Time for an update then. The biggest addition is the updated rightclick menu of the editor, which features a variable browser. Also, the guys at MinGW have ported a newer branch of GCC, so I've added it to Dev.
Changes - Version 5.0.0.4 - 25 September 2011
- Upgraded the compiler to GCC 4.6.1.
- The main tab control now has the optional (check Environment options) ability to spread tabs over multiple lines.
- Added Goto Declaration and Goto Definition buttonsto the editor right click menu. These buttons make code navigation much easier.
- The old code for Ctrl+Click now uses the new code for the buttons and uses two-way navigation.
- Added Profiling to the Compiler toolbar.
- Removed the useless About/Help toolbar.
- The splash screen now shows what Dev is currently loading.
- Code Completion Cache can now also be browsed through.
- Added icons for profiling options and added them to the compiling toolbar.
- Updated the Incremental Search window: it now has a previous and next button.
- Corrected a few regression bugs in the class scanner.
- Fixed a bug in the tooltip content generator.
- And much more minor stuff...
Important notices
- The options format has changed. If you want to reuse an old pre-4.9.9.3 config file (NOT recommended), or, more importantly, when you're overriding Compiler Options in your project, you need to re-set these project settings once and save the project. You'll then have an updated 4.9.9.3+ project file.
- This version has GCC built-in instead of being an aditional package. It also contains D3D9/10/11, GDI, Win32 and OpenGL headers and libraries in that flavor.
- This version is now fully portable. If you also don't want Dev to leave anything behind in the registry, please select "Portable" or "Minimal" in the setup options.
- For ultimate portable programming, please launch devcppPortable.exe located in the main folder of the portable zip download. This will make dev save its configuration files in the same folder as the executable.
Download
The setup can be downloaded here. The Portable zip version can be downloaded here. The source code can be found here.
Beta update
The 5.0.0.5 Beta 12 update can be found here. Its source code can be found here.
Problems / Upcoming changes / TODO
- TODO: Implement C++0x suggestions by Xazax.
- TODO: Fix the ParentID bug in the tooltip and goto menu items.
- FIXED: Fix the bug reported by Anonymous (first comment on 4.9.9.8).
Will Dev-C++ Portable copy font files in order to make the font available on other computers?
ReplyDelete_bfd_window_internal *i --this error causes the write behind buffer, which takes place during parsing the header file...a memory leak..
ReplyDelete@cin.getline:
ReplyDeleteNope, it doesn't, but I sticked with MS Sans Serif, which should be available on 99% of users' PC's.
It's exactly the reason why I didn't use that fancy Segoe UI font Windows 7 uses for its UI: Segoe is unavaible on PC's that don't have Office 2007 or Vista or later installed...
@Anonymous:
I don't really understand. Can you rephrase that?
As a human, I must commend you on your hard word for restoring Dev-C++ back to its former glory. Great job! :D
ReplyDeleteWell, sounds terrible.
ReplyDeleteUm... I have confronted another problem. It pop out "[Warning] command line option '-std=c++0x' is valid for C++/ObjC++ but not for C [enabled by default] " in the case of the source code was written by C......
I know it's nothing to do with Dev-C++, but the setting does confuse gcc.
Could it be fixed???
And still another problem is that if I type spaces in a line that has contained nothing, as long as I switch to other line and switch back, the spaces will disapear... @_@
I think it's not a bug, but it does make Dev-C++ less convenient...
@cin.getline:
ReplyDeleteYou probably set Project >> Project Options >> Compiler >> Code Generation >> Language Standard to something C++0x-related. I suggest to ISO or GNU C++ here. Version 5.0.0.5 wil add some C99 options.
This option can also be found in the global compiler options (Tools >> Compiler Options >> Settings >> ...), but these are ignored if a project is open.
The other bug/feature can be disabled by Tools >> Editor Options >> Keep Trailing Spaces.
@Anonymous:
The 4.9.9.8 bug is fixed in 5.0.0.5 RC1 which I'll upload next week:
http://gamerneeds.org/bestanden/newmouseover.png
@orwells
ReplyDeletethat is a good news!!!!!
i hope 5.0.0.5 release well stable as 4.9.9.9
did try my own build?
Oh, I haven't seen that.
