Re: [PATCH] minor bibitem tweaks
Martin Vermeer wrote: Yes, I know... I can live with this for 1.4. I'll commit the minimalist version. This is clearly an improvement, everything else is ambiguous. Jürgen
Re: C++ question
Lars Gullik Bjønnes wrote: | I feel like a bit of an idiot asking this, but what's the C++ way to | insert some data into an existing file? How would you do it in any other language? At end is easy. At start is more tricky... If it were a memory buffer how would you do it then? (memmove perhaps?) You need to use some temporary (file or memory buffer) Unless you mmap the file, then you might be able to use memmove. Thanks, Lars. I suspected that was going to be the answer. Angus
Re: Windows installer --- software probe
--- Ursprüngliche Nachricht --- Von: Asger Ottar Alstrup [EMAIL PROTECTED] An: unknown Kopie: lyx-devel@lists.lyx.org Betreff: Re: Windows installer --- software probe Datum: Thu, 02 Jun 2005 19:45:06 +0200 One comment: The reason I did not install minsys was that I did not know which file to download and install. There are listed hundreds on that web-site. Maybe you could list what file the user should grab? Regards, Asger I guess you mean msys - minimal gnu system: MSYS-1.0.10.exe on http://www.mingw.org/download.shtml BTW: I'm w/o access to my mailbox until June, 23. Regards Carl -- Weitersagen: GMX DSL-Flatrates mit Tempo-Garantie! Ab 4,99 Euro/Monat: http://www.gmx.net/de/go/dsl
Re: Windows installer --- software probe
Carl Dr. Kleffner wrote: One comment: The reason I did not install minsys was that I did not know which file to download and install. There are listed hundreds on that web-site. Maybe you could list what file the user should grab? Regards, Asger I guess you mean msys - minimal gnu system: MSYS-1.0.10.exe on http://www.mingw.org/download.shtml That's correct. I want the user to grab msys-1.0.x.exe which they can get from http://prdownloads.sf.net/mingw/ (the download.shtml page takes you to the sf.net pages when you select some file for download.) I've changed the label on the radio button to: Download MinSYS (MSYS-1.0.X.exe) which I trust will allow the user to grab the correct binary. Angus
Re: Windows installer --- software probe
Angus == Angus Leeming [EMAIL PROTECTED] writes: Angus Carl Dr. Kleffner wrote: One comment: The reason I did not install minsys was that I did not know which file to download and install. There are listed hundreds on that web-site. Maybe you could list what file the user should grab? Regards, Asger I guess you mean msys - minimal gnu system: MSYS-1.0.10.exe on http://www.mingw.org/download.shtml Angus That's correct. I want the user to grab msys-1.0.x.exe which Angus they can get from http://prdownloads.sf.net/mingw/ (the Angus download.shtml page takes you to the sf.net pages when you Angus select some file for download.) What about this instead? http://sourceforge.net/project/showfiles.php?group_id=2435package_id=82721release_id=158803 I think it is much simpler for newcomers. The only thing I am not sure about is whether the release_id tag is stable with new versions. JMarc
Re: Windows installer --- software probe
Jean-Marc Lasgouttes wrote: What about this instead? http://sourceforge.net/project/showfiles.php?group_id=2435package_id=82721release_id=158803 I think it is much simpler for newcomers. The only thing I am not sure about is whether the release_id tag is stable with new versions. Thanks, Jean-Marc. Let's do this for LyX 1.3.6 anyway. We can always reconsider for future releases. Angus
Re: [PATCH 1.4] de.po update
Michael == Michael Schmitt [EMAIL PROTECTED] writes: Michael Hello, here comes another update for de.po. Please apply. Done. JMarc
Re: [Patch] Re: [patch] key event queue
Martin Vermeer [EMAIL PROTECTED] writes: | On Wed, Jun 01, 2005 at 03:12:05PM +0300, Martin Vermeer wrote: On Wed, Jun 01, 2005 at 01:31:55PM +0200, Lars Gullik Bjønnes wrote: Martin Vermeer [EMAIL PROTECTED] writes: | ... | OK, that settles it then for me: put in the event queue and reverse the new stuff in workAreaKeyPress to a simple showCursor call. Then everything will be consistent, if slow. I verified that this works properly using an old version of your event queue patch: doesn't lose keystrokes (except formally repeated ones), doesn't change their order, and doesn't produce cursor artefacts. Please commit your current version. | Lars? Please beat me too it... Too many balls in the air right now... (send the patch here first, and I'll do a quick review) -- Lgb
Re: [PATCH] Bug 961: reintroduce LFUN_BIBDB_ADD/DEL
Juergen == Juergen Spitzmueller [EMAIL PROTECTED] writes: Juergen Juergen Spitzmueller wrote: I'm not even sure I understand what LABEL_GOTO does (or is it broken?) Juergen OK, now I do. This is really odd. Odd in which sense? JMarc
Re: [PATCH] Bug 961: reintroduce LFUN_BIBDB_ADD/DEL
Juergen == Juergen Spitzmueller [EMAIL PROTECTED] writes: Juergen Juergen Spitzmueller wrote: I'm not even sure I understand what LABEL_GOTO does (or is it broken?) OK, now I do. This is really odd. Juergen I think this menu entry is extremely irritating. How about Juergen the following patch, that does enable the entry only when it Juergen actually makes sense? Yes, definitely good, although I would write: + case LFUN_LABEL_GOTO: { + flag.enabled(!cmd.argument.empty() + ||getInsetByCodeInsetRef(cursor_, InsetBase::REF_CODE)); + break; JMarc
Re: Script alternative - Ch
Angus == Angus Leeming [EMAIL PROTECTED] writes: Angus I really don't see why such a mini-dist would be an advantage. Angus What will happen is what has happened with Ruurd's equivalent Angus packaging of python. It fails to work as people add more Angus functionality to the scripts. Moreover, it will become one Angus more thing to maintain. It depends on whether we can automate the process, I guess. In this case it is just a new makefile target. JMarc
Windows installer finished
Find it at http://www.devel.lyx.org/~leeming/lyx_setup_136.exe It's 5.9MB in size. It's fully internationalized although its messages still need to be translated. If you're interested in doing so, you can grab the source from http://www.devel.lyx.org/~leeming/packaging.tar.bz2 It's 82kB in size. Have a look at installer/lyx_languages/{french,german}.nsh which need translated versions of the 53 strings. Adding other languages is straightforward. The archive unpacks to packaging/ README clean_dvi.py package_lyxwin.sh dtl/ README Makefile dt2dv.c dt2dv.man dtl.doc dtl.h dv2dt.c dv2dt.man dvi.doc edited.txt example.tex hello.tex man2ps tripvdu.tex icons/ lyx_32x32_16colors.png lyx_32x32.png lyx_doc_32x32_16colors.png lyx_doc_32x32.png lyx_doc.svg lyx.svg lyx_32x32.ico lyx_doc_32x32.ico installer/ lyx_languages/ english.nsh french.nsh german.nsh abi_util_fileassoc.nsh download.nsh ioDownload.ini ioSummary.ini lyx_installer.nsi lyx_path_prefix.C is_user_admin.nsh lyxfunc.nsh strtrim.nsh Angus
Re: Windows installer finished
Angus Leeming wrote: The archive unpacks to I propose to add this tree as-is to development/Win32. OK? Angus packaging/ README clean_dvi.py package_lyxwin.sh dtl/ README Makefile dt2dv.c dt2dv.man dtl.doc dtl.h dv2dt.c dv2dt.man dvi.doc edited.txt example.tex hello.tex man2ps tripvdu.tex icons/ lyx_32x32_16colors.png lyx_32x32.png lyx_doc_32x32_16colors.png lyx_doc_32x32.png lyx_doc.svg lyx.svg lyx_32x32.ico lyx_doc_32x32.ico installer/ lyx_languages/ english.nsh french.nsh german.nsh abi_util_fileassoc.nsh download.nsh ioDownload.ini ioSummary.ini lyx_installer.nsi lyx_path_prefix.C is_user_admin.nsh lyxfunc.nsh strtrim.nsh
Re: [Patch] Re: [patch] key event queue
On Mon, 2005-06-06 at 11:46, Lars Gullik Bjnnes wrote: Martin Vermeer [EMAIL PROTECTED] writes: | On Wed, Jun 01, 2005 at 03:12:05PM +0300, Martin Vermeer wrote: On Wed, Jun 01, 2005 at 01:31:55PM +0200, Lars Gullik Bjnnes wrote: Martin Vermeer [EMAIL PROTECTED] writes: | ... | OK, that settles it then for me: put in the event queue and reverse the new stuff in workAreaKeyPress to a simple showCursor call. Then everything will be consistent, if slow. I verified that this works properly using an old version of your event queue patch: doesn't lose keystrokes (except formally repeated ones), doesn't change their order, and doesn't produce cursor artefacts. Please commit your current version. | Lars? Please beat me too it... Too many balls in the air right now... (send the patch here first, and I'll do a quick review) Attached. - Martin Index: ChangeLog === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/ChangeLog,v retrieving revision 1.2193 diff -u -p -r1.2193 ChangeLog --- ChangeLog 6 Jun 2005 08:30:19 - 1.2193 +++ ChangeLog 6 Jun 2005 11:53:47 - @@ -1,3 +1,8 @@ +2005-06-06 Martin Vermeer [EMAIL PROTECTED] + + * BufferView_pimpl.C: revert to showCursor in connection with + Lars's front-end stack + 2005-06-06 JSpitzm [EMAIL PROTECTED] * text.C (redoParagraph): move cursor right after a bibitem Index: BufferView_pimpl.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/BufferView_pimpl.C,v retrieving revision 1.584 diff -u -p -r1.584 BufferView_pimpl.C --- BufferView_pimpl.C 31 May 2005 14:40:23 - 1.584 +++ BufferView_pimpl.C 6 Jun 2005 11:53:48 - @@ -520,12 +520,8 @@ void BufferView::Pimpl::workAreaKeyPress * of the cursor. Note we cannot do this inside * dispatch() itself, because that's called recursively. */ - if (available()) { - screen().prepareCursor(); - cursor_timeout.setTimeout(100); - cursor_timeout.restart(); - cursor_timeout.setTimeout(400); - } + if (available()) + screen().showCursor(*bv_); } Index: frontends/qt2/ChangeLog === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/ChangeLog,v retrieving revision 1.789 diff -u -p -r1.789 ChangeLog --- frontends/qt2/ChangeLog 3 Jun 2005 09:25:45 - 1.789 +++ frontends/qt2/ChangeLog 6 Jun 2005 11:53:48 - @@ -1,3 +1,10 @@ +2005-06-06 Lars Gullik Bjonnes [EMAIL PROTECTED] + (Committed Martin Vermeer [EMAIL PROTECTED]) + + * QContentPane.[Ch]: implement keystroke event stack with + repeat key regulation. Key are not sent to LyX core faster + than rendering can handle. + 2005-06-03 JSpitzm [EMAIL PROTECTED] * QContentPane.[Ch]: Limit the QIM*Events to X11. They are not Index: frontends/qt2/QContentPane.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QContentPane.C,v retrieving revision 1.34 diff -u -p -r1.34 QContentPane.C --- frontends/qt2/QContentPane.C 3 Jun 2005 09:25:46 - 1.34 +++ frontends/qt2/QContentPane.C 6 Jun 2005 11:53:48 - @@ -19,7 +19,6 @@ #include qapplication.h #include qpainter.h -#include qtimer.h #include boost/bind.hpp @@ -90,6 +89,8 @@ QContentPane::QContentPane(QWorkArea * p boost::bind(QContentPane::generateSyntheticMouseEvent, this)); + connect(step_timer_, SIGNAL(timeout()), SLOT(keyeventTimeout())); + setFocusPolicy(QWidget::WheelFocus); setFocus(); setCursor(ibeamCursor); @@ -101,6 +102,9 @@ QContentPane::QContentPane(QWorkArea * p // stupid moc strikes again connect(wa_-scrollbar_, SIGNAL(valueChanged(int)), this, SLOT(scrollBarChanged(int))); + + // Start the timer, one-shot. + step_timer_.start(25, true); } @@ -251,9 +255,35 @@ void QContentPane::wheelEvent(QWheelEven void QContentPane::keyPressEvent(QKeyEvent * e) { + keyeventQueue_.push(boost::shared_ptrQKeyEvent(new QKeyEvent(*e))); +} + + +void QContentPane::keyeventTimeout() +{ + bool handle_autos = true; + + while (!keyeventQueue_.empty()) { + boost::shared_ptrQKeyEvent ev = keyeventQueue_.front(); + + // We never handle more than one auto repeated + // char in a list of queued up events. + if (!handle_autos ev-isAutoRepeat()) { + keyeventQueue_.pop(); + continue; + } + boost::shared_ptrQLyXKeySym sym(new QLyXKeySym); - sym-set(e); - wa_-workAreaKeyPress(sym, q_key_state(e-state())); + sym-set(ev.get()); + + wa_-workAreaKeyPress(sym, q_key_state(ev-state())); + keyeventQueue_.pop(); + + handle_autos = false; + } + + // Restart the timer. + step_timer_.start(25, true); } Index: frontends/qt2/QContentPane.h === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QContentPane.h,v retrieving revision 1.16 diff -u -p -r1.16 QContentPane.h
Re: [Patch] Re: [patch] key event queue
Martin Vermeer [EMAIL PROTECTED] writes: I am fine with this. (I'd better be...) -- Lgb
[patch] fix bug 1665
