After having upgraded to LyX 2.2.0, could it be useful to have also LyX 2.1.5?

[CarLaTeX]

If have already upgraded to LyX 2.2.0 on Windows 10 but I still have also
LyX 2.1.4 on my pc.

Could it be useful to upgrade my 2.1.4 to 2.1.5? For example, in case of
errors in 2.2.0, just to compare the behaviour in 2.1.5?
Or is it unadvisable?

Thank you, bye!
Carla

Re: lyx-cvs messages: show file names changed

[Scott Kostyshak]

On Sat, Apr 16, 2016 at 01:43:18PM -0400, Scott Kostyshak wrote:
> On Sat, Apr 16, 2016 at 08:35:54AM -0400, Richard Heck wrote:
> > On 04/15/2016 05:45 PM, Scott Kostyshak wrote:
> > > On Tue, Mar 29, 2016 at 06:09:56PM -0400, Scott Kostyshak wrote:
> > > > On Tue, Mar 29, 2016 at 10:49:55AM +0200, Jean-Marc Lasgouttes wrote:
> > > > > Le 28/03/2016 00:04, Scott Kostyshak a écrit :
> > > > > > It seems we currently use "git show ${hash}" in our script. Instead 
> > > > > > of
> > > > > > doing two commands like I naively suggested above, what about the
> > > > > > following?
> > > > > What about "git show --stat"?
> > > > Yes that is better than what I proposed. We should do
> > > > 
> > > >git show --patch-with-stat
> > > > 
> > > > so that both the patch and the stat are shown.
> > > Richard are you able to change this? I think it would mean changing the
> > > following in git_show:
> > > 
> > >   p = subprocess.Popen(['git', 'show', hash], stdout=subprocess.PIPE)
> > > 
> > > to
> > > 
> > >   p = subprocess.Popen(['git', 'show', '--patch-with-stat', hash], 
> > > stdout=subprocess.PIPE)
> > > 
> > > git show is also used in get_commit_info():, but if I understand
> > > correctly we don't need to change it there (I'm not confident about this
> > > though).
> > 
> > I will be able to do so, but I'll need to read this thread through and
> > understand exactly what we are trying to do.
> 
> OK it is not urgent.

Just pinging Richard in case he happens to find time to take a look.
It's still not urgent of course.

Scott


signature.asc
Description: PGP signature

Re: Fwd: Re: [LyX/2.2.x] Document change to behavior of charstyles.

[Richard Heck]

On 06/21/2016 05:12 PM, Jean-Pierre Chrétien wrote:
> Le 21/06/2016 20:55, Richard Heck a écrit :
>> commit 76bf4bbc9ea32c3ac2e43efa9da835666bc280ce
>> Author: Richard Heck 
>> Date:   Tue Jun 21 15:55:15 2016 -0400
>
>>
>> -You can download LyX 2.2.0 from https://www.lyx.org/Download/.
>> +You can download LyX 2.1.4 from http://www.lyx.org/Download/.
> ^
> 2.2.1 I guess

Yes, thanks.

rh

Fwd: Re: [LyX/2.2.x] Document change to behavior of charstyles.

[Jean-Pierre Chrétien]

Le 21/06/2016 20:55, Richard Heck a écrit :

commit 76bf4bbc9ea32c3ac2e43efa9da835666bc280ce
Author: Richard Heck 
Date:   Tue Jun 21 15:55:15 2016 -0400




-You can download LyX 2.2.0 from https://www.lyx.org/Download/.
+You can download LyX 2.1.4 from http://www.lyx.org/Download/.

^
2.2.1 I guess

--
Jean-Pierre

Re: LyX 2.2.0 Linux Compilation Error

[Kornel Benko]

Am Dienstag, 21. Juni 2016 um 20:35:55, schrieb Georg Baum 

> Joel Kulesza wrote:
>
> > Still the same error as before. :-( However, I appreciate your time
> > considering this!
>
> OK, then I have no idea anymore. Hopefully the new install will fix it.
>
>
> Georg

Maybe the attached mail from the cmake-list could be of some help?

