürgen Spitzmüller a écrit : Am Freitag, dem 11.07.2025 um 17:48 +0200 schrieb Jean-Marc Lasgouttes: I will finish the warnings you sent. Just FTR I am getting now: LayoutFile.cpp: In destructor ‘lyx::LayoutFileList::~LayoutFileList()’: LayoutFile.cpp:53:17: warning

Re: Worth fixing warnings about deprecated copy constructors and copy assignment operators?

2025-07-11 Thread Jean-Marc Lasgouttes
Le 11/07/2025 à 17:12, Scott Kostyshak a écrit : I did a few more. Not sure I see these. Did you push? I pushed now :) For InsetBibitem and InsetCitation, the destructor is useful, so that one needs to declare the two others (rule-of-3). Is it getting clearer? The concept makes sense. Bu

Re: Worth fixing warnings about deprecated copy constructors and copy assignment operators?

2025-07-11 Thread Jean-Marc Lasgouttes
Le 11/07/2025 à 15:14, Jean-Marc Lasgouttes a écrit : Le 07/07/2025 à 13:11, Scott Kostyshak a écrit : Thanks for taking a look Riki! I don't think I will address these myself since I don't understand the issue well. Someone can do it, or I think it's fine to leave it as is. I&#

Re: Worth fixing warnings about deprecated copy constructors and copy assignment operators?

2025-07-11 Thread Jean-Marc Lasgouttes
Le 07/07/2025 à 13:11, Scott Kostyshak a écrit : Thanks for taking a look Riki! I don't think I will address these myself since I don't understand the issue well. Someone can do it, or I think it's fine to leave it as is. I'll see if I have time in the future to study the relevant issue to gain s

Re: [PATCH] Make Qt6 the default

2025-07-11 Thread Jean-Marc Lasgouttes
Le 10/07/2025 à 22:11, Thibaut Cuvelier a écrit : QMake is mostly deprecated: only bugfixes for Qt 6, removal for Qt 7. CMake is the way forward: you can only build Qt 6 with CMake. (No official or direct announcement, though. Have a look at the development history: https://github.com/qt/qtbase

Re: [PATCH] Make Qt6 the default

2025-07-11 Thread Jean-Marc Lasgouttes
Le 10/07/2025 à 21:24, Richard Kimberly Heck a écrit : On 7/10/25 12:16 PM, Jean-Marc Lasgouttes wrote: Hi there, Would that patch be OK? It makes Qt6 the default for configure-based builds. It does seem time to default to Qt 6. I pushed it. There is no mail because I changed a latin1

[PATCH] Make Qt6 the default

2025-07-10 Thread Jean-Marc Lasgouttes
When qmake fails, one sees the following warning (where qt_major is 5 or 6): Unable to configure Qt$qt_major with qmake. Please report. I have not yet seen any report about error. JMarc >From e59040a150564fbf589c62258f21e4f994f0aa63 Mon Sep 17 00:00:00 2001 From: Jean-Marc Lasgouttes Date: Thu

Re: Metrics Crash/Metrics Loop

2025-07-08 Thread Jean-Marc Lasgouttes
Le 08/07/2025 à 12:25, Jürgen Spitzmüller a écrit : Am Dienstag, dem 08.07.2025 um 12:05 +0200 schrieb Scott Kostyshak: Looks good here. The only issue is that no matter what zoom I set I can't get rid of the vertical red dotted lines (indicating the document is wider than my screen). I think

Re: Fixing the only two -Wdouble-promotion warnings

2025-07-04 Thread Jean-Marc Lasgouttes
Le 04/07/2025 à 17:48, Scott Kostyshak a écrit : The -Wdouble-promotion warning gives two warnings on Clang The two patches look good, but I still do not understand why QString::toqreal does not exist. JMarc -- lyx-devel mailing list lyx-devel@lists.lyx.org https://lists.lyx.org/mailman/list

Re: LyX 2.5.0

2025-07-04 Thread Jean-Marc Lasgouttes
Le 04/07/2025 à 10:02, Jürgen Spitzmüller a écrit : Am Donnerstag, dem 03.07.2025 um 17:26 -0400 schrieb Richard Kimberly Heck: At what point should we start thinking about translations? After beta? It depends on when you want to announce feature freeze (I suggest: now) and string freeze (I s

Re: Two macros undefined (-Wundef)

2025-07-03 Thread Jean-Marc Lasgouttes
Le 03/07/2025 à 13:17, Scott Kostyshak a écrit : #if defined(USE__func__) && USE__func__ I removed the use of USE__func__. Nothing of value has been lost. Great! That warning's gone now. Only remaining one is: src/support/FileName.cpp:687:7: warning: 'MKDIR_TAKES_ONE_ARG' is not defined

