Slashdot Mirror


Borland Releases Kylix 2

Tal Cohen writes "Borland Kylix 2 is now available. Most new features are geared at Enterprise-level developers; the Open edition is still available for free download. The CLX (cross-platform component library) is covered under both GNU and Borland's license." The new features list is interesting - a fair number of buzzwords, but it also looks like they are supporting a lot of the new stuff. The white papers have some interesting topics - including gcc vs. Kylix.

34 of 227 comments (clear)

  1. Maybe too late by lekter1 · · Score: 3, Interesting

    I think it's too late for Borland. Maybe two years before ...

    But now I think it's too late. Kdevelop and the recently released kdestudio 3.0 gold is playing hard.

    --

    greetings,
    lekter
    http://www.hispacluster.org
    1. Re:Maybe too late by uradu · · Score: 5, Insightful

      That's funny in a way. KDevelop is trying hard to emulate Visual Studio, which itself has done its damndest to incorporate any Delphi features it could pilfer (esp. ever since Heijlsberg got on-board), while also introducing a slew of new languages and a new component model. Yet Delphi is still considered the black sheep, because it uses Object Pascal. Well, happy C# coding--never mind that it's a veritable semantic OP clone.

    2. Re:Maybe too late by Junks+Jerzey · · Score: 5, Insightful

      Kdevelop and the recently released kdestudio 3.0 gold is playing hard.

      You've never used Borland's Object Pascal compiler. For all intents and purposes, even going back to the early days of the Pentium, it compiles instantaneously. Give it a large project on a 333MHz PII and--bang--it is compiled and linked before you lift your finger off of the Build key. This is a huge, huge productivity boost.

      Does the compiler do as much optimization as gcc? No. But it's still an optimizing compiler that gets within the "I don't care about the difference" range.

      When I see people talking about needing dual Athlons to get their gcc compile times down to the single digit minutes, then I'm appalled. With Delphi you're at *zero*. That's liberating beyond belief.

    3. Re:Maybe too late by iplayfast · · Score: 3, Informative

      I've found exactly the same thing. I program in several languages, and Delphi definititly is the fastest compile time. The IDE is also nice and allows quick development time. Also has a good debugger. It's worth the investment of 1 or 2 days to learn pascal.

    4. Re:Maybe too late by bitflip · · Score: 3, Funny

      I don't like it. When am I supposed to take a smoke break?

    5. Re:Maybe too late by Junks+Jerzey · · Score: 3, Informative

      Am I right in assuming that this is due to the equivalent of pre-compiled headers and overall better modularization than C(++) allows? Paraphrased: is it the compiler or the language that gives opportunity for these short compile times?

      Sure, there are some things about Object Pascal that make it simpler to compile than C: no macro preprocessor, module info isn't contained in huge text files that need to be compiled over and over again, syntax is generally cleaner, and so on. But there are clones of Object Pascal and they don't get anywhere near the amazing compile times that Borland gets.

      The big implementation wins, as best I can tell, are:

      1. They removed the need for a traditional, general linker. Modules are written to disk in a very simple format, what seems to be a binary image that just needs a small bit of patching.

      2. The compiler is simple recursive descent, without separate lexing and parsing phases.

      3. Much of the core of the compiler is highly optimized, with much of the critical portion written in finely tuned assembly language.

  2. e-buzz word whip lash... by killthiskid · · Score: 5, Funny

    I just read the "what's new" page and I got e-buzz word whip lash...


    They should put up some damn warnings or something.


    i.e. : Build Web Services-enabled database middleware with DataSnap(TM) that scales and interoperates with your complete e-business solution...


    Ok, now a serious question... is ANYONE out there using this? I've read the reviews, I read some tutorials, and my interest is sparked, but I want to here some testimony.

    1. Re:e-buzz word whip lash... by O2n · · Score: 4, Funny

      That phrase alone is "bingo" for everybody in the audience...

    2. Re:e-buzz word whip lash... by borgboy · · Score: 5, Informative

      Yep. Lots of people are using this. Lots. DataSnap is the new term for what Borland previously branded as their MIDAS technology, but due to trademark collisions, yada yada...

