Goodbye and thanks for the fishes lazarus / 8-Other ---------------------------------------------------------------------- ik · 2013-03-08 17:01 Hello, After too much thinking I'ved decided to leave here. I don't know if this is permanent or not, but I can't for now continue to be here. I'ved wrote about it here: http://idkn.wordpress.com/2013/03/08/goodbye-pascal/ Goodbye and thanks for the fishes Ido ---------------------------------------------------------------------- aminer · 2013-03-08 18:12 Hello, Sorry for my English, i don't speak very well english , but i want to give my opinion, i have programmed in Perl that was inspired also by the C programming language , and i have also played with Java, and C# a little bit, but i have found them much harder to learn and to work with than Object Pascal, the only language that is as easy to learn and work with as Object Pascal is Javascript, what i want  to be enhanced in FreePascal and Delphi is the speed, Freepascal must optimize more the code for speed , they must bring us something as fast as GCC compiler or near as fast as GCC compiler. The Object Pascal in FreePascal and Delphi has helped me a lot, they are easy to maintain and easy to understand and to work with and FreePascal brought us portability, Object Pascal is  much easier than Java and C# and C , it's by experience that i am telling you this, take a look for exemple at my Parallel archiver or my other libraries, Object Pascal have made it easy for me to implement them and i have implemented Parallel archiver in only 3 days and thanks for that also to Object Pascal  , I can classify Object Pascal  in the same group as the Javascript language , it's is easy to learn to work with and FreePascal has brought us also portability to oher systems . But as i said the FreePascal implementors  must make it as fast as GCC , cause it's not fast and as Fast as GCC actually. Thank's, Amine Moulay Ramdane. ---------------------------------------------------------------------- aminer · 2013-03-08 18:22 Hello, Although FreePascal and Delphi are not as fast as GCC, you will find my Parallel archiver and Parallel compression library very fast , cause they both using compiled GCC binary objects and DLLs inside Parallel Zlib and Parallel Bzip, and also they are using assembler code inside Parallel LZO and Parallel LZ compression algorithm. Thank you, Amine Moulay Ramdane. ---------------------------------------------------------------------- Nebula · 2013-03-08 19:32 It's sad to see someone give up, rather than keep on contributing, but you've gotta do what you've gotta do. Here's hoping you do return one day ---------------------------------------------------------------------- Laksen · 2013-03-08 20:29 Good luck with any non-standardized language. Python comes to mind(is there Python 3.0 support on windows yet? I can't build Qemu where Python is used as a build system).. As for Freepascal, I think it's just the way things work. I think one of the basic vices of Free Pascal is: "Don't break existing code, and be very conservative about including new language/RTL features". As such I think the people who have been core maintainers have done a good job. "I was told that if I raised an issue, it is actually attacking the people" - Simply wrong if you weren't asking people to implement stuff for you. There has always been a hard line against certain new features in the compiler/RTL, but it's never a dictatorship. If the inclusion of new code is sound and you are willing to chip in then you'll probably be met with understanding or open ears at least. If you are against new development(I couldn't quite tell from your post), such as generics and all that, then you probably have Delphi to blame. But in case you stay with Pascal despite this dramatic post you could try to write a new mode for FPC which encouraged strict pascalisms(like {$mode objfpc} but with another pascal dialect without all the new stuff) ---------------------------------------------------------------------- Leledumbo · 2013-03-09 03:00 This is the original thread that makes him away, in case you people want to know the background story: http://free-pascal-general.1045716.n5.nabble.com/Does-FPC-2-8-0-can- actually-still-be-called-Pascal-tp5713161.html QuoteBut as i said the FreePascal implementors  must make it as fast as GCC , cause it's not fast and as Fast as GCC actually.That's one of the effect of pure OSS project vs paid OSS project (GCC has some people getting paid to work on it). If you have enough compiler optimization knowledge and FPC architecture, please improve the optimization for us. One easy thing (in my mind, I can't do it yet either) is the array indexing optimization. For instance: Code: [Select]for i := Low(a) to High(a) do begin   .. do operation with a[i] .. 1st optimization: a[ i ] is computed on each iteration, while this is can simply be detected as pointer incrementation. GCC (and Oracle Java) can detect this already, FPC cannot yet. 2nd optimization: a[ i ] is computed on each occurence. AFAIK the CSE optimizer should already do this, but I haven't confirmed yet. This simple optimization is enough to boost your code quite significantly, since for loop with array usage is very common, including in RTL/FCL/LCL. ---------------------------------------------------------------------- BigChimp · 2013-03-09 06:28 Ido, Best of luck with your programming, thanks for your open source code & hope you'll be back with FPC again. Thanks, BigChimp ---------------------------------------------------------------------- gulyone · 2013-03-09 13:34 to me , it would be the same as saying "omg im gona stop Qt programming !!! because they introduced signals and slots and that is not in the spirit of C++" what a joke ! when i need Qt  I Use Qt ! when i need Lazarus I use Lazarus ! prefer Lazarus, for the below reasons : - i dislike distributing 3rd libs with my soft, Qt = you must - on big projects , C++ multiple inheritance => Death Diamond = ewww - Lazarus components are