Lazarus revolution lazarus / 13-General ---------------------------------------------------------------------- Arminius · 2009-04-26 15:57 I hope that I am in the right place, if not please move my topic to the right one. Well as the titel already says, I think it's the time for a kind-of "Lazarus-revolution". On every developpers place in Linux I hear C++/C/C++/C/C++/C (and maybe sometimes Java). I first stared to developp in Delphi and recently tried C++. But it's just not the same... I really don't understand why everyone is using it (besides the fact that everyone uses it, so there are a lot of ressources :p). So what we really need is to get some base of programmers that use Object Pascal/Lazarus. That would have two positive effects. More people interested in Lazarus, resulting in more support on every topic. Furthermore it will result in more people developping Lazarus -> faster and potentially better developpement. So what do you think  Is there any IRC channel for Lazarus? (btw we all know that most Linux distros are based on C++, I guess it would be possible to do it in object pascal too right?) ---------------------------------------------------------------------- Marc · 2009-04-26 16:46 Quote from: Arminius on April 26, 2009, 05:57:55 pm... So what we really need is to get some base of programmers that use Object Pascal/Lazarus. why do you think there isn't ? QuoteIs there any IRC channel for Lazarus? Yes: for lazarus: freenode #lazarus-ide for fpc: freenode #fpc Quote(btw we all know that most Linux distros are based on C++, I guess it would be possible to do it in object pascal too right?) What do you mean by C++ based distros ? FPC and Lazarus are already part of some distros. ---------------------------------------------------------------------- Arminius · 2009-04-26 17:24 Quote... So what we really need is to get some base of programmers that use Object Pascal/Lazarus. why do you think there isn't ?Because I talked to the KDE and Kubuntu devels on their channel an they said no one there was developping in Object Pascal... On 3 (popular!) forums I asked, everyone just told me: "Switch to C++, it's better for Linux." So that's kind of obvious... There is nothing for Pascal like kde-apps.org (btw I put an app programmed in Lazarus up there http://www.kde- apps.org/content/show.php/Wiper?action=content&content=103268). Thanks for the channels C++ means that all it's developpement is done in C++. What distros are developped in Lazarus then? ---------------------------------------------------------------------- Chronos · 2009-04-26 20:00 You can start development of Free Pascal based OS where could be Free Pascal primary language. Until than there will be always secondary... ---------------------------------------------------------------------- Arminius · 2009-04-26 20:20 Yes that would be a great idea, I always wanted to do that... to be more precise Atlantis OS (based on Stargate Altantis :p) with an emphasis on secrurity and exeternal devices plug-in (maybe with a small AI? (just kidding ))... but on my own, I think it's close to no chance, even with linux from scratch. But if you want to help, we are already  two :p The thing I wanted to say is that there should be a website like kde- apps for Lazarus, and that it should work tightly with the most well knows OS to diffuse the apps largely. ---------------------------------------------------------------------- Troodon · 2009-04-27 14:48 My 2 cents: "of course" Delphi and Lazarus are better than C++ but the threshold is still set too high for Delphi (Object Pascal) to be adopted by the masses. First, Delphi is a strictly commercial product, the Turbo version is buggy and not fully featured. Lazarus/FPC is free but requires some experience to figure out a few things (previous experience with Delphi is an asset) and documentation is limited. Second, university students learn C++ and Java, mainly because there are strong incentives from the mainstream vendors for the universities to stick to these languages. Third, for better or for worse the programming world is slowly moving towards .Net, because thus spoke M$ and because Novell has some success with Mono (particularly on the iPhone). Now, there is no reason why Lazarus could not be "adapted" to generate .Net code but I think that is a different programming paradigm and I am not sure it should be pursued. Finally (again, this is just my opinion), there may be reasons related to the Object Pascal language syntax that keep some potential new commers away. I would like to suggest a simple experiment: create a new FPC mode where begin...end can be replaced with {...} to make the code more "readable". I know, there are arguments both ways. Nevertheless, I think that if there was a separate "Lazarus/FPC Documentation" project, with a few dedicated people working exclusively (non-developers) on writing detail-rich, up-to-date Beginner's and Developer's Guides that would be also available as print-on-demand books, then the chances of Lazarus/FPC becoming more popular would increase manifold. The documentation project and team should be at least as well structured and perceived at least as important as the development team. No offense meant. Some of you might remember how well written was the Turbo Pascal 5.5 OOP Guide. ---------------------------------------------------------------------- Arminius · 2009-04-27 15:30 I know that few years back Delphi was the language they used, both for programming their devices and for teaching students. Now it's C++, and I really don't like it... (compared to Delphi I learned 2-3 years before). Delphi is stricly commercial, I agree, but Delphi 6 is offered completely free in it's personal version. For the {}, it's widely used, even in php if I recall correctly, but I don't really like it. For me  begin and end is much much clearer! But it would be good to give the choice (because a lot of people like it that way). As long as you have the choice between both, it's ok. What kind of incentives form the vendors are there? But don't you think that making an OS based on it would help diffusing Object Pasxcal and Lazarys no? And