The 08 september 2006 build is out.

[orefa]

What would be like, superb, is a message underneath that "Latest official release", reading: "Latest nightly build"?

Yes, new wording would help! Technically savvy people are notoriously poor at communicating with the public at large, a normal hazard of the profession.

Along these lines I suggest not to use the term "build" when refering to the latest version. Just use "version". When I first read "nightly build" I wondered why developpers were reporting on their workouts. I have programmed for years but never used that term to refer to a version. Well, maybe that's just me. But also the term may not be clear to a number of visitors who just started programming in C++ and don't know what it refers to.

While we're on the subject, since RC2 turned out not to be a release candidate after all, why not take a small step back on nomenclature and a giant step forward in useability: publish a new Beta! Forget about the so-called RC2 from over a year ago. Release today's version as a new "Beta 20060908" for example. Then provide the same link to lead visitors to the nightly builds (as daily Beta versions). The world needs to know!  :lol:

[tiwag]

I can confirm this, probably you have some special characters in your file.

@Yiannis : in windows (Xp) : open up the Cb project and open the tinyxml sources --> tinystr.cpp will show some of those strange blocks

you can get around this when you do "Edit->File Encoding->ASCII" and confirm the reload dialog

then CB loads the file and shows correctly.

i had this issue since a long time, when i got source code files with swedish characters in it ...
with forced ascii encoding, some of them get probably mismatched, but it mostly doesn't harm
because they are all in comments (which i even can't read)
but you can save the file and work with it.

annoying is only, that you have to do this everytime when opening this file ...
maybe we could add a file encoding property to the *.layout file
when the user had choosen a specific encoding from the available encodings ?

brgds, tiwag

[Kazade]

I don't know what changed last night, but now the save bug (where it crashes on CTRL+S) on linux is back! It was fixed in the previous build.

[Poobah]

-File->New Project
-Pick any project
-Project->Properties
-Select "Targets" tab"
-Output filename reads "\test.exe" assuming that the project was called "test".

Shouldn't it default to ".\test.exe" in this case?

[torikx]

Many people seem to think so and I must say I 'm confused as to "why"...
Mind telling us why you thought the project had stalled?

If you dont mind me barging in, I'd like to comment as well:

I think the most prominent reason is the absence of a stable release for almost a year following something labeled as a Release Candidate.
It's my understanding that an RC is supposed to have no more features added, have no grave bugs to be fixed, have a life no longer than several months during which its developers focus solely on elimination of showstoppers and rough edges until final release.
It's also my understanding that a Nightly Build conveys a bleeding-edge, alpha state of software. I find the statement "pretty stable nightly builds" contradictory, it's like an oxymoron. If you think it's stable/usable, then take it out of that cage and release under a name that'll attract people (no other RCs, thank you).

Correct me if wrong, but it seems C::B is like trapped in a cycle of adding features, fixing bugs, adding more features, fixing more bugs... I think it's time to stop adding features, focus on bugs, drop features for now whose bugs need lots of effort to fix and prepare for a final release.
Please take no offense, just my 2c.

[Game_Ender]

Currently I believe CB is working toward 1.0 based on this roadmap.  There is no developers mailing list (and I don't watch the IRC channel) so I can't tell exactly what is planned, but from the roadmap we are pretty close to feature freeze.  Mandrav has basically finished the Code-Completion issues and I believe Thomas is slaving away at the XML based compilier system.

I too agree that we need to chnage the version number scheme but this topic has been argued to death here on the forums.  After RC1, we basically went back to alpha because Mandrav thought there were to many "show stoppers".  It would almost make sense to release Code::Blocks 0.5.0 (or 0.9.0 or any 0.x.0) on the website with an explanation that its a developement release but much better than RC1.

Why Nightly Builds?  Because that's what they are called, they are not new versions because a version implies a release.  Nightly builds are basically for users who don't want to have to compile from SVN but want to try out all the new wizbang features.  Firefox does the same thing.

[mandrav]

Thank you all for sharing your thoughts about the main website.
We have now replaced the link to 1.0rc2 in the front page with a link to the downloads page, which has been also updated with a few words describing what's going on.
Let's hope people won't get confused about the project's activity anymore .

[Encryptor]

Very nice update of the Download page. A long and clear explanation of what's going on. This should reduce dramatically the number of confused users.

[torikx]

In editor settings, mouse drag scrolling fonts are terribly small (this is linux xfce desktop, wxgtk-2.6.3)
Anyone having this problem?


[attachment deleted by admin]

[Pecan]

In editor settings, mouse drag scrolling fonts are terribly small (this is linux xfce desktop, wxgtk-2.6.3)
Anyone having this problem?

On my windows and ubuntu linux, the dragScroll dialog font is the same size as the icon titles font.
 
Yet, the fonts look different sizes on your system.

I don't know what to do about that.

[takeshimiya]

In editor settings, mouse drag scrolling fonts are terribly small (this is linux xfce desktop, wxgtk-2.6.3)
Anyone having this problem?

On my windows and ubuntu linux, the dragScroll dialog font is the same size as the icon titles font.
 
Yet, the fonts look different sizes on your system.

I don't know what to do about that.

Hint, look at dragscrollcfg.cpp, you're calling SetFont() in various places.

[Pecan]

Hint, look at dragscrollcfg.cpp, you're calling SetFont() in various places.

Well... yes. That's the original wxSmith code.
Are you suggesting to not use SetFont() at all?

Or are you suggesting to set larger fonts for xfce linux, but have it look huge on windows and ubuntu?

I'm not very font savey.


 

[takeshimiya]

Hint, look at dragscrollcfg.cpp, you're calling SetFont() in various places.

Well... yes. That's the original wxSmith code.
Are you suggesting to not use SetFont() at all?

Or are you suggesting to set larger fonts for xfce linux, but have it look huge on windows and ubuntu?

I'm not very font savey.

I'm suggesting that if you're goint to use SetFont(), take system font's settings (ie. don't hardcode "Arial").
But in this case it's better to remove all of them, same applies for SetBackgroundColour(), etc, it is not required and it's using a different color than what should be.

[tiwag]

Windows ANSI build is here.

this version doesn't work (at least on my) Win98-SE,
because it wants SHGetfolderPath exported from shell32.dll,
which isn't supported on W98

i can't understand why people create, announce and host such builds
and are not doing a simple testrun if it works  :?

really sloppy ...

btw. i solved the issue and will post the solution in another thread ...

[Pecan]

I'm suggesting that if you're goint to use SetFont(), take system font's settings (ie. don't hardcode "Arial").
But in this case it's better to remove all of them, same applies for SetBackgroundColour(), etc, it is not required and it's using a different color than what should be.

Thanks, I'll work on that.