ReplyDeleteThank you!
try devc++ in solaris 10 it runs
ReplyDelete@Anonymous1:
ReplyDeleteI would really like to see your build, but you keep sending me links to it I can't access.
Please, try emailing it!
Thank you so much for upgrading and pumping life into it. You should link this page on the main page of bloodshed.net so we can find it easy :) I had to dig on wikipedia to find your site. keep up the good work.
ReplyDeleteMrJake
5.0.0.5 beta 7 -- very bad build now ide can no longer see the variable declared in int, long, double, unsigned expt. char and alike.
ReplyDeletethe repository of the parsed header location -- bug is located.
ReplyDeleteide 5.0.0.5 can no longer locate ::cout, ::endl, and other templetized identifier, upon crtl+right+click/mouse override..did you remove the repository stack in devc++.map.
ReplyDelete@Anonymous:
ReplyDeleteIt's exactly the reason why it's a beta build.
Everything can be found/goto'd, except:
1) function/loop-local variables
2) template/namespace variables
Oh, devcpp.map contains debugging information, not header scans.
This code-reading code is troubling me. :(
@MrJake:
I've tried to contact the old team a couple of times, but they aren't responding unfortunately...
Two problems:
ReplyDelete1) When I changed the language, it tell me it have got in a trouble.
2)Anytime I check the "Use fast but imperfect dependency generation." off, it will be recovered next time.
THX for your support!
@cin.getline:
ReplyDelete1) I've tried changing it in the first time config thing and in Environment Options, but I can't produce a crash. Can you elaborate?
2) Looks like I forgot about that one and compile delay. Looks like I fixed it in Beta 8. :P
@cin.getline:
ReplyDeleteBeta #8 is up now, which contains the promised fixed (at least, I hope so, because I thought I fixed these in 4.9.9.3).
And for anyone wondering when the real 5.0.0.5 gets released, here's the current TODO list:
1) statement list index out of bounds when reparsing...
2) local scope variable finding support
3) enable support for non-class function parameter finding, use mousepos, not cursorpos
4) tooltip enable/disable switch
5) autosave
---for convenient search for dev c++ upgraded version i added this link to brothersoft.com
ReplyDeletein news.txt please include your link
ReplyDeletehttp://orwellengine.blogspot.com/ for every update you made.
May I request something?
ReplyDeleteAdding "-std=c++0x" for a command line option by default, so that anyone who uses this can get in the C++11 awesomeness.
@imnotanerd
ReplyDeletetools/compiler option/code generation/language standard --there you can select.
Application version: 5.0.0.4
ReplyDeleteMachine info
---------
Platform : Windows NT
OS version : version 6.0 (build 6002)
Additional info: Service Pack 2
The following error occured in version 5.0.0.4:
Failed to set data for '' (at address 0x00148D82)
Please include a description of what you were doing before the error occured (please give as much precisions as possible) :
I changed the language, and crashed.
State information follows:
Stack trace:
------------
00288C86 (00187C86): ShowExceptionInfo (ExceptionsAnalyzer - 564)
00288E7B (00187E7B): TExceptionsAnalyzer.GlobalExceptionHandler (ExceptionsAnalyzer - 572)
00209F53 (00108F53): RegisterFiletype (FileAssocs - 195)
00209BE8 (00108BE8): Associate (FileAssocs - 133)
00209D24 (00108D24): Associate (FileAssocs - 151)
0022E78F (0012D78F): TEnviroForm.btnOkClick (Envirofrm - 358)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
002602E6 (0015F2E6): TCompiler.WriteMakeObjFilesRules (compiler - 403)
00251A0E (00150A0E): TMainForm.actEnviroOptionsExecute (main - 3572)
00251AE5 (00150AE5): TMainForm.actEnviroOptionsExecute (main - 3592)
00251B02 (00150B02): TMainForm.actEnviroOptionsExecute (main - 3592)
0028B297 (0018A297): (devcpp - 222)
0028B2B5 (0018A2B5): (devcpp - 222)
0028AAF0 (00189AF0): (devcpp - 131)
0028AAF0 (00189AF0): (devcpp - 131)
Please, in the compiler option to add:
ReplyDeleteProcessor - i7
Optimization - 0fast
Please ALL option processor:
ReplyDeletegeneric
native
i386
i486
i586,
pentium-mmx
pentiumpro
i686
pentium2
pentium3
pentium-m
pentium4, pentium4m
prescott
nocona
core2
corei7
corei7-avx
atom
k6
k6-2, k6-3
athlon, athlon-tbird
athlon-4, athlon-xp, athlon-mp
k8, opteron, athlon64, athlon-fx
k8-sse3, opteron-sse3, athlon64-sse3
amdfam10, barcelona
winchip-c6
winchip2
c3
c3-2
geode
@imnotanerd:
ReplyDeleteYou can indeed edit Tools >> Compiler option >> code generation >> language standard, which will also set the standard option for any new projects.
