Slashdot Mirror
Delphi for Linux
Thanks to several readers who wrote to us with the news that Inprise apparently will be porting its Delphi development tool, and others to Linux. This comes on the heels of the recent survey of the Linux developer community.
← Back to Stories (view on slashdot.org)
There is a good Pascal compiler for Linux. Compatible with Turbo Pascal 7.0 and some Delphi extensions. There are also GTK bindings available. It's called
Free Pascal Compiler.
Posted by Synsthe:
That's just silly, not everything has to be GPL to be good. You're living in a world where you just don't want to pay for anything, that's all. Get over it.
Money makes the world go around, and it's not going away anytime soon. GPL and open source are not going to make it go away. Linux (and I'm not saying this is a good thing, but it's a true thing) needs more commercial support to make an even bigger path for itself in the long run, and such developments as this could be very, very good for that.
If you don't want to use it, nobody is forcing you into it. When you're left behind playing your psx games and everybody else is out there making the big bucks working on the platform they like best, we'll try and remember you. =)
--
Mark Waterous (mark@projectlinux.org)
I do not agree with some of your assertions. Perhaps the Delphi code you have come across is bad, but that could simply mean you hang around with bad programmers.
I find Delphi to be a very nice GUI IDE, and OP provides most of what C++ does, and some things it does not. Personally, I like Object Pascal (OP) better then C++. But that may be because I learned Turbo Pascal before I learned C, back in the bad old days of MS-DOS.
You are right, in that the lack of C++ style templates is OP's biggest problem. However, the Delphi approach of making lists of TObject is not as bad as it seems. Delphi provides *much* better runtime type support then C++, so TObject is easily converted back to what it was.
Additionally, this lets you compile your collection classes once, and also replace them using dynamic linking. C++ requires full source for everything, and templates must be compiled for each use. Both approaches have advantages and disadvantages. I hope Borland decides to put the choice in the hands of the designer, and adds templates to OP.
Yes, Delphi uses exceptions a lot. This is because they are Delphi's error handling mechanism. Unlike C++, where it is anybody's guess how an error is reported, Delphi will *always* throw an exception. With this in mind, properly designed code is clean and *very* robust. Well done VCL code will recover from errors automatically.
Yes, the IDE gives you a global variable for your forms by default. It does this to make it easier for the VB-weenies who are writting hobby code. I always delete the global declaration immediately. If you do not, well, what can I say, you are not a very good programmer.
If you try to force C, C++, Eiffel, or some other language's mindset on Delphi, yes, you will get ugly code. But that is your fault, not the fault of Delphi. Since you snuck in a plug for Eiffel at the end, and are posting anonymously, I suspect you have an ulterior motive. In any event, I know Delphi to be a good system, and recommend that others developing for Windows (and soon, Linux) check it out.
dragonhawk@iname.microsoft.com
I do not like Microsoft. Remove them from my email address.
an X client is a GUI, the server is not. GNOME and KDE are desktops, Yes, they are accessed by a GUI, but Netscape has a GUI too. The application (client) has a user interface that uses an X server to make it graphical.
Opinionated Law Student Strikes Again!
Seriously. Sure, they can expand the market of C coders on Linux, but why would people who want GPL or free compilers switch?
I will quite happily use C++ Builder myself, but then, I've got apps to write.
Will in Seattle
Exactly. There are a lot of Motif programmers out there. Motif has been around for what, better than ten years now? And the free Lesstif has most of it covered.
True, Motif tends toward some ugly defaults (mostly in component spacing and apparent widget "thickness"), but these are easily overridden. There are even class libraries around to encapsulate Motif for the C++ hackers that really can't figure out how (or don't want to spend the time) to roll their own.
Not to say that {Mo,Less}tif is perfect, but it's silly to ignore the talent pool that exists for it.
-- Alastair
http://www.borland.com/about/hr/99083.html
:)
check that out. Official job opening.
I had included this hard core evidence when I submitted the article, but Hemos didn't post it... I knew about it for a week but was too lazy
Go Borland!
----------
"They misunderestimated me." --George W Bush, Nov. 6, 2000
> Besides, if they chose Motif they'd have to licence it, adding it to the price of Delphi.
Or use Lesstif instead. (Modulo licensing issues with that, since I believe Lesstif is GPL'd.)(Er, I just checked, and Lesstif is LGPL'd. No problem.) Which would be great if it helps improve Lesstif.
-- Alastair
I do.
Writing gui apps with VI is why so many X apps look like a 3yr old with a crayon layed them out.
---
Joseph Foley
InCert Software Corp.
You missed the most important letter in GUI, the *I*, as in INTERFACE. X doesn't have one without a window manager and/or shell. Joseph
---
Joseph Foley
InCert Software Corp.
Hmmm, this could be a different situation for Borland/Inprise (I have been a Borland supporter for some time now - I hated the name change). Will they support architectures other than x86? PowerPC? Alpha? MIPS? Burning questions...
--
"All that is visible must grow and extend itself into the realm of the invisible."
I swear by MacOS X. Although I use to swear *at* MacOS 9...
Delphi is not only a fine IDE, it is also a fine language. If it wern't for portability issues it would be my language of choice even for non gui apps. Free Pascal is getting close to being as good but not quite there yet.
re: the toolkit
They might do something similar to the way they did their database abstraction. That would let them use GTK+, Qt or whatever. This would let people pick which toolkit they want and possibly even change during devopment.
I guess it's one way to do things, but I don't really like it. I like to know exactly how my app is going to look. Changing gui toolkits is a good way to make your app suffer from java ugly app syndrome. I'm not much of a themes person.
I'd like to see something like GLUI, a gui toolkit written on top of GLUT. Then you can have apps that run identically on Linux win9x Os/2 BeOs etc.
Don't think it'll happen though.
-- That which does not kill us has made its last mistake.
I don't think they will choose Qt, because you can't develop proprietary software with it unless you pay hefty licence fees. Not only would Borland have to pay up, but possibly also every Delphi developer. Unless they cut a deal with Troll Tech, of course.