creating something like KDE-Apps for Lazarus would be a real help too! ---------------------------------------------------------------------- Troodon · 2009-04-27 16:03 Quote from: Arminius on April 27, 2009, 05:30:59 pmWhat kind of incentives form the vendors are there? Visual Studio for free. Quote from: Arminius on April 27, 2009, 05:30:59 pmBut don't you think that making an OS based on it would help diffusing Object Pasxcal and Lazarys no? Absolutely. However, that's a tall order compared to the Lazarus/FPC project. Quote from: Arminius on April 27, 2009, 05:30:59 pmAnd creating something like KDE-Apps for Lazarus would be a real help too! Yes, for sure. The more (commented) examples are available, the better the chances of adoption. ---------------------------------------------------------------------- Arminius · 2009-04-27 16:19 Great we agree  Who could kind-of "decide" that? Is there some kind of "leader" that could mobilise the troops? ^^ ---------------------------------------------------------------------- Troodon · 2009-04-27 17:26 For the OS, I think one way to do it is to create an open-source project at SourceForge. Then see if you can get some people interested to join you. For the apps portal you would need a host. Perhaps you can discuss that with the forum admin. It would be like an extension to this, this, and this. ---------------------------------------------------------------------- Marc · 2009-04-27 17:48 an OS is not a distro ---------------------------------------------------------------------- Arminius · 2009-04-27 17:54 Well ok if you take it strictly, Unix is the OS and for example Kubuntu the distro? Yes the finding people is the more difficult part  For the apps protal, would "Marc" be the right person? ---------------------------------------------------------------------- Chronos · 2009-04-27 18:53 Pascal based OS (and lately more distros) could be really significant achievement for pascal community. But currently there is no chance to do that. Object Pascal language have all needed capabilities but Free Pascal is not right choice to do the job. But who can do it better nowadays. Technology is get more and more complicated and complexity get greater in exponencial manner. Who can create pascal compiler with support for vast major processor and microcontroller architectures. Who can keep step with commercial or major open source compilers and IDEs and have full support for parametrized types, aspect oriented programming, full featured stable IDE with data intelligent modeling, advanced source code management, integrated versioning, component management, internationalization support, and many many other complex features. There can be no more simple pascal language text compiler but heavy infrastructure with bindings to surrounding technology. Who can lead this huge complex project for free? Who can get together such many developers and manage them to work effective and in synergy. Task of this size needs to leadership and support of some strong open source company. Delphi is already dead because of commercial only nature and ignoring open source community. Turbos is useful but based on old BDS 2006 with significant restrictions and there is targeted only to windows platform. Free Pascal + Lazarus is pretty nice and nearly useful but lacks some important advanced features so it can play only second league. Pascal as in my opinion is beautiful abstract clear language is doomed. C/C++/C#/Java will be successful as there are many commercial companies behind its. And there are many major products like windows, linux, a many others and advanced free to use IDEs like Visual Studio Express, Netbeans and many others. But there is no reason to stop Pascal effort because Pascal based languages are simply better than others Lazarus is nice open source equivalent of Delphi 7 with even some improvements like cross platform and utf8 support for example. So what we need is good pascal to C/C++/C#/Java compiler to overcome todays technology complexity and future demands I don't think that will be some good Free Pascal/Lazarus based apps in KDE or any similar linux project. Everything is mainly C/C++ here. ---------------------------------------------------------------------- Arminius · 2009-04-27 19:24 I understand the problem... So all we need is a big corporation behind us? No problem! Oh wait... that's a problem  What's about CodeGear? That would be good for them because helping to promote it. Or maybe Skype or an other corporation like that? (apparently Skype used Delphi so...?) That could give them "good publicity". Btw, I talked to the KDE and Kubuntu devels, the only reason why they don't use Pascal is that they don't have anybody that would uses it... The question might sound stupid but why did the big companies decide do support C++ and not the clearer and better language Object Pascal? Because C++ sounds better? :p For the all overoptimistic and completely idealistic OS thing  Using Linux From Scratch wouldn't that help making a distro? I might say stupid thing so sorry, but is that build in C++? And couldn't we put a layer of Pascal of it? ^^ ---------------------------------------------------------------------- Chronos · 2009-04-27 20:22 Who know why C is more popular than Pascal. Reasons could be many for example somebody can say pascal is only toy for learn algorithms. Maybe Pascal was too abstract at that time and people prefer low level language closed to macro assembler in these old times (compare some macro assembler and preprocessor directives in C). Maybe C have better authors and company support in opposite of academic use of Pascal. Maybe academic research of pascal and shifting to next levels like modula, modula 2, oberon, oberon 2 and component pascal wasn't do any good for glory of pascal. Maybe Pascal wasn't mature enough as in beginning original basic pascal standard and even extended pascal was not so good as nowadays Object Pascal. Possibly Pascal is really not so good as we can imagine. Who knows what have greater historic influence. Interesting is that Apple used Pascal for development of their OS in old times. For good of