Re: Two macros undefined (-Wundef)

2025-07-03 Thread Jean-Marc Lasgouttes
Le 03/07/2025 à 12:17, Kornel Benko a écrit : Am Thu, 3 Jul 2025 12:06:42 +0200 schrieb Scott Kostyshak : I get multiple warnings (from GCC and Clang) when compiling with -Wundef about two main issues: src/support/FileName.cpp:687:7: warning: 'MKDIR_TAKES_ONE_ARG' is not defined, evaluates

Re: 5 warnings from -Woverloaded-virtual (Koji?)

2025-07-02 Thread Jean-Marc Lasgouttes
Le 03/07/2025 à 00:10, Jean-Marc Lasgouttes a écrit : src/mathed/InsetMathHull.h:135:14: warning: ‘virtual void lyx::InsetMathHull::write(std::ostream&) const’ was hidden [- Woverloaded-virtual=]    135 | void write(std::ostream & os) const override;    |  ^

Re: 5 warnings from -Woverloaded-virtual (Koji?)

2025-07-02 Thread Jean-Marc Lasgouttes
Le 02/07/2025 à 18:16, Scott Kostyshak a écrit : Compiling with GCC and -Woverloaded-virtual gives 5 warnings. Are they worth addressing? From what I understand, the main way to address them is via a "using" statement (https://stackoverflow.com/questions/18515183/c-overloaded-virtual-function-w

Re: One last (I think) casting away of qualifiers

2025-07-02 Thread Jean-Marc Lasgouttes
Le 02/07/2025 à 11:17, Scott Kostyshak a écrit : The attached patch just initializes demangled without const to begin with. Is it OK? The side effect is of course that in the call to .arg(), 'demangled' is no longer const. There was no patch, so I took the liberty to commit my own take on it (

Re: patch: make a const cast explicit

2025-07-01 Thread Jean-Marc Lasgouttes
Le 01/07/2025 à 10:54, Scott Kostyshak a écrit : Thanks, JMarc! Please commit. (and also thanks for the explanation). Done. JMarc -- lyx-devel mailing list lyx-devel@lists.lyx.org https://lists.lyx.org/mailman/listinfo/lyx-devel

Re: LyX 2.5.0

2025-07-01 Thread Jean-Marc Lasgouttes
Le 01/07/2025 à 15:32, Pavel Sanda a écrit : Large part of bugs is in fact zombies with no one having time/real interest to fix them. Let us know once you start seriously moving towards 2.5, after those are dropped we will have realistic estimation what's ahead for us wrt 2.5. Was it to be expe

Re: patch: make a const cast explicit

2025-07-01 Thread Jean-Marc Lasgouttes
Le 30/06/2025 à 23:40, Scott Kostyshak a écrit : Clang gives the following warning: src/frontends/qt/GuiAlert.cpp:55:33: warning: cast from 'const char *' to 'char *' drops const qualifier [-Wcast-qual] 55 | QApplication app(argc, (char**)argv); |

Re: Float equality warnings

2025-06-30 Thread Jean-Marc Lasgouttes
Le 30/06/2025 à 13:49, Jürgen Spitzmüller a écrit : Am Montag, dem 30.06.2025 um 13:42 +0200 schrieb Jean-Marc Lasgouttes: AFAIR, we had that with the default font sizes in preferences. But that problem wasn't with the comparison, but the construction (calculation) of values, which wa

Re: Float equality warnings

2025-06-30 Thread Jean-Marc Lasgouttes
Le 30/06/2025 à 13:41, Jürgen Spitzmüller a écrit : Am Montag, dem 30.06.2025 um 12:30 +0200 schrieb Scott Kostyshak: Can we ever trigger a case where the values are within a few epsilons of difference apart? :-) Interesting. So you are suggesting that even if we remove the entire condition "l

Re: Float equality warnings

2025-06-30 Thread Jean-Marc Lasgouttes
Le 30/06/2025 à 12:30, Scott Kostyshak a écrit : On Sun, Jun 29, 2025 at 06:04:44PM +0100, José Matos wrote: On Sun, 2025-06-29 at 15:39 +0200, Scott Kostyshak wrote: Should we use some kind of tolerance value, e.g., based on std::numeric_limits::epsilon and std::numeric_limits::epsilon ? /hom

Re: What do you think about shadow warnings for LyX code by compilers [-Wshadow] ?