      DataSnap is based around a couple of classes:
      1. a TClientDataset class, which is responsible for representing an in-memory dataset which can be persisted as XML and has the capability to record offline dataset changes and post them back to the data persistence layer when necessary.
      2. a TDatasetProvider, which links a TClientDataset to a persistence layer, such as a RBMS such as Oracle, MySQL, Interbase, DB2, or even (gasp) MSSQL. There is also an TXMLTransformProvider that can act as a 2 way mapping layer from a dataset to an XML document.
      3. A TCustomRemoteServer descendant which supplies the remoting capability - placing the TClientDataSet and the TDataSetProvider on separate machines. These components can provide connections via HTTP, vanilla sockets, CORBA, etc. There are also load balancing helpers to distribute the load.

      And then there are the Web Services. Yep. That works too. It's SOAP, plain and simple.

      We've got a DataSnap app deployed today, handling payroll data for ~1400 retail outlets. Heck, the TClientDataset class itself is worth the investment, even if you never build a n-tier system with it.

      --
      meh.
  3. Re:Downloaded this a while ago.... by _DMan_ · · Score: 3, Informative

    I spoke to a Borland rep at a trade show 2 weeks ago. She told me the beta would be available early next year.

  4. Waiting for C++ by wiredog · · Score: 4, Informative

    I started programming with TurboC and Turbo Assembler. I'm now using Builder 5. The drag'n'drop interface is very nice, as are the make files (which are XML,btw). Borland has always had very good compilers, and the STL they use is quite nice. They also ship printed documentation, for those of us who actually rtfm. As soon as they have the C++ version done (RSN for about a year now), I'll buy it.

    1. Re:Waiting for C++ by conan_albrecht · · Score: 4, Insightful

      I'm sure a C++ product will follow, much like C++Builder followed Delphi. However, I can understand why they didn't implement this for C++ first: their target audience is not C++ programmers.

      Any C++ programmers who are already programming for Linux are probably using gcc and the many tools available. Borland is trying to woo Windows programmers to Linux. Since Delphi already has a large source code base out there, making it source code compatible with the Windows version helps Windows people feel comfortable in Linux. These people would have a much harder time switching if they had to start using gcc and vim.

      Once they feel they've wooed as many Delphi developers to Kylix for Linux/Windows development, I'll bet we see a C++Builder for Linux as well.

  5. QT Embedded support? by Anonymous Coward · · Score: 3, Interesting
    Does this have support for generating QT Embedded applications?

    If so, then you can write PDA applications for the Sharp device in the previous story using Kylix/Delphi/Pascal++.

    Not that any non-Delphi person would want to, in my opinion ;)

    Of course, I am basing all this on the assumption that Kylix actually currently uses QT as the base GUI component set, via its own intermediary toolkit abstraction. If I am wrong here, please correct me.

  6. Never mind that these are the guys... by uradu · · Score: 5, Interesting

    that introduced the $49 IDE back in 84, and have consistently turned out more standards-compliant C++ compilers than you know who. And your tastes obviously don't count for much if they lead you to that opinion about Object Pascal. It's still considered by many one of the most advanced and elegant natively compiled languages around. OTOH Forte is certainly not considered the greatest IDE by many. But then again, those are YOUR tastes.

  7. Re:Turbo Pascal by truthsearch · · Score: 3, Insightful

    I don't see much competition among the all-in-one development environments for Linux. Kylix is an attempt to become the MS Visual Studio of Linux. I don't see anything else as broad as this. KDevelop is kind of a Turbo C app. I'm not familiar with KDE Studio, but it seems to be the only real competition.

    Developers, and especially companies, really want the grand development environment that nicely brings everything together. That's exactly why every Windows developer eventually dropped Borland's array of products and went to MS Visual Studio.

    Borland here is trying to keep up with the times and learn from the past. I think this is a great step for Linux.

  8. delphi portability? by juraj · · Score: 3, Interesting

    Has someone of you used this to port Delphi app to
    Linux? I had a nice free LGPL-covered application, that
    I wanted to compile using kylix open edition. But a lot of things are different.I see a lot of units, like QDialog, QForm, etc. under Linux, but they're counterparts in Windows are Dialog, Form, etc. So is there any sourcecode compatibility? Is there a tool for doing this?