-- Ed Avis ed@membled.com
I'm not surprised about Delphi being so far in the lead within the Linux community. With sites like the Delphi Super Page and Torry's Delphi Pages, many Delphi users create new components and distribute them openly (sometimes under the GPL) for others to use and, yep, modify.
This is different from the VB community, who appear to follow the Microsoft philosophy. I have been aided by these "free" Delphi components and even applications, and I have also contributed back.
As to those who criticize Delphi as not a "real language", I'd like an explanation. Many excellent applications have been written in Delphi/Pascal. What makes a "real language"?
Well, here is one data point that said,
"is it important whether Inprise GPLs the source for Delphi or not", I voted, "Do not care".
Do I care whether the source for Oracle is GPLd? No. Do I care whether the source for DB2 is GPLd? No. Do I care whether the source for Call To Power, for Linux, is GPLd? No.
As long as the base is GPLd, we're all protected (as long as the base does not get marginalized, as MS Office for Linux could do, because you KNOW that it will only work with other MS Office for Linux/Windows, which you'll have to pay for).
MSO2K says "we're using XML for our documents". Great. But most of the useful "features" are locked up in...ActiveX stuff.
Unless there's something other than X Windows, I'm not sure what this statement is referring to.
(IMHO)Well, it is. I've used the C/c++ tools from Borland and had great success, going from all the way back to their original Turbo C compiler up through the C++ Builder, which is Borland's C++ version of Delphi. Remember, these tools have been going toe to toe with M$'s Visual toolsets for years, with a high percentage of developers preferring the Borland/Inprise tools.
Within the realm of the C++ and the Delphi products, the real question is more which set of GUI components will be supported, i.e., Gnome, KDE, X, or perhaps a new set of components from Borland(?).
...Open Source isn't the only answer -- but it's almost always a better value than the alternatives...
This world keeps getting better each day that passes ;))
Delphi is atleast a respectable tool, but I wonder how long it will take before Microsoft ports VB? VB is funny to work with... just cut and paste your apps together.
---Got Coffee?---
RAD is extremely useful. Being able to draw out the UI alone is wonderful. Java is an excellent example of this. Coding up the UI manually is really annoying (even the code is a lot cleaner than the code a GUI Builder Tool will generate). But they save SOOO much time. I spent perhaps 20 hours on UI code for a recent Java project. The entire rest of the project, including writing a user guide and documenting the code, took only 60 hours. Thats a long time to write a User Interface. If I had a good RAD tool at the time, I could have easily cut that time in half, and probably down to only 4 or 5 hours to get the same interface.
Also, RAD goes much futher than just Ui development. Properly written delphi code is veyr easy to understand. For one reason, a lot of the code that you don't really care about (and is bug prone) you never write; I'm talking about code to set up screen controls, and getting backend stuff working. In Delphi, I just drop some components in, set some properties, and then make some event handlers to actually do the work. I can't think of anything sweeter than that. I never have to mess with the UI code (unless of course I want to for some reason) and setting up the backend (datasources, etc) is as easy as connecting a few components together. This is what RAD is all about. getting rid of the stuff you just don't need to worry about so you can get the product done quicker. I can't see many problems with this philosophy.
First, I'm not complaining. More big software for Linux is a Good Thing(tm).
;-)
But wasn't the survey kind of wierd? I mean, I would never have expected Delphi to have *that* big of a demand...perhaps Borland rigged the survey?
(Maybe I'm just completely wack here...I know my brain isn't 100% there....damn waking up early for classes
--------------------------
Lazarus is a Delphi clone based on Free Pascal, a GPL'd multi-platform Pascal compiler already working under Linux x86, Win32, OS/2 etc. and compatible with Delphi 2. It is currently in alpha stage and under heavy development, a very nice open source project.
Lazarus is aimed at working with different widget sets like GTK etc. Go to the homepage.
Maybe I can get a deal with the Delphi 3 I have for Win9x.. And Finnally a programming tool for Linux where you don't have to hand code all the GUI Stuff.. =-)
Your link is just what I mean. No official anouncment, just a release of the survey results. I think people are engaging in wishful thinking at this point.
RAD for Linux means more applications. Means more acceptance. Means Linux takes over the world!
Then you make your choice, as is always true in an open market.
No, your tone was not Gestapo, but the notion of anything which is not open source being pointless is cut from that cloth. If there is no room for commercial products to run on Linux, then there are many market sectors which will remain with Windows.
My view is that all Linux-based applications benefit the furtherance of the OS, whether free or commercial.
Delphi gives me freedom to code my apps without being a slave to the GUI. I like that. Also, at the levels that count (the components and rtl), Delphi ships source code (in Professional and above).
Inprise/Borland has always had one of the best outlooks on licensing and on source access. To think that I might either debug or otherwise improve their compiler would be foolishly egotistical. But to have the source to the components protects me from obscure bugs interfering with my applications. The best of both worlds, for my money.
--- Bill
Yes! I was real heavy into Borland products and Delphi in specific before I morphed into my current state of Java developer. I few days ago I started up Delphi 3 and got that warm fuzzy feeling. Cool cool cool. I have to admit it's getting betta...getting betta all the time...
It's 10 PM. Do you know if you're un-American?
Agreed, in every particular. I spend a good deal of time exploring alternatives to MS OSes, because I am tired of the time I spend (even with Delphi) fighting undocumented features of the OS. I'm also tired of weird problems with kernel level services and/or drivers getting in my way.
As to MS tools, I recently had occasion to explore VC (again.) It makes me appreciate even more how well designed a product Delphi is. The notion that anyone could consider VC a visual development environment is unbelievable. After spending 10 days on it I concluded I would rather go back to assembly coding on embedded processors.
I have reservations about Linux, but 98% of them would diminish or disappear if I could use Delphi. My number one concern is the productivity hit I will take in making the transition to Linux.
One of the best aspects, of course, is that most components shold come across with little or no change needed. Some will require change, such as the serial I/O component I use every day.
I doubt that there are many Linux users who can comprehend what a sudden increase in the developer community will occur when Delphi for Linux ships.
--- Bill
I read the anouncement at it-director. So where is the official announcement from Inprise? Anyone have a link to an 'offical' press release?