2025-06-28 Thread Jean-Marc Lasgouttes
Le 28/06/2025 à 18:22, Richard Kimberly Heck a écrit : On 6/28/25 9:07 AM, Scott Kostyshak wrote: I'm wondering if you think it would be a (even trivial) improvement to fix warnings about shadowing. In my experience, shadowed declarations can give rise to hard to find bugs. So I'd be happy to

Re: [LyX/master] Memory overlap

2025-06-27 Thread Jean-Marc Lasgouttes
Le 27 juin 2025 13:45:43 GMT+02:00, Kornel Benko a écrit : >commit 9d527385c68ed73f0b2635671d9fca736a87b97e >Author: Kornel Benko >Date: Fri Jun 27 13:41:45 2025 +0200 > >Memory overlap > >Fix overlap spotted with '-fsanitize=address' >--- > src/support/gzstream.cpp | 4 +++- > 1 fil

Re: patch: have configure.py print to STDOUT, not STDERR

2025-06-27 Thread Jean-Marc Lasgouttes
Le 27 juin 2025 10:26:32 GMT+02:00, Scott Kostyshak a écrit : >Thanks, José for your follow-up commit. These last days I've been >focused mainly on compiler warnings. And also seeing if anything >interesting from fsanitize. > >Scott You may want to play with cppcheck (--enable=style, for example)

Re: [LyX/master] Upgrade boost to version 1.88.0

2025-06-25 Thread Jean-Marc Lasgouttes
Le 25/06/2025 à 23:56, Pavel Sanda a écrit : Well, why we should keep it? We consistently cut feature after feature as the most interesting cases already landed in standard C++ libs. How many new uses we came up during last 10 years? I am all for it. I will nuke all traces of boost. JMarc --

Re: [LyX/master] Upgrade boost to version 1.88.0

2025-06-25 Thread Jean-Marc Lasgouttes
Le 25/06/2025 à 18:03, Jean-Marc Lasgouttes a écrit : Le 25/06/2025 à 17:39, Pavel Sanda a écrit : In the following, I have simulated an assertion. This is what you see at the top. Then there is the back trace gibberish, and then the super useless BOOST_ASSERT output. Do I read it correctly

Re: [LyX/master] Upgrade boost to version 1.88.0

2025-06-25 Thread Jean-Marc Lasgouttes
Le 25/06/2025 à 16:28, Pavel Sanda a écrit : I was already checking in that direction :) But my shallow understanding is that - we get std::source_location only in C++20 - custom messages in C++26 (do we use them? parhaps not) - we lose current function information altogether. The last point loo

Re: [LyX/master] Upgrade boost to version 1.88.0

2025-06-25 Thread Jean-Marc Lasgouttes
Le 25/06/2025 à 17:39, Pavel Sanda a écrit : In the following, I have simulated an assertion. This is what you see at the top. Then there is the back trace gibberish, and then the super useless BOOST_ASSERT output. Do I read it correctly that all we get from current implementation of BOOST_ASSE

Re: [LyX/master] Upgrade boost to version 1.88.0

2025-06-25 Thread Jean-Marc Lasgouttes
Le 25/06/2025 à 17:39, Pavel Sanda a écrit : I always tend to look into RETURN VALUE or EXIT STATUS of man page. This section currently does not exists for lyx, but this might be the right time to add it? OK, but where? I do not find so many of them in the man pages I looked at. JMarc -- lyx-d

Re: [LyX/master] Upgrade boost to version 1.88.0

2025-06-25 Thread Jean-Marc Lasgouttes
Le 25/06/2025 à 14:05, Jean-Marc Lasgouttes a écrit : commit 02760b3269a45bc21c7f32c44cacd7301f42217e Author: Jean-Marc Lasgouttes Date: Wed Jun 25 12:50:38 2025 +0200 Upgrade boost to version 1.88.0 Also remove boost/crc.hpp, which is not used anymore since 197ea636

Re: Opened preview overlaps text in 2.5 alpha

2025-06-25 Thread Jean-Marc Lasgouttes
Le 25/06/2025 à 15:23, Pavel Sanda a écrit : On Mon, Jun 23, 2025 at 05:52:29PM +1200, Andrew Parsloe wrote: In LyX 2.5.0-alpha, when a preview inset is opened by clicking on it, it can overlap surrounding text -- see the attachment. The problem arises when a different font is used within the pr