    1. Re:delphi portability? by robinjo · · Score: 3, Interesting

      I've ported a pretty big non-gui app. It was pretty easy. Moving from Delphi's old sockets to Indy was the biggest thing. Then I just wrote higher class classes, wrapped all the Windows api stuff in those and made a WinStuff-unit out of them. A similar one for Linux. Then some ifdefs for units and done. Now the project compiles without changes in Delphi 5 and Kylix 1.

      I hear it's even easier with Delphi 6 but haven't felt like upgrading yet.

  9. Re:Delphi? by Codifex+Maximus · · Score: 3, Insightful

    And just what is wrong with Pascal? I happen to think it's a wonderful language - it can be depended on. I'd rather use Delphi/Pascal ANY DAY rather than program in any version of BASIC.

    --
    Codifex Maximus ~ In search of... a shorter sig.
  10. They definitely don't make it easy... by Genom · · Score: 3, Informative

    They definitely don't make it easy to "register" for the download -- first they want all sorts of personal info that really shouldn't be required (phone #, street address, etc...) - THEN, if you don't fill out EVERY field (including those not marked as required), and say "YES" to all of the spam checkboxes at the bottom, their javascript form handler balks at you.

    Not to mention that you *have* to have javascript enabled to even register...

    I was going to check it out -- but I *refuse* to give them free reign to spam me by phone, fax, email, and snal mail for the privilege of doing so.

  11. Kylix isn't Klyx by Tha_Zanthrax · · Score: 3, Interesting

    They won't be sued. Kylix and Klyx are completly different apps. The name Kylix isn't stolen from Klyx. Like Delphi is also the name of a greek goddess, a Kylix is a greek fruitbowl of some sort.

  12. Measuring developer use of OSS by pubjames · · Score: 4, Interesting

    It has always been common knowledge that a key to Microsoft's dominance is making things easy for programmers so that they develop for the Windows platform. Is it just me has there been a real drop-off of interest by developers in all things Microsoft?

    I remember there was a time about five years ago when most developers wouldn't even consider developing for anything other than Windows technologies and developer's magazines reflected that. These days, however, I see very little excitement about Microsoft technologies, for instance, I don't see lot of enthusiasm amongst developers about .Net and C#. Surely with Windows being the dominant platform, and .Net being Microsoft's new technical strategy, you'd expect some excited discussion about it amongst developers, but it's just not happening.

    This is just a feeling I have, and I have been trying to think of a way to quantify it, if nothing else to prove to myself that this sea-change is actually occurring and not just because I now take my information from different sources. The simple metric I have come up with is this - the number of times a word occurs on Google:

    Linux - 30,100,000
    Microsoft - 20,100,000

    This crude metric seems to suggest that Linux has 10m more pages than all of Microsoft's products put together. Seeing as Microsoft has such a dominant position in the desktop space and is still much more of a household name than Linux, I think this is quite a clear demonstration that there is a lot more material about Linux out there than about Microsoft's products.

    This came as a suprise:

    "Linus Torvalds" - 640,000
    "Bill Gates" - 649,00

    I would have expected Bill Gates (who's a household name) to occur a lot more than Linus.

    This is also suprising:

    "Internet Explorer" - 2,730,000
    Mozilla - 2,730,000

    "Linux developer" - 20,600
    "Windows developer" - 12,200

    Is it just me, or do these figures suggest that Microsoft should be very worried indeed?

    1. Re:Measuring developer use of OSS by kin_korn_karn · · Score: 3, Interesting

      It's not so much a lack of interest in windows as a focus on the web. The majority of software development nowadays (judging by job postings, which indicate growth) is 2 or 3 tier client-server stuff with a web interface.

      MS has the worst web server technologies, period. Nobody in their right mind wants to run IIS for a large commercial site, it's too vulnerable to DoS/hacks. Apache on Win32 would fix this but people who are timid enough to run Windows as an internet server are going to be scared to death of the prospect of software that doesn't give them the option of running to mommy (read: tech support) when it breaks.

      Also, NT/2k x86 machines don't scale enough to handle the load of a high traffic web site. Load balancing (LocalDirectors, etc.) helps, but why fool with a room full of 1 or 2 processor Xeons when you can buy one really expensive Sun chassis and have room to grow for years without going to the trouble of integrating a new server into the network? Not to mention that x86 machines don't run so well when you try to implement failover stuff like redundant power supplies and hotswap drives on their bus.