much more nice and numerous than Qt ones - my projects dont need speed they need to be stable and safe - if you are a pro you can write pascal code as fast as C code - finally = C++ builder , basicaly , was plugged on top of a Pascal VCL so WHY would I talk to a saint when I can directly talk to God .. that demonstrates that Pascal is God ! Ha ! if you had any doubt (and God doesn't need a ISO aprobation, as all the shorty saints do ! lmfao ) no ! you leave because you are tired of , you leave because of you , the reasons making you leave are yours and inside you , you don't have to find fake justifications or anything else ! so if you want to leave , just leave , and don't "T"roll ! and... if you want to stay , or to come back , you are welcome ! no need trolling either ! TY ---------------------------------------------------------------------- vfclists · 2013-03-09 18:56 I think you are looking at it in simplistic terms. Whenever you use a programming a language you are also joining a community. One which shares your ideas, your passion, your interests etc. It isn't just about writing your code. There is an intellectual, emotional, and a time investment involved and Ik feels that the current FPC/Lazarus community is not one that shares is outlook. Of course some people are more passionate and sensitive in such issues than others. This mailing list thread should put his view in perspective, however we hope he will back sooner rather than later, because in our own ways we are all working toward the same things - to help make the world a better place (I hope  ) These mailing list threads should help put his views into perspective. I agree with the blogging part, the Object Pascal part is more one of taste http://free-pascal-general.1045716.n5.nabble.com/OT-a-bit-Planet- Object-Pascal-td5712879.html http://free-pascal-general.1045716.n5.nabble.com/Does-FPC-2-8-0-can- actually-still-be-called-Pascal-tp5713161.html Quote from: gulyone on March 09, 2013, 02:34:59 pm "omg im gona stop Qt programming !!! because they introduced signals and slots and that is not in the spirit of C++" what a joke ! when i need Qt  I Use Qt ! when i need Lazarus I use Lazarus ! prefer Lazarus, for the below reasons : ---------------------------------------------------------------------- Blaazen · 2013-03-09 20:55 One of the criticized things are the type helpers. They work only with basic types now. Although I agree it is not very pascalish (Integer.ToString;) and I don't plan to use them, I consider them as a first step to class-type helpers and it is another case. Such helpers would allow to add not only new methods but also a new fields to existing classes. I hope they will come. ---------------------------------------------------------------------- taazz · 2013-03-09 21:08 So you say that instead of having an hierarchy of components we could do things like TDataExtenter = class helper (TFieldDataLinkHelper) for TStringgrid; and create a TDBStringgrid or what ever else one has in mind like patterns visitor bridge etc? That would be a big plus and will isolate the various smaller tasks from the hierarchy making easier to write once and use everywhere that is a big step forward if only was not restricted to one helper per class. ---------------------------------------------------------------------- marcov · 2013-03-09 21:10 Quote from: Blaazen on March 09, 2013, 09:55:03 pmOne of the criticized things are the type helpers. They work only with basic types now. Although I agree it is not very pascalish (Integer.ToString;) and I don't plan to use them, I consider them as a first step to class-type helpers and it is another case. Such helpers would allow to add not only new methods but also a new fields to existing classes. I hope they will come. Class helpers were implemented before the type helpers. See e.g. fcl- base/src/streamex.pp (in trunk). But class helpers are only that, helpers. They don't really add to classes, and need to be in scope to do any good. ---------------------------------------------------------------------- taazz · 2013-03-09 21:25 I know about the existing class helpers I've even written one for the Tbitmap to help with some color calculations but they are extremely limited if on the other hand a type helper allows a class helper to support fields as well as methods it could be thought of a pure mans multiple inheritance for example although not quite technically correct terminology it could help on enhancing the delegation of tasks to external classes making easier to mix what ever functionality you might need at the moment of course the only one helper per class per instance is a bit of a pain. ---------------------------------------------------------------------- Chronos · 2013-03-10 11:31 Yes, definitely present way of developing FPC and Lazarus is going to wrong direction. And there are many reasons for that. We all know that core problem is that only small group of developers work on these projects only occasionally in their free time or maybe some in working time in their job. But really helping with Lazarus/FPC development partially in daily job means that you are loosing precious time which you could use to build great product which could make your company best choice on market. Then there will be always argue about use of professional fully featured payed IDE or use free but incomplete product. To defend Delphi/Pascal from use of other languages we are really in weak position. It is even worse in commercial and professional open source area then in hobby area. One big problem was drop in Delphi usage during "dark times" when Borland decides to get rid of Delphi. Many of us didn't know what to expect from future. If Delphi will be dead or live. Many left Delphi development during these years. And you could tell to somebody how great Pascal/Delphi when producer itself don't want their own child. Thanks god that Embarcadero gave Delphi new momentum and Delphi caught new breath. But maybe these XE1, XE2, XE3 versions are not