Attached is a tested, working 1.3 patch that fixes bug 1665, following an idea of Jean-Marc (see http://bugzilla.lyx.org/show_bug.cgi?id=1665). I know that this is no crash, but it is very annoying for me (because we produce many of our documents via commandline export), so I fixed it nevertheless. I propose to apply this both to 1.3 and 1.4. OK? GeorgIndex: src/mathed/ChangeLog === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/mathed/ChangeLog,v retrieving revision 1.299.2.21 diff -u -p -r1.299.2.21 ChangeLog --- src/mathed/ChangeLog 26 Apr 2005 10:00:58 - 1.299.2.21 +++ src/mathed/ChangeLog 6 Jun 2005 12:06:10 - @@ -1,3 +1,10 @@ +2005-06-06 Georg Baum [EMAIL PROTECTED] + + * math_parser.h (latexkeys): new member requires + * math_factory.C (initSymbols): fill the requires field for fonts + * math_symbolinset.C (validate): use sym_-requires instead of + sym_-inset (fixes bug 1665) + 2005-04-26 Georg Baum [EMAIL PROTECTED] * math_cursor.C (interpret): treat as special character (fixes Index: src/mathed/math_factory.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/mathed/math_factory.C,v retrieving revision 1.70.2.2 diff -u -p -r1.70.2.2 math_factory.C --- src/mathed/math_factory.C 7 Dec 2004 10:49:53 - 1.70.2.2 +++ src/mathed/math_factory.C 6 Jun 2005 12:06:11 - @@ -144,6 +144,12 @@ void initSymbols() // tmp.inset _is_ the fontname here. // create fallbacks if necessary + // store requirements as long as we can + if (tmp.inset == msa || tmp.inset == msb) +tmp.requires = amssymb; + else if (tmp.inset == wasy) +tmp.requires = wasysym; + // symbol font is not available sometimes string symbol_font = lyxsymbol; Index: src/mathed/math_parser.h === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/mathed/math_parser.h,v retrieving revision 1.52.2.1 diff -u -p -r1.52.2.1 math_parser.h --- src/mathed/math_parser.h 7 Dec 2004 10:50:01 - 1.52.2.1 +++ src/mathed/math_parser.h 6 Jun 2005 12:06:11 - @@ -39,6 +39,8 @@ struct latexkeys { string extra; /// how is this called as XML entity? string xmlname; + /// required LaTeXFeatures + string requires; }; Index: src/mathed/math_symbolinset.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/mathed/math_symbolinset.C,v retrieving revision 1.52.2.4 diff -u -p -r1.52.2.4 math_symbolinset.C --- src/mathed/math_symbolinset.C 19 Jan 2005 11:46:53 - 1.52.2.4 +++ src/mathed/math_symbolinset.C 6 Jun 2005 12:06:11 - @@ -122,10 +122,8 @@ bool MathSymbolInset::takesLimits() cons void MathSymbolInset::validate(LaTeXFeatures features) const { - if (sym_-inset == msa || sym_-inset == msb) - features.require(amssymb); - else if (sym_-inset == wasy) - features.require(wasysym); + if (!sym_-requires.empty()) + features.require(sym_-requires); }
Re: Windows installer finished
Angus == Angus Leeming [EMAIL PROTECTED] writes: Angus Angus Leeming wrote: The archive unpacks to Angus I propose to add this tree as-is to development/Win32. Be careful with binary status of the files in cvs. Also, binary files mean that patch-1.3.6 will not work reliably. JMarc
Re: [patch] fix bug 1665
Georg == Georg Baum [EMAIL PROTECTED] writes: Georg Attached is a tested, working 1.3 patch that fixes bug 1665, Georg following an idea of Jean-Marc (see Georg http://bugzilla.lyx.org/show_bug.cgi?id=1665). I know that this Georg is no crash, but it is very annoying for me (because we produce Georg many of our documents via commandline export), so I fixed it Georg nevertheless. I propose to apply this both to 1.3 and 1.4. OK? This is OK for 1.3.x. JMarc
Re: [Patch] Re: [patch] key event queue
On Mon, 2005-06-06 at 15:17, Lars Gullik Bjnnes wrote: Martin Vermeer [EMAIL PROTECTED] writes: I am fine with this. (I'd better be...) Done. - Martin signature.asc Description: This is a digitally signed message part
Re: [PATCH] Bug 961: reintroduce LFUN_BIBDB_ADD/DEL
Jean-Marc Lasgouttes wrote: I'm not even sure I understand what LABEL_GOTO does (or is it broken?) Juergen OK, now I do. This is really odd. Odd in which sense? Let me put it like this: It's not what I would have expected (though I see its use). Jürgen
Re: Windows installer finished
Jean-Marc Lasgouttes wrote: Angus I propose to add this tree as-is to development/Win32. Be careful with binary status of the files in cvs. OK, but that warning applies only to these files in the icons directory. lyx_32x32_16colors.png lyx_32x32.png lyx_doc_32x32_16colors.png lyx_doc_32x32.png lyx_32x32.ico lyx_doc_32x32.ico Also, binary files mean that patch-1.3.6 will not work reliably. I don't follow. Could you explain further? -- Angus
Re: Windows installer finished
Angus == Angus Leeming [EMAIL PROTECTED] writes: Also, binary files mean that patch-1.3.6 will not work reliably. Angus I don't follow. Could you explain further? Binary file do not appear in patches. Therefore LyX 1.3.5 + patch-1.3.6 != LyX 1.3.6 JMarc
Re: Windows installer finished
Jean-Marc Lasgouttes [EMAIL PROTECTED] writes: Angus == Angus Leeming [EMAIL PROTECTED] writes: Also, binary files mean that patch-1.3.6 will not work reliably. | Angus I don't follow. Could you explain further? | Binary file do not appear in patches. Therefore | LyX 1.3.5 + patch-1.3.6 != LyX 1.3.6 That would require something like xdelta. -- Lgb
Re: [PATCH] Bug 961: reintroduce LFUN_BIBDB_ADD/DEL
Jean-Marc Lasgouttes wrote: Yes, definitely good, although I would write: + case LFUN_LABEL_GOTO: { + flag.enabled(!cmd.argument.empty() + ||getInsetByCodeInsetRef(cursor_, InsetBase::REF_CODE)); + break; I did so and committed. Thanks, Jürgen
Re: [PATCH] Bug 961: reintroduce LFUN_BIBDB_ADD/DEL
Juergen Spitzmueller wrote: http://bugzilla.lyx.org/show_bug.cgi?id=961 The attached patch reimplements those lfuns. It also fixes InsetBibtex::delDatabase, which is broken in 1.3. The only drawback is that the lfuns are not working when the cursor sits behind the inset. The same applies to 1.3. The reason is that getInsetByCode stops at the end of the document. Attached is a third (and IMHO final) version that fixes all drawbacks and additionally also fixes a problem in InsetBibtex::addDatabase, which did not add a database when its name was a substring of another database (e.g. mybib vs. mybib2). OK to apply? Jürgen Index: BufferView_pimpl.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/BufferView_pimpl.C,v retrieving revision 1.586 diff -u -r1.586 BufferView_pimpl.C --- BufferView_pimpl.C 6 Jun 2005 13:11:29 - 1.586 +++ BufferView_pimpl.C 6 Jun 2005 13:21:07 - @@ -51,6 +51,7 @@ #include undo.h #include vspace.h +#include insets/insetbibtex.h #include insets/insetref.h #include insets/insettext.h @@ -128,9 +129,20 @@ { T * inset = 0; DocIterator it = cur; - if (it.nextInset() - it.nextInset()-lyxCode() == code) { - inset = static_castT*(it.nextInset()); + T * first_inset = static_castT*(it.nextInset()); + while (it) { + if (!it.nextInset()) { + // try from the beginning, but break after one loop + it.pit() = 0; + it.pos() = 0; + if (!it.nextInset() || it.nextInset() == first_inset) +break; + } + if (it.nextInset()-lyxCode() == code) { + inset = static_castT*(it.nextInset()); + break; + } + it.forwardInset(); } return inset; } @@ -982,6 +994,8 @@ case LFUN_MARK_ON: case LFUN_SETMARK: case LFUN_CENTER: + case LFUN_BIBDB_ADD: + case LFUN_BIBDB_DEL: case LFUN_WORDS_COUNT: flag.enabled(true); break; @@ -1212,6 +1226,24 @@ case LFUN_CENTER: center(); break; + + case LFUN_BIBDB_ADD: { + InsetBibtex * inset = + getInsetByCodeInsetBibtex(cursor_, + InsetBase::BIBTEX_CODE); + if (inset) + inset-addDatabase(cmd.argument); + break; + } + + case LFUN_BIBDB_DEL: { + InsetBibtex * inset = + getInsetByCodeInsetBibtex(cursor_, + InsetBase::BIBTEX_CODE); + if (inset) + inset-delDatabase(cmd.argument); + break; + } case LFUN_WORDS_COUNT: { DocIterator from, to; Index: LyXAction.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/LyXAction.C,v retrieving revision 1.206 diff -u -r1.206 LyXAction.C --- LyXAction.C 8 May 2005 10:02:37 - 1.206 +++ LyXAction.C 6 Jun 2005 13:21:08 - @@ -188,6 +188,8 @@ { LFUN_INSERT_LABEL, label-insert, Noop }, { LFUN_INSET_OPTARG, optional-insert, Noop }, { LFUN_INSERT_BIBITEM, bibitem-insert, Noop }, + { LFUN_BIBDB_ADD, bibtex-database-add, Noop }, + { LFUN_BIBDB_DEL, bibtex-database-del, Noop }, { LFUN_INSERT_LINE, line-insert, Noop }, { LFUN_INSERT_PAGEBREAK, pagebreak-insert, Noop }, { LFUN_LANGUAGE, language, Noop }, Index: lfuns.h === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/lfuns.h,v retrieving revision 1.41 diff -u -r1.41 lfuns.h --- lfuns.h 8 May 2005 10:02:37 - 1.41 +++ lfuns.h 6 Jun 2005 13:21:08 - @@ -355,6 +355,8 @@ // 270 LFUN_WORDS_COUNT, LFUN_OUTPUT_CHANGES, // jspitzm 20050121 + LFUN_BIBDB_ADD, + LFUN_BIBDB_DEL, LFUN_LASTACTION // end of the table }; Index: insets/insetbibtex.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/insets/insetbibtex.C,v retrieving revision 1.54 diff -u -r1.54 insetbibtex.C --- insets/insetbibtex.C 17 May 2005 11:11:45 - 1.54 +++ insets/insetbibtex.C 6 Jun 2005 13:21:13 - @@ -271,7 +271,7 @@ bool InsetBibtex::addDatabase(string const db) { string contents(getContents()); - if (!contains(contents, db)) { + if (tokenPos(contents, ',', db) == -1) { if (!contents.empty()) contents += ','; setContents(contents + db); @@ -283,16 +283,17 @@ bool InsetBibtex::delDatabase(string const db) { - if (contains(getContents(), db)) { + string contents(getContents()); + if (contains(contents, db)) { + int const n = tokenPos(contents, ',', db); string bd = db; - int const n = tokenPos(getContents(), ',', bd); if (n 0) { - // Weird code, would someone care to explain this?(Lgb) - string tmp(, ); - tmp += bd; - setContents(subst(getContents(), tmp, , )); + // this is not the first database + string tmp = ',' + bd; + setContents(subst(contents, tmp, )); } else if (n == 0) - setContents(split(getContents(), bd, ',')); + // this is the first (or only) database + setContents(split(contents, bd, ',')); else return false; }
Re: Windows installer finished
Jean-Marc Lasgouttes wrote: Also, binary files mean that patch-1.3.6 will not work reliably. Angus I don't follow. Could you explain further? Binary file do not appear in patches. Therefore LyX 1.3.5 + patch-1.3.6 != LyX 1.3.6 `diff -a` ? That's true for more than just the files I've just added however. All of the following are flagged as binary in the 1.3.x tree: development/Win32/packaging/icons/lyx_32x32.ico development/Win32/packaging/icons/lyx_32x32.png development/Win32/packaging/icons/lyx_32x32_16colors.png development/Win32/packaging/icons/lyx_doc_32x32.ico development/Win32/packaging/icons/lyx_doc_32x32.png development/Win32/packaging/icons/lyx_doc_32x32_16colors.png development/lyxserver/perl/LyX-Client-0.01.tar.gz development/lyxserver/perl/LyX-Polite-0.01.tar.gz lib/images/banner.ppm Moreover, the problem will only get worse when the 498 .xpm files in lib/images are converted to .png format the day we drop XForms. -- Angus
Re:Windows installer finished
Angus Leeming wrote: Find it at http://www.devel.lyx.org/~leeming/lyx_setup_136.exe Hi Angus, first many thanks for your work. I tested it and have the following annotations: - Your installer asks for several programs. If one of these are not present, the installer opens the download/install webpage of the programs but jumps to the next step without waiting until the missing program is installed. - Is it really necessary to install msys, perl and phthon fully? Its download size is huge (23 MB for Perl + Python) for modem/ISDN users. Ruurd only used the executables perl.exe, python.exe and sh.exe. Is it possible to include them to the installer package? Because it would ease the installation a lot. - While LyX reconfigures, it doesn't recognize acrobat: +checking for acrobat... no altough it is installed on my system. The reason is that the acrobat.exe istn't in the PATH but in the registry (as most of Win programs) under HKEY_LOCAL_MACHINE\Software\Classes\Applications\ Is it possible to search the registry for this entry? Acroread under Win is stored in the same registry path. - I installed LyX with admin permissions and it appeared with german menus like I wanted. But when I change to a user account it appeared with english menus. To get german ones for the user account I had to create a new environment variable named LANG with the value de_DE as described at Ruurds LyX-webpage. Is it possible to add this variable automatically for every user account? - At last a crash report of your LyX 1.3.6CVS version: Go to Edit - Preferences, choose Paths and press one of the Browse buttons for the working directories. Result CRASH: lyx: SIGSEGV signal caught Sorry, you have found a bug in LyX. Please read the bug-reporting instructions i n Help-Introduction and send us a bug report, if necessary. Thanks ! Bye. abnormal program termination That's all for the moment. thanks for the installer again and regards Uwe p.s. Please CC all replies to my email adress as I'm not on the lyx-devel list.