Pascal someone have to establish stable open source company. I think this is pretty important as some developers have to work on various pascal project every day and for use in commercial open source application good technical support is needed as well as wiki, forum, irc chat, source code repository and another supporting stuff. You can make distro based on FPC/Lazarus tools but there will be needed pretty huge programmer effort to that. And even after success there will be only C based linux kernel and bunch of subsidiary tools. Core will be still pretty old and POSIX based. Who want to build castle on swamp? Object or future Pascal could be surely significant better than old C based Linux kernel... ...but here we are, always only academic discussions as always. You can try to download FPC source and explore all files and figure out how compiler works and than you can go back and write your ideas. ---------------------------------------------------------------------- Troodon · 2009-04-28 03:25 The fact is, C was created for writing operating systems. Pascal, on the other hand, was created for writing applications. Unfortunately, some people (many of them actually, with some encouragement) became confused and thought that all applications should be written in C/C++. That is what I call the First Catastrophe in Application Programming. The Second one occurred when Java became the new programming paradigm. But Java is basically an interpreted subset of C++ (okay, we might disagree on this). The only advantages these two languages have over Object Pascal is the number of followers and, consequently, the amount of online help available for each. And the helping hand lent by their respective sponsor companies. Object Pascal / Delphi / Lazarus/FPC would only stand a chance at mass appeal if they were taught in universities and if they underwent some changes that would make them appear more "modern" at the syntax level. I am and will always remain a TP/Delphi fan but I must admit that sometimes I find sections of C# code easier to read than Delphi because... they contain more spaces. ---------------------------------------------------------------------- Ñuño_Martínez · 2009-04-28 08:15 Actually there's a kernel written in Pascal: Toro. By the way I don't know how an OS written in Pascal will help to increase its popularity. AFAIK old MacOS GUI was written in Object Pascal. I think current language popularity is defined by money. The most popular languages are the most popular because there's a big company with lot of money paying people to say it's popular. That's the only reason why C# and ASP are still popular when anybody can see that Objective C, Object Pascal, Java and PHP are much better languages. But that's only my opinion. ---------------------------------------------------------------------- Leledumbo · 2009-04-28 11:07 My little baby FPOS is also one of many efforts to create a Pascal based kernel. You might want to check it out (and help developing it ). ---------------------------------------------------------------------- sauer2 · 2009-04-28 12:08 Just another OS Kernel won't make it. What is needed, is a real killer application written in Lazarus/FP. ---------------------------------------------------------------------- Chronos · 2009-04-28 14:13 What would be that "killer application" ? Another calculator, audio or movie player or even complete office suite? All of this already exists and it still runs on top of platform based of another language. And if you want to made for example fast application you should make it with same langugage as OS and libraries. C/C++ on Windows for win32 and C/C++ on Linux as well. You will choose java on java based system and assembler on assembler based system. So to fully take advantage of our lovely language we need go to roots. After that somebody can judge this language and make comparison which system is better. Delphi was pretty advanced IDE and langugage with wheir component model but it was running only on top of poor procedural win32 and todays C# and their object oriented windows api are throwing delphi to dark as delphi is only addition to base windows platform. Skype is made in Delphi and who from users cares about that. I can't imagine better pascal satisfaction than good open source OS with fast, advanced and stable kernel and with many good aplication. But it cant be done without money and leadership of strong company but there is not much sense of doing that now or in near future. ---------------------------------------------------------------------- sauer2 · 2009-04-28 16:16 Well, maybe the Pixel Image Editor could be a chance, it was announced for the day after tomorrow.                        Even if it's commercial, it runs on many platforms and looks on all platforms the same, because it uses a selfmade toolkit. And that's all written in Freepascal. I just hope, it's worth its money, because the beta 699 was full with bugs and disfunctions. EDIT: If Skype is written using Delphi, how did they made the linux ports? With retarded Kylix? I hope not... ---------------------------------------------------------------------- Arminius · 2009-04-28 18:08 From Wikipedia: Products developed with Delphi Among the many products developed with Delphi are:     * Accounting: Intuit Lacerte     * Business: SME:Service Management Enterprise Service management software     * CodeGear products: CodeGear Delphi, CodeGear C++ Builder, Borland JBuilder versions 1 and 2     * Communication: Globestar Systems Connexall (Event Notification and Presence Management Platform     * Database management/development: MySQL Tools (Administrator, Query Browser), TOAD     * Engineering Software: Altium Designer/Protel (electronics design), OrcaFlex (offshore structural analysis)     * Image viewers: FastStone Image Viewer, FuturixImager, Photofiltre, BioniX Wallpaper Changer     * Internet messaging: Skype (VoIP and IM), QIP (IM), The Bat! (e-mail client), PopTray (e-mail checker), FeedDemon (feed viewer), XanaNews, Xnews (newsgroup reader)     * Music production: FL Studio     * Point of Sale and Retail management: Infinity RMS by