> ... they were existing Borland-on-Windows developers looking to develop for Linux! That
> explains the toolkit confusion, etc.
Yes, which makes these results even more significant. If 40% of existing Linux developers said OK to Inprise tools on Linux, then it would not be so significant a result. It would actually be a little ho-hum.
The respondents here are mostly Windows developers who want to move to Linux and take their favourite tools with them. In this light, the numbers are far more telling.
-M
Some code won't work. I find delphi bulid-in function is not enough, sometimes have to call win32 libraries directly.
Motif would be the obvious choice for a big corporate like Inprise/Borland, especially if they are dabbling in Unix for the first time. There are far more *experienced* Motif hackers out there that they could employ, and other big corporates would be keener to get a Motif capable Delphi on their proprietary Unices.
...
The whole Unix world doesn't run Linux yet, and toolkits like GTK+ and Qt aren't as widely accepted in places like Hewlett Packard, Sun, etc as Motif.
And finally, with judicial tweaking Motif applications can be made to look as pretty as GTK+ and Qt ones.
Still run like a dog though
Chris Wareham
There's an interesting article called "How Programmers Stole The Web on Bruce Tognazzini's user interface site, Ask Tog. One point it puts across is that when programming languages and environments were relatively simpler (he says in the '70's, I say in the '70s and '80s), there was an explosion in programming creativity as many people who'd never even touched a computer before were creating all kinds of interesting and even ground-breaking programs. Consider the spreadsheet, which Dan Bricklin cobbled together in BASIC, or MYST, created by the Robyn and Rand Miller using HyperCard. I've seen non-programmers write small (but often-used, even institutionalized) applications that met their own particular needs using HyperCard, Toolbook, FileMaker, Visual Basic, REALBasic and Director -- all without having to run to the local geek for anything more than a little help. In the article, Tog says that Pascal is as difficult as C, but anyone who's had to chase pointers or manage memory (or used Delphi) will probably disagree. The important thing is that the presence of a simpler (than C/C++) programming language couple with a simpler programming environment (no makefiles, no sepearate source/header files) should encourage similar development in Linux. I also think that giving people -- that means anyone, not just the code gurus -- the ability to "do it themselves" fits perfectly with the Linux philosophy. If a "serious" programmer is free to write a utility or driver if it meets some need, a non-geek should also be able to put together apps that meet their workaday needs.
It's also good news for experienced developers. I work in tandem with a hard-core human interface guy, and what I usually do is hand him VB/Delphi forms and have him lay out the interface, after which I attach the code. Having someone who's actually concerned about the interface and giving him the ability to build it and see it all at once is a great timesaver for me (as I don't have to do it) and for the users (because someone's taking their workflow into consideration). Yes, we still always start off with pen and paper when designing, but giving a person with human interface skills (but not programming skills) the ability to build interfaces is efficient and benefits the users too.
A library of ready-made interface and other components is also handy. VCL is so much nicer to deal with than MFC, and I think it'll be a considerable boon to Linux app developers. Yes, I like doing things for myself, but sometimes it's nice (and less expensive in terms of effort and debugging) when some stuff has already been done for you.
However, the most important benefit of Delphi for Linux applies to both newbies and gurus -- RAD enviroments, simpler langauges and libraries like VCL let you have a life! I can't even begin to measure the value of such a feature.
Borland didn't rig it. The majority of respondents are Borland product users. What else would you expect? The survey was not intended or presented as anything other than a marketing tool for Borland. What it clearly showed is that among Borland tool users there is a great deal of interest in Borland tools for Linux. Quit griping: it means more growth for Linux.
If Jave becomes the defacto tool for Linux, then the apps will suffer. But for those of us who champion Borland tools, it doesn't matter: we will use Delphi, and will be productive, and our users will enjoy performance benefits over Java-built apps.
What gives you the idea that Object Pascal is "pseudo-object"? There are many ways of approaching OOP, and Delphi is as valid as any other, and more than some. At least it avoids the catastrophe called multiple-inheritance! Your comment smacks of ignorance of language.
--- Bill
Ethical cancer? In the USA? this is, in case it escaped your notice, a capitalistic country. Although some of Stallman's notions are excellent, the great difficulty with the Open Source movement is that it bids fair to become a fanatical cult.
Your ignorance of Delphi is impressive. With the Professional package and above, you do get source to RTL and VCL. I sincerely doubt that anyone in the Linux community is at great risk of improving the core of the compiler, so that really is a non-issue.
I have no idea what you're spouting with respect to "fictitious" things, but if you have any proof of your allegations, speak up. Otherwise you are merely trading in Microsoftian FUD.
--- Bill
I hate it when people bash Delphi because it's not free or becuase its Pascal or whatever. Borland/Inprise are going to generate more Linux related jobs. Once some of the bashers enter the job market they will appreciate this.
There are two main camps of people developing software for businesses. The VB/Backoffice losers and the Delphi Client-Server crowd. One just got a huge competitive advantage over the other because they can deliver a better, faster, more stable product with no OS licensing fees and Macro virus bullshit. If the market truly is free one of these is going to get trashed RSN.
So with all these business apps running on Linux the demand for good Linux support people will go through the roof. And all these Linux support people are going to get paid to write little GPLed tools in gcc/perl etc that benefit the whole community. Delphi isn't a threat to Linux - it improves the job market and increases the size of the community.
I agree. While I run Linux at home, and I perfer coding under C/C++, at work I write custom DB applications in VB (with COM/DCOM Server backends written in C++) and I can say that while I am not a big fan of Basic as a language, I can get software out the door quicker, with fewer bugs under VB than if I was doing the front ends in C++.
I think people need to look at what Delphi is targeted for, which is custom DB apps, which typically are not released on a Wide-spread basis. But if the ability to rapidly deploy custom apps under Linux is there, a lot of companies will be much more inclined to use Linux other than server only.
You are not going to see apps such as Word Processors written in delphi I think.
You're living in a world where you just don't want to pay for anything, that's all. Get over it.