Re: [LyX/master] Return const references when it makes sense

2025-06-24 Thread Jean-Marc Lasgouttes
Le 24/06/2025 à 18:07, Richard Kimberly Heck a écrit : -    std::string docbookAttr() const; +    std::string docbookAttr() const { return docbook_attr_; } Lots of good catches. It seems that I cannot move the definition to the header and think about adding the & at the same time! The only

Re: [LyX/master] Use const reference in range-base for loop

2025-06-22 Thread Jean-Marc Lasgouttes
Le 22/06/2025 à 20:03, Jean-Marc Lasgouttes a écrit : commit a7910f75c754ef1a7b00bec15a247580625f3e72 Author: Jean-Marc Lasgouttes Date: Sun Jun 22 20:02:43 2025 +0200 Use const reference in range-base for loop Spotted by cppcheck. These 3 commits based on cppcheck output

Re: An instance of -Wnested-anon-types

2025-06-11 Thread Jean-Marc Lasgouttes
Le 07/06/2025 à 01:14, Scott Kostyshak a écrit : With Clang and -pedantic, I get the following: /home/scott/lyxbuilds/master-master/repo/src/TexRow.h:73:4: warning: anonymous types declared in an anonymous union are an extension [-Wnested-anon-types] 73 | struct {} b

Re: A few instances of -Wdeprecated-copy-with-user-provided-copy

2025-06-11 Thread Jean-Marc Lasgouttes
Le 07/06/2025 à 01:00, Scott Kostyshak a écrit : When I compile with -Wextra with Clang, I get the following: Actually, autotools already compiles with -Wextra, but -Wno-deprecated-copy is used to avoid tons of warnings for Qt < 5.13. I just pushed a series of commits that only disables the

Re: Minor Website Typo

2025-06-11 Thread Jean-Marc Lasgouttes
Le 11/06/2025 à 19:45, Joel Kulesza a écrit : All, I hope you're doing well.  I'm sorry to be silent for so long, but I've enjoyed following the list and many developments discussed within. Hi Joel, Nice to hear from you. I'm still happily using 2.3.2 but want to begin using the 2.4 series,

Re: Implementing \intertext and \shortintertext

2025-06-11 Thread Jean-Marc Lasgouttes
Le 11/06/2025 à 16:29, Koji Yokota a écrit : I tried implementation of warnings when a command is incompatible with particular math environments to see how it looks like. Personally, I like this as it makes broader range of users accessible to the commands. Also, it makes a similar check possi

Re: An instance of -Wnested-anon-types

2025-06-09 Thread Jean-Marc Lasgouttes
Le 07/06/2025 à 01:14, Scott Kostyshak a écrit : With Clang and -pedantic, I get the following: /home/scott/lyxbuilds/master-master/repo/src/TexRow.h:73:4: warning: anonymous types declared in an anonymous union are an extension [-Wnested-anon-types] 73 | struct {} b

Re: A few instances of -Wdeprecated-copy-with-user-provided-copy

2025-06-09 Thread Jean-Marc Lasgouttes
Le 07/06/2025 à 01:00, Scott Kostyshak a écrit : When I compile with -Wextra with Clang, I get the following: /home/scott/lyxbuilds/master-master/repo/src/Bullet.h:53:11: warning: definition of implicit copy constructor for 'Bullet' is deprecated because it has a user-provided copy assignment

Re: 2.4.4 Tarballs

2025-06-07 Thread Jean-Marc Lasgouttes
Le 6 juin 2025 00:00:35 GMT+02:00, Richard Kimberly Heck a écrit : >Here: > >http://ftp.lyx.org/pub/lyx/devel/lyx-2.4/ > >Please prepare binaries! > Hi Rini, I think that you forgot to tag 2.4.4 and to reset version back to 2.4.5. JMarc -- lyx-devel mailing list lyx-devel@lists.lyx.org https

Re: LyX 2.5-alpha1 Release

2025-06-07 Thread Jean-Marc Lasgouttes
Le 7 juin 2025 04:51:38 GMT+02:00, Andrew Parsloe a écrit  : >On 6/06/2025 9:28 am, Richard Kimberly Heck wrote: >> The LyX teams is pleased to announce the release of a PREVIEW VERSION of LyX >> 2.5 for TESTING PURPOSES ONLY. Tarballs and binaries for Windows and OSX can >> be found here: >> >

Re: Deprecated warning

2025-05-23 Thread Jean-Marc Lasgouttes
Le 19/05/2025 à 11:17, Kornel Benko a écrit : Using external hunspell, I see these warnings /usr2/src/lyx/lyx-git/src/HunspellChecker.cpp:372:9: warning: 'spell' is deprecated [-Wdeprecated-declarations] /usr2/src/lyx/lyx-git/src/HunspellChecker.cpp:434:35: warning: 'suggest' is deprecated [-W