Triquestra International     * Software development: Dev-C++ (IDE), DUnit (unit testing), Help & Manual (help system authoring), Inno Setup (installer engine), Game Maker     * Web authoring: Macromedia HomeSite (HTML editor), TopStyle Pro (CSS editor), Macromedia Captivate (screencast)     * Web browsers (MSIE shells): Avant Browser, Netcaptor     * Utilities: Spybot - Search & Destroy, Ad-Aware (anti-spyware), jv16 PowerTools, Total Commander (file manager), Copernic Desktop Search, PowerArchiver, Beyond Compare(file comparison)], WordWeb (dictionary/thesaurus). (http://en.wikipedia.org/wiki/CodeGear_Delphi) I think the thing about a Pascal based OS is that if people wanted to developp it, they would be kinde-of "forced" to get to know Pascal better, right? I'd love to help developping such a thing but... that's far beyond anything I touched... I tried doing some TCP (which somehow worked, but now doesn't anymore  ), UDP (that never worked, and couldn't find anybody able to tell me why) and even 3D (GLScene didn't work 10 times, worked once and then didn't work again, while following point by point the tutorial, no idea why). Now I was wondering about something, with a Pascal Kernel, would the use of KDE (or Gnome) be possible? I guess I could ask the devels if they would help inserting it... Btw wouldn't it make sense to fusion those 2 seperate projects in one "big"? ---------------------------------------------------------------------- LazaruX · 2009-04-28 20:25 ...and don't forget that Resource Hacker was also made in Delphi... ---------------------------------------------------------------------- Legolas · 2009-04-28 21:45 IMHO the real killer applications written in Lazarus/FPC are Lazarus and FPC themselves ---------------------------------------------------------------------- Troodon · 2009-04-28 23:10 Quote from: Legolas on April 28, 2009, 11:45:43 pmIMHO the real killer applications written in Lazarus/FPC are Lazarus and FPC themselves  I totally agree! ---------------------------------------------------------------------- Leledumbo · 2009-04-29 03:48 QuoteIf Skype is written using Delphi, how did they made the linux ports? With retarded Kylix? I hope not...AFAIK, the Linux port is written in C/C++ and the Mac port is written in Objective-C (3 different languages for 3 different platforms, what a waste... ) ---------------------------------------------------------------------- Arminius · 2009-04-29 14:09 Quote from: Troodon on April 29, 2009, 01:33:55 amWhat do you think? While I think that the idea is kind of revolutionnary, it don't know if it would work. I know that there is already one that was written in Java (I think). But I don't understand, what are the advantages of such a type of OS... And I find it a bit creepy the part "lazons". for my part, I wouldn't like to see my PC part of some network, I would like kind of "slave" with no idea what data those lazons would exchange. But in the same spirit, would some kind of AI not be good idea? (or has that already been done on other OS?). The thing is I am wondering... what kind of support would we need from corporations? Promotion? Other things? I certainly don't have the authority to do that, but I think we don't have anything to loose to ask do we? For Skype for example we could contact them a propose that "only one free and multi platform lazarus" might help. Actually, who is financing the FPC/Lazarus project right now? I still think that a Pascal OS would help spreading the word about Lazarus, thus get more people to want to develop it, and of course I think it's vital to develop FPC/Lazarus.  I asked (just to know) the KDE Devels and they said that they would help integrating it (they did it already once), usually the distro devels manage to do it themselves. But that's only my opinion ... What would you think about a reunion on the IRC channel to do some kind of brainstorming? ---------------------------------------------------------------------- Chronos · 2009-04-29 19:43 Its better to help get Lazarus better and more useful. For example make Lazarus look like Delphi 2009 with docked windows and better component list. Or help implement advanced parametrized types features or something other like type interfaces, records with procedures, classes with nested types, for-in construction and iterators, better reflection than RTTI, interpreted or run time compiled code, live templates, refactoring, ... I don't know what of these features are already implemented in free pascal. Or you can try make adapt compiler to some new processor or whole platform. Time is precious. ---------------------------------------------------------------------- Arminius · 2009-04-29 20:30 Well I think that Delphi 2009 doesn't look as good as Delphi 7... What do you mean by docked windows? Sure I would love to help, but I have no idea where and how. I programmed a lot in Delphi, but always personal, small project, where I knew what I did, even when I explored completely new domains. Here I have no idea where everything is. Is there a tutorial? Btw, I think a good place to make publicity for the project is www.3dbuzz.com (where I learned Delphi from). Some of them might/could help. And the creators of the website are skilled persons, that might (I say might), agree to help. But again, I don't have the authority to ask for that. Would an admin do it? Please? ---------------------------------------------------------------------- Leledumbo · 2009-04-30 02:26 QuoteWhat do you mean by docked windows?In simple term, windows embedded in another. For example, try ManualDocker component. ---------------------------------------------------------------------- Arminius · 2009-04-30 14:29 Ok I see now Are there any infos for that? About structure etc...? But first I got to get those errors out of the recompile of Lazarus. Btw did you get my PM? ---------------------------------------------------------------------- Leledumbo · 2009-05-01 04:54 QuoteOk I see now Wink Are there any infos for that? About structure etc...?Docking QuoteBtw did you get my PM?Yup, I'm still writing a Dia diagram for it and some docs about the design structure, so