Re: Windows installer finished
On Monday 06 June 2005 14:38, Uwe Sthr wrote: Angus Leeming wrote: Find it at http://www.devel.lyx.org/~leeming/lyx_setup_136.exe Hi Angus, first many thanks for your work. I tested it and have the following annotations: - Your installer asks for several programs. If one of these are not present, the installer opens the download/install webpage of the programs but jumps to the next step without waiting until the missing program is installed. Yes. But once you've grabbed everything it asks you to restart the LyX installer so that it can find your newly installed packages. - Is it really necessary to install msys, perl and phthon fully? No. You need Perl only if you're going to use reLyX. I tried to explain that in the accompanying commentary. Its download size is huge (23 MB for Perl + Python) for modem/ISDN users. Ruurd only used the executables perl.exe, python.exe and sh.exe. Is it possible to include them to the installer package? Just thinking out loud... Perhaps: [x] Download the official Python distribution [ ] Install a stripped-down and maybe good enough Python.exe But this will involve me in a lot more work with NSIS getting to grips with optional installation of components. I'm running out of LyX-time. I'm also running out of NSIS-enthusiasm. Because it would ease the installation a lot. - While LyX reconfigures, it doesn't recognize acrobat: +checking for acrobat... no altough it is installed on my system. The reason is that the acrobat.exe istn't in the PATH but in the registry (as most of Win programs) under HKEY_LOCAL_MACHINE\Software\Classes\Applications\ Is it possible to search the registry for this entry? Acroread under Win is stored in the same registry path. I could, but where do I stop? Acroread certainly isn't required to get LyX running and we already have 6 download pages and a complaint that we're downloading too much stuff... It's the sort of thing that really belongs in the configure script (which needs to be rewritten in some other language if it is to understand the registry...) - I installed LyX with admin permissions and it appeared with german menus like I wanted. But when I change to a user account it appeared with english menus. To get german ones for the user account I had to create a new environment variable named LANG with the value de_DE as described at Ruurds LyX-webpage. Is it possible to add this variable automatically for every user account? It may be possible, but I don't know how... - At last a crash report of your LyX 1.3.6CVS version: Go to Edit - Preferences, choose Paths and press one of the Browse buttons for the working directories. Result CRASH: Thank you. I see it on linux too. That's all for the moment. thanks for the installer again and regards Uwe p.s. Please CC all replies to my email adress as I'm not on the lyx-devel list. Angus
Re: Windows installer finished
Angus == Angus Leeming [EMAIL PROTECTED] writes: - While LyX reconfigures, it doesn't recognize acrobat: +checking for acrobat... no altough it is installed on my system. The reason is that the acrobat.exe istn't in the PATH but in the registry (as most of Win programs) under HKEY_LOCAL_MACHINE\Software\Classes\Applications\ Is it possible to search the registry for this entry? Acroread under Win is stored in the same registry path. Angus I could, but where do I stop? Acroread certainly isn't required Angus to get LyX running and we already have 6 download pages and a Angus complaint that we're downloading too much stuff... Angus It's the sort of thing that really belongs in the configure Angus script (which needs to be rewritten in some other language if Angus it is to understand the registry...) What should be done if possible is to check whether there is an association for .pdf files. In this case, start (with appropriate options) should be used as a viewer. Actually, it may be possible to do that on win 2k/xp from a shell script. Does NSIS offer support for querying such things? Another solution would be to have code in LyX packaging to do this kind of things (and for MacOSX too). JMarc
Re: Windows installer finished
On Monday 06 June 2005 14:38, Uwe Sthr wrote: - At last a crash report of your LyX 1.3.6CVS version: Go to Edit - Preferences, choose Paths and press one of the Browse buttons for the working directories. Result CRASH: Fixed in 1.3.x CVS. Angus
Re: Windows installer finished
On Monday 06 June 2005 14:38, you wrote: - Your installer asks for several programs. If one of these are not present, the installer opens the download/install webpage of the programs but jumps to the next step without waiting until the missing program is installed. - Is it really necessary to install msys, perl and phthon fully? Also, why not point the installer to the folder containing Ruurd's stripped down versions? Or something equivalent. Basically, I don't want to commit myself to creating, testing and maintaining these things. If some Windows user has the motivation, then that's great. I can certainly adapt the installer to open up a web page pointing to the stripped down version in addition to the official version. Angus
Re: Windows installer finished
Angus Leeming wrote: Its download size is huge (23 MB for Perl + Python) for modem/ISDN users. Ruurd only used the executables perl.exe, python.exe and sh.exe. Is it possible to include them to the installer package? Just thinking out loud... Perhaps: [x] Download the official Python distribution [ ] Install a stripped-down and maybe good enough Python.exe I prefer the latter one as it saves a lot of installation time and hard disk space. But why should it be installed? Isn't it possible to add an actual version of perl.exe, python.exe and sh.exe (and some dll's if needed) to the folder LyX\bin\ and use these files for lyxrc.defaults by default (as Ruurd did)? This solution shouldn't produce NSIS troubles. Is it possible to search the registry for this entry? Acroread under Win is stored in the same registry path. I could, but where do I stop? Acroread certainly isn't required to get LyX running and we already have 6 download pages and a complaint that we're downloading too much stuff... My proposal for python etc. above would reduce it to 3 download pages. And you don't need to ask for installing acrobat/acroread as it is installed on 99% of Win PC's. It is enough to check if the program is installed and add the excutable to the viewers for PDF in the preferences. --- I forgot to ask you how you handle the math fonts. I didn't see that they were installed. Just copy them to the \Windows\Fonts directory isn't enough, but I don't know how to invoke the correct installation by the command line. So it is OK if the user has to install them manually before installing LyX. Btw. Ekkehardt Schlicht created a special font package using the BaKoMa math fonts: http://wiki.lyx.org/uploads/LyX/WindowsSetup/bakoma4lyx.zip But they are not free for commercial use as written in the license. regards Uwe
Re: Windows installer finished
Jean-Marc Lasgouttes wrote: - While LyX reconfigures, it doesn't recognize acrobat: +checking for acrobat... no altough it is installed on my system. The reason is that the acrobat.exe istn't in the PATH but in the registry (as most of Win programs) under HKEY_LOCAL_MACHINE\Software\Classes\Applications\ Is it possible to search the registry for this entry? Acroread under Win is stored in the same registry path. Angus I could, but where do I stop? Acroread certainly isn't required Angus to get LyX running and we already have 6 download pages and a Angus complaint that we're downloading too much stuff... Angus It's the sort of thing that really belongs in the configure Angus script (which needs to be rewritten in some other language if Angus it is to understand the registry...) What should be done if possible is to check whether there is an association for .pdf files. In this case, start (with appropriate options) should be used as a viewer. See development/Win32/packaging/installer/lyx_path_prefix.C which adds shell code to the configure script to generate a path_prefix entry in lyxrc.defaults and which goes on to use start to run sh.exe configure. It would be quite simple to add more functions to this .dll. Perhaps: add_path_prefix() add_viewer() run_configure() but I'm getting a little edgy about all this. It appears that the goal is getting wider and wider. That means more work (primarily for me, it would appear), more chance of bugs and no great benefit AFAICS. I also don't think that we have to do everything this time around. Moreover, I think we're forgetting that our users are capable individuals who can do stuff themselves too... Actually, it may be possible to do that on win 2k/xp from a shell script. Does NSIS offer support for querying such things? Another solution would be to have code in LyX packaging to do this kind of things (and for MacOSX too). JMarc -- Angus
Re: Windows installer finished
Angus == Angus Leeming [EMAIL PROTECTED] writes: Angus See development/Win32/packaging/installer/lyx_path_prefix.C Angus which adds shell code to the configure script to generate a Angus path_prefix entry in lyxrc.defaults and which goes on to use Angus start to run sh.exe configure. That's evil, isn't it? ;) As far as adding prefs for windows is concerned, one possibility is to provide a preference file that contains everything needed by LyX/Win32. Generating this file is probably easier than hacking lib/configure. LyX/Mac follows this route (provide a preconfigured preference file). Angus but I'm getting a little edgy about all this. It appears that Angus the goal is getting wider and wider. That means more work Angus (primarily for me, it would appear), more chance of bugs and no Angus great benefit AFAICS. Yes. Angus I also don't think that we have to do everything this time Angus around. Moreover, I think we're forgetting that our users are Angus capable individuals who can do stuff themselves too... I think you have to draw a line at some point. Fell free to do it. JMarc
Re: Windows installer finished
Angus Leeming wrote: - Is it really necessary to install msys, perl and phthon fully? Also, why not point the installer to the folder containing Ruurd's stripped down versions? Or something equivalent. That's a good solution. Basically, I don't want to commit myself to creating, testing and maintaining these things. I understand you. I can certainly adapt the installer to open up a web page pointing to the stripped down version in addition to the official version. I'll create a wiki page with all needed files. I think we need - perl.exe - perlx.dll - phython.exe - pythonxx.dll - sh.exe What other files are needed? regards Uwe
Re: Windows installer finished
On Monday 06 June 2005 15:54, Uwe Sthr wrote: I can certainly adapt the installer to open up a web page pointing to the stripped down version in addition to the official version. I'll create a wiki page with all needed files. I think we need - perl.exe - perlx.dll - phython.exe - pythonxx.dll - sh.exe What other files are needed? That's precisely my point. I don't know. Off the top of my head: sed, tr, cut, grep... Angus
Re: Windows installer finished
On Monday 06 June 2005 15:39, Uwe Sthr wrote: Angus Leeming wrote: Its download size is huge (23 MB for Perl + Python) for modem/ISDN users. Ruurd only used the executables perl.exe, python.exe and sh.exe. Is it possible to include them to the installer package? Just thinking out loud... Perhaps: [x] Download the official Python distribution [ ] Install a stripped-down and maybe good enough Python.exe I prefer the latter one as it saves a lot of installation time and hard disk space. But why should it be installed? I think we have a terminology confusion. To me, installed just means unpack it somewhere usable. The registry is a [EMAIL PROTECTED] nightmare for which I see little need. Isn't it possible to add an actual version of perl.exe, python.exe and sh.exe (and some dll's if needed) to the folder LyX\bin\ and use these files for lyxrc.defaults by default (as Ruurd did)? This solution shouldn't produce NSIS troubles. This solution will require 1. Extensive testing to ascertain which modules are needed by the scripts that LyX packages. 2. Ongoing maintenance. Moreover, if I already have the official stuff installed why would I want to use the unoffical stuff? Adding code to NSIS to conditionally install something is most certainly a NSIS trouble that I'm keen to avoid right now. Is it possible to search the registry for this entry? Acroread under Win is stored in the same registry path. I could, but where do I stop? Acroread certainly isn't required to get LyX running and we already have 6 download pages and a complaint that we're downloading too much stuff... My proposal for python etc. above would reduce it to 3 download pages. No it wouldn't. It would just change three of them to: Would you prefer to use these stripped down things that people have been discovering don't work (see countleess reports on the lyx-users list), or would you prefer to use an official version which definitely will work but which requires a big download committment. And you don't need to ask for installing acrobat/acroread as it is installed on 99% of Win PC's. It is enough to check if the program is installed and add the excutable to the viewers for PDF in the preferences. I'd rather reuse the code I do have than rewrite something especially for acroroead... I forgot to ask you how you handle the math fonts. I didn't see that they were installed. Just copy them to the \Windows\Fonts directory isn't enough, but I don't know how to invoke the correct installation by the command line. So it is OK if the user has to install them manually before installing LyX. I do nothing about math fonts. Nor do I plan to do anything about math fonts before LyX 1.3.6 is released. I think that we should focus on getting what we have to work rather than trying to get it to do more and more and more. Btw. Ekkehardt Schlicht created a special font package using the BaKoMa math fonts: http://wiki.lyx.org/uploads/LyX/WindowsSetup/bakoma4lyx.zip But they are not free for commercial use as written in the license. I seem to remember that the consencus was that the licence was too restrictive for us to distribute them. Angus
Re: LyX meeting in Paris. What about July 14-18?
