PDA

View Full Version : [DevQt] General discussion



GreyGeek
10th February 2006, 20:55
Hi boys!

You don't want any help from members of the female gender? ;)
This forum has them, you know!

elcuco
11th February 2006, 00:49
You don't want any help from members of the female gender? ;)
This forum has them, you know!

Females... i heard about that rumor... like hobits... or elves... ;-)
Next time start writing in pink or something :D :D :cool:

wysota
11th February 2006, 02:44
Females... i heard about that rumor... like hobits... or elves... ;-)
Next time start writing in pink or something :D :D :cool:

Don't underestimate the power of the Dark Side, young Padawan!

abdulhaq
11th February 2006, 14:50
A few screenshots would be nice, and what about KDevelop? I'm sure they will like some competition but it's a lot to bite off....

michel
11th February 2006, 21:36
Oh, cripes! No wonder you're making an IDE. They took it out of 4.0! Nooooooo

Well, I guess you better hurry up and finish it then :)

wysota
11th February 2006, 23:13
Oh, cripes! No wonder you're making an IDE. They took it out of 4.0!

There never was an IDE in Qt. Designer 3 is not an IDE (but it's often confused as being such).

rh
12th February 2006, 02:31
There never was an IDE in Qt. Designer 3 is not an IDE (but it's often confused as being such).

Well there sort of was. The Designer had project management and a built in text editor. It lacked integrated compiling and debugging, which i guess you could argue is what constitutes an IDE. The editor and project manager have since been removed in Qt4.x.

wysota
12th February 2006, 11:16
Johan Thelin added the possibility to compile directly from Designer (http://www.digitalfanatics.org/e8johan/projects/deside/index.html) to make it work more like an IDE. But let's be honest, without this and class management, we can't say it was an IDE. The editor was so simple, it didn't give any tools to work with the code (except syntax highlighting and code completion).

jamadagni
17th February 2006, 11:34
"QT"? Because the latter has nothing to do with this site
What is QT then?

elcuco
17th February 2006, 11:46
QuickTime.

.. and it's funny seeing people discussing about a non existing project.
If you want this to get up, open a site, svn and start developing.

jacek
17th February 2006, 11:47
What is QT then?
Shortcut for QuickTime?

vitaly
19th February 2006, 22:50
What is the purpose to create new IDE from beginning? There are hundreds of attemps to create new IDE but no one is finished. Why emacs or vi are worse? I think that it will be better if one will try to create a perfect tool for refactoring purposes. May be it will be better to make Qt support in emacs?

vitaly
20th February 2006, 06:40
You may look at http://www.volkoeditor.com. It is being developed with use of Qt library. It have GNU license. May be you should start from this one because of some fatures are already implemented there.

michel
20th February 2006, 08:47
What is the purpose to create new IDE from beginning? There are hundreds of attemps to create new IDE but no one is finished. Why emacs or vi are worse? I think that it will be better if one will try to create a perfect tool for refactoring purposes. May be it will be better to make Qt support in emacs?

I think elsewhere fmc had mentioned the main reason for this as being a need to learn the Qt library. And yes, there are editors already which are great for writing source code in. The problem is finding one which is tightly integrated and geared specifically towards Qt development rather than a full-featured source code text editor which maybe has the ability to do debugging output (like VolkoEditor). Does VE let you view the class hierarchy of the program? Does it let you view signals and slots? I wouldn't know. There's too many problems trying to compile it on my machine using FreeBSD 6.0 and Qt 4.1 (about three "ISO C++ forbids...blahblah" errors, several structs used which aren't in any of the includes he put in the headers, and a Server.cpp that says:



#ifdef Q_WS_WIN

#include <utime.h>
#include <stdlib.h>
#include <winsock2.h>

#define sleep(x) _sleep(x)

const int MAXHOSTNAMELEN = 256;

#else // UNIX

#include <netdb.h>
#include <unistd.h>
#include <utime.h>
#include <rpc/types.h> // MAXHOSTNAMELEN
#endif // Q_WS_WIN


which assumes that if one isn't using Windows, then MAXHOSTNAMELEN will be in rpc/types.h even though: 1) afaik this is only true for Sun and Linux; and 2) it would make more sense to just write



#include <utime.h> // why does he include it twice in both the if and else statements?