@cin.getline:
Well, I've tried editting the language a couple of times in a couple of places, but I can't generate a crash. Are you able to reproduce the crash? If so, please tell me how.
@Anonymous:
You could add these to the compiler option textbox (like '-march=whatever'.
I've added the missing -march and the Ofast options to 5.0.0.5.
With some luck I'll get a stable enough 5.0.0.5 done this weekend, so you folks will be able to torture it.
I look forward to programming the hell out of 5.0.0.5! :D
ReplyDeleteAuto-complete pl0x?
Progress report:
ReplyDelete1) Local variable finding is working now (95% of cases).
2) Editor tooltips now also show up on first start.
3) Added more flags to Compiler Options.
TODO:
1) Add tooltip on/off option (ez).
2) Fix the bug mentioned by cin.getline (?).
3) Autosave.
4) Function finder for regular functions outside classes.
TODO:
ReplyDelete1) Fix the bug mentioned by cin.getline (?).
2) Autosave.
3) Function finder for regular functions outside classes.
Important Update!
ReplyDeletehttp://gamerneeds.org/bestanden/SOURCE%205.0.0.5%20RC9.7z
http://gamerneeds.org/bestanden/Dev-Cpp%205.0.0.5%20RC9%20Overwrite%20Update.7z
I *know* this version contains the so-called ParentID bug (keep an eye on the statusbar when using variable hovering with the mouse). Dev will warn you (but NOT crash) about/because of the error! It's safely encapsulated in an if/else branch.
This bug is exactly the reason why I politely request anyone using this to confirm Dev won't crash or do anything not-so-nice because of it.
http://gamerneeds.org/bestanden/Dev-Cpp%205.0.0.5%20RC9%20Overwrite%20Update.7z
ReplyDelete--exception overflow during debugging...
Problem signature:
ReplyDeleteProblem Event Name: APPCRASH
Application Name: devcpp.exe
Application Version: 5.0.0.5
Application Timestamp: 2a425e19
Fault Module Name: KERNELBASE.dll
Fault Module Version: 6.1.7601.17514
Fault Module Timestamp: 4ce7b8f0
Exception Code: 0eedfade
Exception Offset: 0000b760
OS Version: 6.1.7601.2.1.0.256.1
Locale ID: 1033
Additional Information 1: d1a0
Additional Information 2: d1a047955475a709b95e8bce27fa2a6e
Additional Information 3: a4cd
Additional Information 4: a4cde0b8f597860c6bb144cbad2b2f76
--orwell are you the only one tweaking this package?coz if it is you maybe need to build a community for this project in order to make is more stable..2 heads are better than one.
ReplyDeleteFantastic!!!
ReplyDeletePlease option generic code of "Language standard (-std)" to add "None".
In "Optimize level" in the Russian -> "Уровень оптимизации"
In "Optimize for march..." -> "Оптимизация под процессор"
Is NEW options? -> (-m32) for compiling in 64bit to 32bit?
ReplyDeleteAnd -> (-m64) for compiling in 32bit for 64bit.
ReplyDeleteEros:
ReplyDeleteis original compiler GCC to make eror of
# Makefile created by Dev-C++ 5.0.0.5 Beta 9
RM = rm -f
a see of:
RM = rm.exe -f
else
ReplyDelete${RM} $(OBJ) $(BIN)
to
$(RM) $(OBJ) $(BIN)
cc1plus.exe: error: invalid option argument 'Ofast' ((( not compiller((
ReplyDelete@Anonymous:
ReplyDeleteUhm, GNU C++ (-stdgnu++98) is the actual default for GNU GCC. This is 100% equivalent to sending no -std flag to GCC.
I'll add these Russian translations! Thanks.