I already bought my tickets today, so most probably we will meet up in Paris #:O) ~Jürgen -- -._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._ Dr. Jürgen VignaE-Mail: [EMAIL PROTECTED] Mitterstrich 151/A I-39050 SteineggWeb: http://www.lyx.org/~jug -._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._
Re: Windows installer: NSIS vs .msi
Angus Leeming wrote: But this will involve me in a lot more work with NSIS getting to grips with optional installation of components. I'm running out of LyX-time. I'm also running out of NSIS-enthusiasm. While having any sort of installer is better than none at all, I'd suggest future effort re: Windows installers aim for an .msi target (Microsoft Installer) format. Then, those of us in AD environments can enjoy pain-free deployment to (all) client machines. (-: -- Rex p.s. I'll be doing it myself for lyx-1.3.6 as a proof-of-concept (using the Advanced Installer tool).
Re: LyX meeting in Paris. What about July 14-18?
Juergen == Juergen Vigna [EMAIL PROTECTED] writes: Juergen I already bought my tickets today, so most probably we will Juergen meet up in Paris #:O) ~Jürgen Great! When? JMarc
Re: LyX meeting in Paris. What about July 14-18?
What do you mean my flight dates/times? Arrival in Paris evening the 14th, departure evening the 18th :) I will check back with you as soon as I have all plans ready, for the moment I just booked the flight (77 ). ~Jrgen Jean-Marc Lasgouttes wrote: Juergen == Juergen Vigna [EMAIL PROTECTED] writes: Juergen I already bought my tickets today, so most probably we will Juergen meet up in Paris #:O) ~Jrgen Great! When? JMarc -- -._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._ Dr. Jrgen VignaE-Mail: [EMAIL PROTECTED] Mitterstrich 151/A I-39050 SteineggWeb: http://www.lyx.org/~jug -._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._
Re: Windows installer: NSIS vs .msi
Rex Dieter wrote: Angus Leeming wrote: But this will involve me in a lot more work with NSIS getting to grips with optional installation of components. I'm running out of LyX-time. I'm also running out of NSIS-enthusiasm. While having any sort of installer is better than none at all, I'd suggest future effort re: Windows installers aim for an .msi target (Microsoft Installer) format. Fine. Having struggled with NSIS for the last month, I would love to see it die :( However, I think that Asger's point that we should stick with something that's both Open Source and actively developed. Notice also that you might like MSI, but others think that NSIS is better: http://www.peterprovost.org/archive/2004/01/19/1040.aspx I'm noticing that the install and uninstall runs so much faster than the Installshield or VS.NET msi files that I have been using. I mentioned NSIS a few days/weeks ago in one of my Geek Notes columns. Because my customer was very size sensitive, the project that we shipped last week used NSIS for its installer. The VS.NET/MSI installer we created was almost 4MB while the NSIS installer was less than 520K. And like Matt, I find the NSIS installers are much faster. The downside is that these installers are not MSI based. In fact, the reason the VS.NET installers are so big is that they bundle two copies of the Windows Installer 2.0 (ANSI and Unicode) within the installer. If you switch the deployment project properties from Windows Installer bootstrapper to No bootstrapper, you can alleviate this, but you also can only run the installer on machines that are already running Windows Installer 2.0. Then, those of us in AD environments can enjoy pain-free deployment to (all) client machines. (-: At the cost of increased (quadrupled ?) download size. Can you explain why you can't just compile the NSIS installer on your weird windows box? How hard is $ path to/makensis lyx_installer.nsi ? p.s. I'll be doing it myself for lyx-1.3.6 as a proof-of-concept (using the Advanced Installer tool). Look forward to it. -- Angus
Re: Windows installer finished
Jean-Marc Lasgouttes [EMAIL PROTECTED] writes: | Angus I also don't think that we have to do everything this time | Angus around. Moreover, I think we're forgetting that our users are | Angus capable individuals who can do stuff themselves too... | I think you have to draw a line at some point. Fell free to do it. Yeah. You have made the initial happen. Now pass it on to those that actually will use it. -- Lgb
Re: LyX python package
Dear Jose, dear developers, I uploaded my (still not finally tested) LyX python package to http://wiki.lyx.org/Tools/PyClient Hopefully this can become a base for further discussion. On 31.05.05, José Matos wrote: G. Milde wrote: On 20.05.05, Jose' Matos wrote: On Friday 20 May 2005 10:04, G. Milde wrote: Now a serious question: is this [__init__.py] a trivial (empty) file or do you have any initialization step going inside? Maybe we should put some backwards compatibility code there. Currently I am writing for Python 2.3. At some stage, there might be code to add a directory with user-provided python scripts to the packages __path__, so the package becomes extensible. (Remember, my main concern is a scripting support for LyX.) I am not sure if it is worth to have a local script directory. In most of the points below this is my main objection. I think to reach my goal of a user-extensible lyx behaviour, such a directory is mandatory. A user who wants e.g. a permanent history for search-replace actions or automatic abbreviations expansion could then write a Python function that does just (using LyX.lfuns) and place it in a special plugins directory. Starting LyX with a python client should then make this functions available to a keybinding (via server-notify). --- lyx2lyx directory: I have been working on the lyx2lyx directory because cvs looses the history if you rename a directory. If I remember right, there is some strange and complicated way (described in the CVS book) how the administrator can rename/move a directory and preserve history. OTOH, you can also move the old one to the attic and create a new with a bit of documentation saying that there is a predecessor on the attic and how to get it... If that was not the case this directory would have been renamed LyX for more than a year ago. So at this level, at least for me lyx2lyx == LyX In the CVS, lyx2lyx (or pyLyX) is still a more transparent name than lyx-cvs/devel/lib/LyX/. Once installed, python-site-lib/LyX/ is preferable. I agree. For cvs a python directory (called python naturally seems a good bet). My suggestion is to put it under scripts as lyx-cvs/devel/lib/scripts/python/ so lyx-cvs/devel/lib/ doesnot become too crowded. when you say a LyX/lyx2lyx sub-module, there is no need for it, we already have a LyX.File that does all the work for you. With a LyX package this would become LyX.LyX.File As I have told you I would remove the need for the middle LyX I.e. you would rename/move your LyX.py to LyX/File.py? (I would prefer lowercase for modules.) - So clearly what we have to do is to agree on the python api to be used. Which means class names, file names and placement of files and sub-directories (== sub-modules) under the site-lib/LyX/ package directory. Yes. :-) As soon as the code starts to get shape those issues will be solved easily. :-) So I hope http://wiki.lyx.org/uploads/Tools/PyClient/pylyx0.1.tar.bz2 will help solving the remaining issues. It is still experimental code and feedback is welcome. Günter -- G.Milde web.de
Re: tex2lyx + international characters = redundant braces in lyx (1.3)?
On Tue, May 31, 2005 at 08:24:17PM +0200, Sven Schreiber wrote: Hi, I just stumbled over what I think is a little quirk in tex2lyx. When I convert a file that contains {\ss} latex codes, using the tex2lyx - lyx2lyx - lyx 1.3.5 procedure from the wiki (on windows), the braces appear as ERT in the final lyx file, although they are redundant. The braces don't do any harm, they're only redundant. Now just hang on for a second and try to figure out the kind of effort needed to prove that a certain construnct in a .tex file is redundant. Example: the latex input So{\ss}e should appear as Soße in the lyx file. (Hope the German es-zett is preserved in the mail.) What appears is: Soert{/ertßert}/erte. Tried to workaround with So\ss{}e, which gives Soßert{/ertert}/erte (still two different ert insets) Well, why not write Soße in your .tex. Works since the end of the previous millenium... So is it possible for tex2lyx to recognize a pair of braces that simply serves to delimit latex codes from surrounding text, and remove them in the lyx file? That would be great. It is 'possible'. However, it is almost impossible in this area to get anything right in all cases, so I am somewhat reluctant to have too many 'clever' additions to tex2lyx, especially if these additions are specific to certain laguages only. Btw, note that all the {\ss} in the input file were caused by the WinEdt (very popular windows latex shell) function to automatically save non-ascii input as ascii-latex-codes. So I guess this (minor) problem may appear more often in the future when lyx on windows is more widespread, and many people want to import their non-English latex files into lyx. As far as I am concerned this is rather ugly output of WinEdt. People could always run the thing through perl -e 's:{\\ss}:ß:' if they want to have an automated conversion. Or convince the WinEdt folk to produce 'suitable' output. Andre'
Re: [PATCH] bug 1890: insetoptarg triggers assertion
On Wed, Jun 01, 2005 at 03:26:47PM +0200, Jean-Marc Lasgouttes wrote: Martin == Martin Vermeer [EMAIL PROTECTED] writes: Martin Hmmm. This Martin 43 InsetCollapsable::InsetCollapsable Martin 44 (BufferParams const bp, CollapseStatus status) Martin 45 : InsetText(bp), label(Label), status_(status), openinlined_(false) Martin 46 { Martin suggests that a new textinset is being created at construction Martin of the collapsable. I do not think so. It means ``use my InsetText personality to call the constructor''. Lars, Angus anyone, would you care to comment? That's the constructor of the base class. Since we are using single inheritance, the address of the InsetCollapsable is the same as that of the base class. Andre'
Re: Graphics path with :
On Tue, May 31, 2005 at 11:12:26AM +0200, Jean-Marc Lasgouttes wrote: Angus == Angus Leeming [EMAIL PROTECTED] writes: Angus Reading http://wiki.lyx.org/LaTeX/FilesWithSpecialChars and Angus http://wiki.lyx.org/uploads/LaTeX/FilesWithSpecialChars/filenames.tex, Angus I don't see why a ':' in a file name won't work, but you're the Angus one using this stuff... I guess ':' is a problem in a windows file name if it is not the drive separator. 'Problem' is the wrong word. They serve a purpose IIRC. Try writing and reading to a.txt:1. Andre'
backporting lyx2lyx to 1.3.6
I had a look at how files with the wrong version are read in 1.4. It is different than in 1.3, newer files are not read at all and not treated as if the format was current. The change is here: http://www.lyx.org/cgi-bin/viewcvs.cgi/lyx-devel/src/buffer.C.diff?r1=1.512r2=1.513 Lars, was that intended or only a side effect? If this is what we want I'll replace lyx2lyx in 1.3 by the 1.4 version and adapt the code in buffer.C to it, if not I am going to implement a commandline switch of lyx2lyx that tells it to convert to the best matching format if the desired format is not possible an use that in 1.3 and 1.4. Georg
Re: backporting lyx2lyx to 1.3.6
Georg Baum [EMAIL PROTECTED] writes: | I had a look at how files with the wrong version are read in 1.4. It is | different than in 1.3, newer files are not read at all and not treated as | if the format was current. The change is here: I think this was intended, but it would not be intended for 1.3.x. With 1.4 we don't know of anything newer, but with 1.3 we do. ...but perhaps we should change 1.4? -- Lgb
Re: Windows installer finished
Angus Leeming wrote: What other files are needed? That's precisely my point. I don't know. Off the top of my head: sed, tr, cut, grep... Looking at Ruurds Version, I know what is needed. I uploaded the files as .rar to http://wiki.lyx.org/uploads/Tools/LyXWinTest/lyxutils.rar The files are the official actual versions. Taken from MSYS, Perl, and Python. Isn't it possible to add an actual version of perl.exe, python.exe and sh.exe (and some dll's if needed) to the folder LyX\bin\ and use these files for lyxrc.defaults by default (as Ruurd did)? This solution shouldn't produce NSIS troubles. This solution will require 1. Extensive testing to ascertain which modules are needed by the scripts that LyX packages. Why extensive testing? We know which files are needed and provide the official actual versions (the same that are used in the actual MSYS version). 2. Ongoing maintenance. Only if some users have trouble. I remember only one problem since LyXWin exists. Moreover, if I already have the official stuff installed why would I want to use the unoffical stuff? The files are the official ones. They are tested and under full support of the authors. Adding code to NSIS to conditionally install something is most certainly a NSIS trouble that I'm keen to avoid right now. I don't understand why this is an NSIS topic? I would just add the files to the package and use them as default. Then you don't have to search for the path to the files as they are part of the LyXWin package. Would you prefer to use these stripped down things that people have been discovering don't work (see countleess reports on the lyx-users list), or would you prefer to use an official version which definitely will work but which requires a big download committment. The countless reports concerned a problem with a sed.exe version. The solution was to replace it by a newer one. But this problem can appear everytime, because MSYS can also use a broken version. As the stripped files are taken from MSYS, installing MSYS instead of using the stripped files makes no difference. But we save the user from downloading and installing packages he don't need. The same is for the python/perl files. For example Openoffice delivers the python23.dll and don't install python, so why shouldn't we do the same? I installed your LyXWin version using the stripped files (I previously deinstalled python, perl, and MSYS) and it works correctly. OK I have not much time to test it roughly but I think it's stable. There is now only one problem: If I use the stripped files, the folder of the files (if you include them to your package it is then LyX's \bin folder) must be in the PATH. But that could easily be fixed. regards Uwe