#ifdef Q_WS_WIN
#include <stdlib.h>
#include <winsock2.h>
#define sleep(x) _sleep(x)
#else
#include <netdb.h>
#include <unistd.h>
#include <rpc/types.h>
#ifndef MAXHOSTNAMELEN
#define MAXHOSTNAMELEN 256
#endif
#endif


). BTW, there's another conversation in these DevQt threads about why it shouldn't use Regexps for syntax highlighting. Afaict from the source code for VE, it uses them exclusively for doing this. DevQt's syntax highlighter is still full of bugs, but I'll be willing to bet €1.000.000 when it's finished it will be done highlighting the code much faster than VolkoEdit.

vitaly
20th February 2006, 11:17
I think elsewhere fmc had mentioned the main reason for this as being a need to learn the Qt library. And yes, there are editors already which are great for writing source code in. The problem is finding one which is tightly integrated and geared specifically towards Qt development rather than a full-featured source code text editor which maybe has the ability to do debugging output (like VolkoEditor).
What is integration with Qt? Modern projects may use more that one language and library. If you create "tightly integrated with Qt" editor, you will feel some difficulties with other languages and libraries.


Does VE let you view the class hierarchy of the program?
I don't know. I think no. But DevQT doesn't support it. It is complex enought. You possibly should create C++ parser to manage class hierarchy. It is difficult enought. If you create such parser you may implement some refactoring tools (look at Fauler's book about refactoring).


Does it let you view signals and slots?
What do you mean?


I wouldn't know. There's too many problems trying to compile it on my machine using FreeBSD 6.0 and Qt 4.1 ...
May be. It is in stage of development.



BTW, there's another conversation in these DevQt threads about why it shouldn't use Regexps for syntax highlighting. Afaict from the source code for VE, it uses them exclusively for doing this. DevQt's syntax highlighter is still full of bugs, but I'll be willing to bet €1.000.000 when it's finished it will be done highlighting the code much faster than VolkoEdit.
Highlight is one of hundreds features which are really needed. The editor with best highlight engine but without some other features is useless. I use emacs. I've no problems with highlighting. And I'm happy with it.

soul_rebel
20th February 2006, 11:56
What is the Point of DevQt?
kdevelop will be ported to windows once the kdelibs are and i doubt DevQt will be able to compete.
it would be much better to help the kdevelop team...

michel
20th February 2006, 13:36
What is integration with Qt? Modern projects may use more that one language and library. If you create "tightly integrated with Qt" editor, you will feel some difficulties with other languages and libraries.


An IDE with Qt integration doesn't mean "Use only qt libs or else your program will break." It means the environment is specifically set up to be efficient for Qt developers who are, 99% of the time, making Qt programs. I don't use KDevelop, but I'm sure you could make Tcl/Tk programs with absolutely no Qt/KDE code or libraries included in them whatsoever if you want. It doesn't mean KDevelop isn't tightly integrated with KDE/Qt.


I don't know. I think no. But DevQT doesn't support it. It is complex enought. You possibly should create C++ parser to manage class hierarchy. It is difficult enought. If you create such parser you may implement some refactoring tools (look at Fauler's book about refactoring).

Of course it's complex. But then, MS Visual C++ wasn't written in a few weeks (or months) either. So what is the profit of pointing out what DevQt does or doesn't support yet? How long have they been working on it? A month? Less? It's like going to the construction site of an apartment building, looking at the big sandy hole full of wet cement that will be the basement, and saying, "God, I wouldn't want to live here. This place doesn't even have central heating or cable television."



>> Does it let you view signals and slots?
What do you mean?


Are you really asking me what signals and slots are, or just wondering what I mean exactly by "view?" As I am not really part of the DevQt project you would have to ask them what they have in mind. But personally, I liked the way they were handled in the old 3.x Designer more than they are in 4.1 Designer. FYI, I was being rhetorical. VE will let you view the source code for signals and slots, but it certainly won't know what they are or care. Just like it won't care what qmake is, what a qmake project is, what moc or uic are, etc. That's the problem.