Hmm, I don't think MinGW can compile for 64bit yet, because MinGW itself isn't 64bit either.
About {} and (): they both work as variable inserters, but can you explain why () would be better (the {} should be there for a reason)?
Uhm, Ofast works just fine here...
Ofast, of automatic configure Make error, else it can.
ReplyDelete"Hmm, I don't think MinGW can compile for 64bit yet, because MinGW itself isn't 64bit either."
ReplyDeleteI tried. Collects.
Specify let for bin to a folder and all.
Install http://tdm-gcc.tdragon.net/download
end option compiler NO -m34, other in 32bit for 64bit. ;)
*-m32
ReplyDelete"Hmm, I don't think MinGW can compile for 64bit yet, because MinGW itself isn't 64bit either."
ReplyDeleteI tried. Collects.
Specify let for bin to a folder and all.
Install http://tdm-gcc.tdragon.net/download
end option compiler NO -m32, other in 32bit for 64bit. ;)
Some translations from English to Chinese(TW).
ReplyDeleteThe translations below are in the "Tools" → "Compiler Options" → "Settings".
"Code warnings" → "程式碼警告"
"Show most warnings" → "顯示所有警告"
"Show some more warnings" → "顯示額外警告"
"Check ISO C/C++/C++0x Conformance" → "嚴格遵守 ISO C/C++/C++0x 標準"
"Strip executable" → "移除所有除錯資訊"
"Optimization Level" → "最佳化等級"
"Language standard" → "語言標準"
The translations below are in the "Editor options".
"Ensure that file ends with new line." → "確保檔案以換行結束"
"Append closing braces" → "自動加上靠近的括號"
"Highlight current line" → "強調目前所在的行"
"Show inherited member" → "顯示繼承的成員"
THANK YOU!!
(You can ignore the bug I mentioned serveral days ago. Actually, it won't cause any fatal errors. )
@Anonymous (I take you're the same guy):
ReplyDeleteNo offence, but I honestly don't really understand what you're trying to explain, except from pointing me to a 64bit MinGW.
@All who are supplying non-western translations:
Dev-C++ uses Microsofts 8bit-character based sets, like the russian set called Windows-1251. I'm having some trouble editting these files with my Windows-1252 OS...
I'd be delighted if anyone would:
1) Open up English.lng
2) Find the sentence that isn't translated to your language. Write down the number before it.
3) Open up YourLanguage.lng.
4) Find that same index. It probably isn't there. If it IS, replace the text behind it by your translation.
5) If it's not there, create it and append '=(translation)' to it.
If to replace the compiler on 64 bit, it will work also in 32 bit system for it 32 bit, but by default generates a code for 64 bit cars.
ReplyDeleteIn this case to generate for 32 bit it is necessary to add-m32.
I have mailed a new translation to you. Do you copy?
ReplyDelete@cin.getline:
ReplyDeleteI inserted the file in the lang folder. Seems to work fine.
@Anonymous:
That x64 version of MinGW is quite interesting. However, I first want 5.0.0.5 to be stable.
Here's a TODO list:
1) Fix the language change bug reported by cin.getline.
2) Let Anonymous translate some russian text.
3) It'd be great if I could fix the ParentID bug.
Thanks big!!!
ReplyDeleteNow all works, I think it is possible to let out stable release.
To you it is very grateful!!!
Some text can't be found in the English.lng like "Append closing braces".
ReplyDeleteBag: it is impossible to create the new project!
ReplyDeleteI just wish to express a Thank You for your time and efforts in developing on this IDE !
ReplyDeleteI switched from MSVC++ 2008 to Dev-C++ for more than a year although I knew that developing for Dev-C++ stopped for several years. Just accidentally I stumbled over your blog here and am very pleased at your job!
P.S.: Is there a recent help file for all MinGW compiler command line options, too?
Thank you.
Wieland
Yes, thank you very much indeed!
ReplyDeleteHere is a small bug I found:
Whenever I type a closing normal bracket, i.e. ")", the program adds a second one which I then have to delete.
Since this wasn't mentioned before I fear I could be the only user affected. Wasn't able to find anything wrong with the code insertion settings, though.
And another thing: did you think of contacting the people over at "wxDevCpp" (wxdsgn.sourceforge.net)? They, too, are updating the original DevCpp source code, as it seems.