Re: backporting lyx2lyx to 1.3.6
Lars Gullik Bjønnes wrote: Georg Baum [EMAIL PROTECTED] writes: | I had a look at how files with the wrong version are read in 1.4. It is | different than in 1.3, newer files are not read at all and not treated | as if the format was current. The change is here: I think this was intended, but it would not be intended for 1.3.x. I can see no difference in this respect between 1.3 and 1.4 after applying the patch that I sent earlier: The patch sends newer lyx files through lyx2lyx, too and only refuses to load files that can't be converted. Then LyX 1.3.6 would be able to read current 1.4 files, but not anymore if we change the format again. New file formats could be added by simply replacing lyx2lyx, without the need to update the lyx executable. With 1.4 we don't know of anything newer, but with 1.3 we do. Not now, but at some point in time we will. ...but perhaps we should change 1.4? That was exactly my question ;-) Georg
Re: Windows installer finished
Uwe Stöhr wrote: Angus Leeming wrote: What other files are needed? That's precisely my point. I don't know. Off the top of my head: sed, tr, cut, grep... Looking at Ruurds Version, I know what is needed. I uploaded the files as .rar to http://wiki.lyx.org/uploads/Tools/LyXWinTest/lyxutils.rar The files are the official actual versions. Taken from MSYS, Perl, and Python. Isn't it possible to add an actual version of perl.exe, python.exe and sh.exe (and some dll's if needed) to the folder LyX\bin\ and use these files for lyxrc.defaults by default (as Ruurd did)? This solution shouldn't produce NSIS troubles. This solution will require 1. Extensive testing to ascertain which modules are needed by the scripts that LyX packages. Why extensive testing? We know which files are needed and provide the official actual versions (the same that are used in the actual MSYS version). I'm afraid that I don't believe that this is so. Several scripts have changed since Ruurd released LyX/Win 1.3.5 and their requirements have changed too. Witness the reports that people have been making that the lyx2lyx packaged with tex2lyx doesn't work with Ruurd's python. 2. Ongoing maintenance. Only if some users have trouble. I remember only one problem since LyXWin exists. I remember a whole stack of reports about missing python modules... Moreover, if I already have the official stuff installed why would I want to use the unoffical stuff? The files are the official ones. They are tested and under full support of the authors. Adding code to NSIS to conditionally install something is most certainly a NSIS trouble that I'm keen to avoid right now. I don't understand why this is an NSIS topic? I would just add the files to the package and use them as default. Then you don't have to search for the path to the files as they are part of the LyXWin package. Because I don't believe that Ruurd's stripped down versions are sufficient and I'm not willing to create *and test* their replacement. Would you prefer to use these stripped down things that people have been discovering don't work (see countleess reports on the lyx-users list), or would you prefer to use an official version which definitely will work but which requires a big download committment. The countless reports concerned a problem with a sed.exe version. Not really. It was a problem with the DOS line endings of the generated chkconfig.sed. But other scripts have problems too. I installed your LyXWin version using the stripped files (I previously deinstalled python, perl, and MSYS) and it works correctly. Ha! Scripts are by their nature dynamic. Have you tested that every single script works? Of course not! OK I have not much time to test it roughly but I think it's stable. There is now only one problem: If I use the stripped files, the folder of the files (if you include them to your package it is then LyX's \bin folder) must be in the PATH. But that could easily be fixed. ??? Then tell the installer where to find this folder and it'll add it to the path_prefix variable in lyxrc.defaults. -- Angus
Re: Windows installer: NSIS vs .msi
Angus Leeming wrote: Rex Dieter wrote: Angus Leeming wrote: But this will involve me in a lot more work with NSIS getting to grips with optional installation of components. I'm running out of LyX-time. I'm also running out of NSIS-enthusiasm. While having any sort of installer is better than none at all, I'd suggest future effort re: Windows installers aim for an .msi target (Microsoft Installer) format. Fine. Having struggled with NSIS for the last month, I would love to see it die :( However, I think that Asger's point that we should stick with something that's both Open Source and actively developed. Like http://sourceforge.net/projects/wix/ ? Notice also that you might like MSI, but others think that NSIS is better: http://www.peterprovost.org/archive/2004/01/19/1040.aspx I humbly disagree with these others. I'm noticing that the install and uninstall runs so much faster than the Installshield or VS.NET msi files that I have been using. True. I mentioned NSIS a few days/weeks ago in one of my Geek Notes columns. Because my customer was very size sensitive, the project that we shipped last week used NSIS for its installer. The VS.NET/MSI installer we created was almost 4MB while the NSIS installer was less than 520K. A proper msi installer should have very little overhead. I'd say was created incorrectly (probably bundling some extra crud). The downside is that these installers are not MSI based. In fact, the reason the VS.NET installers are so big is that they bundle two copies of the Windows Installer 2.0 (ANSI and Unicode) within the installer. If you switch the deployment project properties from Windows Installer bootstrapper to No bootstrapper, you can alleviate this, but you also can only run the installer on machines that are already running Windows Installer 2.0. I make .msi packages with MSI-2.0 requirements too, but they auto-download/install msi-2.0 from microsoft.com if not already present. Then, those of us in AD environments can enjoy pain-free deployment to (all) client machines. (-: At the cost of increased (quadrupled ?) download size. Not in my experience. Can you explain why you can't just compile the NSIS installer on your weird windows box? How hard is $ path to/makensis lyx_installer.nsi With software packaged as .msi's, I can deploy to my 100's of client machines with a few clicks with AD Group Policy. Can't do that with any nsis installer (please correct me if I'm wrong). -- Rex
Re: backporting lyx2lyx to 1.3.6
Georg Baum [EMAIL PROTECTED] writes: | Lars Gullik Bjønnes wrote: Georg Baum [EMAIL PROTECTED] writes: | I had a look at how files with the wrong version are read in 1.4. It is | different than in 1.3, newer files are not read at all and not treated | as if the format was current. The change is here: I think this was intended, but it would not be intended for 1.3.x. | I can see no difference in this respect between 1.3 and 1.4 after applying | the patch that I sent earlier: The patch sends newer lyx files through | lyx2lyx, too and only refuses to load files that can't be converted. Then | LyX 1.3.6 would be able to read current 1.4 files, but not anymore if we | change the format again. | New file formats could be added by simply replacing lyx2lyx, without the | need to update the lyx executable. With 1.4 we don't know of anything newer, but with 1.3 we do. | Not now, but at some point in time we will. ...but perhaps we should change 1.4? | That was exactly my question ;-) I read it the other way: make 1.3 behave like 1.4 -- Lgb
Re: LyX meeting in Paris. What about July 14-18?
Jean-Marc Lasgouttes wrote: Hello Michael, Do you have a more precise idea of when you will be coming? I just bought my ticket. I will arrive at Gare du Nord on 16.07. at 13:05 and leave on 17.07. at 17:55. It is a very short visit but I think it's worth the effort! Eventually, you should give me some instructions on how to get to your apartment. Is there an affordable hotel around the corner? Regards, Michael
Re: backporting lyx2lyx to 1.3.6
On Monday 06 June 2005 17:51, Georg Baum wrote: I had a look at how files with the wrong version are read in 1.4. It is different than in 1.3, newer files are not read at all and not treated as if the format was current. The change is here: http://www.lyx.org/cgi-bin/viewcvs.cgi/lyx-devel/src/buffer.C.diff?r1=1.5 12r2=1.513 And why should them be read if they are newer? It was me who motivated Lars to do this change. The argument is simple, every knowledge about the lyx file format should go into lyx2lyx. Lars, was that intended or only a side effect? If this is what we want I'll replace lyx2lyx in 1.3 by the 1.4 version and adapt the code in buffer.C to it, if not I am going to implement a commandline switch of lyx2lyx that tells it to convert to the best matching format if the desired format is not possible an use that in 1.3 and 1.4. I would prefer this solution. After all how do you decide the file is newer? Knowing the format of every file header? But then this is exactly the reason why lyx2lyx was created in first place, to relieve lyx from knowing every possible lyx format Georg -- José Abílio
Re: Windows installer: NSIS vs .msi
Rex Dieter wrote: Can you explain why you can't just compile the NSIS installer on your weird windows box? How hard is $ path to/makensis lyx_installer.nsi With software packaged as .msi's, I can deploy to my 100's of client machines with a few clicks with AD Group Policy. Can't do that with any nsis installer (please correct me if I'm wrong). Hey! My knowledge is the sum of four weeks graft, that's all. I look forward to it all being superceeded. Angus
Prerelease of LyX 1.3.6 on WIndows?
JMarc, all. I just added a page to the wiki http://wiki.lyx.org/Windows/LyX136pre describing a prerelease of LyX 1.3.6 on Windows. Christian, can I get you to move the binary http://www.lyx.org/~leeming/lyx_setup_136.exe to the wiki upload area. I guess it should be moved to http://wiki.lyx.org/uploads/Windows/LyX136pre/lyx_setup_136.exe ? It's too big (6MB) for me to do this myself, I believe. Feel free to offer suggestions on how to improve the page, or just edit it yourself :) If you think that this thing is ready to fly, then I'll make an announcement on the lyx-users list. Angus
Re: Windows installer finished
Angus Leeming wrote: Why extensive testing? We know which files are needed and provide the official actual versions (the same that are used in the actual MSYS version). I'm afraid that I don't believe that this is so. Several scripts have changed since Ruurd released LyX/Win 1.3.5 and their requirements have changed too. Witness the reports that people have been making that the lyx2lyx packaged with tex2lyx doesn't work with Ruurd's python. That's why I don't use Ruurds versions but the actual ones. I would just add the files to the package and use them as default. Then you don't have to search for the path to the files as they are part of the LyXWin package. I don't believe that Ruurd's stripped down versions are sufficient and I'm not willing to create *and test* their replacement. Scripts are by their nature dynamic. Have you tested that every single script works? Of course not! You are of course right that every script could cause troubles. But as I told, the files are the same as if the user installed the full version of python etc. But as you and I don't have the time to test it roughly. I propose the following: You (or I) could create a wiki page with a description of the new LyXWin installer version. I'm sure that after a mail to the lyx-users list many Win users will test it. If I use the stripped files, the folder of the files (if you include them to your package it is then LyX's \bin folder) must be in the PATH. But that could easily be fixed. ??? Then tell the installer where to find this folder and it'll add it to the path_prefix variable in lyxrc.defaults. I mean that the folder of the stripped files have to be in the system environment variable PATH so that every script can find them. In my opinion it would be a good solution if the installer asks for ghostscript, imagemagick, and latex (in this order because imagemagick needs ghostscript installed!) as it does in the current version. Then the user chooses an install directory for LyX and the LyX-files are extracted there. The stripped versions of python etc. are included in the install package. Assuming the install directory is C:\LyX they are extracted in C:\LyX\bin (the same directory like the lyx.exe). Now this directory is added to the PATH using Win's command line command set path=C:\LyX\bin;%path% Now the menu language is set for all users with e.g. the comand line command set LANG=de_DE At last LyX is first configured and started. (I'm not sure if the command line commands are correct. But I can ask Michael Schmidt who knows them.) thanks and regards Uwe
Re: backporting lyx2lyx to 1.3.6
Jose' Matos [EMAIL PROTECTED] writes: | On Monday 06 June 2005 17:51, Georg Baum wrote: I had a look at how files with the wrong version are read in 1.4. It is different than in 1.3, newer files are not read at all and not treated as if the format was current. The change is here: http://www.lyx.org/cgi-bin/viewcvs.cgi/lyx-devel/src/buffer.C.diff?r1=1.5 12r2=1.513 | And why should them be read if they are newer? It was me who motivated | Lars to do this change. The argument is simple, every knowledge about the | lyx file format should go into lyx2lyx. Yes, that is right. Now I remember. lyx2lyx solves that. LyX only cares about its own format. (there should be a way to query the lyx binary about what format version it needs) -- Lgb
RE: Prerelease of LyX 1.3.6 on WIndows?