what many expected from further Delphi development. Ok, customers will tell them what they are willing to pay and this is how things are going in commercial environment. Another big problem is missing free Delphi version for open source development. This is hurting Delphi a lot against other languages as Java. High price for professional edition isn't helping either. Even if Delphi would be best technology on entire world, pricing is set badly and it will not be widely used anytime in future for that. Free pascal is going own way trying to mimic Turbo Pascal and Delphi. But even it is free and even usable for building commercial product, something is still bad. Maybe it is concept itself, maybe it is lack of community, maybe it lack of active developers. But I don't think that more developers will solve problems. Community in forum is pretty active but it is not sufficient for success. Something is still missing. Just take some great products as example. Let's say LibreOffice. Very nice product, usable, documented, actively developed, supported, multiplatform, community is growing. What lays behind such success? Why is so vital? Is somebody pushing it all time? Is foundation helping? Is just interest of big companies as Red Hat, SUSE and others? Is it cool product with bright future? To stay in development area let's say Eclipse IDE. Just compare Lazarus and Eclipse and you will find many differences. Many have huge impact to usability and potential community grow. Again some foundation is stand behind. Concept is set as modular open source development platform that even commercial companies have interest and use Eclipse and contribute to the project. I love pascal but many of us use Lazarus/FPC just because there isn't better open source alternative. Then we can stick to incomplete messy and not so perfect Lazarus or go on and use other languages like C, C++, Java, C#, Python, Ruby, Perl, Javascript, e.g. I would be good to have one fully featured IDE and language for everything include system programming and embedded applications, GUI applications, database applications, web applications, smartphone apps, e.g. But so far non of to me known languages offer this. Say goodbye to pascal and say welcome other language... http://www.tiobe.com/index.php/content/paperinfo/tpci/index.html ---------------------------------------------------------------------- airpas · 2013-03-10 11:57 QuoteSay goodbye to pascal and say welcome other language... http://www.tiobe.com/index.php/content/paperinfo/tpci/index.html other language that do every things !! what is it ? ---------------------------------------------------------------------- Leledumbo · 2013-03-10 12:45 Quoteother language that do every things !! what is it ?It's Pascal for me I develop kernel in Pascal, I develop console/GUI/web/dbcentric/mobile apps in Pascal. You can do anything with it as long as you know how to do it ---------------------------------------------------------------------- Chronos · 2013-03-10 12:57 Maybe Java? That with Java you can't do? It is running on many mobile phones for ages. It can be used to create great GUI applications. It is multiplatform. It can be compiled to native code and used in embedded devices. So no JVM required. It is used for web applications as well. It is used for business applications and multitier database applications. Some ERP like Compiere use it. It can be used even for multiplatform 3D game development for example Minecraft. As everything in this world Java is not perfect but Lazarus/FPC either. Somebody would say that you should use different language for different purpose like PHP, Perl, Python, Ruby for web, C for system programming, C# for Windows applications, Javascript for client-side and web programming, etc. I speak for myself. Just ask many programmers which left Delphi community on behalf of other languages. Just try to find good delphi programmer for daily job nowadays. Maybe something is wrong and change is needed. ---------------------------------------------------------------------- gulyone · 2013-03-10 13:13 Quote from: Chronos on March 10, 2013, 12:31:35 pmYes, definitely present way of developing FPC and Lazarus is going to wrong direction. And there are many reasons for that. We all know that core problem is that only small group of developers work on these projects only occasionally in their free time or maybe some in working time in their job. But really helping with Lazarus/FPC development partially in daily job means that you are loosing precious time which you could use to build great product which could make your company best choice on market. Then there will be always argue about use of professional fully featured payed IDE or use free but incomplete product. To defend Delphi/Pascal from use of other languages we are really in weak position. It is even worse in commercial and professional open source area then in hobby area. One big problem was drop in Delphi usage during "dark times" when Borland decides to get rid of Delphi. Many of us didn't know what to expect from future. If Delphi will be dead or live. Many left Delphi development during these years. And you could tell to somebody how great Pascal/Delphi when producer itself don't want their own child. Thanks god that Embarcadero gave Delphi new momentum and Delphi caught new breath. But maybe these XE1, XE2, XE3 versions are not what many expected from further Delphi development. Ok, customers will tell them what they are willing to pay and this is how things are going in commercial environment. Another big problem is missing free Delphi version for open source development. This is hurting Delphi a lot against other languages as Java. High price for professional edition isn't helping either. Even if Delphi would be best technology on entire world, pricing is set badly and it will not be widely used anytime in future for that. Free pascal is going own way trying to mimic Turbo Pascal and Delphi. But even it is free and even usable for building