Wait, Wait, Wait, I buy don't care if it is free of charge. As for calling me cheap thats a stupid insult, you don't even know who I am. I never said GPL is the equivilant of digital sweetness. I have seen rather sh**ty software in released in GPL, I have also seen some rather great software released in closed source.
If it is a good product, GPL'd, and cost a reasonable price I would buy it. I'm insulted by the idea that you think I am cheap and only use software for free. I am just interested in having businesses sell open source software. I want the OSS paradigm to work in the commercial environment. I want when I buy hardware I can see the source, and use it and modify it as I see fit. I know GPL is not the only open source license on earth. I also know that many corporations cannot release source code yet because they may use APIs that are not open. This is understandible. Hell I will admit I use netscape although I cannot see the source. (I will on the other hand say I would be alot less buggy if programmers could see the source). Having all software free or commercial OSS is my ideal, I don't think society is ready for it yet. I just don't understand why I am taking so much slack for it.
Money makes the world go around
Money does not make the world go round gravity does. >=)
I'm not sure I agree that Linux needs a good Rapid Application Developement tool.. In my experience, all RAD gets you is quickly rolled out, badly implemented, non-source havin', bloated applications, that don't have a good design.
If you try and develop an application for the here and now, for the short term.. you will end up with the mess Microsoft has. They needed an OS for the IBM PC... they bought a quick and dirty (RAD) hack called QDOS. Then they RAD'ed a GUI ontop of it called Windows, instead of going with a better design they were working on with IBM.. a little thing called OS/2... then they patched a 32-bit API ontop of DOS, and called that Windows 95... then patched thier patched Win32 API ontop of a VMS clone, and called that NT.
But they did it rapidly didn't they?
You can bet your life there are a small bunch of techies under an NDA somewhere in Redmond, WA, working on that one right now. :)
"Cake or death!" (E. Izzard)
--
"This isn't the post you're looking for. Move along."
"Joy!"
- Stimpy
I had forgotten about that! That's REALLY good news!
--
"All that is visible must grow and extend itself into the realm of the invisible."
I swear by MacOS X. Although I use to swear *at* MacOS 9...
How much do you bet they will not GPL the code. It is cool that Inprise was thoughtfull enough to code for Linux. It would be cooler if it is GPL'd. If they will not open the source then it is pointless. With a few exceptions like (PSX games, netscape, etc) I try to avoid using closed source programs.
A RAD tool like Delphi will bring tons of new apps to Linux. I've worked for two big companies where the internal use software was developed with Delphi...If the compatibility is at source code level, many companies will have no excuse to use Linux internally. That's the desktops, yes, not server-side only. So...we will be one step closer to world domination.
Really, a linux desktop would be a dream for tech support. Imagine having a generic installation for every department and not having to deal with regitery, slightly different ne200 cards with different win drivers. User profiles changing location, tighter security, flexible permisions for users...I can wait to see that!
Because many computers at big companies just run 2 or 3 softwares every day. Many just run 1 (tel. assistance, accounting, word-processor...)
Well, I hope to see that one day
If the Windows Share/Cripple-ware crowd thinks (which seems to be the case here) that this will somehow cause a demand for their crap to be ported over to linux, they are sadly mistaken. Quite frankly I don't think shareware has much of a future within Unix/Linux because most of it's programers (especally Windows programers) have given the concept such a bad rep over the years, so much so that I suspect most people interested in Unix/Linux want little or nothing to do with it anymore. Opera may be able to make the crossover, but the rest of that crap? Rather doubtful.
GUIs are intended to make life easier for end-users, who clearly are not at ease with the command line (else Windows would not currently own the market.) The immediate side-effect of the GUI on the programming side is to engulf the programmer in the details of constructing the visual interface. The notion that such an interface should be built in any way but with a visual editor is truly bizarre. And painful.
Delphi combines a very highly productive IDE and visual toolset with a very capable language and compiler. Into the bargain, it introduces components which are arguably the first example of a successful software component. I do not count OCXs, which incur a large overhead, nor any sort of similar element in an interpreted language (sorry, Smalltalk.)
Inprise/Borland, in committing to port to Linux, confer on the OS another level of credibility, and a crucial one, in my view: that Linux can be the foundation for commercial software success, as well as for the free stuff.
As to those who froth at the mouth about the evil nature of commercial software, I fail to understand why having more choices is a bad thing. If I would rather pay for a good tool which I already know how to use, why should I be forced to use gcc instead? Free software is nice, but free choice is better.
--- Bill
Yes, it would be cool to have such a full featured development environment in Linux, but that's not the point.
The point is to have a entirely cross 'platform' Windows/Linux development environment. Most of the people here do not seem all that interested in the cross platform benefits of this announcement - but a large percentage of the people who answered the survey were very interested in being able to port apps simply back and forth.
-josh
All objects in Java are references (much like C pointers) but with the pointer semantics of C hidden. I know many people who claim you cannot code a binary tree, linked list, etc in Java because it lacks pointers. This is simply not true. Java's primitive types (int, float...) are, of course, not pointers.
Just a minor couple of points though:
http://get.to/neuralabyss.software - Visit NeuralAbyss Software
- NeuralAbyss
~^~~~^~~~^~~~^~~~^~~~~^^^~~~~~~~~~~~~~~~
Real programmers don't comment their code.
No, they still won't *officially* commit to Delphi on Linux. But they're not that far from it:
At the official launch of Delphi 5 here in Finland last Thursday (Aug 19), when talking about the *next* version... the presenter twice did the old "I'm not really talking about CoughDelphiForLinuxCough" trick. (OK, actually it wasn't so much a "cough" as holding his hand before his mouth and mumbling it; I think he must have squeezed a nostril shut with his thumb too, because it sounded really nasal -- but that definitely was what he said.)
I think they're pretty sure that this *IS* something they want to do.
Christian R. Conrad
MY opinions, not my employer's - Hedengren, Finland.
Christian R. Conrad
mail me at iki.fi ; same user ID as here
Nice that there are such tools, but I am experienced in Delphi and would like to move some apps to Linux. As a practical matter, if I have to switch tools, I will be very slow to make such a move.
--- Bill
I would be perfectly pleased to see Borland port their own components. They cover all my basic needs, and I don't see any conflict with any of the formats in Qt/KDE.