Re: Implementing \intertext and \shortintertext

2025-05-22 Thread Jean-Marc Lasgouttes
Le 22/05/2025 à 18:28, Pavel Sanda a écrit : On Thu, May 22, 2025 at 05:51:26PM +0200, Jean-Marc Lasgouttes wrote: Le 22/05/2025 ?? 16:23, Pavel Sanda a écrit : On Thu, May 22, 2025 at 10:58:29PM +0900, Koji Yokota wrote: I???m not sure if this is the good timing to propose, but I???d like to

Re: Implementing \intertext and \shortintertext

2025-05-22 Thread Jean-Marc Lasgouttes
Le 22/05/2025 à 16:23, Pavel Sanda a écrit : On Thu, May 22, 2025 at 10:58:29PM +0900, Koji Yokota wrote: I???m not sure if this is the good timing to propose, but I???d like to introduce a support for \intertext and \shortintertext commands in an appropriate LyX version. I have a patch for it

Re: Implementing \intertext and \shortintertext

2025-05-22 Thread Jean-Marc Lasgouttes
Le 22/05/2025 à 16:23, Pavel Sanda a écrit : On Thu, May 22, 2025 at 10:58:29PM +0900, Koji Yokota wrote: I???m not sure if this is the good timing to propose, but I???d like to introduce a support for \intertext and \shortintertext commands in an appropriate LyX version. I have a patch for it

Re: Implementing \intertext and \shortintertext

2025-05-22 Thread Jean-Marc Lasgouttes
Le 22/05/2025 à 15:58, Koji Yokota a écrit : Dear All, I’m not sure if this is the good timing to propose, but I’d like to introduce a support for \intertext and \shortintertext commands in an appropriate LyX version. I have a patch for it as attached to this mail which is also available on a

Re: Crash while searching (and displaying table?) in master

2025-05-19 Thread Jean-Marc Lasgouttes
Le 16/05/2025 à 21:11, Pavel Sanda a écrit : On Wed, May 14, 2025 at 11:52:24AM +0200, Jean-Marc Lasgouttes wrote: I can reproduce this issue. Does the patch here work for you? I can confirm that your patch fixes the crash here. I am actually frustrated because I do not really understand

Warning: dvipng failed to generate images from lyxpreviewSZhsFp.dvi... fallback to legacy method

2025-05-19 Thread Jean-Marc Lasgouttes
Dear all, With UserGuide, I get a lot of output concerning previews: Warning: dvipng failed to generate images from lyxpreviewSZhsFp.dvi... fallback to legacy method Warning: epstopdf failed on page 1, file lyxpreviewSZhsFp.001 Warning: epstopdf failed on page 2, file lyxpreviewSZhsFp.002 [...

Re: Towards #13132 - Extend the default length of pdftex log lines

2025-05-17 Thread Jean-Marc Lasgouttes
Le 17 mai 2025 22:07:46 GMT+02:00, Pavel Sanda a écrit : >Hi, > >I would like to get your feedback wrt #13132. > >TeXLive > 2020 implements option for extending the default length of >la/pdf/lua/xe/tex log lines >(e.g. pdflatex -cnf max_print_line=999) which would be helpful - our log >parser wo

Re: How to create an empty macro?

2025-05-16 Thread Jean-Marc Lasgouttes
Le 16 mai 2025 19:40:07 GMT+02:00, Pavel Sanda a écrit : >I'd keep CT only in master. We do not coordinate translations across minor >releases and there is a high chance the CT-on in branch manual would make it >into actual release, because no one is checking. Ok, I'll fix that. JMarc -- lyx

Re: Crash while searching (and displaying table?) in master

2025-05-16 Thread Jean-Marc Lasgouttes
Le 16 mai 2025 21:11:59 GMT+02:00, Pavel Sanda a écrit : >Don't know if it's related, but with your fix, the second time I hit Find >button >the crash is not there anymore, but the Idx inset uncollapses with broken >geometry -- the line gets broken. That might cause subsequent problems? I saw t

Re: Compilation Failure in Master

2025-05-16 Thread Jean-Marc Lasgouttes
Le 16/05/2025 à 16:31, Kornel Benko a écrit : External libs: Yes, it was anyway the default for prerelease|release build-types. While debugging the default was OFF for development|gprof build-types. Now the default would be always ON except for Windows. This was also done with autoconf when gcc