you can understand easily. ---------------------------------------------------------------------- skalogryz · 2009-05-02 18:49 btw, i was thinking about 'killing application'. there's really no need to invent a new OS. for example, may be, it's better to use existing OSes? Make a Linux distrubutive based on FPC/Lazarus developed applications? It's not about rewritting every possible application to FPC, but rather implementing basic GUI applets in pascal, like: system configuration, basic OS utils, etc. Infact, they can be as simple as just GUI wrappers over command-line tools. ---------------------------------------------------------------------- Arminius · 2009-05-03 08:18 On one side, you might be right, but on the other, to get people interested in Object Pascal, it must be a "cool", and not too small project to work on. Something they really desire to work on. The problem with the command wrap-up is that sometimes it is usefull to stay in command line (to pastebin any problems to IRC instead of describing what happend). Again, that problem might be solved if we output what command was used+what response the programm got. But what kind of distribution were your thinking of? Lazbuntu? xD (I talked about Ubuntu because it's the mostly widerspread distro, and we want to get a lot of people, and maybe canonical behind us  :p) The main problem is that OF COURSE what needs to be developped is Lazarus/FPC. But if we don't get peoples interest in Lazarus, there is no chance of that happening. And who would spontaniously want to developp a compiler for a language you don't use?   Who would be interested in some kind of "Promote Lazarus/FPC - Object Pascal"? Which would include floding ( ) the "market" with apps, maybe distro partially/entirely based on Object Pascal (and of course contacting several websites/organisations (nothing to loose))? ---------------------------------------------------------------------- sauer2 · 2009-05-03 12:54 QuoteAnd who would spontaniously want to developp a compiler for a language you don't use? Summer of Code participants ---------------------------------------------------------------------- Marc · 2009-05-04 11:06 why do you the the Summer of Code is interested in pascal ? ---------------------------------------------------------------------- Arminius · 2009-05-04 14:54 Yeah I didn't find any Pascal related project in gsc. ---------------------------------------------------------------------- Marc · 2009-05-04 17:13 (we already tried twice to apply) ---------------------------------------------------------------------- Arminius · 2009-05-04 17:25 And what reason of refusal did they give you? ---------------------------------------------------------------------- Vincent Snijders · 2009-05-04 17:41 AFAIK, none. ---------------------------------------------------------------------- LazaruX · 2009-05-05 09:35 I think a killing application made in Lazarus should be a Cross Compatible desktop environment. Like I proposed here: http://forum.lazarus.freepascal.org/index.php/topic,4558.0.html I am not talking about a Window Manager, but a DE. So a: - start menu - application bar - file explorer - ... But I also think the Lazarus community should not become that huge like some Linux distro's. ---------------------------------------------------------------------- bee · 2009-05-05 09:47 As web application become more demanded nowadays, I think a killing "application" for Lazarus (and FreePascal) is an application framework for web application. It had been discussed here: http://forum.lazarus.freepascal.org/index. php/topic,2489.msg31223/topicseen.html#new What I imagine is we could build a RIA right from Lazarus IDE (more or less like IntraWeb) and the web application automatically cross platform (thanks to FreePascal) and easy to deploy. By providing this framework, we could build web application visually, rapidly, and much easier and faster (both on development and execution) than using PHP. ---------------------------------------------------------------------- bee · 2009-05-05 10:00 > What I imagine is we could build a RIA right from Lazarus IDE (more or less like IntraWeb) and the web application automatically cross platform (thanks to FreePascal) and easy to deploy. Or maybe even more! We could build a full-blown web OS using the framework. The OS itself (start menu, task bar, global events, activity monitor, control panel, theme, etc) is provided by the framework. An "application" that is build by the framework can be "installed" or "uninstalled" into the OS, just like a desktop OS. If the OS isn't available, the application can be run independently as a single web application (but without any features from the OS). We could also make two versions of LCL, one for desktop application, one for web application. With just some configurations, we could compile any applications become a desktop application or web application, from single source code (ok, with minimal IFDEFs). Well.. in theory, it's very possible and doable. If it's really done, then FPC/Lazarus could become a truly cross platform development tool. Both as desktop application (on any OSes and widgets) and as web application (on any [web] servers and browsers). Isn't wonderful? *dreaming* ---------------------------------------------------------------------- duncanparsons · 2009-05-05 14:07 Quote from: bee on May 05, 2009, 12:00:01 pm> What I imagine is we could build a RIA right from Lazarus IDE (more or less like IntraWeb) and the web application automatically cross platform (thanks to FreePascal) and easy to deploy. Or maybe even more! We could build a full-blown web OS using the framework. The OS itself (start menu, task bar, global events, activity monitor, control panel, theme, etc) is provided by the framework. An "application" that is build by the framework can be "installed" or "uninstalled" into the OS, just like a desktop OS. If the OS isn't available, the application can be run independently as a single web application (but without any features from the OS). We could also make two versions of LCL, one for desktop application, one for web application. With just some