      Last but not least, NT doesn't run EJB very well. If you want to get performance out of EJB, which is the current buzzword-compliant technology for large-scale software projects, you need a Sun box, period. I'm convinced that Sun and IBM are conspiring to cripple Java on Windows, which is just fine with me.

      In most cases, the only necessary MS-based piece I've seen in recent software development is the browser.

      MS may have the browser market tied up (with good reason, IE is nice now and XP, I hate to say it, is a great desktop OS), but they've already written a browser and have said under penalty of perjury that they'll give it away for free. They're in the process of eating their lunch on the server side in the big-ticket markets, which is where all the important stuff happens anyway.

      Stop worrying :)

    2. Re:Measuring developer use of OSS by platypus · · Score: 3, Insightful

      Also, one thing to note may be that by necessity there could be more documentation about Linux online than there is about MS products (lack of paper manuals for ISO downloaders, etc. could be reasons), which would lead to more hits for Linux-related pages. Still, it's interesting.

      Yes it is, but I doubt that it's clear that there is more documentation for linux than for MS products.
      What we should consider is the linux howtos are mirrored a bazillion times throughout the internet, while in MS's case a lot of documentation is concentrated at support.microsoft.com or other microsoft sites.

      A search for "linux networking howto" yields 2770 hits for instance.

  13. Re:Turbo Pascal by bradintheusa · · Score: 3, Insightful

    Are you nuts. This product kicks butt compared to all the other IDE wannbes. Borland is the only alternative to Microsoft and doing a great job producing the best tools. It's revenues and share price (and profits) are all up up up.
    What is there not to like?

  14. Re:Wine and Kylix by stewart.hector · · Score: 3, Informative

    Kylix used WineLibs, not wine - which are different.
    Wine is to run windows applications
    winelibs are to link against winelibaries during compilation - ie, you have a windows application, and you want to compile it under linux - easier porting.

    Yes, in Kylix 1 the IDE was sluggish.

    Kylix 2 has less dependecy on WINELIBS, but it is still there. You can see postings on borlands news server - http://newsgroups.borland.com .

    Hopefully K2 IDE will be alot faster.

    --
  15. Good for business by gorillasoft · · Score: 3, Insightful


    I don't see where anyone has mentioned that this should be good for some businesses. In the case of a shop with split MS/Linux computers, they can write one internal business app in a RAD environment that will run on both of their systems. This allows for the good productivity of a RAD tool with the portability of Java, C, etc.

    It could also be an incentive to switch over to Linux - they could have their apps written in Delphi on Windows, and then move as slowly/quickly as they want when converting to Linux without necessitating major code porting. In a slow economy, cost savings are of more obvious importance to management.

    Of course, all previously MS-only code (VB, etc) would still need to be reworked, but there are benefits to be had for businesses looking at Kylix.

    If Kylix takes off, it could really be a boon for Linux.

  16. Re:Borland again... by Ronin+Developer · · Score: 5, Insightful

    Alas, you are equating apples to oranges.

    First, I am not a Java developer. But, I am aware that one of the biggest problems facing the JBuilder team was the diversity of Java VMs (i.e. some worked and some don't). JBuilder allows you to target multiple Java VMs with ease and that was a bane to releasing a "stable" product. They worked with the various Java VM teams (Blackdown is one that comes to mind) to make it compatible. JB3 was a blacksheep product. JB4 fixed many issues and JB5 is now the current product.

    After having coded in C, C++, the forsaken VB and a slew of other languages, I discovered Delphi in 1994. I haven't looked back since.

    Why? Because it has enabled me and my teams to develop applications in a fraction of the time that C++ would have required and substantially more stable and reliable than VB has ever been. My Delphi apps came in on or ahead of schedule and don't crash. IMO, Delphi is a secret weapon when you need to get high quality, database applications out to the market place when under a tight schedule.