Highlight is one of hundreds features which are really needed. The editor with best highlight engine but without some other features is useless. I use emacs. I've no problems with highlighting. And I'm happy with it.

Yes, that is precisely my point. Other than the highlighter being currently "better" than DQt's (that is, it works without bugs; when DQt's is fixed of the bugs, VE's regexps scanner will probably be left in the dust), what possible benefit would there to be to using VolkoEditor to create a Qt IDE? It's little more than Kate with FTP support, and unlike VE at least Kate and FTP compile on my system without me having to modify the source code in 4 files. On the other hand, DevQt already has some support for qmake .pro files. VE doesn't even know what they are. To go and make a new IDE around VE would be more work at this point than for fmc/Cesar/et al to just continue with what they have already done.


SoulRebel:
What is the Point of DevQt?
kdevelop will be ported to windows once the kdelibs are and i doubt DevQt will be able to compete.
it would be much better to help the kdevelop team...

Perhaps the most obvious reason to have a Qt IDE and not rely on KDevelop is simply that KDevelop requires you have the KDE libraries installed on your computer. Qt in itself does not. All it requires of you is that you have an X-Server running that it can connect to and have Qt installed where it can find it. I have KDE installed, but what if I change my mind later because I want something that takes up less space? Well, I guess I have to go back to programming Qt applications in a text editor because everybody who had the idea to make a Qt IDE which doesn't require anything but Qt took your advice and joined the KDevelop team instead. Then I get the idea to make my own IDE for Qt, someone tells me to just use KDevelop, and the whole cycle starts all over again.

Asking why to make a Qt IDE and not just use KDevelop begs the question why should anybody make programs in Qt and not just use the KDE API (it also begs the question why anybody should have invented the automobile instead of working with horse breeders to create better horses)? Oh wait, that's right! Not everybody wants to make a program which only works with KDE. By the same logic, not everybody wants to use a program (KDevelop) which only works with KDE. :) An IDE which only needs the Qt libraries to run can be used by somebody using GNOME or some other desktop manager for X, a raw X-server without a desktop, MS Windows, MacOS, or whatever else you can manage to build the Qt library for and have the resources to load the IDE. KDevelop needs BOTH Qt AND KDE, which means even if you don't want to use KDE you have to have the libraries installed regardless. That's very Microsoft-esque to me. I don't think the KDevelop team's intentions are so malevolent, mind you. It's just that KDevelop is meant to make developing under KDE--particularly programs using the KDE API--easier. It's not meant to be a complete solution for everybody everywhere. And if you want to make Qt programs but don't want to install KDE (or don't have the space for it), it definitely isn't the solution for you.

As for the comment on "not being able to compete," not everything in life is a competition. It is often the case, particularly in science, that multiple people, or multiple groups of people, are developing the same or similar ideas completely without interest in what the other is doing, or how much more "popular" the other person/group's work is compared to their own. Once in university I made my own database program with its own storage scheme, compression algorithms, etc. I didn't do it because I wanted to make money off it and compete against MySQL. I did it for the benefit of the learning experience and a sense of accomplishment. It is the same reason people paint pictures or build little model airplanes or do almost any other hobby or study. Sure, millions of people already do it too, and most do it better than they ever will, but that doesn't mean there is still no reward for doing it. Another example: I use FreeBSD because I like it and I didn't like the Linux distros I tried in university in the late 90s. But I don't want everyone who uses Linux to be destroyed or to stop using what they like to use and what is best for them. They aren't my "competitor" just because they like to use something else. BTW, using the same argumentum ad populum, since all of the *nix users combined can't "compete" with the number of Microsoft users, I guess we should all disregard our own personal wants and needs and collectively install Windows XP instead.

PS: Love the Brecht quote. :) I am listening to the 1930 recording of Die Dreigroschenoper on CD, where Brecht sings the Moritat himself, as we speak. He sounds like a dying cat.