commercial product, something is still bad. Maybe it is concept itself, maybe it is lack of community, maybe it lack of active developers. But I don't think that more developers will solve problems. Community in forum is pretty active but it is not sufficient for success. Something is still missing. Just take some great products as example. Let's say LibreOffice. Very nice product, usable, documented, actively developed, supported, multiplatform, community is growing. What lays behind such success? Why is so vital? Is somebody pushing it all time? Is foundation helping? Is just interest of big companies as Red Hat, SUSE and others? Is it cool product with bright future? To stay in development area let's say Eclipse IDE. Just compare Lazarus and Eclipse and you will find many differences. Many have huge impact to usability and potential community grow. Again some foundation is stand behind. Concept is set as modular open source development platform that even commercial companies have interest and use Eclipse and contribute to the project. I love pascal but many of us use Lazarus/FPC just because there isn't better open source alternative. Then we can stick to incomplete messy and not so perfect Lazarus or go on and use other languages like C, C++, Java, C#, Python, Ruby, Perl, Javascript, e.g. I would be good to have one fully featured IDE and language for everything include system programming and embedded applications, GUI applications, database applications, web applications, smartphone apps, e.g. But so far non of to me known languages offer this. Say goodbye to pascal and say welcome other language... http://www.tiobe.com/index.php/content/paperinfo/tpci/index.html hmm on the other hand >> see what we are able to do with no money while the opponents have millions , and we are still standing for years straight in our boots ! now imagine , if we have the same money they have , ... to do this , we need 3 things 1) no more mimic 2) inovate and BE the target of them mimicing us 3) following these 2 1st steps , listen to sponsors and make them trust ---------------------------------------------------------------------- Chronos · 2013-03-10 13:15 Quote from: Leledumbo on March 10, 2013, 01:45:44 pmI develop kernel in Pascal, I develop console/GUI/web/dbcentric/mobile apps in Pascal. You can do anything with it as long as you know how to do it This could be told about nearly every languages. Even about assembler or machine core. You can do anything you can imagine in assembler. Or even better, you can write kernel in BrainFuck. And if you can't, you probably don't know language well enough. So this is poor argument. What you could compare is for example learning curve, availability of documentation, project coordination (roadmap, deadlines, feed-back, ...), platform support, result code speed, size of project which could be handled, development efficiency, generalization, openness, free or commercial support, stability, features, back and forward compatibility, attractiveness, compatibility with third party products, and many more... Kernel in pascal? Even I am aware that it would be possible to some extent I am not aware of any working and usable kernel written in pascal. There are only some unfinished hobby projects http://wiki.freepascal.org/Operating_Systems_written_in_FPC For the record, I appreciate your involvement in this project. Community needs more people like you. ---------------------------------------------------------------------- Chronos · 2013-03-10 13:28 Quote from: gulyone on March 10, 2013, 02:13:32 pmhmm on the other hand >> see what we are able to do with no money while the opponents have millions , and we are still standing for years straight in our boots ! Yes, slow and steady. Great effort was put to these projects. But maybe train is run away already. What would you answer to forum questions for Windows/Metro support? Or Qt5 or GTK3? That we are not late because others don't have it either? Or something like that anybody can do it and contribute with code? It will be ready then it will be ready? Well that it may not be enough to be general accepted IDE. How to make current and new potential community members happy. How to develop solid quality product with limited or no money? Do hobby open source project or general usable IDE for good commercial applications?  Should I give somebody half of my wage to be able to work on Lazarus/FPC development for full time and bring to us new desired features on time ? ---------------------------------------------------------------------- typo · 2013-03-10 13:44 Quote from: Chronos on March 10, 2013, 02:28:47 pmHow to make current and new potential community members happy? Giving them power. Quote from: Chronos on March 10, 2013, 02:28:47 pmHow to develop solid quality product with limited or no money? Making the software good for academic use. Quote from: Chronos on March 10, 2013, 02:28:47 pmIs hobby open source project or general usable IDE good for commercial applications?  It is good enough for small business. But bugs are the main problem. ---------------------------------------------------------------------- Leledumbo · 2013-03-10 13:44 QuoteThis could be told about nearly every languages.Not really, some languages do have limited area of applications. QuoteEven about assembler or machine core. You can do anything you can imagine in assemblerThis one is true. QuoteOr even better, you can write kernel in BrainFuckNo, you can't. Despite BrainFuck is turing complete, it can't be used to access specific register or call external code. QuoteAnd if you can't, you probably don't know language well enoughNot true, proven with BrainFuck above. QuoteKernel in pascal? Even I am aware that it would be possible to some extent I am not aware of any working and usable kernel written in pascalYes, I know. Because mostly are just hobby projects, and no serious projects that use it. Toro is an exception though, it's fairly complete and can fulfill its purpose (as a realtime kernel for