Re: Compilation Failure in Master

2025-05-16 Thread Jean-Marc Lasgouttes
Le 16/05/2025 à 15:40, Scott Kostyshak a écrit : From what I understand, this would set BOOST, Z, DTL, HUNSPELL, and ICONV to default to ON? I don't have any personal preference since it's easy for me to change my build script accordingly, but I don't know if JMarc's original argument of being

Re: segfault in master

2025-05-16 Thread Jean-Marc Lasgouttes
Le 16/05/2025 à 12:00, Pavel Sanda a écrit : On Thu, May 15, 2025 at 04:44:43PM -0400, Alexander Dunlap wrote: I have been trying out the master branch and am somewhat frequently getting segfaults with the error below in the dialogue box. It often seems to happen when typing math commands that L

Re: Compilation Failure in Master

2025-05-16 Thread Jean-Marc Lasgouttes
Le 16/05/2025 à 10:13, Kornel Benko a écrit : Am Thu, 15 May 2025 17:50:26 +0200 schrieb Scott Kostyshak : Do you mean CMake should turn set LYX_EXTERNAL_ICONV passed on the version of the compiler? I still don't know why it succeeds for me (with the default of LYX_EXTERNAL_ICONV=OFF) and not

Re: Compilation Failure in Master

2025-05-15 Thread Jean-Marc Lasgouttes
Le 15/05/2025 à 17:50, Scott Kostyshak a écrit : I still don't know why it succeeds for me (with the default of LYX_EXTERNAL_ICONV=OFF) and not for Riki. Could be different compiler versions or could be different othe CMake flags. Kornel does it compile successfully (although with warnings) for y

Re: Compilation Failure in Master

2025-05-15 Thread Jean-Marc Lasgouttes
Le 15/05/2025 à 12:31, Scott Kostyshak a écrit : On Thu, May 15, 2025 at 08:49:32AM +0200, Jean-Marc Lasgouttes wrote: Le 15 mai 2025 01:46:37 GMT+02:00, Richard Kimberly Heck a écrit : I was just using defaults: mkdir build cd buildcmake ..; and that's it. If I set LYX_EXTERNAL_ICO

Re: Compilation Failure in Master

2025-05-14 Thread Jean-Marc Lasgouttes
Le 15 mai 2025 01:46:37 GMT+02:00, Richard Kimberly Heck a écrit : > >I was just using defaults: > >mkdir build >cd buildcmake ..; and that's it. If I set LYX_EXTERNAL_ICONV to ON, then it >builds. The error seems to be in building iconv itself. > We ship libinconv 1.15. It might be that the la

Re: How to create an empty macro?

2025-05-14 Thread Jean-Marc Lasgouttes
Le 14/05/2025 à 17:06, Richard Kimberly Heck a écrit : On 5/14/25 9:32 AM, Jean-Marc Lasgouttes wrote: Le 13/05/2025 à 17:45, Richard Kimberly Heck a écrit : I'd just remove it. This is way too complicated for the User Guide anyway. I did that. Should I backport to branch? What am I sup

Re: How to create an empty macro?

2025-05-14 Thread Jean-Marc Lasgouttes
Le 13/05/2025 à 17:45, Richard Kimberly Heck a écrit : I'd just remove it. This is way too complicated for the User Guide anyway. I did that. Should I backport to branch? What am I supposed to do in terms of change tracking ? JMarc -- lyx-devel mailing list lyx-devel@lists.lyx.org https://li

Re: Crash while searching (and displaying table?) in master

2025-05-14 Thread Jean-Marc Lasgouttes
8ae346c2a5abfc93c6d68cf147df05fd4d9104e7 Mon Sep 17 00:00:00 2001 From: Jean-Marc Lasgouttes Date: Wed, 14 May 2025 11:38:43 +0200 Subject: [PATCH] Fix crashes related to simple search and metrics update These crashes are a consequence of commit 7f85024f, which improved performance by not caching position of insets

Re: How to create an empty macro?

2025-05-12 Thread Jean-Marc Lasgouttes
Le 12/05/2025 à 16:57, Richard Kimberly Heck a écrit : On 5/12/25 5:47 AM, Jean-Marc Lasgouttes wrote: Hi there, I tried to replace the example in section 6.9 Branches of the user's guide by proper math macros, because the example as it is breaks previews. So I would like to d

How to create an empty macro?