As far as GUI kits, I'm agnostic. I just want to be productive, and very soon after moving to Linux. Coding GUI by hand doesn't cut it, and never will.
--- Bill
Linux is anarchy; there is no industry. And Unix lost their franchise. What counts is a solid and easy to use interface. All the rest is for hobbyists.
--- Bill
PimpBot writes:
/.ers who took the survey, most of whom OBVIOUSLY (judging from their comments in this and the other recent forum) wouldn't know Delphi from a hole in the ground, would -- *did*! -- just yell "if it isn't GPLd it's useless!" Sure, I'd love to see that too (and as soon as I have my Linux machine up and running again, I'll go to the Lazarus project and see if I can help out), but that's useless news to Borland -- they're a commercial company.
:-)
"I would never have expected Delphi to have *that* big of a demand...perhaps Borland rigged the survey?"
No they didn't.
It's just that there's a *LOT* of Delphi developers in the world, who really really like their tool, and would want to see it on Linux (I'm one of them; see my user profile). And the survey was publicised on... surprise, surprise: Web sites about *Delphi*!
Which is the logical thing to do, of course: Borland wanted to know if they could *sell* their tool. So the logical people to ask are those who already have proven they're willing to pay for it; existing Delphi developers. Come on, admit it: The
So did you expect all of us to scream louder for JBuilder or C++ Builder than Delphi, or what? Well, there are several reasons why we didn't. Reason #1: Delphi, Object Pascal-based Delphi, is Borland's greatest seller -- not C++, not Java. Reason #2: There are already a lot of C/C++ and Java-based tools available on Linux, but nothing for us OP fans. Reason #3: With C++ Builder's VCL being identical to Delphi's, we *know* that porting Delphi is the key -- do that, and C++ Builder is a good as ported, too.
There, hope your brain is "in whack" now; good luck with your classes, now that you are awake!
Christian R. Conrad
MY opinions, not my employer's - Hedengren, Finland.
Christian R. Conrad
mail me at iki.fi ; same user ID as here
First, I don't think KFK2's "Maybe I can get a deal with the Delphi 3 I have for Win9x" meant "Wow, maybe I can sell my Delphi / Windows CD to someone"; I think it meant "Maybe Borland will recognize it as a valid ground for quoting me an 'upgrade' price, not the full one, for Delphi / Linux".
Second, what "all the computer magazines" have you seen D3 CDs on? Danish ones?!? Here in Finland, I've only ever seen Delphi 2 -- repeatedly, actually -- on *British* magazine _PC Plus_.
(Yeah, D5 is officially out; the first official launch in Europe was here in Finland last Thursday, August 19.)
Christian R. Conrad
MY opinions, not my employer's - Hedengren, Finland.
Christian R. Conrad
mail me at iki.fi ; same user ID as here
I have a large investment in development which has been for Windows. As much as I look forward to moving apps to Linux, I dread the thought of being forced into C++. And I dread still further the thought of having to code for Linux as I would have before Delphi saved me from hard coding for the WinAPI.
I will be shocked if Delphi for Linux makes cross-OS coding transparent. But if it makes it practical with compiler conditionals, I will be extremely pleased.
Even more than the matter of cross-OS compatibility is the cross-OS components issue. If we are able to immediately make use of most, if not all, of our existing collection of components, then we, as developers, win big time.
--- Bill
Inprise helping the enemy, huh? Guess Redmond will be starting to forget to send Inprise their beta OSs and API docs ... Inprise dev tools will start mysteriously crashing on new OSs ...
...
IBM, Lotus, Novell, et al. have all been there before
Regards, Ralph.
Try again..
:)
You PAY for the commercial QT 2.0 (which is also available for Windows and can make life easier for them moving Delphi from Windows to Linux).
They could easily include the runtime price in the total end user price..
It might be interested to see if they will release a free version "for non commercial use" which will allow users (students, home users) to compile open source Delphi programs. I would call that - cool
Hetz (Heunique)
>I predict that within the next year, Microsoft will announce similar >support for Linux with their major development products,
8 16S0049
Rather doubtful. In fact I just saw something over on LinuxToday saying Microsoft intends to inist that vendors must agree to tightly integrate their products with Microsoft's Visual Studio development suite of tools, which includes VisualBasic, Visual C++ and Visual J++. The article is entitled "CRN: Protecting Windows -- Microsoft launches developer program to counter competitors" and can be found here: http://linuxtoday.com/stories/9018.html. It can also be found here: http://www.techweb.com/se/directlink.cgi?CRN19990
According to question 29 on the survey (All the development tools I/we currently use are...), 50.9% of people use Delphi. This in itself indicates that the results of the survey are not valid. Perhaps 50.9% of Borland users use Delphi???
According to question 28 (The primary development tool I/we currently use is...), 39.9% of people primarily develop with Delphi, and 6.3% with Mickey$oft Visual C++. The expression that comes to mind is: "What a load of bollocks!"
Ok, Q32 (My company's annual revenues are...): 37.9% of people who responded to the email are obviously hobbyists, students, unemployed people, dreamers, coders who work in their spare time, or people who live somewhere for free. If this were a survey of *real* working software engineers, then the results would be very different - you have to pay the bills somehow.
Borland/Inprise: fix you bugs in JBuilder and port that before you lose the market to the other vendors who have already ported tools (e.g. IBM. Metrowerks). Port your C++ tool and give it to students to lock them into Borland and thus help keep them away from Microsoft. Then consider Delphi.
Then again, the survey was supposed to be global. I've heard that Delphi use is greater outside of North America... but that much so???
...
Only if you don't know your toolkits. The editor you use will only affect the look of your program if you don't care about the graphical presentation. You can make beautiful applications with Emacs or vi and absolutely disgusting applications with VB or Delphi.
In the end, companies care about productivity in programming
...
Although they have a funny way of showing it at times
If you mean ease of GUI creation, then I have yet to see a GUI builder that comapres to some of the Motif ones. However, Kdevelop is shaping up nicely. Just a pity my knowledge of C++ is fairly bad, which means I find myself sticking to GTK+ and Motif, despite the enticing out-of-the-box Windows & Unix portability of Qt.