Kornel

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


[CMake] VS2015 and toolset v110.mbox
Description: application/mbox

Re: Compiling with Microsoft Visual C++

[Georg Baum]

racoon wrote:

> On 20.06.2016 22:10, Georg Baum wrote:
>>
>> Good, that brings us a step further. I assume you have the Windows
>> Platform SDK installed? If not, please install it from
>> https://developer.microsoft.com/en-us/windows/downloads/windows-8-1-sdk
>> and try again. If it is installed, it might be broken:
>> http://stackoverflow.com/questions/22188919/windows-sdk-registry-variable-not-found/25466317
>> contains some hints how to repair a broken SDK installation. It might
>> also be a cmake problem (although I doubt that), since a user had exactly
>> the same issue with a cmake project here:
>> http://stackoverflow.com/questions/36362469/cmake-does-not-find-windows-sdk-8-1
> 
> I could not find the SDK in installed programs. So I installed it. But
> it still does not work. Same error. And I still can't find the SDK under
> programs. Perhaps they are somewhere else to be found?

The standalone SDK should be installed below C:\Program Files\Microsoft 
SDKs\ or C:\Program Files (x86)\Windows Kits\. If there is nothing below one 
of these folders then it is probably not installed.

Unfortunately the SDK stuff is quite confusing. Sometimes the SDK is 
included with MSVC, sometimes you need a standalone version. After some 
googling it looks like MSVC 2015 comes with a newer SDK that should be used, 
but the installion is optional. Some more things you could try:

- Make sure the MSVC included SDK is installed, as described at 
http://stackoverflow.com/questions/31619296/cmake-does-not-find-visual-c-compiler.
 Then create the MSVC project files with cmake again.
- In contrast to my earlier advice, select MSVC 2010 in cmake as generator, 
and have MSVC convert the projects during loading (this would help in case 
there is a MSVC 2015 related bug in cmake).


Georg

Re: LyX 2.2.0 Linux Compilation Error

[Georg Baum]

Joel Kulesza wrote:

> Still the same error as before. :-( However, I appreciate your time
> considering this!

OK, then I have no idea anymore. Hopefully the new install will fix it.


Georg

[ANNOUNCE] LyX 2.1.5 Released

[Richard Heck]

We are pleased to announce the release of LyX 2.1.5. This is the fifth
and final maintenance release in the 2.1.x series.

LyX 2.1.5 is the result of on-going efforts to make our stable version
even more reliable and stable. We have fixed a number of bugs and made a
number of improvements. Many of these were minor, but there were a few
crashes we fixed, too.These are detailed in the announcement
.

Please note that this is the final release in the 2.1.x series. It is
intended for users who for some reason do not wish to upgrade to 2.2.x
at this time. We strongly encourage all such users to upgrade to this
version. Most users should upgrade to 2.2.0.

Note also that LyX 2.1.5 is the last release that will support Windows
Vista. This is due to the fact that the Windows releases in the 2.2.x
series require Qt 5.6, which is not available for Vista.

One of the main features of 2.1.5 is its ability to read and write files
in the 2.2.x format. Since the conversion process back and forth usually
will not leave one with an identical file, however, we do not recommend
attempting to collaborate with users of 2.2.x. We strongly encourage all
LyX users to upgrade to this version.

You can download the sources in tar.gz format
 (also tar.xz
). There are
also some binaries  available.

Re: LyX 2.1.5 Tarballs

[Richard Heck]

On 06/19/2016 08:16 PM, Uwe Stöhr wrote:
> Am 13.06.2016 um 17:33 schrieb Richard Heck:
>
>>
>> The tarballs for the release of LyX 2.1.5 can be found here:
>> ftp://ftp.lyx.org/pub/lyx/devel/lyx-2.1/2.1.5/
>> Please prepare binaries.
>
> Here is the Windows build:
> http://ftp.lyx.de/LyX%202.1.5/
>
> Please add to the announcement that this will be the last LyX version
> with support for Windows Vista.