    Yes, Delphi has had its bogus releases (Delphi 4 in particular). D3 was very stable and D5 fixed D4. D6 now offers cross platform development capability (if you use the CLX library) via Kylix. The language, Object Pascal, is not the same as Pascal just as C++ is not C. I suggest you take a few hours and learn the differences between Object Pascal and Pascal. The only real downside to Object Pascal is it is more verbose than the equivalent C/C++ code. But, then again, it's also a lot easier to understand and maintain (a side benefit of its Pascal roots).

    Kylix has bugs (just as any major new tool does). Blame that on both errors in the Kylix tool itself as well as buggy Linux distros (RH in particular). To me, the only bug that really affects me in Kylix is the fact that TThread is broke. Did they fix it in K2? Let's hope so.

    My point being, is that you had a bad experience with a single Borland tool. Never mind the fact that many other development shops gave it high praise...you had a bad experience.

    Before you blast Borland for putting out "inferior products at insane prices", I suggest you learn more about their products, read some serious critical reviews and then try the products yourself. You may come to realize just how far off base your statements really are.

    As for the high prices...well, I can't dispute that. They did it to keep pace with Microsoft. Why? Because how can a product be good when it costs so little? Surely that other product that costs twice as much must be twice as good. Right? By that reasoning, I guess that means that Open Source and Free software must really suck. We know that's not true. FWIW, Borland now appears to be reviewing their pricing structure.

  17. C++ is too a priority by fm6 · · Score: 3, Informative
    However, I can understand why they didn't implement this for C++ first: their target audience is not C++ programmers.
    Not true. Most programmers are like you -- they like Borland's RAD technology, but they want to program in C++. Obviously Borland has to cater to them. The thing is that all of Borland's RAD products are based somehow on Delphi. (This even includes early versions of JBuilder, though the current version is 100% Java.) You don't need to know Delphi to use C++Builder, but all the core software -- the IDE, the object framework, etc. -- is implemented in Delphi. So major upgrades to C++Builder are always implemented in Delphi first. The same goes for porting the whole schmeer to Linux!
  18. Borland is a no-spam company by MarkusQ · · Score: 3, Informative
    I was going to check it out -- but I *refuse* to give them free reign to spam me by phone, fax, email, and snal mail for the privilege of doing so.

    I've been using Borland products since the early eighties, and the most I've gotten is a few mailings telling me about events in my area, product updates, and an occasional bit of free stuff. I typically tweek my address to catch/track spammers (e.g. misspell my name), and I've never had anyone else send me something using the address I gave Borland. In short, I'd trust them.

    -- MarkusQ

  19. Give me C++ any day by UnknownSoldier · · Score: 4, Interesting

    > And just what is wrong with Pascal?

    Not trolling or meaning to start a "holy war", as this is just a personal opinion:

    - Declarations are backwards, due to the awkward grammer: name type
    Compared to the more logical C/C++ way: type name

    - The distinction between functions and procedures (the language sports an artifical difference.) The lack of parenthesis in the declaration make it difficult to quickly visually spot functions.

    - Operators, or the lack of them (no bit shifts, scope operator, namespaces?) i.e. := for assign? Just give me the dam equal sign already! :)

    - Too wordy. { } are don't clutter my code whitespace, like 'begin' 'end' do.

    In short, I just hate how the language looks.

    It's the same as a person liking one spoken language over another. Sure they both can explain concepts, but which one is more compact, and is "fluent" for the person?

    Pascal is a great teaching language, and Delphi is very impressive (Borland has always had lightning fast compiles on their Pascal languages, due to the grammar.) But I'd rather take a language I hate less (C/C++) then one that gives me a grammer that I hate (Pascal & sons.)

    I like the multiparadigm support of C++.
    i.e. procedural, object orientated, and generic programming paradigms.

    For me, Pascal++ is just plain wrong, but if you're productive at using it, hey, more power to you!

    Cheers

    1. Re:Give me C++ any day by CaptJay · · Score: 3, Informative

      - The distinction between functions and procedures (the language sports an artifical difference.) The lack of parenthesis in the declaration make it difficult to quickly visually spot functions.

      That's not actually an artificial difference. The choice behind two reserved words for procedures and functions allows the language's grammar to keep its LALR form, which means you can compile it much much faster. The form of Pascal's grammar is one of the biggest reason it compiles so damn fast compared to C++: there is no operator ambiguity. The same goes for the ":=" operator.