full usage of CPU and memory). Building kernels from scratch for existing desktop platforms are quite uncommon, which is why it would never work. Today, focusing on kernel stability and adding features through kernel modules are much more appreciated. Learning to build or simply as a proof that Pascal is capable building it is good though, at least it opens the gate for everyone to open their eyes and jump to kernel development with it. In embedded world, the situation is a bit different. Pure Pascal is not uncommon, especially in microcontrollers. Both commercial and free/open source solutions are available and growing. ---------------------------------------------------------------------- gulyone · 2013-03-10 13:45 Quote from: Chronos on March 10, 2013, 02:28:47 pm How to make current and new potential community members happy. Well noone can evaluate if a whole community is happy or unhappy , the only thing i can say is that I feel u r very unhappy and sad, reading you. and I suppose noone but you is able to change this so...... as im not here for psy purpose in anyway , as im as passionate as my friend leledumbo from who i've read many pages of code , im gona simply tell you sumthing with all the possible respect : GOOOOOODDDDD BBYYYYYYYYEEEEEEEEEEEEEE ---------------------------------------------------------------------- Chronos · 2013-03-10 16:19 Quote from: gulyone on March 10, 2013, 02:45:47 pmQuote from: Chronos on March 10, 2013, 02:28:47 pm How to make current and new potential community members happy. Well noone can evaluate if a whole community is happy or unhappy , the only thing i can say is that I feel u r very unhappy and sad, reading you. and I suppose noone but you is able to change this so...... as im not here for psy purpose in anyway , as im as passionate as my friend leledumbo from who i've read many pages of code , im gona simply tell you sumthing with all the possible respect : GOOOOOODDDDD BBYYYYYYYYEEEEEEEEEEEEEE You should read all previous posts in this topic. I am not who is going away. Not now as I haven't in general better tool for my little hobby projects. And it is not about me but about entire community. For example try to build operating system with Pascal or C language. I promise you will see difference. Same is valid for modular ERP software or maybe Office suite or other big project. Where are big first class 3D games in pascal? Why non exists today? You can answer by yourself. As this is discussion forum I wrote my perspective in topic where one community member gave up fpc. You simply close your eyes or think about it. ---------------------------------------------------------------------- Chronos · 2013-03-10 16:35 Quote from: Leledumbo on March 10, 2013, 02:44:30 pmQuoteOr even better, you can write kernel in BrainFuckNo, you can't. Despite BrainFuck is turing complete, it can't be used to access specific register or call external code. QuoteAnd if you can't, you probably don't know language well enoughNot true, proven with BrainFuck above. This is matter of what you understand of operating system. In fact just because it is turing complete and it have input and output it is possible. Nobody done this yet. Same is valid for Free Pascal. Nobody build operating system in free pascal yet but this doesn't mean that it is impossible. It is simply hard and time consuming and it would need some additional support on compiler side. Quote from: Leledumbo on March 10, 2013, 02:44:30 pmBuilding kernels from scratch for existing desktop platforms are quite uncommon, which is why it would never work. Today, focusing on kernel stability and adding features through kernel modules are much more appreciated. Learning to build or simply as a proof that Pascal is capable building it is good though, at least it opens the gate for everyone to open their eyes and jump to kernel development with it. In embedded world, the situation is a bit different. Pure Pascal is not uncommon, especially in microcontrollers. Both commercial and free/open source solutions are available and growing. Yes using pascal in microcontrollers is possible but FPC as compiler is limited in such support. ARM is somehow supported but there are lots of other microcontrollers out there. Some are 16-bit some are 8-bit with limited memory. Doors are open. We are taking back to topic. Everything is possible. But we are still not flying in spaceships between solar systems. We have to evaluate what is available today. If anybody want to defend Lazarus/FPC by practical usability there should not be matter of emotions. Some comparison should be done. Maybe we could do that in wiki. To see where weak spots are. ---------------------------------------------------------------------- gulyone · 2013-03-10 18:20 Quote from: Chronos on March 10, 2013, 05:35:16 pmQuote from: Leledumbo on March 10, 2013, 02:44:30 pmQuoteOr even better, you can write kernel in BrainFuckNo, you can't. Despite BrainFuck is turing complete, it can't be used to access specific register or call external code. QuoteAnd if you can't, you probably don't know language well enoughNot true, proven with BrainFuck above. This is matter of what you understand of operating system. In fact just because it is turing complete and it have input and output it is possible. Nobody done this yet. Same is valid for Free Pascal. Nobody build operating system in free pascal yet but this doesn't mean that it is impossible. It is simply hard and time consuming and it would need some additional support on compiler side. Quote from: Leledumbo on March 10, 2013, 02:44:30 pmBuilding kernels from scratch for existing desktop platforms are quite uncommon, which is why it would never work. Today, focusing on kernel stability and adding features through kernel modules are much more appreciated. Learning to build or simply as a proof that Pascal is capable building it is good though, at least it opens the gate for everyone to open their eyes and jump to kernel development with it. In embedded world, the situation