2025-05-12 Thread Jean-Marc Lasgouttes
Hi there, I tried to replace the example in section 6.9 Branches of the user's guide by proper math macros, because the example as it is breaks previews. So I would like to define \newcommand[1]{\question}{#1} \newcommand[1]{\answer}{} However, I cannot construct a math macro that does not us

Re: TextMetrics error

2025-05-11 Thread Jean-Marc Lasgouttes
Le 11/05/2025 à 01:13, Scott Kostyshak a écrit : TextMetrics.cpp (1157): MacroContext not initialised! Going through the buffer again and hope the context is better then. Please report This is fixed now. Looks good. I extended the fix to moving rows/columns at a7f81701. Thanks. Basical

Re: Scrolling assertion with big ERT inset

2025-05-10 Thread Jean-Marc Lasgouttes
Le 08/05/2025 à 13:39, Scott Kostyshak a écrit : In the attached document, if I scroll with mouse wheel or two-finger scroll (not with dragging the scrollbar) down to the end of the document, and then I two-finger scroll up to the beginning, I get an assertion at some point along the way. Can any

Re: TextMetrics error

2025-05-10 Thread Jean-Marc Lasgouttes
Le 09/05/2025 à 16:20, Jean-Pierre Chrétien a écrit : In fact, I have it readable in the local git repo : jp@pelican:/ext/lyx/master$ ls -l lib/unicode_alphanum_variants -rw-r--r-- 1 jp jp 8675 25 mars  17:55 lib/unicode_alphanum_variants but not in the local/share tree : $ find /usr/local/sha

Re: TextMetrics error

2025-05-10 Thread Jean-Marc Lasgouttes
Le 06/05/2025 à 11:13, Scott Kostyshak a écrit : On Tue, May 06, 2025 at 08:33:36AM +0200, Jean-Marc Lasgouttes wrote: Le 01/05/2025 à 18:28, Jean-Pierre Chrétien a écrit : Excellent. I made it visible to understand when it triggers. Can you describe what you do? The message occurs when I

Re: Another icon set?

2025-05-09 Thread Jean-Marc Lasgouttes
Le 12/03/2025 à 14:18, Jürgen Spitzmüller a écrit : Am Mittwoch, dem 12.03.2025 um 13:39 +0100 schrieb Jean-Marc Lasgouttes: I see. There are also the icons of the outliner that stand out. But they should be taken from the set! But it is very nice (although I'll need time to get used

Re: OSU OSL possible shutdown

2025-05-09 Thread Jean-Marc Lasgouttes
Le 03/05/2025 à 22:05, Pavel Sanda a écrit : It seem that recent US budget cuts might affect our infra provider and we'll need to think where to move next... https://osuosl.org/blog/osl-future/ What a nightmare... For their sake and ours, let's hope that they can survive this tempest. JMarc

Re: Scrolling assertion with big ERT inset

2025-05-09 Thread Jean-Marc Lasgouttes
Le 08/05/2025 à 13:39, Scott Kostyshak a écrit : In the attached document, if I scroll with mouse wheel or two-finger scroll (not with dragging the scrollbar) down to the end of the document, and then I two-finger scroll up to the beginning, I get an assertion at some point along the way. Can any

Re: TextMetrics error

2025-05-09 Thread Jean-Marc Lasgouttes
Le 06/05/2025 à 11:30, Jean-Pierre Chrétien a écrit : I tried to insert online math, I get this message (unrelated, I guess): Could not find variant symbols file This one is strange. Do you have a lib/unicode_alphanum_variants file? Yes I do not understand what happens, the code is very

Re: tex2lyx and svjour3

2025-05-07 Thread Jean-Marc Lasgouttes
Le 7 mai 2025 18:30:16 GMT+02:00, "Jürgen Spitzmüller" a écrit : >Yes, this problem has been reported several times, but it hasn't been >addressed yet. One solution would be to rename the layout (but I think >there are other layouts with the same problem), another to make tex2lyx >aware of the di

Re: TextMetrics error

2025-05-05 Thread Jean-Marc Lasgouttes
Le 01/05/2025 à 18:28, Jean-Pierre Chrétien a écrit : Excellent. I made it visible to understand when it triggers. Can you describe what you do? The message occurs when I open a table. Better now? I tried to insert online math, I get this message (unrelated, I guess): Could not find varia

Re: [LyX/master] Avoid code duplication related to PREEDIT

2025-05-01 Thread Jean-Marc Lasgouttes
Le 01/05/2025 à 03:26, Stephan Witt a écrit : thank you for the revert - I’ve checked the cursor movement now and it works again. I’m away from home and couldn't act earlier. The bug is indeed trivial (I found it know), I'll re-apply ASAP. JMarc -- lyx-devel mailing list lyx-devel@lists.lyx.

