Page 1 of 1
Lazarus v0.9.30.4 + ZEOSDBO-6.6.6-stable not work
Posted: 13.04.2012, 15:28
by nenodias
When I try install zcomponent.lpk
Have a error on the \src\core\ZClasses.pas...
below my message error
Code: Select all
D:\Programas\Zeos Simone\ZEOSDBO-6.6.6-stable\src\core\ZClasses.pas(84,22) Error: No matching implementation for interface method "IUnknown.QueryInterface(constref TGuid,out <Formal type>):LongInt; StdCall;" found
D:\Programas\Zeos Simone\ZEOSDBO-6.6.6-stable\src\core\ZClasses.pas(183,14) Warning: An inherited method is hidden by "TZAbstractObject.Equals(const IUnknown):Boolean;"
D:\Programas\Zeos Simone\ZEOSDBO-6.6.6-stable\src\core\ZClasses.pas(186,14) Warning: An inherited method is hidden by "TZAbstractObject.ToString:AnsiString;"
D:\Programas\Zeos Simone\ZEOSDBO-6.6.6-stable\src\core\ZClasses.pas(190,1) Fatal: There were 1 errors compiling module, stopping
Any suggestion for this install...
I discovery the problem...
The Lazarus 0.9.30.4 use the FPC 2.6, probably ZEOSDBO not have suport.Only one update would solve my case
They need UPDATE the zeosdbo-6.6.6-stable for the LAZARUS compatible with FPC 2.6 (Free Pascal Compiler 2.6) for that the zeosdbo-6.6.6-stable work on version Lazarus
Posted: 16.04.2012, 09:39
by MYSQL889
Hi!
Lazarus forum, there was an open issue about this ...
http://www.lazarus.freepascal.org/index ... l#msg90552
Posted: 16.04.2012, 20:51
by EgonHugeist
Zeos7 is ready for FPC 2.6+ and stable for FPC. The only missing thing for the Beta/Stable state is to fullfill the Full Unicode IDE support for Delphi12_UP.
I don't know anything about the 6.6.6. Component. I must admit.
Michael
Posted: 16.04.2012, 22:10
by MYSQL889
Hi EgonHugeist,
This obvious, that the future is Zeos 7
. Take this opportunity to thank you for the great job lately with Zeos .
Courage and go!
Posted: 17.04.2012, 09:02
by Stilgar
Sorry, but for Lazarus (64bit) is ready too Zeos7?
Stilgar
Posted: 17.04.2012, 17:26
by EgonHugeist
MYSQL889,
thank for acknowledgement of the work i've done for the comunity. Hehe and i get some things back too: It's an good training for my untrained school english instead of donated money. And it's a question of hornor for me.
I'm hopefull my patches will change the state. But it's a completely changed design of the string handling internals. Thought, it's possible that the FPC also goes the UnicodeString way. So Zeos must be prepared for such possibilities. So what's comming or not the Forum shows us. Some criticts hint's or positive/negative news are always welcome (On my thread for example).
Stilgar,
what can i say Zeos7 is ready and 6.6.6 i don't know. But i must admit the Win64 support is actually only available in the \testing(since yesterday night r1160) and \testing-egonhugeist(since the last weekend r1158). For *nix systems it should be possible since last winter (patch of papelhiego as far as i know).
Michael
Posted: 17.04.2012, 19:34
by MYSQL889
EgonHugeist, Mdaems,........
All the latest work your are doing, is very positive for the project ZeosLib.
Thank you all!
Posted: 18.04.2012, 02:11
by Stilgar
Thanks for replay.
I wait the stable version for Lazarus (Win64).
Stilgar
Posted: 18.04.2012, 18:58
by EgonHugeist
That's what we've to find out! And only users can bring us suitable results...
Posted: 15.05.2012, 13:56
by marcov
Note that this continues to pop up regularly in Lazarus forums and maillists.
Is there a chance to respin 6.6.6 with all FPC specific ifdefs removed from that file (they shouldn't be necessary anymore for at least 3 releases).?
The problem is in Zeos in that it has outdated own definitions of aggregated and contained objects that were added to FPC years ago under a not version dependent IFDEF.
Either the IFDEF should be updated to exclude recent FPCs (anything 2.4.0 and later), or removed altogether (which IMHO is better)
Posted: 16.05.2012, 12:41
by EgonHugeist
Hmm that should be possible, Marco. What is here the download location for the old 6.6:6 components? 6.6.6-patches?
And the next issues i see. Fix is the x64 support, bad updatebehavior if no key is available, Null-Timestamps, QuotedIdentifiers, deprecated MetaInformations and so on.. We've fixed a really veery long list of reported bugs in Z7..
That means for us: do all jobs three times.
Why don't they realize that the 6 series are more buggy then the 7alpha concerning FPC or AnsiDelphi. Just run our testsuites..
Posted: 16.05.2012, 19:41
by marcov
The whole point of that is that it is hard to explain on the Lazarus forum, if at the same time the Zeos site still advocates 6.6.6 for download as "stable".
Some strong warnings should be added.
Posted: 25.05.2012, 21:49
by EgonHugeist
marcov,
what is a general x64 define? I'm not willing to fix everything on this deprecated component. But you right with fixing this object. But what about the TBookMark issue?
Do you mean warnings for FPC2.6.0 and x64 directly on LazarusZeos.inc?
Michael
Posted: 29.05.2012, 19:33
by marcov
ifdef cpu64 (all 64-bit CPUs) or ifdef x86_64, but I'm not aware of a specific 64-bit issues.
Afaik there is no tbookmark issue, 2.6.0 and 2.6.2 will be mostly as 2.4.x, and 2.7.x will be like D2009+ which you already support, but nobody can complain that stable zeos doesn't work with unstable FPC. IIRC I already adapted zeoslib/testing for that.
This thread is mostly about zclasses defining own aggregateobject and containedobject variants that have been in FPC since years of lore. To my best knowledge, it is not just unnecessary with 2.6.0, but also for the entire 2.4.0 series and probably even 2.2.4 (it just doesn't error with those)
The 2.6.0 change that "const" has been splitted into "const" and "constref" in 2.6.0+.
This is easily fixed in zclasses. Just remove all FPC specific code relating to Aggregateobject from ZClasses.