is a bit different. Pure Pascal is not uncommon, especially in microcontrollers. Both commercial and free/open source solutions are available and growing. Yes using pascal in microcontrollers is possible but FPC as compiler is limited in such support. ARM is somehow supported but there are lots of other microcontrollers out there. Some are 16-bit some are 8-bit with limited memory. Doors are open. We are taking back to topic. Everything is possible. But we are still not flying in spaceships between solar systems. We have to evaluate what is available today. If anybody want to defend Lazarus/FPC by practical usability there should not be matter of emotions. Some comparison should be done. Maybe we could do that in wiki. To see where weak spots are. can you tell me who built an operating system with gnu C or with Qt ? i cant wait !!!l furthermore , u should have start another thread / post , because im confused and i dont know what is doing ur blah blah on a 'good bye' message. However , i can tell you sumthing , i lived these years when Pascal (and not only ONE compiler ) was in use almost everywhere .. AND YES , Professor Wirth didnt stop his Activities on Pascal , 'H'e also invented Modula , Modula 2, Oberon , wich are Pascal cousins and very similar on many aspects ... Finally Oberon 2 became the OS !!! of the NASA.. so you are a person who doesn't know what he is talking about on a post where that is not the subject COOL ! go ahead lmao its sunday , ive nothing much to do and you made my day , definitely haha ---------------------------------------------------------------------- marcov · 2013-03-10 19:04 Quote from: Chronos on March 10, 2013, 02:28:47 pmWhat would you answer to forum questions for Windows/Metro support? Or Qt5 or GTK3? That we are not late because others don't have it either? What alternate course do you propose that does amount to forever running after the latest and greatest, and never quite reaching it? Quote Or something like that anybody can do it and contribute with code? It will be ready then it will be ready? Well that it may not be enough to be general accepted IDE. I don't think that that can be done under 100k/annum. QuoteHow to make current and new potential community members happy. How to develop solid quality product with limited or no money? Do hobby open source project or general usable IDE for good commercial applications?  Lazarus is still somewhat compatible to Delphi. That is more than a whole lot of OSS IDEs can say. QuoteShould I give somebody half of my wage to be able to work on Lazarus/FPC development for full time and bring to us new desired features on time ? Of course. ---------------------------------------------------------------------- Chronos · 2013-03-10 21:42 Quote from: gulyone on March 10, 2013, 07:20:37 pmfurthermore , u should have start another thread / post , because im confused and i dont know what is doing ur blah blah on a 'good bye' message. However , i can tell you sumthing , i lived these years when Pascal (and not only ONE compiler ) was in use almost everywhere .. AND YES , Professor Wirth didnt stop his Activities on Pascal , 'H'e also invented Modula , Modula 2, Oberon , wich are Pascal cousins and very similar on many aspects ... Finally Oberon 2 became the OS !!! of the NASA.. so you are a person who doesn't know what he is talking about on a post where that is not the subject COOL ! go ahead lmao its sunday , ive nothing much to do and you made my day , definitely haha You should read first post of this topic and even more please read article on writer blog with his view. He start this topic for some reason which you are probably missing or simply don't want to hear. It's your choice. I don't want to start another topic for similar issue because I am member of this forum for some years and I am aware of situation around this project and I know what answers I could expect from core team. That's pity that another community member is leaving and saying goodbye to us and even worst to say goodbye to pascal itself. As I already wrote here. We all know that he is not first either last. You can think about it and suggest some actions or live on and thing that community is small just because they don't know how the project good is. Some people are quite happy with current status, other have bigger expectations. ---------------------------------------------------------------------- gulyone · 2013-03-11 00:08 Quote from: Chronos on March 10, 2013, 10:42:27 pmYou should read first post of this topic and even more please read article on writer blog with his view. He start this topic for some reason which you are probably missing or simply don't want to hear. It's your choice. hmmm NooooOOOooo ! I don't want to ! because ima Pascal Integrist with a long beard and a logical bomb in my pocket ! ---------------------------------------------------------------------- Chronos · 2013-03-11 18:16 Quote from: marcov on March 10, 2013, 08:04:55 pmWhat alternate course do you propose that does amount to forever running after the latest and greatest, and never quite reaching it?Implementation can be done in limited time. No need to have never finished component and libraries. Yes, money could solve this. To hire some external programmers maybe freelancers. Large active community could do that as well, but many are giving up and you can't dictate community size. Quote from: marcov on March 10, 2013, 08:04:55 pmI don't think that that can be done under 100k/annum. Ok, then what about try to collect such money? If main advantage of free & open source is that you don't need to pay for something which is already done, then this doesn't mean that we are not willing to pay for features what are not here yet. What about to establish foundation which would cover development and collect money and spend them for greater good? Something like fundrising if money are needed? Feature will be implemented, users will be happy and it better position against other language evengelists. Examples