In my book, Angus is a superstar. Angus, do not worry about all the flack. That is only because there is a lot of interest in a windows installer. And if the itch scratches enough for the whiners, they can pick up the ball and improve it. You've made a great contribution, and deserve some credit. Regards, Asger
Re: [PATCH] minor bibitem tweaks
Martin Vermeer wrote: > Yes, I know... I can live with this for 1.4. I'll commit the minimalist version. This is clearly an improvement, everything else is ambiguous. Jürgen
Re: C++ question
Lars Gullik Bjønnes wrote: | I feel like a bit of an idiot asking this, but what's the C++ way to | insert some data into an existing file? How would you do it in any other language? At end is easy. At start is more tricky... If it were a memory buffer how would you do it then? (memmove perhaps?) You need to use some temporary (file or memory buffer) Unless you mmap the file, then you might be able to use memmove. Thanks, Lars. I suspected that was going to be the answer. Angus
Re: Windows installer --- software probe
> --- Ursprüngliche Nachricht --- > Von: Asger Ottar Alstrup <[EMAIL PROTECTED]> > An: unknown > Kopie: lyx-devel@lists.lyx.org > Betreff: Re: Windows installer --- software probe > Datum: Thu, 02 Jun 2005 19:45:06 +0200 > > One comment: > > The reason I did not install minsys was that I did not know which file > to download and install. There are listed hundreds on that web-site. > Maybe you could list what file the user should grab? > > Regards, > Asger > I guess you mean msys - minimal gnu system: MSYS-1.0.10.exe on http://www.mingw.org/download.shtml BTW: I'm w/o access to my mailbox until June, 23. Regards Carl -- Weitersagen: GMX DSL-Flatrates mit Tempo-Garantie! Ab 4,99 Euro/Monat: http://www.gmx.net/de/go/dsl
Re: Windows installer --- software probe
Carl Dr. Kleffner wrote: One comment: The reason I did not install minsys was that I did not know which file to download and install. There are listed hundreds on that web-site. Maybe you could list what file the user should grab? Regards, Asger I guess you mean msys - minimal gnu system: MSYS-1.0.10.exe on http://www.mingw.org/download.shtml That's correct. I want the user to grab msys-1.0.x.exe which they can get from http://prdownloads.sf.net/mingw/ (the download.shtml page takes you to the sf.net pages when you select some file for download.) I've changed the label on the radio button to: Download MinSYS (MSYS-1.0.X.exe) which I trust will allow the user to grab the correct binary. Angus
Re: Windows installer --- software probe
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes: Angus> Carl Dr. Kleffner wrote: >>> One comment: >>> >>> The reason I did not install minsys was that I did not know which >>> file to download and install. There are listed hundreds on that >>> web-site. Maybe you could list what file the user should grab? >>> >>> Regards, Asger >> I guess you mean msys - minimal gnu system: MSYS-1.0.10.exe on >> http://www.mingw.org/download.shtml Angus> That's correct. I want the user to grab msys-1.0.x.exe which Angus> they can get from http://prdownloads.sf.net/mingw/ (the Angus> download.shtml page takes you to the sf.net pages when you Angus> select some file for download.) What about this instead? http://sourceforge.net/project/showfiles.php?group_id=2435_id=82721_id=158803 I think it is much simpler for newcomers. The only thing I am not sure about is whether the release_id tag is stable with new versions. JMarc
Re: Windows installer --- software probe
Jean-Marc Lasgouttes wrote: What about this instead? http://sourceforge.net/project/showfiles.php?group_id=2435_id=82721_id=158803 I think it is much simpler for newcomers. The only thing I am not sure about is whether the release_id tag is stable with new versions. Thanks, Jean-Marc. Let's do this for LyX 1.3.6 anyway. We can always reconsider for future releases. Angus
Re: [PATCH 1.4] de.po update
> "Michael" == Michael Schmitt <[EMAIL PROTECTED]> writes: Michael> Hello, here comes another update for de.po. Please apply. Done. JMarc
Re: [Patch] Re: [patch] key event queue
Martin Vermeer <[EMAIL PROTECTED]> writes: | On Wed, Jun 01, 2005 at 03:12:05PM +0300, Martin Vermeer wrote: >> On Wed, Jun 01, 2005 at 01:31:55PM +0200, Lars Gullik Bjønnes wrote: >> > Martin Vermeer <[EMAIL PROTECTED]> writes: > | ... | >> OK, that settles it then for me: put in the event queue and reverse the >> new stuff in workAreaKeyPress to a simple showCursor call. Then everything >> will be consistent, if slow. >> >> I verified that this works properly using an old version of your event >> queue patch: doesn't lose keystrokes (except formally repeated ones), >> doesn't change their order, and doesn't produce cursor artefacts. >> >> Please commit your current version. > | Lars? Please beat me too it... Too many balls in the air right now... (send the patch here first, and I'll do a quick review) -- Lgb
Re: [PATCH] Bug 961: reintroduce LFUN_BIBDB_ADD/DEL
> "Juergen" == Juergen Spitzmueller <[EMAIL PROTECTED]> writes: Juergen> Juergen Spitzmueller wrote: >> I'm not even sure I understand what LABEL_GOTO does (or is it >> broken?) Juergen> OK, now I do. This is really odd. Odd in which sense? JMarc
Re: [PATCH] Bug 961: reintroduce LFUN_BIBDB_ADD/DEL
> "Juergen" == Juergen Spitzmueller <[EMAIL PROTECTED]> writes: Juergen> Juergen Spitzmueller wrote: >> > I'm not even sure I understand what LABEL_GOTO does (or is it >> broken?) >> >> OK, now I do. This is really odd. Juergen> I think this menu entry is extremely irritating. How about Juergen> the following patch, that does enable the entry only when it Juergen> actually makes sense? Yes, definitely good, although I would write: + case LFUN_LABEL_GOTO: { + flag.enabled(!cmd.argument.empty() + ||getInsetByCode(cursor_, InsetBase::REF_CODE)); + break; JMarc
Re: Script alternative - Ch
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes: Angus> I really don't see why such a mini-dist would be an advantage. Angus> What will happen is what has happened with Ruurd's equivalent Angus> packaging of python. It fails to work as people add more Angus> functionality to the scripts. Moreover, it will become "one Angus> more thing" to maintain. It depends on whether we can automate the process, I guess. In this case it is just a new makefile target. JMarc
Windows installer finished
Find it at http://www.devel.lyx.org/~leeming/lyx_setup_136.exe It's 5.9MB in size. It's fully internationalized although its messages still need to be translated. If you're interested in doing so, you can grab the source from http://www.devel.lyx.org/~leeming/packaging.tar.bz2 It's 82kB in size. Have a look at installer/lyx_languages/{french,german}.nsh which need translated versions of the 53 strings. Adding other languages is straightforward. The archive unpacks to packaging/ README clean_dvi.py package_lyxwin.sh dtl/ README Makefile dt2dv.c dt2dv.man dtl.doc dtl.h dv2dt.c dv2dt.man dvi.doc edited.txt example.tex hello.tex man2ps tripvdu.tex icons/ lyx_32x32_16colors.png lyx_32x32.png lyx_doc_32x32_16colors.png lyx_doc_32x32.png lyx_doc.svg lyx.svg lyx_32x32.ico lyx_doc_32x32.ico installer/ lyx_languages/ english.nsh french.nsh german.nsh abi_util_fileassoc.nsh download.nsh ioDownload.ini ioSummary.ini lyx_installer.nsi lyx_path_prefix.C is_user_admin.nsh lyxfunc.nsh strtrim.nsh Angus
Re: Windows installer finished
Angus Leeming wrote: > The archive unpacks to I propose to add this tree as-is to development/Win32. OK? Angus > packaging/ > README > clean_dvi.py > package_lyxwin.sh > dtl/ > README > Makefile > dt2dv.c > dt2dv.man > dtl.doc > dtl.h > dv2dt.c > dv2dt.man > dvi.doc > edited.txt > example.tex > hello.tex > man2ps > tripvdu.tex > icons/ > lyx_32x32_16colors.png > lyx_32x32.png > lyx_doc_32x32_16colors.png > lyx_doc_32x32.png > lyx_doc.svg > lyx.svg > lyx_32x32.ico > lyx_doc_32x32.ico > installer/ > lyx_languages/ > english.nsh > french.nsh > german.nsh > abi_util_fileassoc.nsh > download.nsh > ioDownload.ini > ioSummary.ini > lyx_installer.nsi > lyx_path_prefix.C > is_user_admin.nsh > lyxfunc.nsh > strtrim.nsh
Re: [Patch] Re: [patch] key event queue
On Mon, 2005-06-06 at 11:46, Lars Gullik Bjønnes wrote: > Martin Vermeer <[EMAIL PROTECTED]> writes: > > | On Wed, Jun 01, 2005 at 03:12:05PM +0300, Martin Vermeer wrote: > >> On Wed, Jun 01, 2005 at 01:31:55PM +0200, Lars Gullik Bjønnes wrote: > >> > Martin Vermeer <[EMAIL PROTECTED]> writes: > > > | ... > | > >> OK, that settles it then for me: put in the event queue and reverse the > >> new stuff in workAreaKeyPress to a simple showCursor call. Then everything > >> will be consistent, if slow. > >> > >> I verified that this works properly using an old version of your event > >> queue patch: doesn't lose keystrokes (except formally repeated ones), > >> doesn't change their order, and doesn't produce cursor artefacts. > >> > >> Please commit your current version. > > > | Lars? > > Please beat me too it... > > Too many balls in the air right now... > (send the patch here first, and I'll do a quick review) Attached. - Martin Index: ChangeLog === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/ChangeLog,v retrieving revision 1.2193 diff -u -p -r1.2193 ChangeLog --- ChangeLog 6 Jun 2005 08:30:19 - 1.2193 +++ ChangeLog 6 Jun 2005 11:53:47 - @@ -1,3 +1,8 @@ +2005-06-06 Martin Vermeer <[EMAIL PROTECTED]> + + * BufferView_pimpl.C: revert to showCursor in connection with + Lars's front-end stack + 2005-06-06 J�Spitzm� <[EMAIL PROTECTED]> * text.C (redoParagraph): move cursor right after a bibitem Index: BufferView_pimpl.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/BufferView_pimpl.C,v retrieving revision 1.584 diff -u -p -r1.584 BufferView_pimpl.C --- BufferView_pimpl.C 31 May 2005 14:40:23 - 1.584 +++ BufferView_pimpl.C 6 Jun 2005 11:53:48 - @@ -520,12 +520,8 @@ void BufferView::Pimpl::workAreaKeyPress * of the cursor. Note we cannot do this inside * dispatch() itself, because that's called recursively. */ - if (available()) { - screen().prepareCursor(); - cursor_timeout.setTimeout(100); - cursor_timeout.restart(); - cursor_timeout.setTimeout(400); - } + if (available()) + screen().showCursor(*bv_); } Index: frontends/qt2/ChangeLog === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/ChangeLog,v retrieving revision 1.789 diff -u -p -r1.789 ChangeLog --- frontends/qt2/ChangeLog 3 Jun 2005 09:25:45 - 1.789 +++ frontends/qt2/ChangeLog 6 Jun 2005 11:53:48 - @@ -1,3 +1,10 @@ +2005-06-06 Lars Gullik Bjonnes <[EMAIL PROTECTED]> + (Committed Martin Vermeer <[EMAIL PROTECTED]>) + + * QContentPane.[Ch]: implement keystroke event stack with + repeat key "regulation". Key are not sent to LyX core faster + than rendering can handle. + 2005-06-03 J�Spitzm� <[EMAIL PROTECTED]> * QContentPane.[Ch]: Limit the QIM*Events to X11. They are not Index: frontends/qt2/QContentPane.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QContentPane.C,v retrieving revision 1.34 diff -u -p -r1.34 QContentPane.C --- frontends/qt2/QContentPane.C 3 Jun 2005 09:25:46 - 1.34 +++ frontends/qt2/QContentPane.C 6 Jun 2005 11:53:48 - @@ -19,7 +19,6 @@ #include #include -#include #include @@ -90,6 +89,8 @@ QContentPane::QContentPane(QWorkArea * p boost::bind(::generateSyntheticMouseEvent, this)); + connect(_timer_, SIGNAL(timeout()), SLOT(keyeventTimeout())); + setFocusPolicy(QWidget::WheelFocus); setFocus(); setCursor(ibeamCursor); @@ -101,6 +102,9 @@ QContentPane::QContentPane(QWorkArea * p // stupid moc strikes again connect(wa_->scrollbar_, SIGNAL(valueChanged(int)), this, SLOT(scrollBarChanged(int))); + + // Start the timer, one-shot. + step_timer_.start(25, true); } @@ -251,9 +255,35 @@ void QContentPane::wheelEvent(QWheelEven void QContentPane::keyPressEvent(QKeyEvent * e) { + keyeventQueue_.push(boost::shared_ptr(new QKeyEvent(*e))); +} + + +void QContentPane::keyeventTimeout() +{ + bool handle_autos = true; + + while (!keyeventQueue_.empty()) { + boost::shared_ptr ev = keyeventQueue_.front(); + + // We never handle more than one auto repeated + // char in a list of queued up events. + if (!handle_autos && ev->isAutoRepeat()) { + keyeventQueue_.pop(); + continue; + } + boost::shared_ptr sym(new QLyXKeySym); - sym->set(e); - wa_->workAreaKeyPress(sym, q_key_state(e->state())); + sym->set(ev.get()); + + wa_->workAreaKeyPress(sym, q_key_state(ev->state())); + keyeventQueue_.pop(); + + handle_autos = false; + } + + // Restart the timer. + step_timer_.start(25, true); } Index: frontends/qt2/QContentPane.h === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QContentPane.h,v retrieving revision 1.16 diff -u -p -r1.16 QContentPane.h ---
Re: [Patch] Re: [patch] key event queue
Martin Vermeer <[EMAIL PROTECTED]> writes: I am fine with this. (I'd better be...) -- Lgb
[patch] fix bug 1665
