Delphi 2005 a flop, IMOHO >:P

[WILL]

Well ladies[size=9px](maybe?)[/size] and Gents, the verdict is in...

I have tried the latest flavor of Delphi and I have to say that at first I was intrigued... then put off [size=9px](but the sheer size of the installation; over a gig)[/size]... then dismayed by the ugly and irritating tweeks of the new editor... finally disappointed by the fact that CLX support is not included.

I have to ssay that Borland has made a serious miss on this one. Not only have they pulled 'an Adobe' [size=9px](known for it's every other version having to 'correct' a mistake in interface design that the previous 'latest and greatest' left users with that familiar bad taste in their mouths)[/size], BUT they completely let down a recently new [size=9px](and created by their own hands mind you!)[/size] technology that was still fighting to grow to what it's original intention was in the first place.

What is Borland thinking :x


The only 'good' thing to it, which I have no personal interest in, is the .NET(which it's that great either) and C#(which is no really Pascal either).

Personally the last compiler that was any good that they made was Delphi 7 and I doubt I'll be switching until either a WORKING version of Kylix 4, or whatever, comes out OR a much 'nicer/easier to install on Debian' IDE comes out for Free Pascal [size=9px](fully supported Object Pascal language will help sway me too)[/size].



Of course this is just one man's oppinion.

[cairnswm]

I was at a prelaunch presentation of Delphi 2005 and while I was impressed with quite a lot I was very dissapointed in that Delphi didn;t seem to be advancing anymore.

Delphi 8 was only .Net which put me off. Delphi 2005 seems to have joined Delphi 8 and Delphi 7 into a single package.

I agree that Borland has missed a trick with Kylix but lets hope our friends at FreePascal and Lazarus sort this problem out.

WILL - Doesn't Lazarus run on your version of Linux?

[czar]

Delphi 2005 seems to have many unnecessary changed made to IDE. The only thing I like so far is the retractable code feature, you can collapse procedures. But that is hardly a must. I think we will stick to Delphi 7 for our development requirements. I still haven't been convinced about .NET so that is not a selling point at all.

I am not sure why they had to go and radically alter the way Delphi looks and behaves. If it isn't broken why fix it?

[{MSX}]

a much 'nicer/easier to install on Debian' IDE comes out for Free Pascal [size=9px](fully supported Object Pascal language will help sway me too)[/size].

That's my idea
We need a new IDE. I don't like Kylix and Lazarus..
When will we start working on it ?

[Lightning]

See, i told you Borland has bad ideeas lately, they officially stated that Kylix won't be upgraded at least until 2005, it doesn't even work with latest kernels 2.6.x, and those ugly big QT libs ...
They probably thought will convince people to use it if it looked like VS, Adobe, Macromedia, etc. IDE's, but i find floating forms easier to use.
When Macromedia upgraded Dreamweaver's IDE, you could chose wich version you like MX or old version.
Code collapse might be nice but it's not a must.
A developement ide should be floating not like a graphic ide, or at least you should be able to chose.

MSX do you have any ideea how hard it is to make a crossplatform IDE like Lazarus, have a look at it and it's code, BTW the Lazarus synedit is customised to work in Lazarus and isn't an official version from the synedit team so using it in another IDE is quite hard without the LCL.

Lazarus does work on Debian, but it seems Debian has some font problems, talk to other Debian users, and also get the latest 1.9.2.3 sources/packages/installer since it has many bug fixes and you no longer need to rename for recompilation.
For small apps use the folowing:
strip --strip-all yourapp
upx --best yourapp
This compresses the win32 IDE(lazarus.exe) from 21 MB to 1.5 MB and projects are 3-500 KB big after compression.

[{MSX}]

it doesn't even work with latest kernels 2.6.x, and those ugly big QT libs ...

I'm using it on my 2.6.8 kernel Just need to hack it a little..

MSX do you have any ideea how hard it is to make a crossplatform IDE like Lazarus,

Yeah I't real hard.
But my idea was to implement only the 2 "big" features: debugging and code completition. For the first, it need only to interface with gdb (which is great), a thing that is already done in lazarus, and could possibly be extracted with some hack. For the second, it's a bit harder, but the sources of FPC includes (obviously) the parser and the scanner, so that one could do a code aware editor (with some job, of course). Also, if necessary, a ppu readed is included in the fpc sources as a separate program: ppudump.
All the rest is simply file handling, editing, settings etc. Easy (even if time-consuming) things to do.
What do you think ? (insults are welcome )

[Clootie]

Well, Delphi2005 editors & designers can be changed to "floating" behaviour by [View->Desktops->Classic Undocked]

[Lightning]

Lazarus already has debugging and code completion, however debugging is still not completed on win32 but code tools work, code hints are a little buggy on win32 but work on Linux.

MSX if you have some knowlege then why don't you help Lazarus since it has full FPC support (the FPC team supports it and viceversa).
Why try to swim against the watter flow ?
You all say Lazarus is buggy, bad, makes big apps., why don't you make it better, you have something, why try to reinvent the wheel ?
The Lazarus project needs developers and because it's the best such project many chose it, it will soon be the best IDE for Linux.

The ideea is we should help the existing Lazarus project, not start a new one and use Lazarus code for it, it's just wrong.

BTW if you do some contribution you might see your name in the contributors list. They also have two IRC channels #fpc and #lazarus-ide so you can discuss features you would like, bugs, etc.

[WILL]

I'd have to agree with Lightning on the idea that the perverbial wheel need not be reinvented. Lazaraus has great potential and so far [size=9px](with a few peives of mine, that are a bit picky)[/size] installs 'ok' as a win32 app. And compiles quite well, that I have seen. With any good tool though, it only comes with hard work and great amount of work. Especially an opensource or indie-based project like Lazarus, Free Pascal and even Linux it's self [size=9px](at one time it was...)[/size].

Saying that. It needs a lot of work! And a lot more people testing it out and possibly some minor to major modifications to appeal to the majority of Pascal programmers. I can pick it apart right here, but I'd be getting way off topic [size=9px](and besides the point)[/size].

Sadly most of the programmers that had gotten into game programming when DelphiX had made it all the more popular by adding an extra boost as it did. After it got abandoned, all of these DirectX suites came out and people got SO focused on trying to make their own and ultimately all trying to do the same thing and all that lost time on projects that could have been spent on other moer practical endevours. I'd not like to see that happen hereas we have so many directions the Pascal game scene can go from here.


Getting back to the original topic; The way it's looking these are our solutions to the hightening fall of the great and mighty Borland. Who is only showing us time and time again that it can make great products, only to lead them into the shit-pile not too long after it starts to see the fruits of their success.



cairnswm: I have tried it on Debian[size=9px](Woody)[/size] and couldn't understand the installation instructions well enough to install it fully. Documentation is sorely addressed for the Debian distribution and I was directed to the Lazarus Wiki Manual when I had mentioned it on the forums. Unfortunately I don't have it installed on my system now, but I have VMWare and can make an installation when I get the time.

[cairnswm]

[size=7px](Remind me to avoid Debian when I get a PC for linux)[/size]