exists, Eclipse foundation to name one. It is about priorities of core team. Quote from: marcov on March 10, 2013, 08:04:55 pmLazarus is still somewhat compatible to Delphi. That is more than a whole lot of OSS IDEs can say. There are always some compatibility is open source projects which imitate some commercial one. But it is also matter of opinion if it should go own way or try to be semi-compatible. I could say that it is basically plus for newcomers but not so good for advanced coders who have detailed knowledge of code and saw design and implementation flaws easily. For example trying to make general components to handle dockable forms is nightmare and we still don't have any really usable. Quote from: marcov on March 10, 2013, 08:04:55 pmQuoteShould I give somebody half of my wage to be able to work on Lazarus/FPC development for full time and bring to us new desired features on time ? Of course. Ok, who are willing to work fulltime on the project instead of his current job? How much money is enough to work daily as new job? Such person have to be already part of core team or could join the team? Is there something like roadmap? I only found http://www.freepascal.org/future.var and http://wiki.lazarus.freepascal.org/Roadmap But how decisions are made. Original author of this topic said something about future of project. We all have own opinions to various parts of the project. QuoteThe problem is that I'm no longer believe in the path of the language, you can call it Delphi, you can call it Object Pascal, it does not matter. I think it goes the wrong way, and instead of able to talk about it with fellow developers, there is constantly arguments, and bickering about it.Maybe there should more information about decisions and future needs. Current situation is that few developers are implementing what they need or what they think is useful. There isn't any feed back from community. Model as "Pay no money => then shut up and keep going." I have many ideas what should be done but I don't know what is allowed and would be accepted to be done. I am not talking about implementing some components which overall effect will be insignificant. I am talking about concept and platform itself. So maybe it can't be helped. ---------------------------------------------------------------------- marcov · 2013-03-11 22:21 Quote from: Chronos on March 11, 2013, 07:16:48 pmQuote from: marcov on March 10, 2013, 08:04:55 pmWhat alternate course do you propose that does amount to forever running after the latest and greatest, and never quite reaching it?Implementation can be done in limited time. You'd be surprised. You can do something proof of concept quite quickly, but something supportable and end-user ready is something totally else. QuoteNo need to have never finished component and libraries. Yes, money could solve this. To hire some external programmers maybe freelancers. No, even money does not solve that. Even with a beginning of money, you can pursue a handful more options. Not all of them. It starts with making choices. Then following these choices, don't waver, and FINISH them. If it takes one or ten year, popular or not. QuoteLarge active community could do that as well, but many are giving up and you can't dictate community size. Actually for a community driven project, the FPC/Lazarus project is quite large, specially if you think that most are actually coders, not merely packagers (as with e.g. Linux distros). QuoteQuote from: marcov on March 10, 2013, 08:04:55 pmI don't think that that can be done under 100k/annum. Ok, then what about try to collect such money? There is nothing wrong with that. But of course any fund raising is a cost/benefit matter, and realism crashes in. Currently, most chance on commercial sponsoring seems to  be situated in Core members getting brief contracts (typically say, half an year) for specific bits. This has already happened several times. While many open source projects have foundations, they mostly had sponsors BEFORE the foundations, and set up the foundations for tax purposes, and when the number of sponsors increased. QuoteIf main advantage of free & open source is that you don't need to pay for something which is already done, If you believe that, I suggest reading what RMS has written on the subject. I might not agree with all of it, but there is a big difference between a freeloader and free software. Quotethen this doesn't mean that we are not willing to pay for features what are not here yet. What about to establish foundation which would cover development and collect money and spend them for greater good? It would be great. And a miracle. QuoteSomething like fundrising if money are needed? Feature will be implemented, users will be happy and it better position against other language evengelists. Examples exists, Eclipse foundation to name one. It is about priorities of core team. Eclipse had sponsors before the money. Among others IBM plediging a Billion to open source. Guarantee me a billion, and I personally set up a foundation. Hell, I will walk on my hands to the chamber of commerce to set it up :_) QuoteQuote from: marcov on March 10, 2013, 08:04:55 pmLazarus is still somewhat compatible to Delphi. That is more than a whole lot of OSS IDEs can say. There are always some compatibility is open source projects which imitate some commercial one. True. But the level of compatibility (not just language, but libraries too!) is rare. QuoteBut it is also matter of opinion if it should go own way or try to be semi-compatible. I could say that it is basically plus for newcomers but not so good for advanced coders who have detailed knowledge of code and saw design and implementation flaws easily. For example trying to make general components to handle dockable forms is nightmare and we still don't have any really usable. It's actually most important for the professional users. They have large codebases, favour stability over features, and the tie-in with Delphi's commercial component market is important. QuoteQuote from: marcov on March 10, 2013, 08:04:55 pmQuoteShould I give somebody half of my wage to be able to work on Lazarus/FPC development for full time and bring to us new desired features on time ? Of course. Ok, who are willing to work fulltime on the project instead of his current job? Some of the contributors (not just core) are self employed, and could be persuaded to take projects.  