On Fri, 24 Apr 2009, Ron Pinkas wrote: > I've learned to accept that Harbour developers reserve the rights to rename > function names they borrow, to "rewrite from scratch" borrowed ideas and > extensions, to introduce borrowed yet modified syntax, all with complete > disregard to API or user code compatibility. Obviously the claim is that > the reasons are always noble, well justified, and have nothing to do with > xHarbour.
Can you precisely define what you are talking about? I was talking things like xbase++ multivalue macros which you started in Harbour then continued your work on then in xHarbour and you never finished ot so they are still broken after over 6 years of xHarbour life. But in Harbour they are implemented in absolutely different way and they are working. I do not agree that someone will discard my work saying that it comes from something what never worked correctly and cannot work well due to fatal design decision. I had to create it from scratch because alternative was dropping completely this idea to not keep broken and not working code. I was talking about simplex which were and still is broken because it's wrongly designed from the beginning what Ryszard shows before xHarbour started. In Harbour there is no even single line from this idea and the same will have to happen in xHarbour if someone will want to fix it. I was talking about FOR EACH, WITH OBJECT implementation which does not come from xHarbour at all. The Harbour version does not have even single line from xHarbour code and the names does not comes from xHarbour too. The only one thing which comes directly from xHarbour is : inside WITH OBJECT what causes some serious problem to other code and PP rules and we have to leave with them :-( I was talking about TRY / CATCH which still does not work correctly, etc. BTW Your idea to use ALWAYS clause inside BEGIN SEQUENCE was great for me and I implemented it in Harbour. In Harbour all of the above were implemented in different way which does not cause side effects and runtime overheads. The results you can see in speedtst.prg - Harbour which now supports more features in which some "xHarbour extensions: works but do not work in xHarbour is also nearly twice faster in pure PCODE evaluation then xHarbour. And I'm talking about ST mode. In the MT mode the difference is much bigger but the speed is the less important problem of xHarbour MT mode ... > As far as I know, xHarbour does not include any idea borrowed from Harbour, Not true. A lot of my code I 1-st committed to Harbour and later was ported to xHarbour. PP, -gc3, many modifications in compiler and HVM, RDDs. Later a lot of my code was borrowed from Harbour but it's the fact that Miguel does not use to inform about the source of his commits, f.e. if we are talking about "xHarbour ideas" some basic fixes for FOR EACH / WITH OBJECT / TRY CATCH in compiler like: 2007-08-22 11:45 UTC+0100 Miguel Angel Marchuet <miguelangel/at/marchuet.net> 2007-08-22 12:12 UTC+0100 Miguel Angel Marchuet <miguelangel/at/marchuet.net> were committed without any information about their source. Do you see any information that the solution is copied from Harbour? Or: 2007-10-31 10:00 UTC+0100 Miguel Angel Marchuet <miguelangel/at/marchuet.net> Do you think the the line Synced with harbour as possible. is enough in this case? In this case 99% of the commit is my code copied from Harbour and there is no even single note about it. And then you are saying that you do not know anything about code borrowed from Harbour. You also do not like to refer to the source of some of your modifications, f.e.: 2009-03-13 23:58 UTC-0430 Ron Pinkas <ron.pinkas/at/xharbour.com> it's my Harbour solution: 2006-08-19 01:10 UTC+0200 Przemyslaw Czerpak (druzus/at/priv.onet.pl) adopted to xHarbour source code and it was in few days after my message to this list where I informed you what is wrong and that you should refer to Harbour source code for proper solution. Here we have yet another funny thing. In this modifications you used extension which also comes from Harbour source code but probably where wrongly ported and in the past you modified. Of course everything without information from where it comes and what it does. But at leas in your last commit you haven't copied Harbour source code as is. Unlike FOR EACH / WITH OBJECT etc. which are not patented xHarbour names and such construction were added to xHarbour I'm talking about direct coping of source code or algorithms without even single credit note to original authors or projects. I understand that sometime is hard to check who exactly committed sth so I understand that it's hard to refer original author. But I do not understand that it's possible to copy ~250 lines of my Changelog entry removing my name from it. I do not talk about your or mine imaginations or feelings but only about the facts. > that was implemented in an intentionally incompatible manner. I hope that > Harbour developers will decide to compromise their infinite need to improve > on ideas borrowed from xHarbour, at least when such improvement breaks > compatibility. I also hope that Harbour developers will understand the > value of a unified user community, even if development efforts are not > unified. It was my goal in the past to keep support for both compilers and I was making it for really long time but in current days IMHO it's not worth to longer invest time in xHarbour. It does not offer any interesting functionality, it's much slower, some important things does not work at all and it's very hard to force some deeper modifications by main door. > Please, let's move on to better things. OK. I also end this subject. I only hope that Miguel with update his ChangeLog entries from last years. best regards, Przemek ------------------------------------------------------------------------------ Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects _______________________________________________ xHarbour-developers mailing list xHarbour-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xharbour-developers