Done.

rh

Re: Fixed in Master + Milestone 2.3.0

[Richard Heck]

On 06/21/2016 09:23 AM, Guenter Milde wrote:
> On 2016-06-20, Richard Heck wrote:
>> Note that this is different from having the *resolution* be fixed: These 
>> would still have no resolution and so would still turn up the right way in 
>> searches. 
> Having a keyword "fixed" and a resolution with the same name is confusing.

Yes, it is, but I'm not sure it is easy to change it.

We can leave the rest as is.

Richard

Re: Layout Changes Inside CharStyles?

[Richard Heck]

On 06/21/2016 10:45 AM, José Abílio Matos wrote:
> On Tuesday, June 21, 2016 4:07:25 PM WEST Helge Hafting wrote:
>> I might use a charstyle to set large amounts (several paragraphs or an 
>> entire chapter) in a different color. I may still want headings & 
>> enumerations inside of that!
>>
>> It can perhaps be argued that I ought to use a paragraph style for this, and 
>> nest the other stuff inside. Currently, that is more cumbersome – especially 
>> the creation of a custom paragraph style.
> Then the solution here is to ease the creation of custom paragraph styles and 
> not to mix two different concepts.

You can also do a custom inset, which is the same as a character style,
except that it appears in a different place on the menu.

Note also that this proposal does not *prohibit* anything. It just makes
"ForcePlain 1" the default for charstyles.

Richard

Re: provides "package" 1

[Richard Heck]

On 06/21/2016 06:58 AM, PhilipPirrip wrote:
> I was wondering if there's a special reason why
> provides fontenc 1
> provides inputenc 1
> provides amsmath 1
>
> prevent the corresponding packages from loading, but
>
> provides fontspec 1
> provides listings 1
> provides polyglossia 1
>
> -for example- don't?
>
> Should I report a bug?

Yes, that's a bug. This should work for all LyX-loaded packages.

I think

diff --git a/src/BufferParams.cpp b/src/BufferParams.cpp
index 373eee5..592833a 100644
--- a/src/BufferParams.cpp
+++ b/src/BufferParams.cpp
@@ -1589,7 +1589,8 @@ bool BufferParams::writeLaTeX(otexstream & os,
LaTeXFeatures & features,
os << from_ascii(ams);
 
if (useNonTeXFonts) {
-   os << "\\usepackage{fontspec}\n";
+   if (!features.isProvided("fontspec"))
+   os << "\\usepackage{fontspec}\n";
if (features.mustProvide("unicode-math")
&& features.isAvailable("unicode-math"))
os << "\\usepackage{unicode-math}\n";
@@ -2115,7 +2115,7 @@ bool BufferParams::writeLaTeX(otexstream & os,
LaTeXFeatures & features,
}
if (features.isRequired("bicaption"))
lyxpreamble += "\\usepackage{bicaption}\n";
-   if (!listings_params.empty() || features.isRequired("listings"))
+   if (!listings_params.empty() || features.mustProvide("listings"))
lyxpreamble += "\\usepackage{listings}\n";
if (!listings_params.empty()) {
lyxpreamble += "\\lstset{";

fixes the case of listings and fontspec. But I'd like to hear from
others, as I'm not expert on this.

I'm not sure how to handle polyglossia.

Richard

Re: Layout Changes Inside CharStyles?

[José Abílio Matos]

On Tuesday, June 21, 2016 4:07:25 PM WEST Helge Hafting wrote:
> Why do we generally not want to change paragraph style just because some 
> char style is in effect?

The purpose of the char styles is to be used for inline elements, that is 
elements that are placed inside a paragraph.

Currently every char style is reset after the paragraph end, and if necessary 
set at the begin of the new. Why should the custom char styles behave in a 
different way, or if they do it call it a different name.
 