Soldaten wohnen
auf den Kanonen!

soul_rebel
20th February 2006, 14:57
not everything in life is a competition :)
i know and i didnt mean that. (although there are people that try to convince one that that is all that life is about)

I did it for the benefit of the learning experience thats why i code too :D

or whatever else you can manage to build the Qt library for ; KDevelop needs BOTH Qt AND KDE yes thats the way it is right now, but i as soon as kdelibs-4.0 are out i expect them to be more portable and slimmer, in the end they will be as portable as the qtlibs. also i do not think that somebody developing apllications cannot spare a couple of megs for kdelibs.
btw. there was a (afaik refused) google summer of code project of porting kdevelop to either kdelibs-svn or qtlibs-4.x (and removing dependency on kdebase and other uneccessary stuff). nevertheless this will come and it does provide much more features than devqt can possibly provide then...
i am not saying this because i am against choice or because i think devqt "can't stand the competiton" (the post was over-simple); i just think it would benefit all, if human resources were organised in a better way. now of course "better" is a very personal view, but lets look a kdevelop' qt features now:
- c++ qt support
- python qt support
- ruby qt support
- java qt support
- support for perl-qt-bindings is being worked on i think
- seamless integration of the designer
when will devqt have these features just for qt?
in addition kdevelop supports tons of other languages, now you might say: "i only want to code qt-c++" but sooner or later you might want to try other qt bindings or kdelibs (after all it really is only additional classes for qt), or maybe even a totally different language; then you will have to get used to a completely different environment.
i dont want to say: dont do it, devqt-team, i just think it would be better for everyone if they tried to organise.
sometimes forks of projects are right, but if they only dublicate work that is already done, i regard them as wrong.

I use FreeBSD me too :)

michel
21st February 2006, 11:03
There is a post on here somewhere, you will have to look it up because I don't remember what thread it was in (maybe newbie, I think I told some newbie what he should and shouldn't try to do when he first learns Qt, and then fmc described the IDE and how it was a learning experience), where fmc stated the reasons why they started this project. I am not sure why the others are helping, but I assume it is either because they have a similar intention or simply because they consider it a good idea and want it to be realized. Either way, if somebody wants to do something, even if it's already been done before, sometimes it's best just to not tell them they can't or shouldn't do it and try to be supportive. It's quite possible DevQt will never be a rival for KDevelop, and it's also quite possible that it will end up like thousands of other projects on SourceForge, Freshmeat, etc. that were some college kid's two month long excitement that spiral down into a project that never reached even a beta version. But the mental exercise, the personal satisfaction and sense of achievement that comes from actually doing something instead of just thinking about it, and the experience that can be used later on more serious projects (in any field)--imo the sum of these parts is greater than the whole, regardless of whether or not anything of practical use to the public comes out of it. As for joining an existing project team, sometimes projects need volunteers, and sometimes they already have plenty and only want you to submit bug reports or people to write technical documentation for them. Maybe the people involved in this want something more than that.

ccf_h
8th March 2006, 11:49
Sorry, if I was missing something, but I'm just wondering.:confused:
The Qt4 IDE at http://frenchbreaker.free.fr/qt4ds/index.php
compiles fine is GPLed and looks really promising. Why did you feel the need to start a second project with exactly the same rationale? I mean it's perfectly legitimate to do so but, if the aim is to have a nice Qt4 IDE, it seems a bit unpractical (at least to me). Or did you contact the other project and the maintainer was unwilling to accept support and cooperation?

KjellKod
8th March 2006, 15:46
Hi,

I get the following error message when I try to checkout the latest DevQt


svn checkout http://svn.berlios.de/svnroot/repos/devqt/trunk
svn: REPORT request failed on '/svnroot/repos/devqt/!svn/vcc/default'
svn: REPORT of '/svnroot/repos/devqt/!svn/vcc/default': 400 Bad Request (http://svn.berlios.de)

KjellKod
8th March 2006, 15:55
I can understand how Qt4Ds was missed since I have looked recently for good IDEs on Windows but did not find Qt4Ds (now I used Bloodshed DevC++)

I have tested the Qt4Ds - after seing it on this thread and I think it looks good but it is definitely a Beta release. I still have not tested the latest version of DevQt (since I can't get the svn to checkout and wwwCVS/SVN to get a file at a time is not an option) but at least DevQt can open up my .pro file without problems.