Chris Wareham
PigleT writes:
;-)
"Me, I hope that delphi becomes one way of doing gtk and qt stuff in pascal, rather than a *rival*."
Oh, I'm fairly sure that that's exactly what Borland wants to do. They'd be stupid to invent a whole new GUI widget set, wouldn't they? Then they'd just have one more headache, developing and maintaining that. No, *two* -- the other would be, developers might be reluctant to use something that didn't automatically give their applications the same "look and feel" as what end-users are using already.
"Would open source still be so cool if there were more that ran on windoze?"
Hell yes! In fact, at http://www.torry.ru/ you'll find some components (plug-in stuff ready to use in the RAD IDE) that are already LGPL'd. The Delphi community -- yes, I think there *is* such a thing -- is actually among the most "Open" you'll find on Windows.
(I think it's because using such a wonderful tool makes you much more happy, and happy people are much more likely to be sociable and generous. No wonder perhaps that Visual Basic code slaves, and the poor souls who have to wrestle with "Visual" [Ha!] C++ / The Abomination Known As MFC to give them their "components", are such grinches...
Christian R. Conrad
MY opinions, not my employer's - Hedengren, Finland.
Christian R. Conrad
mail me at iki.fi ; same user ID as here
"What Borland can do is simply write the delphi compiler that transforms Objective Pascal into C, which can then be compiled through the gcc backend."
Eh, 'simply'?!? What a horrible kludge!
"I believe AT&T did this with their C++ compiler, but I could be mistakened."
No, you're not mistaken; that's what they did.
You may have noticed, though, that it isn't what they (C++ compiler vendors; dunno if AT&T has one any more) do nowadays.
I think there is a reason for that.
Christian R. Conrad
MY opinions, not my employer's - Hedengren, Finland.
Christian R. Conrad
mail me at iki.fi ; same user ID as here
"If the Windows Share/Cripple-ware crowd thinks (which seems to be the case here) that this will somehow cause a demand for their crap to be ported over to linux, they are sadly mistaken."
So you think the license is built into the language; Delphi can only be used to build shareware?
My, what a fuckwit you must be.
Christian R. Conrad
MY opinions, not my employer's - Hedengren, Finland.
Christian R. Conrad
mail me at iki.fi ; same user ID as here
clawson (rhetorically) asks:
"If you use MS-based tech (COM/DCOM, MFC, ADO, etc.) in Windows Delphi apps, how can you expect this to port to an environment that doesn't have these things?"
Yup, that makes it hard.
That may be one partial reason why many Delphi developers use as little as possible of that kind of stuff.
Besides just in general wanting to do as little as possible to promote MS-proprietary "technologies", even on MS platforms, that is.
Christian R. Conrad
MY opinions, not my employer's - Hedengren, Finland.
Christian R. Conrad
mail me at iki.fi ; same user ID as here
"I also hope they use the QT widget set, its object oriented and generally eaiser to program than GTK+"
That's irrelevant.
Whatever they implement Delphi on top of, that implementation *will* be the VCL [Delphi's class library, the Visual Component Library] "on top of" whatever it is.
That's the whole point -- Delphi more or less *is* the VCL; the compiler is just incidental, needed to translate your VCL-using code (and the VCL itself) into machine language.
So if QT is easy and GTK+ hard, or whatever, doesn't really matter to the *Delphi* coder -- Borland's people will get to wrestle with that, porting the VCL. (OK, that was an over-simplification, since one of the main points of Delphi is that you can also write your own components. But basically it's true.)
One possibility, though, is that it would in fact be easier for them (and us, and third-party component developers) to use GTK+ if it is *less* "advanced" than QT; it's one thing to build on top of a low-level API, that can be tedious enough -- but implementing a higher-level class library on top of *another* high-level class library that "goes against the grain" of the one you're trying to build looks like it could be a real struggle.
AFAICS, FWIW, etc.
Christian R. Conrad
MY opinions, not my employer's - Hedengren, Finland.
Christian R. Conrad
mail me at iki.fi ; same user ID as here
"Delphi uses Borland's VCL library which wraps up most of the important elements of the Windows API, but not all of it. That was a concious design decision -- not to try and wrap everything [ . . . ] Because of that, many existing Delphi and C++Builder programs have some Windows API calls sprinkled in with their VCL code. [ . . . ] it won't be as easy as a recompile [ . . . ] even if as much as 20% of the code needed to be tweaked"
Another issue is *where* that Win-API-calling code is located.
If it is in your project itself, then you have work to do -- but then you've bungled on re-use and all those OO benefits already...
If you isolate it in an IDE-installable component, you'll only have to port that once, and as long as the interface is the same, your project code is then freely movable.
For instance, I've bungled like this myself in one current project: I override the Create and CreateParams methods in some of my forms to make them appear as part of another form, the code is taken directly from Charlie Calvert's "RitcheyForm" in _Delphi 2 Unleashed_. But the new "Frames" in Delphi 5 seem to be exactly the same code wrapped up in a component/form type -- port those to Delphi/Linux (which no doubt Borland will do), and all I'd have to do was use those in stead of my own semi-kludge and I'd be done.
The issue of API code hindering portability will be a problem mainly for in-house and third-party *component* developers, not *application* developers, I think.
Christian R. Conrad
MY opinions, not my employer's - Hedengren, Finland.
Christian R. Conrad
mail me at iki.fi ; same user ID as here
Er... have you seen the Qt prices? A single-developer licence for Qt is $1550, for one platform only. I'm sure Delphi sells for a lot less than that.
And there is no 'runtime price' - the price is per developer only.
-- Ed Avis ed@membled.com
As others have pointed out, this is big news indeed, and critical to the future success of Linux.