Attached is a tested, working 1.3 patch that fixes bug 1665, following an idea of Jean-Marc (see http://bugzilla.lyx.org/show_bug.cgi?id=1665). I know that this is no crash, but it is very annoying for me (because we produce many of our documents via commandline export), so I fixed it nevertheless. I propose to apply this both to 1.3 and 1.4. OK? GeorgIndex: src/mathed/ChangeLog === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/mathed/ChangeLog,v retrieving revision 1.299.2.21 diff -u -p -r1.299.2.21 ChangeLog --- src/mathed/ChangeLog 26 Apr 2005 10:00:58 - 1.299.2.21 +++ src/mathed/ChangeLog 6 Jun 2005 12:06:10 - @@ -1,3 +1,10 @@ +2005-06-06 Georg Baum <[EMAIL PROTECTED]> + + * math_parser.h (latexkeys): new member requires + * math_factory.C (initSymbols): fill the requires field for fonts + * math_symbolinset.C (validate): use sym_->requires instead of + sym_->inset (fixes bug 1665) + 2005-04-26 Georg Baum <[EMAIL PROTECTED]> * math_cursor.C (interpret): treat & as special character (fixes Index: src/mathed/math_factory.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/mathed/math_factory.C,v retrieving revision 1.70.2.2 diff -u -p -r1.70.2.2 math_factory.C --- src/mathed/math_factory.C 7 Dec 2004 10:49:53 - 1.70.2.2 +++ src/mathed/math_factory.C 6 Jun 2005 12:06:11 - @@ -144,6 +144,12 @@ void initSymbols() // tmp.inset _is_ the fontname here. // create fallbacks if necessary + // store requirements as long as we can + if (tmp.inset == "msa" || tmp.inset == "msb") +tmp.requires = "amssymb"; + else if (tmp.inset == "wasy") +tmp.requires = "wasysym"; + // symbol font is not available sometimes string symbol_font = "lyxsymbol"; Index: src/mathed/math_parser.h === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/mathed/math_parser.h,v retrieving revision 1.52.2.1 diff -u -p -r1.52.2.1 math_parser.h --- src/mathed/math_parser.h 7 Dec 2004 10:50:01 - 1.52.2.1 +++ src/mathed/math_parser.h 6 Jun 2005 12:06:11 - @@ -39,6 +39,8 @@ struct latexkeys { string extra; /// how is this called as XML entity? string xmlname; + /// required LaTeXFeatures + string requires; }; Index: src/mathed/math_symbolinset.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/mathed/math_symbolinset.C,v retrieving revision 1.52.2.4 diff -u -p -r1.52.2.4 math_symbolinset.C --- src/mathed/math_symbolinset.C 19 Jan 2005 11:46:53 - 1.52.2.4 +++ src/mathed/math_symbolinset.C 6 Jun 2005 12:06:11 - @@ -122,10 +122,8 @@ bool MathSymbolInset::takesLimits() cons void MathSymbolInset::validate(LaTeXFeatures & features) const { - if (sym_->inset == "msa" || sym_->inset == "msb") - features.require("amssymb"); - else if (sym_->inset == "wasy") - features.require("wasysym"); + if (!sym_->requires.empty()) + features.require(sym_->requires); }
Re: Windows installer finished
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes: Angus> Angus Leeming wrote: >> The archive unpacks to Angus> I propose to add this tree as-is to development/Win32. Be careful with binary status of the files in cvs. Also, binary files mean that patch-1.3.6 will not work reliably. JMarc
Re: [patch] fix bug 1665
> "Georg" == Georg Baum <[EMAIL PROTECTED]> writes: Georg> Attached is a tested, working 1.3 patch that fixes bug 1665, Georg> following an idea of Jean-Marc (see Georg> http://bugzilla.lyx.org/show_bug.cgi?id=1665). I know that this Georg> is no crash, but it is very annoying for me (because we produce Georg> many of our documents via commandline export), so I fixed it Georg> nevertheless. I propose to apply this both to 1.3 and 1.4. OK? This is OK for 1.3.x. JMarc
Re: [Patch] Re: [patch] key event queue
On Mon, 2005-06-06 at 15:17, Lars Gullik Bjønnes wrote: > Martin Vermeer <[EMAIL PROTECTED]> writes: > > I am fine with this. > (I'd better be...) Done. - Martin signature.asc Description: This is a digitally signed message part
Re: [PATCH] Bug 961: reintroduce LFUN_BIBDB_ADD/DEL
Jean-Marc Lasgouttes wrote: > >> I'm not even sure I understand what LABEL_GOTO does (or is it > >> broken?) > > Juergen> OK, now I do. This is really odd. > > Odd in which sense? Let me put it like this: It's not what I would have expected (though I see its use). Jürgen
Re: Windows installer finished
Jean-Marc Lasgouttes wrote: > Angus> I propose to add this tree as-is to development/Win32. > > Be careful with binary status of the files in cvs. OK, but that warning applies only to these files in the icons directory. lyx_32x32_16colors.png lyx_32x32.png lyx_doc_32x32_16colors.png lyx_doc_32x32.png lyx_32x32.ico lyx_doc_32x32.ico > Also, binary files mean that patch-1.3.6 will not work reliably. I don't follow. Could you explain further? -- Angus
Re: Windows installer finished
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes: >> Also, binary files mean that patch-1.3.6 will not work reliably. Angus> I don't follow. Could you explain further? Binary file do not appear in patches. Therefore LyX 1.3.5 + patch-1.3.6 != LyX 1.3.6 JMarc
Re: Windows installer finished
Jean-Marc Lasgouttes <[EMAIL PROTECTED]> writes: >> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes: > >>> Also, binary files mean that patch-1.3.6 will not work reliably. > | Angus> I don't follow. Could you explain further? > | Binary file do not appear in patches. Therefore | LyX 1.3.5 + patch-1.3.6 != LyX 1.3.6 That would require something like xdelta. -- Lgb
Re: [PATCH] Bug 961: reintroduce LFUN_BIBDB_ADD/DEL
Jean-Marc Lasgouttes wrote: > Yes, definitely good, although I would write: > > + case LFUN_LABEL_GOTO: { > + flag.enabled(!cmd.argument.empty() > + ||getInsetByCode(cursor_, > InsetBase::REF_CODE)); + break; I did so and committed. Thanks, Jürgen
Re: [PATCH] Bug 961: reintroduce LFUN_BIBDB_ADD/DEL
Juergen Spitzmueller wrote: > http://bugzilla.lyx.org/show_bug.cgi?id=961 > > The attached patch reimplements those lfuns. It also fixes > InsetBibtex::delDatabase, which is broken in 1.3. > > The only drawback is that the lfuns are not working when the cursor sits > behind the inset. The same applies to 1.3. The reason is that > getInsetByCode stops at the end of the document. Attached is a third (and IMHO final) version that fixes all drawbacks and additionally also fixes a problem in InsetBibtex::addDatabase, which did not add a database when its name was a substring of another database (e.g. "mybib" vs. "mybib2"). OK to apply? Jürgen Index: BufferView_pimpl.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/BufferView_pimpl.C,v retrieving revision 1.586 diff -u -r1.586 BufferView_pimpl.C --- BufferView_pimpl.C 6 Jun 2005 13:11:29 - 1.586 +++ BufferView_pimpl.C 6 Jun 2005 13:21:07 - @@ -51,6 +51,7 @@ #include "undo.h" #include "vspace.h" +#include "insets/insetbibtex.h" #include "insets/insetref.h" #include "insets/insettext.h" @@ -128,9 +129,20 @@ { T * inset = 0; DocIterator it = cur; - if (it.nextInset() && - it.nextInset()->lyxCode() == code) { - inset = static_cast(it.nextInset()); + T * first_inset = static_cast (it.nextInset()); + while (it) { + if (!it.nextInset()) { + // try from the beginning, but break after one loop + it.pit() = 0; + it.pos() = 0; + if (!it.nextInset() || it.nextInset() == first_inset) +break; + } + if (it.nextInset()->lyxCode() == code) { + inset = static_cast (it.nextInset()); + break; + } + it.forwardInset(); } return inset; } @@ -982,6 +994,8 @@ case LFUN_MARK_ON: case LFUN_SETMARK: case LFUN_CENTER: + case LFUN_BIBDB_ADD: + case LFUN_BIBDB_DEL: case LFUN_WORDS_COUNT: flag.enabled(true); break; @@ -1212,6 +1226,24 @@ case LFUN_CENTER: center(); break; + + case LFUN_BIBDB_ADD: { + InsetBibtex * inset = + getInsetByCode(cursor_, + InsetBase::BIBTEX_CODE); + if (inset) + inset->addDatabase(cmd.argument); + break; + } + + case LFUN_BIBDB_DEL: { + InsetBibtex * inset = + getInsetByCode(cursor_, + InsetBase::BIBTEX_CODE); + if (inset) + inset->delDatabase(cmd.argument); + break; + } case LFUN_WORDS_COUNT: { DocIterator from, to; Index: LyXAction.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/LyXAction.C,v retrieving revision 1.206 diff -u -r1.206 LyXAction.C --- LyXAction.C 8 May 2005 10:02:37 - 1.206 +++ LyXAction.C 6 Jun 2005 13:21:08 - @@ -188,6 +188,8 @@ { LFUN_INSERT_LABEL, "label-insert", Noop }, { LFUN_INSET_OPTARG, "optional-insert", Noop }, { LFUN_INSERT_BIBITEM, "bibitem-insert", Noop }, + { LFUN_BIBDB_ADD, "bibtex-database-add", Noop }, + { LFUN_BIBDB_DEL, "bibtex-database-del", Noop }, { LFUN_INSERT_LINE, "line-insert", Noop }, { LFUN_INSERT_PAGEBREAK, "pagebreak-insert", Noop }, { LFUN_LANGUAGE, "language", Noop }, Index: lfuns.h === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/lfuns.h,v retrieving revision 1.41 diff -u -r1.41 lfuns.h --- lfuns.h 8 May 2005 10:02:37 - 1.41 +++ lfuns.h 6 Jun 2005 13:21:08 - @@ -355,6 +355,8 @@ // 270 LFUN_WORDS_COUNT, LFUN_OUTPUT_CHANGES, // jspitzm 20050121 + LFUN_BIBDB_ADD, + LFUN_BIBDB_DEL, LFUN_LASTACTION // end of the table }; Index: insets/insetbibtex.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/insets/insetbibtex.C,v retrieving revision 1.54 diff -u -r1.54 insetbibtex.C --- insets/insetbibtex.C 17 May 2005 11:11:45 - 1.54 +++ insets/insetbibtex.C 6 Jun 2005 13:21:13 - @@ -271,7 +271,7 @@ bool InsetBibtex::addDatabase(string const & db) { string contents(getContents()); - if (!contains(contents, db)) { + if (tokenPos(contents, ',', db) == -1) { if (!contents.empty()) contents += ','; setContents(contents + db); @@ -283,16 +283,17 @@ bool InsetBibtex::delDatabase(string const & db) { - if (contains(getContents(), db)) { + string contents(getContents()); + if (contains(contents, db)) { + int const n = tokenPos(contents, ',', db); string bd = db; - int const n = tokenPos(getContents(), ',', bd); if (n > 0) { - // Weird code, would someone care to explain this?(Lgb) - string tmp(", "); - tmp += bd; - setContents(subst(getContents(), tmp, ", ")); + // this is not the first database + string tmp = ',' + bd; + setContents(subst(contents, tmp, "")); } else if (n == 0) - setContents(split(getContents(), bd, ',')); + // this is the first (or only) database + setContents(split(contents, bd, ',')); else return false; }
Re: Windows installer finished
Jean-Marc Lasgouttes wrote: >>> Also, binary files mean that patch-1.3.6 will not work reliably. > Angus> I don't follow. Could you explain further? > Binary file do not appear in patches. Therefore > LyX 1.3.5 + patch-1.3.6 != LyX 1.3.6 `diff -a` ? That's true for more than just the files I've just added however. All of the following are flagged as binary in the 1.3.x tree: development/Win32/packaging/icons/lyx_32x32.ico development/Win32/packaging/icons/lyx_32x32.png development/Win32/packaging/icons/lyx_32x32_16colors.png development/Win32/packaging/icons/lyx_doc_32x32.ico development/Win32/packaging/icons/lyx_doc_32x32.png development/Win32/packaging/icons/lyx_doc_32x32_16colors.png development/lyxserver/perl/LyX-Client-0.01.tar.gz development/lyxserver/perl/LyX-Polite-0.01.tar.gz lib/images/banner.ppm Moreover, the problem will only get worse when the 498 .xpm files in lib/images are converted to .png format the day we drop XForms. -- Angus
Re:Windows installer finished
Angus Leeming wrote: > Find it at http://www.devel.lyx.org/~leeming/lyx_setup_136.exe Hi Angus, first many thanks for your work. I tested it and have the following annotations: - Your installer asks for several programs. If one of these are not present, the installer opens the download/install webpage of the programs but jumps to the next step without waiting until the missing program is installed. - Is it really necessary to install msys, perl and phthon fully? Its download size is huge (23 MB for Perl + Python) for modem/ISDN users. Ruurd only used the executables perl.exe, python.exe and sh.exe. Is it possible to include them to the installer package? Because it would ease the installation a lot. - While LyX reconfigures, it doesn't recognize acrobat: +checking for "acrobat"... no altough it is installed on my system. The reason is that the acrobat.exe istn't in the PATH but in the registry (as most of Win programs) under HKEY_LOCAL_MACHINE\Software\Classes\Applications\ Is it possible to search the registry for this entry? Acroread under Win is stored in the same registry path. - I installed LyX with admin permissions and it appeared with german menus like I wanted. But when I change to a user account it appeared with english menus. To get german ones for the user account I had to create a new environment variable named "LANG" with the value "de_DE" as described at Ruurds LyX-webpage. Is it possible to add this variable automatically for every user account? - At last a crash report of your LyX 1.3.6CVS version: Go to Edit -> Preferences, choose "Paths" and press one of the "Browse" buttons for the working directories. Result CRASH: lyx: SIGSEGV signal caught Sorry, you have found a bug in LyX. Please read the bug-reporting instructions i n Help->Introduction and send us a bug report, if necessary. Thanks ! Bye. abnormal program termination That's all for the moment. thanks for the installer again and regards Uwe p.s. Please CC all replies to my email adress as I'm not on the lyx-devel list.