configurations, we could compile any applications become a desktop application or web application, from single source code (ok, with minimal IFDEFs). Well.. in theory, it's very possible and doable. If it's really done, then FPC/Lazarus could become a truly cross platform development tool. Both as desktop application (on any OSes and widgets) and as web application (on any [web] servers and browsers). Isn't wonderful? *dreaming* Isn't that pretty much what Morfik is about? ---------------------------------------------------------------------- Chronos · 2009-05-05 14:40 Anything is possible but who would implement it. Who have enough motivation to do huge projects like everyday job for free. For the web OS. Some already exists using PHP + java script like lovely http://www.eyeos.com/. Try demo. It could be done with Free Pascal instead of PHP but it can't be killer application now as many other projects exists. And again who will lead big project pro many years. Graphic components and widgets for web development is really not necessary as many PHP programmers make projects simply using text IDE and writing own HTML code. For web developing with FPC we can use Powtils (http://z505.com/powtils/idx.shtml) already and simple procedural code with Write procedure can be similarly attractive as echo function in PHP. It's your choice to use complex web classes or simple set of procedures. Both approach Powtils and Intraweb have its advantages. ---------------------------------------------------------------------- Troodon · 2009-05-05 15:32 I think FPC/Lazarus is already a truly cross platform development tool. As for the "Web" application proposal, I really don't know what a Web application is. The client-server paradigm, based on the concept of time-sharing, is alive and kicking, and always will be. Whether it's Web or J2EE or .Net or SaaS or whatever, it's still client & server softwares working together. The native code-based application is the fastest and most complex thing end-users can run on their desktop. The server can benefit from more computational power and doesn't need a GUI. The critical element is the client-server communication approach and that can be as simple as a database. RDBMS systems are optimized data-exchange systems, there is no point reinventing the wheel. Ten years ago experts argued that Web-based software was "the next big thing" and it turned out not to be because of security concerns and because the client (the browser) is either non-standard or too slow or not complex enough. What we need for Lazarus/FPC at this time is excellent tutorials on database usage (in particular, for ZeosLib) and a strong platform- independent multimedia component library (I never have been able to compile ACS). IMO the "killer app" includes a Lazarus client desktop application that can take maximum advantage of the local hardware, a local or remote database system (preferably MySQL for performance reasons), and a remote FPC server application (service/daemon). ---------------------------------------------------------------------- Arminius · 2009-05-05 20:00 I created this page: http://wiki.lazarus.freepascal.org/Lazarus_revolution If you are interested in helping a project, please write me so I can add your name, or if you have a wiki account, just add yourself. Please tell what kind of project you would like to join, if any specific. Yes, I know, Lazarus is the most important one, but you need to have something that evolves around it too. ---------------------------------------------------------------------- bee · 2009-05-06 05:28 Quote from: duncanparsons on May 05, 2009, 04:07:00 pmIsn't that pretty much what Morfik is about? No, it's not. Morfik's web OS paradigm is making the browser as the OS. While what I meant web OS here is a web application that acts like an OS (within a browser). Chronos had given the example: eyeOS. ---------------------------------------------------------------------- bee · 2009-05-06 05:48 Quote from: Troodon on May 05, 2009, 05:32:55 pmI think FPC/Lazarus is already a truly cross platform development tool. Yes, it is. I have no doubt about it as I use it in daily basis on many platforms, mainly Mac and Linux. But, I wanted to extend a bit what platform meaning.  While in native code, platform is more to OS and processor, I also want to include web and desktop as a "platform". So, a truly cross platform, by my "new" meaning, also supports web programming out-of-the-box. Quote from: Troodon on May 05, 2009, 05:32:55 pmAs for the "Web" application proposal, I really don't know what a Web application is. An application that is called by browser through http protocol. Quote from: Troodon on May 05, 2009, 05:32:55 pmTen years ago experts argued that Web-based software was "the next big thing" and it turned out not to be because of security concerns and because the client (the browser) is either non-standard or too slow or not complex enough. Really? How about google? yahoo? facebook? friendster? orkut? twitter? wordpress? etc, or even this very forum. Don't they big enough to you? Not yet to mention thousands of web applications that run in private network. Also thousands of framework for web application out there, they wouldn't be alive if there's no big demands. I think both desktop and web application has its own dis/advantages, but none of them is "bigger" than the other. It all depends on the requirement and demand. Browser incompatibility is yesterday story. There are many frameworks out there that hide those incompatibilities. It's like LCL that hide incompatibilities among many widgetsets. Quote from: Troodon on May 05, 2009, 05:32:55 pmWhat we need for Lazarus/FPC at this time is excellent tutorials on database usage (in particular, for ZeosLib) and a strong platform-independent multimedia component library (I never have been able to compile ACS). IMO the "killer app" includes a Lazarus client desktop application that can take maximum advantage of the local hardware, a local or remote database system (preferably MySQL for performance reasons), and a remote FPC server application (service/daemon). Well, it's more as a mandatory task but