However both do look promising - we'll see who wins the race :cool:
Too bad though that Qt4Ds forum was in French :( that is discouraging to say the least for non-french speaking

wysota
8th March 2006, 19:06
I get the following error message when I try to checkout the latest DevQt

There is an extra "!" sign in the url. Someone probably misstyped it. Did you try a www access?

KjellKod
8th March 2006, 19:22
The WWW access works (both CVS and SVN) but then you would have to download the files one at a time, right? In that case I rather wait till the error is fixed

KjellKod
8th March 2006, 19:30
Ah! I think I found the error. I believe it is because I am behind a proxy. Apparently the proxy does not approve of REPORT requests. I will try using another port (asking the IT department to change settings of the proxy is futile :eek:

KjellKod
8th March 2006, 19:50
Jikes, now it looks like I am spamming the forum. Sorry :confused: Anyhow. I figured that I am probably not the only one with the proxy problem. The solution was simple once I understood the problem.

When you are behind a proxy server that is not passing the WebDAV methods that subversion uses you can work around the problem by using SSL to hide what you're doing from the proxy:

I.e. svn checkout
https://svn.berlios.de/svnroot/repos/devqt/trunk

Matt Smith
16th March 2006, 01:06
I can understand how Qt4Ds was missed since I have looked recently for good IDEs on Windows but did not find Qt4Ds (now I used Bloodshed DevC++)

I have tested the Qt4Ds - after seing it on this thread and I think it looks good but it is definitely a Beta release. I still have not tested the latest version of DevQt (since I can't get the svn to checkout and wwwCVS/SVN to get a file at a time is not an option) but at least DevQt can open up my .pro file without problems.

However both do look promising - we'll see who wins the race :cool:
Too bad though that Qt4Ds forum was in French :( that is discouraging to say the least for non-french speaking

I second this ... I just tried it out on my Mac and it does look promising, but it seems to have no understanding whatsoever of the Mac application bundle. For example, it attempts to start Qt Designer by simply calling a file called "designer" when that's not how you start a Mac GUI app. You have to type "open" followed by the name of the bundle.

Also, it uses Plastique, even on the Mac. The whole idea of Qt/Mac is that it uses Carbon widgets and looks like a real Mac application. It is not meant to look like a crudely ported X11 app.

Finally, there seems to be no way of changing the font, which in the text editor widget is proportional (Apple system). That is not usual for editing source code: people expect a monospaced font. Whether the same is true on Windows or X11, where I've yet to try it out, I don't know. In their screenshots, the font is the proportional Windows system font, so perhaps they're happy with that.

I'll be glad if this becomes a standard for Qt 4 development because everyone likes an IDE that knows their software, but I cannot use it right now for my project even if it is self-hosting. I'll be sticking to Aquamacs/XEmacs and the command line for the time being.

fullmetalcoder
16th March 2006, 16:28
I second this ... I just tried it out on my Mac and it does look promising, but it seems to have no understanding whatsoever of the Mac application bundle. For example, it attempts to start Qt Designer by simply calling a file called "designer" when that's not how you start a Mac GUI app. You have to type "open" followed by the name of the bundle.


Thanks for those information!!! I have absolutely no knowledge of the way mac paths and processes works and I guess all members of DevQt team are the same. Thus we would greatly appreciate that a Mac user join us so as to get a really cross-platform IDE.



Finally, there seems to be no way of changing the font, which in the text editor widget is proportional (Apple system). That is not usual for editing source code: people expect a monospaced font. Whether the same is true on Windows or X11, where I've yet to try it out, I don't know. In their screenshots, the font is the proportional Windows system font, so perhaps they're happy with that.