I think that is now more realistic than having really fulltimers in fixed jobpositions. QuoteHow much money is enough to work daily as new job? Such person have to be already part of core team or could join the team? I don't know much about wages in the rest of the world, but 70-100k per annum is normal for schooled IT in Western Europe. That is employer costs btw, iow bruto wages, and includes taxes that fall on the employer. As for core, core only matters if you want to have it in the standard distribution. If it conforms with the various core's ideas (Lazarus and FPC have distinct but somewhat interwoven cores) where the project is heading. As in the first half of the message, focus is simply to important. Core is the guardian of that focus, to avoid the long term goals of the project be abandoned for short term external commercial gains and interests. QuoteIs there something like roadmap? I only found http://www.freepascal.org/future.var and http://wiki.lazarus.freepascal.org/Roadmap The roadmap on those pages is mostly from the 2.2.2, early 2.4 series. A lot can be discussed. QuoteBut how decisions are made. That's a lot less central than a lot of people think. Many forum people think that core is mostly busy with plotting evil courses in secret maillinglists. The contrary is true. On the internal maillists global direction is hardly ever discussed, and most is just an internal communication over day-to-day business (and even then, in FPC's case very compiler centric, much less libraries, 90%+ of the messages are about compiler internals). Only during moments that the decision is made to work towards branching trunk to a new stable branch, course is somewhat discussesd The reason for that is fairly simple. Once you get long and deep enough into the project, and carry some responsibility, there are not that many realistic, independent courses, so naturally the noses of core mostly already point into the same direction. Shared upbringing more or less. This is also visible in newer core members (of say the last handful of years). In short, being on core ruins your naivety about what is doable. That can be disillusioning though . :-) QuoteOriginal author of this topic said something about future of project. We all have own opinions to various parts of the project. Yes. But that is mostly based on forum banter, and not really related to the future of the project as seen by the implementers. And it might sound arrogant, but in the end, that is mostly what counts. QuoteQuoteThe problem is that I'm no longer believe in the path of the language, you can call it Delphi, you can call it Object Pascal, it does not matter. I think it goes the wrong way, and instead of able to talk about it with fellow developers, there is constantly arguments, and bickering about it. There is some frustration in your messages though. I had you pegged down as a Java programmer forced to work on Delphi (and maybe a port to FPC) though. QuoteMaybe there should more information about decisions and future needs. It is very hard to make a clear statement of direction if there every contribution is volunteer driven. Most core members can't say how much time they will be able to spend in the next 1,2 year, and if they will do simply maintenance or major new functionality. QuoteCurrent situation is that few developers are implementing what they need or what they think is useful. There isn't any feed back from community. Model as "Pay no money => then shut up and keep going." Actually, community input on the macro scale is rarely realistic. The easiest way to have influence on the project is to become a committer. Once you have managed a part of the project for a year or more, provided bugreports (WITH patches), you'll get a commit bit, and usually after a while that is upscaled from local (your region of interest) to global. (commit everything) Most useful community input comes in on detail or intermediate level. And the intermediate submitters of today are the committers of tomorrow if they persist. But you have to show you can carry responsibility first. There is no sane decision making without carrying responsibility. Even waving with money is hard, since as you already said it probably won't enough to sustain a major developer for years to come. Which means that the people TAKING the money also have to consider it worthwhile, and not just from a money perspective. They have to factor in the risk of leaving paid employment (or existing customers, for the self employed) etc etc. IOW offer money for something unrealistic, and there will be no takers. QuoteI have many ideas what should be done but I don't know what is allowed and would be accepted to be done. I am not talking about implementing some components which overall effect will be insignificant. I am talking about concept and platform itself. So maybe it can't be helped. Start working. The best way is to say it with code. Dreams are a dime a dozen, only code is worth something. ---------------------------------------------------------------------- gulyone · 2013-03-12 05:06 Chronos -10 Marcov +10 000 still thinking that whole discussion needed a separate 'D'edicated post , because It Is an important deep subject but has nothing todo on a 'goodbye' spot ---------------------------------------------------------------------- mas steindorff · 2013-03-12 18:17 Quote from: gulyone on March 12, 2013, 06:06:06 amstill thinking that whole discussion needed a separate 'D'edicated post , because It Is an important deep subject but has nothing todo on a 'goodbye' spot and I think you have all forgoten about the fish