I don't think it'd be the "killer" application. ---------------------------------------------------------------------- bee · 2009-05-06 06:14 Quote from: chronos on May 05, 2009, 04:40:15 pmAnything is possible but who would implement it. Who have enough motivation to do huge projects like everyday job for free. Ouch! Quote from: chronos on May 05, 2009, 04:40:15 pmFor the web OS. Some already exists using PHP + java script like lovely http://www.eyeos.com/. Try demo. It could be done with Free Pascal instead of PHP but it can't be killer application now as many other projects exists. And again who will lead big project pro many years. Yes, many languages or frameworks are getting popular and become "standard" not because they're the best, mostly because simply they're the first. C and PHP are good examples. If you come later and want to get over their popularity, whether you come with big money to support it (M$ style) or you got many significant advantages over the prior (FLOSS style). By the way, if you have way of thinking like that, then I think we don't need to continue this discussion as there are already something for anything. We don't need fpc as gcc had them all already. We don't need Lazarus as Eclipse had them all already. We don't need LCL as wxWidget had them all already. Then why are we here in the first place? Quote from: chronos on May 05, 2009, 04:40:15 pmGraphic components and widgets for web development is really not necessary as many PHP programmers make projects simply using text IDE and writing own HTML code. For web developing with FPC we can use Powtils (http://z505.com/powtils/idx.shtml) already and simple procedural code with Write procedure can be similarly attractive as echo function in PHP. It's your choice to use complex web classes or simple set of procedures. Both approach Powtils and Intraweb have its advantages. Well, I agree each programming approach has its advantages. It's like whether you can make a program using direct API call (procedural) or using LCL (OOP). Powtils is good, no doubt about it. But it's not ready for RAD. The same goes to PHP. It's 80's programming style. Today, I would prefer design my form visually rather than write it in code. IMO, HTML, CSS, and JS, is the "assembler" of web application. All of them can be wrapped within an OOP framework so we could use them effectively and efficiently. It's exactly like the reason why we use Lazarus and its LCL (or Delphi and its VCL). Though nowadays you can still make any programs using assembler, but do you really want to do that? That's why I see an opportunity there for pascal and its (coming) web application framework. Our competitors in this field are not yet many, the big players are google (using java and its GWT), Ruby, Adobe (flash family), and M$ (.net family). But most of them are not native, so performance is an issue to them. Pascal is the only native solution here (of course if we want to join the battle in this field). ---------------------------------------------------------------------- Chronos · 2009-05-06 11:53 Quote from: bee on May 06, 2009, 08:14:12 am Today, I would prefer design my form visually rather than write it in code. IMO, HTML, CSS, and JS, is the "assembler" of web application. All of them can be wrapped within an OOP framework so we could use them effectively and efficiently. It's exactly like the reason why we use Lazarus and its LCL (or Delphi and its VCL). Though nowadays you can still make any programs using assembler, but do you really want to do that? It would be pretty good to have these pascal/delphi components for all sort of languages like different versions of HTML, XHTML, CSS, JavaScript, MathML, SVG. But where are all of these now? Who can or want to maintain these components? And what about SQL components and object database? It takes a lot of human effort to do good and clear components and even that huge universal components will be too slow and too complex for simple task. But better something than nothing... ---------------------------------------------------------------------- Troodon · 2009-05-06 12:16 Quote from: bee on May 06, 2009, 07:48:15 amReally? How about google? yahoo? facebook? friendster? orkut? twitter? wordpress? etc, or even this very forum. Don't they big enough to you? Not yet to mention thousands of web applications that run in private network. No, they are not. Most of them are kiddie applications that we could all live without. The companies that make money from the adds would disagree, of course. But don't mistake profitable for useful. Google's search business was more successful than, say, Altavista's because the Google techs knew how to make many Linux computers work together, but that has nothing to do with the Web browser. I have yet to see a company using Google Docs for mission-critical stuff. Companies do use intranets (private networks) intensively, mostly to provide simple interfaces for their back-end applications (database systems, binary and Java apps). For a developer, browser Javascript incompatibilities become apparent as soon as you want to add too much functionality to your client interface. I know it because I used to develop commercial Web-based software. It's those "little things" that work in Internet Explorer but not in Firefox or Safari and only partly, or differently, in Opera; finding workarounds can drive one crazy. It's a good thing you're enthusiastic about Web applications and I agree they have their niche but there is only so much one can do in a Web browser. By the way, it's nice you put a link in your signature but how many people on this forum do you think can read your wordpress blog? ---------------------------------------------------------------------- bee · 2009-05-06 12:16 Quote from: chronos on May 06, 2009, 01:53:10 pmIt would be pretty good to have these pascal/delphi components for all sort of languages like different versions of HTML, XHTML, CSS, JavaScript, MathML, SVG. But where are all of these now? Who can or want to maintain these components? Well, as I told everyone on the other thread, a group of people had been starting and maintaining the