> I might use a charstyle to set large amounts (several paragraphs or an 
> entire chapter) in a different color. I may still want headings & 
> enumerations inside of that!
> 
> It can perhaps be argued that I ought to use a paragraph style for this, 
> and nest the other stuff inside. Currently, that is more cumbersome – 
> especially the creation of a custom paragraph style.

Then the solution here is to ease the creation of custom paragraph styles and 
not to mix two different concepts.

> Please be very careful about introducing functional limitations. The 
> people who apply a layout change by mistake, can use the excellent 
> "undo" facility.
> 
> Helge Hafting

-- 
José Abílio

Re: Layout Changes Inside CharStyles?

[Helge Hafting]

Den 20. juni 2016 17:34, skrev Richard Heck:

At the moment, it is possible to effect a layout change inside almost
all of our character styles. So far as I can see, this is almost always
a bug. One does not want, e.g., to be able to switch to an enumeration
inside the Code character style. So I propose the following simple patch:

Why do we generally not want to change paragraph style just because some 
char style is in effect?


I might use a charstyle to set large amounts (several paragraphs or an 
entire chapter) in a different color. I may still want headings & 
enumerations inside of that!


It can perhaps be argued that I ought to use a paragraph style for this, 
and nest the other stuff inside. Currently, that is more cumbersome – 
especially the creation of a custom paragraph style.


Please be very careful about introducing functional limitations. The 
people who apply a layout change by mistake, can use the excellent 
"undo" facility.


Helge Hafting

Re: Fixed in Master + Milestone 2.3.0

[Guenter Milde]

On 2016-06-20, Richard Heck wrote:

> I propose that bugs meeting that description (such as #4398) simply be
> marked as having status "Fixed". The "Fixed in Master" status seems
> misleading: 

For me, "fixed in master" is a clear description of the status while
"fixed" would be misleading (because it is only fixed in the development
version).

> They are not going to get fixed in stable. If we mark them
> Fixed, it can mean: We are done with them.

> Note that this is different from having the *resolution* be fixed: These
> would still have no resolution and so would still turn up the right way
> in searches. 

Having a keyword "fixed" and a resolution with the same name is confusing.

> The only difference would be that "fixedinmaster" would
> mean: Fixed in master and (possibly, probably) to be fixed in stable.

For this distinction, I rather propose to use

a) the milestone (set it to 2.2.x, if you want the fix backported), or

b) a new keyword "backport" or similar.

Günter

Re: Compiling with Microsoft Visual C++

[racoon]

On 20.06.2016 22:10, Georg Baum wrote:

On 20.06.2016 10:14, racoon wrote:

On 19.06.2016 20:02, Georg Baum wrote:


Yes. Does cmake still not recognize the compiler if you call it from
that command window? In that case please send the CMakeError.log and
CMakeOutput.log files created by cmake, this would probably mean we
need to
change our cmake compiler detection.


Yes, no recognition still. Files attached.


Good, that brings us a step further. I assume you have the Windows Platform
SDK installed? If not, please install it from
https://developer.microsoft.com/en-us/windows/downloads/windows-8-1-sdk and
try again. If it is installed, it might be broken:
http://stackoverflow.com/questions/22188919/windows-sdk-registry-variable-not-found/25466317
 contains some hints how to repair a broken SDK
installation. It might also be a cmake problem (although I doubt that),
since a user had exactly the same issue with a cmake project here:
http://stackoverflow.com/questions/36362469/cmake-does-not-find-windows-sdk-8-1


I could not find the SDK in installed programs. So I installed it. But 
it still does not work. Same error. And I still can't find the SDK under 
programs. Perhaps they are somewhere else to be found?


Daniel

provides "package" 1

[PhilipPirrip]

I was wondering if there's a special reason why
provides fontenc 1
provides inputenc 1
provides amsmath 1

prevent the corresponding packages from loading, but

provides fontspec 1
provides listings 1
provides polyglossia 1

-for example- don't?

Should I report a bug?