Re: Windows installer finished
On Monday 06 June 2005 14:38, Uwe Stöhr wrote: > Angus Leeming wrote: > > Find it at http://www.devel.lyx.org/~leeming/lyx_setup_136.exe > > Hi Angus, > > first many thanks for your work. I tested it and have the following > annotations: > > - Your installer asks for several programs. If one of these are not > present, the installer opens the download/install webpage of the > programs but jumps to the next step without waiting until the > missing program is installed. Yes. But once you've grabbed everything it asks you to restart the LyX installer so that it can find your newly installed packages. > - Is it really necessary to install msys, perl and phthon fully? No. You need Perl only if you're going to use reLyX. I tried to explain that in the accompanying commentary. > Its download size is huge (23 MB for Perl + Python) for modem/ISDN > users. Ruurd only used the executables perl.exe, python.exe and > sh.exe. Is it possible to include them to the installer package? Just thinking out loud... Perhaps: [x] Download the official Python distribution [ ] Install a stripped-down and maybe "good enough" Python.exe But this will involve me in a lot more work with NSIS getting to grips with optional installation of components. I'm running out of LyX-time. I'm also running out of NSIS-enthusiasm. > Because it would ease the installation a lot. > > - While LyX reconfigures, it doesn't recognize acrobat: > > +checking for "acrobat"... no > > altough it is installed on my system. The reason is that the > acrobat.exe istn't in the PATH but in the registry (as most of Win > programs) under > > HKEY_LOCAL_MACHINE\Software\Classes\Applications\ > > Is it possible to search the registry for this entry? Acroread > under Win is stored in the same registry path. I could, but where do I stop? Acroread certainly isn't required to get LyX running and we already have 6 download pages and a complaint that we're downloading too much stuff... It's the sort of thing that really belongs in the configure script (which needs to be rewritten in some other language if it is to understand the registry...) > - I installed LyX with admin permissions and it appeared with > german menus like I wanted. But when I change to a user account it > appeared with english menus. To get german ones for the user > account I had to create a new environment variable named "LANG" > with the value "de_DE" as described at Ruurds LyX-webpage. Is it > possible to add this variable automatically for every user account? It may be possible, but I don't know how... > - At last a crash report of your LyX 1.3.6CVS version: > Go to Edit -> Preferences, choose "Paths" and press one of the > "Browse" buttons for the working directories. Result CRASH: Thank you. I see it on linux too. > That's all for the moment. > > thanks for the installer again and regards > Uwe > > p.s. Please CC all replies to my email adress as I'm not on the > lyx-devel list. Angus
Re: Windows installer finished
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes: >> - While LyX reconfigures, it doesn't recognize acrobat: >> >> +checking for "acrobat"... no >> >> altough it is installed on my system. The reason is that the >> acrobat.exe istn't in the PATH but in the registry (as most of Win >> programs) under >> >> HKEY_LOCAL_MACHINE\Software\Classes\Applications\ >> >> Is it possible to search the registry for this entry? Acroread >> under Win is stored in the same registry path. Angus> I could, but where do I stop? Acroread certainly isn't required Angus> to get LyX running and we already have 6 download pages and a Angus> complaint that we're downloading too much stuff... Angus> It's the sort of thing that really belongs in the configure Angus> script (which needs to be rewritten in some other language if Angus> it is to understand the registry...) What should be done if possible is to check whether there is an association for .pdf files. In this case, "start" (with appropriate options) should be used as a viewer. Actually, it may be possible to do that on win 2k/xp from a shell script. Does NSIS offer support for querying such things? Another solution would be to have code in LyX packaging to do this kind of things (and for MacOSX too). JMarc
Re: Windows installer finished
On Monday 06 June 2005 14:38, Uwe Stöhr wrote: > - At last a crash report of your LyX 1.3.6CVS version: > Go to Edit -> Preferences, choose "Paths" and press one of the > "Browse" buttons for the working directories. Result CRASH: Fixed in 1.3.x CVS. Angus
Re: Windows installer finished
On Monday 06 June 2005 14:38, you wrote: > - Your installer asks for several programs. If one of these are not > present, the installer opens the download/install webpage of the > programs but jumps to the next step without waiting until the > missing program is installed. > > - Is it really necessary to install msys, perl and phthon fully? Also, why not point the installer to the folder containing Ruurd's stripped down versions? Or something equivalent. Basically, I don't want to commit myself to creating, testing and maintaining these things. If some Windows user has the motivation, then that's great. I can certainly adapt the installer to open up a web page pointing to the stripped down version in addition to the official version. Angus
Re: Windows installer finished
Angus Leeming wrote: Its download size is huge (23 MB for Perl + Python) for modem/ISDN users. Ruurd only used the executables perl.exe, python.exe and sh.exe. Is it possible to include them to the installer package? Just thinking out loud... Perhaps: [x] Download the official Python distribution [ ] Install a stripped-down and maybe "good enough" Python.exe I prefer the latter one as it saves a lot of installation time and hard disk space. But why should it be installed? Isn't it possible to add an actual version of perl.exe, python.exe and sh.exe (and some dll's if needed) to the folder LyX\bin\ and use these files for lyxrc.defaults by default (as Ruurd did)? This solution shouldn't produce NSIS troubles. Is it possible to search the registry for this entry? Acroread under Win is stored in the same registry path. I could, but where do I stop? Acroread certainly isn't required to get LyX running and we already have 6 download pages and a complaint that we're downloading too much stuff... My proposal for python etc. above would reduce it to 3 download pages. And you don't need to ask for installing acrobat/acroread as it is installed on 99% of Win PC's. It is enough to check if the program is installed and add the excutable to the viewers for PDF in the preferences. --- I forgot to ask you how you handle the math fonts. I didn't see that they were installed. Just copy them to the \Windows\Fonts directory isn't enough, but I don't know how to invoke the correct installation by the command line. So it is OK if the user has to install them manually before installing LyX. Btw. Ekkehardt Schlicht created a special font package using the BaKoMa math fonts: http://wiki.lyx.org/uploads/LyX/WindowsSetup/bakoma4lyx.zip But they are not free for commercial use as written in the license. regards Uwe
Re: Windows installer finished
Jean-Marc Lasgouttes wrote: >>> - While LyX reconfigures, it doesn't recognize acrobat: >>> >>> +checking for "acrobat"... no >>> >>> altough it is installed on my system. The reason is that the >>> acrobat.exe istn't in the PATH but in the registry (as most of Win >>> programs) under >>> >>> HKEY_LOCAL_MACHINE\Software\Classes\Applications\ >>> >>> Is it possible to search the registry for this entry? Acroread >>> under Win is stored in the same registry path. > > Angus> I could, but where do I stop? Acroread certainly isn't required > Angus> to get LyX running and we already have 6 download pages and a > Angus> complaint that we're downloading too much stuff... > > Angus> It's the sort of thing that really belongs in the configure > Angus> script (which needs to be rewritten in some other language if > Angus> it is to understand the registry...) > > What should be done if possible is to check whether there is an > association for .pdf files. In this case, "start" (with appropriate > options) should be used as a viewer. See development/Win32/packaging/installer/lyx_path_prefix.C which adds shell code to the configure script to generate a path_prefix entry in lyxrc.defaults and which goes on to use "start" to run "sh.exe configure". It would be quite simple to add more functions to this .dll. Perhaps: add_path_prefix() add_viewer() run_configure() but I'm getting a little edgy about all this. It appears that the goal is getting wider and wider. That means more work (primarily for me, it would appear), more chance of bugs and no great benefit AFAICS. I also don't think that we have to do everything this time around. Moreover, I think we're forgetting that our users are capable individuals who can do stuff themselves too... > Actually, it may be possible to > do that on win 2k/xp from a shell script. Does NSIS offer support for > querying such things? Another solution would be to have code in LyX > packaging to do this kind of things (and for MacOSX too). > > JMarc -- Angus
Re: Windows installer finished
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes: Angus> See development/Win32/packaging/installer/lyx_path_prefix.C Angus> which adds shell code to the configure script to generate a Angus> path_prefix entry in lyxrc.defaults and which goes on to use Angus> "start" to run "sh.exe configure". That's evil, isn't it? ;) As far as adding prefs for windows is concerned, one possibility is to provide a preference file that contains everything needed by LyX/Win32. Generating this file is probably easier than hacking lib/configure. LyX/Mac follows this route (provide a preconfigured preference file). Angus> but I'm getting a little edgy about all this. It appears that Angus> the goal is getting wider and wider. That means more work Angus> (primarily for me, it would appear), more chance of bugs and no Angus> great benefit AFAICS. Yes. Angus> I also don't think that we have to do everything this time Angus> around. Moreover, I think we're forgetting that our users are Angus> capable individuals who can do stuff themselves too... I think you have to draw a line at some point. Fell free to do it. JMarc
Re: Windows installer finished
Angus Leeming wrote: - Is it really necessary to install msys, perl and phthon fully? Also, why not point the installer to the folder containing Ruurd's stripped down versions? Or something equivalent. That's a good solution. Basically, I don't want to commit myself to creating, testing and maintaining these things. I understand you. I can certainly adapt the installer to open up a web page pointing to the stripped down version in addition to the official version. I'll create a wiki page with all needed files. I think we need - perl.exe - perlx.dll - phython.exe - pythonxx.dll - sh.exe What other files are needed? regards Uwe
Re: Windows installer finished
On Monday 06 June 2005 15:54, Uwe Stöhr wrote: > > I can certainly adapt the installer to open up a > > web page pointing to the stripped down version in addition to the > > official version. > > I'll create a wiki page with all needed files. > > I think we need > > - perl.exe > - perlx.dll > - phython.exe > - pythonxx.dll > - sh.exe > > What other files are needed? That's precisely my point. I don't know. Off the top of my head: sed, tr, cut, grep... Angus
Re: Windows installer finished
On Monday 06 June 2005 15:39, Uwe Stöhr wrote: > Angus Leeming wrote: > >>Its download size is huge (23 MB for Perl + Python) for > >> modem/ISDN users. Ruurd only used the executables perl.exe, > >> python.exe and sh.exe. Is it possible to include them to the > >> installer package? > > > > Just thinking out loud... > > Perhaps: > > [x] Download the official Python distribution > > [ ] Install a stripped-down and maybe "good enough" Python.exe > > I prefer the latter one as it saves a lot of installation time and > hard disk space. But why should it be installed? I think we have a terminology confusion. To me, "installed" just means "unpack it somewhere usable". The registry is a [EMAIL PROTECTED] nightmare for which I see little need. > Isn't it possible to add an actual version of perl.exe, python.exe > and sh.exe (and some dll's if needed) to the folder LyX\bin\ and use > these files for lyxrc.defaults by default (as Ruurd did)? This > solution shouldn't produce NSIS troubles. This solution will require 1. Extensive testing to ascertain which modules are needed by the scripts that LyX packages. 2. Ongoing maintenance. Moreover, if I already have the "official" stuff installed why would I want to use the "unoffical" stuff? Adding code to NSIS to conditionally install something is most certainly a "NSIS trouble" that I'm keen to avoid right now. > >>Is it possible to search the registry for this entry? Acroread > >>under Win is stored in the same registry path. > > > > I could, but where do I stop? Acroread certainly isn't required > > to get LyX running and we already have 6 download pages and a > > complaint that we're downloading too much stuff... > > My proposal for python etc. above would reduce it to 3 download > pages. No it wouldn't. It would just change three of them to: Would you prefer to use these stripped down things that people have been discovering don't work (see countleess reports on the lyx-users list), or would you prefer to use an official version which definitely will work but which requires a big download committment. > And you don't need to ask for installing acrobat/acroread as > it is installed on 99% of Win PC's. It is enough to check if the > program is installed and add the excutable to the viewers for PDF > in the preferences. I'd rather reuse the code I do have than rewrite something especially for acroroead... > I forgot to ask you how you handle the math fonts. I didn't see > that they were installed. Just copy them to the \Windows\Fonts > directory isn't enough, but I don't know how to invoke the correct > installation by the command line. So it is OK if the user has to > install them manually before installing LyX. I do nothing about math fonts. Nor do I plan to do anything about math fonts before LyX 1.3.6 is released. I think that we should focus on getting what we have to work rather than trying to get it to do more and more and more. > Btw. Ekkehardt Schlicht created a special font package using the > BaKoMa math fonts: > http://wiki.lyx.org/uploads/LyX/WindowsSetup/bakoma4lyx.zip > But they are not free for commercial use as written in the license. I seem to remember that the consencus was that the licence was too restrictive for us to distribute them. Angus
Re: LyX meeting in Paris. What about July 14-18?
I already bought my tickets today, so most probably we will meet up in Paris #:O) ~Jürgen -- -._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._ Dr. Jürgen VignaE-Mail: [EMAIL PROTECTED] Mitterstrich 151/A I-39050 SteineggWeb: http://www.lyx.org/~jug -._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._-._
Re: Windows installer: NSIS vs .msi
Angus Leeming wrote: But this will involve me in a lot more work with NSIS getting to grips with optional installation of components. I'm running out of LyX-time. I'm also running out of NSIS-enthusiasm. While having any sort of installer is better than none at all, I'd suggest future effort re: Windows installers aim for an .msi target (Microsoft Installer) format. Then, those of us in AD environments can enjoy pain-free deployment to (all) client machines. (-: -- Rex p.s. I'll be doing it myself for lyx-1.3.6 as a proof-of-concept (using the Advanced Installer tool).