Although this is often overlooked, especially by mainstream press, and was certainly missed by 'bad-old-days' Apple execs, the thing that really killed the Mac in business settings was not a lack of productivity apps, cost or bad marketing. It was the lack of a decent RAD environment. Not having tools that allowed you to easily create crossplatform frontends to corporate databases [so 4D doesn't count here] left the Mac in a ghetto for corporate creative professionals, scientists, etc.. [PowerBuilder was too little to late, and just took too many resources.] When standarization really took hold in the mid 90s, it was no wonder that Macs around the country got the boot. Its too bad Delphi was never made available for the Mac. (IIRC, Phillip Kahn hated the Macintosh, Bedrock notwithstanding, or perhaps withstanding!)
It's a fact: more than productivity apps, email, or anything else, corporate machines live and breath on custom database applications. And even with new models like distributed objects, etc.., don't expect mainline 4GL tools to go away _anytime_ soon.
I'm now one of those rah-rah Java distributed objects guys myself. But though I don't use Delphi (don't do much RDBMS development at all anymore), having it available for Linux is just the ticket, and is just as imporant as having a good office suite. This is good news for Borland, who will have the _only_ mainstream object-based RAD 4GL tool that will run on both Wintel and Linux. It is great news for Linux, and might be a significant milestone for an end to Wintel hegemony.
Bravo Borland!
Do you have any evidence for this claim? Or is it just libel?
I guess it's convenient your anonymous.
----------------------------------------------
Stephen Molitor steve_molitor@yahoo.com
The Borland poll specifically referred to Linux; as such, they seem more interested in porting to Linux than to one of the commercial unices. Motif is no standard on Linux, given that the only way to distribute binaries ordinary Linux users can use is to statically link them, and having a tool that leaves lots of redundant Motif crud in one's binaries wouldn't look good.
Besides, if they chose Motif they'd have to licence it, adding it to the price of Delphi. For that, they may as well use Qt (one of the most programmer-friendly toolkits I have seen).
The one thing that's kept me from embracing Linux as my main OS has been that I get paid for writing Windoze apps. Being able to write in Delphi, a tool I find to enhance my productivity greatly, and then simply to do a recompile from Linux to Windoze would allow me to embrace Linux full time.
I'm sure I'm not alone, and I'm sure that I'm not the only Delphi developer who has used the Borland tools out of a disdain for using anything by Microsoft.
From my perspective, a port of Delphi to Linux, whatever GUI(s) might be supported, could easily bring a LOT of desktops into the Linux world - a LOT of Delphi work is in building internal applications for businesses.....and it wouldn't be at ALL hard to write a component set to support MySql databases, or even PHP scripting.....
I'm not feeling that clever this morning.
Although I loathe Pascal, even in it's objectified Delphi version, I've got to say that this great news. At companies I've worked at with Windows programmers, they all sang the virtues of Delphi. Even hardened C++ and Visual Basic programmers loved it.
If the API's are going to be similar (I assume some wrapper around Motif and the Unix libraries) to the Windows version then we could see some interesting cross-pollination of ideas.
I haven't read the article yet to see if they're also porting C++ Builder, but I suppose Unix already has great C/C++ development tools.
Chris Wareham
Enough ranting..
Kenny
If Borland choose Motif as their standard Linux GUI, I will be greatly disappointed. Laziness will have won over good design criteria, and several layers of design mistakes will be entrenched as a Linux standard. Many upcoming Linux applications will then be dependent on Motif, consigning Gtk and Qt to also-ran status.
If Borland want something that promises easy porting from Windows, the obvious choice would be WINE, in its function as a compatibility library. If they want a toolkit for new Linux applications, Gtk or Qt (provided the licensing issues can be worked out) would be best.
Beats me. These are just two great tastes that seem to go together. So yes, replace Wintel with Windows if it'll make you happy!
Someone said "now I don't have to code the GUI by hand." I just wanted to point out that there are already free tools available that help in that regard. There is QtEZ, a very basic GUI builder for Qt development, and KDevelop, a GUI buider and IDE for KDE.
--
Interested in XFMail? New XFMail home page
Actually, you do see such apps written in Delphi. I heard HomeSite was written in Delphi. In fact, Delphi is written in Delphi. While VB and Delphi are often compared against each other, Delphi is actually in a class all of it's own. I don't know of any other RAD product anywhere close to Delphi's quality that can actually generate efficient compiled code.
VB has its advantages, especially when doing some COM stuff, but it doesn't compile to efficient code. If you try to generate a stand-alone executable, then mysteriously the executable grows by the same size as the run-time DLL.
Actually there are a number of commercial apps out there written in our lovely Delphi. MultiEdit is another.
;-)
Oh my! Will you be able to live without vi or Emacs?
- Alfred Ayache
Toronto Delphi User Group
http://www.tdug.com
Those standards are changing pretty fast at the moment.
With Open Source Unix variants constantly gaining market share, the importance of CDE+Motif decreases.
Qt seems to be used by rather large companies (see this interview), so there is obviously room besides Motif.
In the end, companies care about productivity in programming, and that's an area where Qt definitely beats Motif.
The only question is how much the underlying toolkit (e.g. Motif vs Qt) would affect coding in Delphi/Pascal.
If you need Windows API calls, you could just link with the WINE compatibility library which would provide that layer. It wouldn't be as elegant or efficient as writing it natively, but it'd be good enough for quick ports.
Amazes me that the Linux zealots start blowing their party horns without any critical analysis. Borland did a survey that showed, surprisingly, that Borland products are in the most demand on Linux. Notice they list JBuilder highly and don't bother to list VCafe and J++, the leading market share Java IDE products.
Don't hold your breath for Delphi. By the time it gets ported Java will be the defacto Linux choice for OOD, at least for serious commercial development. Fore scripting PERL is fine; who needs pseudo-object Delphi.
Like the previous poster said, I'm not sure you really understand what RAD is.
IMO, RAD was good because it allowed programmers to do an end run around all the software engineering experts that would take a year to approve, specify, analyse, test, etc.., ad naseum, a simple corporate address book implementation, when someone could simply hack one out in a couple of days using a RAD tool, and easily fix any problems noticed after the users get a hold of it.
Not the methodology to use to design an MRI console or a Nuclear power plant, but works just great for some situations.
I'm betting on either 2002 or 2003, myself.
Will in Seattle
Re: cross-platform...
not too likely.