effort in ExtPascal project. It's not as hard as it sounds in fact. We just wrap a mature existing javascript framework (in this case is ExtJS) that handles all of the "assembler" things into pascal framework. It's more or less like the way LCL wrap API calls to existing widgetsets. This approach saves lots of time and effort if we do all the "assembler" things by ourselves. We just need more people to boost progress of this project. Quote from: chronos on May 06, 2009, 01:53:10 pmAnd what about SQL components and object database? It takes a lot of human effort to do good and clear components and even that huge universal components will be too slow and too complex for simple task. I don't think so. It's more or less like the way FCL wraps database API calls on different database backend as sqlDB or FCL-db components. It does requires effort and time to do that but it's not as huge or slow or complex as many people think. ---------------------------------------------------------------------- bee · 2009-05-06 12:31 Quote from: Troodon on May 06, 2009, 02:16:29 pmNo, they are not. Most of them are kiddie applications that we could all live without. The companies that make money from the adds would disagree, of course. But don't mistake profitable for useful. Google's search business was more successful than, say, Altavista's because the Google techs knew how to make many Linux computers work together, but that has nothing to do with the Web browser. I have yet to see a company using Google Docs for mission-critical stuff. Companies do use intranets (private networks) intensively, mostly to provide simple interfaces for their back-end applications (database systems, binary and Java apps). Internet wouldn't be as expansive as today without supports of good applications that run over it. Alright, to some points I may agree with you that most web applications are not "mission-critical" stuffs. But, that's the point, because not all applications is a mission- critical (well, IMO mission-critical applications is minority), web browser (and web application) got its position and attention. For that mission-critical stuff, we may use desktop application. Quote from: Troodon on May 06, 2009, 02:16:29 pmFor a developer, browser Javascript incompatibilities become apparent as soon as you want to add too much functionality to your client interface. I know it because I used to develop commercial Web-based software. It's those "little things" that work in Internet Explorer but not in Firefox or Safari and only partly, or differently, in Opera; finding workarounds can drive one crazy. Have you tried any javascript frameworks out there that hides all those incompatibilities so you may write program (using that framework) without being bothered by those "little things"? This is exactly like the problem that LCL had been trying to overcome. If you believe LCL could overcome incompatibilities among different widgetsets, why can't we apply the same approach to browser incompatibilities issue? We don't even need to do it by ourselves, many people had done it for us, even for free! ---------------------------------------------------------------------- bee · 2009-05-06 12:37 Quote from: Troodon on May 06, 2009, 02:16:29 pmBy the way, it's nice you put a link in your signature but how many people on this forum do you think can read your wordpress blog? Hehehe... I used to write many programming related articles in English in my blog. But, as I think I better to promote FPC/Lazarus to my Delphi developers fellow in my country, I started to write more articles in my own language. It's a good question though. I think I should write in English once in a while. Thanks. ---------------------------------------------------------------------- Troodon · 2009-05-06 14:03 Quote from: bee on May 06, 2009, 02:31:22 pmnot all applications is a mission-critical (well, IMO mission-critical applications is minority) You've never worked in a large corporate environment, have you? ---------------------------------------------------------------------- bee · 2009-05-07 06:47 Quote from: Troodon on May 06, 2009, 04:03:45 pmYou've never worked in a large corporate environment, have you? I don't know what you meant by "large corporate environment", but my company operates national-wide in my country. I got some distributed client-server apps, also some web apps. As I said, both client-server and web app has their own dis/advantages and consequences. None is bigger over the other, it all depends on the demand and requirements. You know, I also may ask the exact same question to you. ---------------------------------------------------------------------- Arminius · 2009-05-08 19:55 I understand that those project require a lot of time, but nobody says it has to be done in a week. So please, if your are by principle interested in promotion via programming, tell me/write me/add yourself. ---------------------------------------------------------------------- Arminius · 2009-05-14 16:56 Ok I made the base. Please tell me what you think: http://wiki.lazarus.freepascal.org/Lazarus_revolution ---------------------------------------------------------------------- Troodon · 2009-05-14 17:52 Speaking of KDE and Qt, has anyone compiled a Lazarus/FPC application with Qt-4.5? Thanks. ---------------------------------------------------------------------- Vincent Snijders · 2009-05-14 17:56 Yes. ---------------------------------------------------------------------- Troodon · 2009-05-14 17:59 Excellent! ---------------------------------------------------------------------- Arminius · 2009-05-16 18:23 I was wondering... How difficult would making a firewall be? Couldn't that be a kind of killer app? It's not like a antivirus that you have to keep frequently up to date with their virus definition. ---------------------------------------------------------------------- sauer2 · 2010-01-07 22:42 Hehe, now (more than a half year later) i know one: Someone should make more nice games like Hedgewars. ---------------------------------------------------------------------- Ñuño_Martínez · 2010-01-14 23:50 I'm on the way, sauer2. Read my next post.