Re: TextMetrics error

2025-04-30 Thread Jean-Marc Lasgouttes
Le 30/04/2025 à 15:24, Jean-Pierre Chrétien a écrit : Dear devs After a fresh compilation of 2.5.0~devel with QT6, I see this : TextMetrics.cpp (1155): MacroContext not initialised! Going through the buffer again and hope the context is better then. Please report So I report :-) Excellent.

Re: [LyX/master] Avoid code duplication related to PREEDIT

2025-04-30 Thread Jean-Marc Lasgouttes
t line… > >BR Stephan > >> Am 29.04.2025 um 13:33 schrieb Jean-Marc Lasgouttes : >> >> commit 2637d81f313d1c94aafdbd8c687d3034e89f70df >> Author: Jean-Marc Lasgouttes >> Date: Tue Apr 29 13:29:54 2025 +0200 >> >>

Re: Search + replace math symbols?

2025-04-29 Thread Jean-Marc Lasgouttes
Le 29/04/2025 à 13:51, Kornel Benko a écrit : The issue is that regex does not identify them as regular chars, that is regex requires them backslashed. So setting them to values > 0xf while searching makes them 'normal'. What are the codepoints involved? In your addAccent thing (why the n

Re: LyX 2.5.0

2025-04-29 Thread Jean-Marc Lasgouttes
Le 29/04/2025 à 14:33, Pavel Sanda a écrit : Another thing I would like to propose to stop endlessly postponing major milestones. Our bug lists grew over reasonable size. In particular we could drop the milestone altogether if bug 1) already had 2.4.0 milestone in the past and 2) does not have

Re: Search + replace math symbols?

2025-04-29 Thread Jean-Marc Lasgouttes
Le 29/04/2025 à 12:02, Kornel Benko a écrit : Am Tue, 29 Apr 2025 08:28:22 + schrieb Bernt Lie via lyx-users : I'm on LyX 2.4.3 + Windows 64bit. I'm trying to replace the bullet ("\bullet") in [cid:image001.png@01DBB8F1.6CF55C50] with a star ("\star"). But the LyX Advanced Find and Replac

Re: Lyx-Devel crash with EmbeddedObjects.lyx

2025-04-29 Thread Jean-Marc Lasgouttes
Le 28/04/2025 à 22:14, Andreas Riedel a écrit : Dear developers, I compiled Lyx, the latest Git version, today under both Debian 11 and 12. I noticed a bug: When I open and close the EmbeddedObjects file, Lyx crashes reproducibly. Thanks for the report, Andreas. Jürgen, I suspect this is a

Re: [LyX/master] Fix handling of \int with external math tools

2025-04-29 Thread Jean-Marc Lasgouttes
Le 21/04/2025 à 12:57, Jean-Marc Lasgouttes a écrit : Le 21/04/2025 à 12:54, Jean-Marc Lasgouttes a écrit : commit 0ed7820843d2b4dd09240d6e6145ddb013d090b8 Author: Jean-Marc Lasgouttes Date:   Mon Apr 21 12:53:34 2025 +0200 Fix handling of \int with external math tools Riki, this is

Re: [LyX/master] Finish MathArray->MathData transition (1/4): rename CoordCache::arrays to cells

2025-04-28 Thread Jean-Marc Lasgouttes
Le 28/04/2025 à 14:41, Jean-Marc Lasgouttes a écrit : commit 0702580ce2532ca7fb2f79b6ae6f40f30bbabf8f Author: Jean-Marc Lasgouttes Date: Mon Apr 28 13:38:17 2025 +0200 Finish MathArray->MathData transition (1/4): rename CoordCache::arrays to cells This is a follow-up of

Re: LyX 2.5.0

2025-04-27 Thread Jean-Marc Lasgouttes
Le 27/04/2025 à 21:12, Richard Kimberly Heck a écrit : With this work, which is fantastic, it really seems to me like we are ready to start thinking about potentially considering maybe moving toward 2.5.0. Not to be too definitive about it! I agree that this is fantastic. Congrats to Jürgen fo

Re: [LyX/master] rename InsetMathHull member: label_ => labels_

2025-04-25 Thread Jean-Marc Lasgouttes
Le 25 avril 2025 21:45:52 GMT+02:00, Kornel Benko a écrit : >Done at 5cb026e5. > > Kornel Thanks Kornel. JMarc -- lyx-devel mailing list lyx-devel@lists.lyx.org https://lists.lyx.org/mailman/listinfo/lyx-devel

  1   2   3   4   5   6   7   8   9   10   >