If you use MS-based tech (COM/DCOM, MFC, ADO, etc.) in Windows Delphi apps, how can you expect this to port to an environment that doesn't have these things?
If you write low-level apps in Linux, how can you expect the Unix-like stuff to work on Windows NT, much less Win9x? [read up on the gotchas of AT&T's UWIN and Cygnus's CygWin as far as Unix vs. MS system behaviors]
Sure, if the GUI stuff is cross-platform that can be a big time saver...
>>Delphi is such a poor language
As opposed to what? C, C++? Ha! Yes, C/C++ suit some people better than Delphi Object Pascal, but I've found that most people who insult OP have never actually used it. They think since they learned Standard Pascal way back when they know what they're talking about.
Delphi Object Pascal is clear, fast, simple, and can do anything you need it to do. It has some quirks, like a funky if..then..else issue with where to put the semicolons, but they allow Delphi to be a very fast single pass compiler.
The OP that Delphi uses is designed 100% to be a RAD, compiled language. Getting C++ to be a RAD language is a hack. There are other RAD languages out there, but as far as I know, they are all INTERPRETED.
There already exist a number of tools that ease the transitioning of code between Windows and Unix. The GUI toolkit examples that I've seen rely on Motif as the common denominator. Given this expertise perhaps it wont come as a surprise if Borland choose to use the Motif libraries.
Not that I'm a Motif advocate, although some of the criticism levelled at it is unfair. The 'designed by committee' aspects do extend to some of the convenience routines, but compared to Xlib it was a dream come true, and it even improved on Xt by having widget creation convenience routines.
The Motif UIL compiler is still an unsurpassed feature in terms of rapid GUI creation. The Glade GUI designer has some similar principles in it's use of XML to describe an interface - I wonder if this was influenced by UIL.
Chris Wareham
Delphi itself is just another fine IDE. By the time Delphi for Linux will be released, we'll have quite a number of IDEs anyway.
The crucial thing is what widget set Borland/Inprise will use.
They could write one of their own (partly using Win code) or use wrapper around existing toolkits.
The conservative (and IMHO stupid) solution would be to take Motif.
The cheapest solution would be a free toolkit like fltk or, more likely GTK+, although Inprise could be hesitant to embrace a fast moving open source 'target'.
A very interesting solution would be Qt. The main complaint at the moment is the price, which makes it more difficult to develop small commercial or shareware programmes on Linux/Unix. This is due to the fact that you get 1 year of support from TrollTech.
A cheaper, no-support version bundled with Delphi would definitely be attractive for many developers, while the high number of sales would ensure TrollTech the necessary revenue.
For you people wondering WHY Delphi is so popular.... try it, you may find you like it. I've been working with it for a couple years now, and have found that it has the ease of (flame suit on ) VB, and the power of C++ (maybe not to the extreeme extent, but pretty damn good!)
Also the fact that the survey was aimed at Borland developers many of which already use Delphi, it is no surprise to me!
For their part it is an extreemly good stratigic move. Since Microsoft will publish VB for Linux WHEN HELL FREEZES OVER, or Microsoft, buys it out... Inprise can grab a HUGE part of the market.
For Linux, there will be a huge influx of developers that already know the ins and outs of Delphi, so there will be a lot more apps! Which is a very good thing!
Just my 2 cents (1 cent US!)
Right now the only way to write crossplatform code is to use Java (which in many cases is not really an option right now). If Borland somehow manages to make delphi programs cross platform (i.e. you can pick up delphi source and compile it on both linux and windows), that would be a huge gain for the Linux community (more apps) but also for Borland because this will attract loads of new customers (windows programmers who like to create programs for Linux too and Linux programmers who like to create programs for windows too).
So, way to go Borland!
Jilles
I predict that within the next year, Microsoft will announce similar support for Linux with their major development products, at the very least as a matter of reflexive vaporware.
> 50.9% of people use Delphi.
The pool didn't show that. It showed that 50.9% of the people who answered the pool use Delphi (or are liars).
What Borland (Inprise) can use the pool for, is
1) Get a vague idea of the interest from the absolute numbers.
2) Get an idea of what those people who are interested in both Borland and Linux would like to see most. These will be their "easiest" customers, so that is quite significant.
Porting Delphi first seems like the brigtest move for Borland. They have a lot of existing customers using that, and unlike Java and C++ they are the only provider.
The article @ IT director is incorrect. InterBase is currently a part of borland.com, it also had always had a completely standalone team of developers. They started as Groton Database Software and then changed their name to InterBase Software Corp. Some time later they were purchased by Borland.
Moreof, IB is availavle in different incarnations: Sun, Muzzdie, NetWare. IB 4 for Linux is free, but comes with no source.
Also, it's known that Delphi's Object Pascal compiler is built around C compiler core and diferences between Builder and Delphi are rather cosmetic. Remember, it took very little time for Borland to ship Builder after they sucessfully launched Delphi.
KuroiNeko
Point: Voice your opinion! Vote for more quality software! Buy Delphi for Linux!
--
"All that is visible must grow and extend itself into the realm of the invisible."
I swear by MacOS X. Although I use to swear *at* MacOS 9...
Choosing Motif would be like shooting your self in both feet with a very fine, new gun.
I don't know if QT is "ready" for it, but that would be a clear possibility. At least version 2.0.
As long as it is not Motif. Motif is to say the least ugly and is non-free.
I really dig the idea of having DelphLin 5.0, because the "Delphi" way of working makes me feel happy and light of mind.
... and Windows definitely doesn't!
Best regards,
Steen Suder
Best regards,
Steen Suder
-- for email: send to
..and glade for GTK+ into C/C++/Ada (sooner or later, perl as well)...
Me, I hope that delphi becomes one way of doing gtk and qt stuff in pascal, rather than a *rival*.
Would open source still be so cool if there were more that ran on windoze?
(Eg qt, gtk, glib, things: I've used ports of the last two on windows before now...)
~Tim
--
~Tim
--
Rushing on down to the circle of the turn
That would be SO nice.. Apps would work under both Linux AND Win32.. Just a simple recompile..
-- I'm the root of all that's evil, but you can call me cookie..