I am quite happy with having both projects installed for different purposes, but since both adopted the original "Dec-Cpp" user folder for config files, it's a big hassle to keep them from overwriting each other's settings. I know DevCpp 5 has the portable executable as an option, but when I bind *.c and *.cpp files to it, and then double-click a file, the portable version is no longer used, it seems.
Once again a big thank you for your efforts: they are very much appreciated. You should consider adding references to all your project pages to the Dev-C++ wikipedia entry, so that you get the attention you deserve.
Philo
wxDev-C++ uses the old version GCC, and I am not familiar with it, and that's the reason why I choose Dev-C++ 5 developed by Orwell. However, I think you should find somebody working with you in order to complete this program. Two man can work better.
ReplyDelete@Caha:
ReplyDeleteYes, it probably is. I'll upload 5.0.0.5 today.
@cin.getline:
I'll add these to the translations. Your previous submit has been implemented and updated too, as you know.
@Anonymous:
I can't fix your bag unfortunately, as I don't have a clue how to replicate it.
@Wieland:
Thank you! ;)
And, yes, there is. Its 99% equal to GCC.
http://gcc.gnu.org/onlinedocs/gcc-4.6.1/gcc/Invoking-GCC.html#Invoking-GCC
@Philo:
It does append a closing brace when typing ( for example, but it doesn't insert anything when typing ) though. You can disable this stuff in Editor Options if you want to test if it is to blame.
Hmm, the file binding problem sounds interesting. I'll have a look at it.
Pleas!
ReplyDeletetbSpecials->Constraint->MaxWidth=500
Not in all languages of a word same are long as in English
source/Vcl:
ReplyDeleteThe instruction for assemblage not the full!
Some *.DCU files which are necessary NOT compiling as are absent in *.dpk
sl.Add('Conroe=core2');
ReplyDeletesl.Add('Nehalem=corei7');
sl.Add('Sandy=corei7-avx');
to:
sl.Add('Conroe(core2)=core2');
sl.Add('Nehalem(corei7)=corei7');
sl.Add('Sandy(corei7-avx)=corei7-avx');
Pleas:
ReplyDeleteNewAllBtn->AutoSize -> True
InsertBtn->AutoSize -> True
ToggleBtn->AutoSize -> True
GotoBtn->AutoSize -> True
To Main Form.
It is necessary to let out release, you have changed a folder with drawings, updating imposed on fair installation DOES NOT WORK correctly.
ReplyDeleteFolders images aren't present (
NOIMG BITMAP ". \IMAGES\noimg.bmp"
"@Philo:
ReplyDeleteIt does append a closing brace when typing ( for example, but it doesn't insert anything when typing ) though. You can disable this stuff in Editor Options if you want to test if it is to blame."
Yes, I do have 'Append closing braces checked', as that is the setting I used in the old DevCpp.
But it no longer seems to work correctly:
When I type an opening brace, nothing happens; when I type a closing brace, the program adds a second one. This occurs even when no opening brace is present.
On another note: you might be interested in checking out this article: http://www.cplusplus.com/articles/36vU7k9E/
It explains why Cplusplus members "deprecated" DevCpp and advise everybody not to use it. Reasons given: too old and buggy, no longer updated.
The interesting part is that the article has last been edited in September, so they are clearly unaware of your endeavours here. I was going to leave a message in their forum pointing to your blog, but then I'm not the person who should do that because I cannot answer any questions about your project or future plans.
How can this be a continuation of the original Bloodshed Dev-C++? Your email and name is no where in the credited developers of the original team. How are you connected to Dev-C++? If you are just some fan working on continuing the ide/compiler then great. Just give the proper credit to the original team.
ReplyDelete@Anonymous:
ReplyDeleteThis is, as I've written myself on the wiki, an unofficial update/continuation of Dev-C++.
I haven't removed the pre-4.9.9.3 developers from the about section or anything either.
And, well, I've given up on contacting the original devs after trying about five times, which more or less implies they don't know me, which leads to me not being on their credit list.
Bottom line: I do give proper credit: I've decided not to replace any text containing 'Bloodshed.net' or anything related from Dev, and the wiki properly lists this one as an unofficial update!
Oh snap, Orwell DevCpp has a sourceforge page now? NICE!
ReplyDeletehttp://sourceforge.net/projects/orwelldevcpp/