Not allowing changing font size is damn stupid most of the times default font are quite ugly! BTW, do you like our last version? The settings dialog looks cool doesn't it? (Don't pay attention to compilation, assistant openning and designer running, cause it hasn't been tested on Mac yet...:o )

pasnox
19th April 2006, 10:52
Originally Posted by Matt Smith
I second this ... I just tried it out on my Mac and it does look promising, but it seems to have no understanding whatsoever of the Mac application bundle. For example, it attempts to start Qt Designer by simply calling a file called "designer" when that's not how you start a Mac GUI app. You have to type "open" followed by the name of the bundle.

Hi,

I really need Mac OS X feedback for my Qt4DS Monkey project, current version is 2006.1.0.0RC2b.
It will be very friendly if you can help me fixing the mac version by testing the last version.
You can get it from http://sourceforge.net/projects/qt4ds-monkey

Thanks P@Nox,

fullmetalcoder
19th April 2006, 13:35
Ahh ! Pasnox finally came here! Nice to see you!
Then we are at least two froggies @QtCentre and, what a coincidence, both of us are working on an Qt4 IDE.

Well, back to Mac OS X : there're not any Mac users in the DevQt team so testing on that OS requires some other people to do it and give feedback. BTW we would welcome you in the team so we wont have that kinda problem with platform specific issues.

@Pasnox : admins of QtFr offered me to open a section dedicated to DevQt but I decided noit to do it until the new base is over (yep I have been rewriting nearly everything...). Anyway, in a week or two I'll do it so I suggest you do the same right here (no dedicated section but just open 2-3 threads and I'm sure the admins will make them sticky or open a "Projects" section and move them all there)

GreyGeek
19th April 2006, 15:56
Sorry, if I was missing something, but I'm just wondering.:confused:
The Qt4 IDE at http://frenchbreaker.free.fr/qt4ds/index.php
compiles fine is GPLed and looks really promising. Why did you feel the need to start a second project with exactly the same rationale? I mean it's perfectly legitimate to do so but, if the aim is to have a nice Qt4 IDE, it seems a bit unpractical (at least to me). Or did you contact the other project and the maintainer was unwilling to accept support and cooperation?
Well, to be brutally honest, one might ask why "frenchbreaker" is being developed, since it looks nearly identical to Kate, the editor which I use for editing my QT4 projects on Linux?:confused:

wysota
19th April 2006, 19:24
Well, to be brutally honest, one might ask why "frenchbreaker" is being developed, since it looks nearly identical to Kate, the editor which I use for editing my QT4 projects on Linux?:confused:

Maybe because Kate is not available for Qt4 yet? ;)

pasnox
20th April 2006, 08:54
Hi,

happy to see you again fullmetaccoder :)

I will create thread when my official webpage will be avalable, because for now , it's very bad :p

Why developping Qt4DS Monkey ?! if you post the question, it's because you don't read the readme.txt... so i repeat :

Qt4DS Monkey is a multiplatform Qt 4 IDE, meaning, working the same way on all platforms that Qt 4 support.

Does kate is working on windows ?! no, so when you switch to another OS you need new IDE, new settings, times and times lost !

With Qt4DS, no time lost because it's the same IDE!

It's OK ?!

PS: frenchbreaker is not the IDE name, it's just an old homepage name where i have put in the past my unofficial homepage for the project...

P@sNox,

ccf_h
20th April 2006, 09:46
Ok, I'm feelin some kind of stupid asking this question, but still can't resist:
If you are two teams developing essentially the same kind of application why don't you throw your lots together and join forces? :confused:

fullmetalcoder
20th April 2006, 09:55
Ok, I'm feelin some kind of stupid asking this question, but still can't resist:
If you are two teams developing essentially the same kind of application why don't you throw your lots together and join forces? :confused:

Kinda stupid question needs kinda stupid answer :
We froggies are kinda stupid as well and we don't like cooperation...:mad:

Ok just kidding.:p
The point is that we both started aproject for ourself, without knowing someone was working on the same kind of thing. Then there'is the team question : DevQt is completely open to coders who want to get involved in its developpement whereas Qt4DS is kinda "one-man-show" (as was DevQt at the beginning). Moreover each of us has already a project account (DevQt is registered at Berlios and Qt4Ds at sourceforge) which also brings the problem of version control : CVS for Qt4Ds and SVN for DevQt.