      - Operators, or the lack of them (no bit shifts, scope operator, namespaces?)

      bitshifts operators: "shl" and "shr". You can apply "and", "or", "not", etc to bits as well.

      Namespace: You can delimit namespace by prefixing the name of the unit. Example: MyUnit.TMyClass.Create works perfectly.

      Language preference is just that, a preference, but I just wanted to explain some of the features which you seemed to dislike. =)

      Regards,

      --
      "I remember Y1K, every abacus had to get another bead"
  20. Re:The secret to the speed by mikera · · Score: 3, Informative

    Actually, single pass compilation makes sod all difference. A pass to scan a parsed file for declarations might take 1% of your time if that. Putting all the declarations at the top is more of a Pascal convention for declaring your interfaces than anything else.

    The main compilation-time advantages Pascal has over C/C++ are a simple, elegant langauge syntax, no complex preprocessor, general avoidance of header files, and the fact that modules ("units" in OP-speak) are almost always pre-compiled into a format that makes linking quick (these also double as pseudo-header files). Add to all that the fact that Borland are simply very good compiler writers.

    Of course, you lose stuff from C++ that some people (including myself) use a lot such as macros, templates, multiple inheritance etc. Whether you actually need these is debatable, but their exclusion certainly makes for a clean language that is pretty beginner and maintainer friendly.

    Last time I used Borland Pascal, it was also pretty good at stuff like dead-code elimination - not sure how GCC compares there.

  21. Re:Turbo Pascal by mikera · · Score: 3, Insightful

    Having used both Visual Studio and Delphi, I can honestly say that Borland is way aheead in the Visual development stakes. If anything, it has been MS playing catch up in terms of tool quality.

    Borland lost out to Visual Studio because of that little thing called the OS monopoly. Companies wanted the assurance that their tools are very closely tied to the OS. Borland's products are still IMO much better, and given the much leveller playing field on Linux I really hope they can succeed.

    Basically, if you haven't played with Delphi/Kylix before I seriously suggest you give it a shot. It's free to try, goddamit, and I'm sure most good coders can suspend their disdain for Pascal for long enough to realise that it's actually pretty damn good.

  22. Re:Borland again... by Ronin+Developer · · Score: 3, Interesting

    You truly are clueless, aren't you? This isn't about the average slashdotter. The need for a high productivity development environment for Linux is essential to move Linux into the mainstream and for its widespread acceptance in the business world supported by the IT community.

    While the argument that there are powerful, free tools for Linux holds. The generally accepted Open Source model for development tools simply will not cut it in a real IT shop. Tools used by IT shops need to be stable, robust, and supported. I like knowing that I can pick up the phone and get somebody when there is a problem that I don't have the immediate time or resources to solve. That's what I pay for.

    Users (business and home) want their favorite applications (or a strong competitor) available for an OS before they will accept it. These two areas are the niches where Delphi 6 and Kylix fill. D6 and Kylix do deliver what they promise. I can attest to that.

    Pascal as a language is generally dead. The modernized Object Pascal is an OOP language that meets many of the requirements of C++ programs. Object Pascal is far from dead. On the Windows platform, it is very viable. Because of the type safety inherent in the language, it's pretty hard to screw up development of simple apps. Yes, for pointers you need to do a few odd things. That's why we have class refernces. These are basically smart pointers. But, the object model is quite robust and capable.

    The underlying language is not dependent upon the visual elements you seem so quick to condemn. However, when a Delphi 6 or Kylix developer can sit down, write, test and deploy a SOAP service in under 10 minutes (yes..it was canned..but demonstrated at Borcon 2001)....well that's pretty amazing. Even more amazing is that with little effort, you have CGI, DSO, or ISAPI modules to fit almost any web server.

    These services have all the native support of Delphi's (and Kylix's) database and internet connectivity, a highly interactive development environment and a fully capable and visual debugger.

    Okay..I sound like a sales person. But, I'll tell you. After working with C, C++, VB, Powerbuilder and a slew of other languages, Delphi kicks butt on the Windows platform. The advent of Kylix and CLX on Linux will enable those same high powered productity and business apps developed using Delphi to be run on Linux. That's what it's all about...it's not a language war...it's about viability.