Finally the problem lies in the code itself : to say the truth Qt4DS looks nice for the user but for the coder it's horrible! I had a look at it and I looked even more confusing that the original DevQt code ( which was not the clearest code ever...:o ).

I don't refuse cooperation but I can hardly think of a way to flawlessly merge the two projects...

pasnox
20th April 2006, 11:58
Hi,

Qt4DS was a kinda "one-man-show", not anymore as some developers join the project.

And as fullmetalcoder says : Joining force was the only thing we would do, but as it says no one know about others project and so, it's nearly impossible now to merge our projects, and starting from nothing in the moment will be a bad idea.

Peraps, It's not impossible that one day, we are working on some standart widgets, project ?! (like QTextEditor etc)

Wait and see.

fullmetalcoder: yes, source is not clean clean, i wait a really stable version 1 (working with no problems) then i will rewrite/review the code.
I see that it''s what you are doing for now, so what do you think about :
What don't you continue your current version of DevQt until a stable V1 or something working good, and thinking together about a new project.
We have a lot of ideas (and feedback help a lot) we can discuse about, it can be positifs points for a really good crossplatform Qt 4 IDE.

I really think that concurrent project (because our one have the same goal, same developpement : Qt 4 library for developping Qt 4 Application) are nothing else time lost.

Tell me what you think, P@sNox

fullmetalcoder
21st April 2006, 16:29
What don't you continue your current version of DevQt until a stable V1 or something working good, and thinking together about a new project.
We have a lot of ideas (and feedback help a lot) we can discuse about, it can be positifs points for a really good crossplatform Qt 4 IDE.


Kinda late!!! :o

The new base is nearly over (zip archive is on Berlios (http://prdownload.berlios.de/devqt/devqt-branch-3-alpha.zip)) and the team agreed that the current version was unmaintainable...

Anyway, if you want to join the team, bringing your experience with you, you'll be welcomed. If you still cling to Qt4DS then we can still cooperate on some common standards (and even some common code) such as project management, editor widget, plugin system(could they be compatible??? would be kinda great deal:) ) and so on ...

pasnox
24th April 2006, 08:36
Hi,

I had a look at your source, some minor compile probleme, but not working under GNU/Linux Mandrake ( default lugin can't be found ), when i rename the plugin named libdefault.so to default.so, no messsagebox appear, but i can't do nothing in the ide.
It's normal ?!

Else source tree and structure are really clean, you make a good job!

P@sNox,

fullmetalcoder
24th April 2006, 21:50
I didn't test it under Linux yet but I'm gonna do it as soon as possible. I don't know how the plugin template is handled under linux but, yes, the default plugin should be named (discarding about the extension) "defaut", not "default" nor "libdefaut"...

If no message box appeared the the plugin as been loaded but what exactly do you mean by :

i can't do nothing in the ide.
:confused:


Else source tree and structure are really clean, you make a good job!
:cool: Thanks a lot but :

I wasn't alone, a few other coders help me a lot:D
The source isn't clean at all, at least not the way I understand that word. It may compile and run fine but the source itslef is unmaintainable because modules are medled in an ugly way that defy any understanding...:eek:

pasnox
25th April 2006, 08:08
Hi,

By saying "Can't do anything, i mean : can't open project ( opendialog doesn't open ), can't open files, etc etc"

P@sNox

fullmetalcoder
25th April 2006, 09:24
Hi,

By saying "Can't do anything, i mean : can't open project ( opendialog doesn't open ), can't open files, etc etc"

P@sNox

This is just normal then. The beta I posted on Berlios is just a very early beta of the next DevQt branch. ATM you can just open new editors with the central widget and close them after editing. Cut, copy, paste, undo, redo, select all are implemented but that's all.

The next beta to be released will feature file loading and saving and lot of improvements in the plugin system (the avaiable beta has a built-in indenter, highlighter, matcher and loader : the next one will do everything inside plugins)

Project management will come afterward because I want to be sure the core work before going further plus I'd like to rewrite it using real Model/View instead of item based tree.