From tomas.hlavaty at knowledgetools.de Thu Mar 4 12:21:35 2010 From: tomas.hlavaty at knowledgetools.de (Tomas Hlavaty) Date: Thu, 04 Mar 2010 13:21:35 +0100 Subject: [Cl-perec-devel] allegro port In-Reply-To: (Attila Lendvai's message of "Mon, 1 Feb 2010 11:42:32 +0100") References: <86sk9zsdc9.fsf@knowledgetools.de> <86fx5zs95j.fsf@knowledgetools.de> <86vdeuqoak.fsf@knowledgetools.de> Message-ID: <86zl2ogtv4.fsf@knowledgetools.de> Hi Attila, > ok, i've done the darcs2 conversion. the following shell fragment may thank you. I have updated the new version with the portability patches. Please see http://www.knowledgetools.de/tmp/temporaer/tomas/ for more details. Regards, Tomas From tomas.hlavaty at knowledgetools.de Thu Mar 4 13:42:57 2010 From: tomas.hlavaty at knowledgetools.de (Tomas Hlavaty) Date: Thu, 04 Mar 2010 14:42:57 +0100 Subject: [Cl-perec-devel] allegro port In-Reply-To: ("Levente =?utf-8?B?TcOpc3rDoXJvcyIncw==?= message of "Fri, 29 Jan 2010 17:04:16 +0100") References: <86sk9zsdc9.fsf@knowledgetools.de> <86fx5zs95j.fsf@knowledgetools.de> <86vdeuqoak.fsf@knowledgetools.de> <86sk9pcfhe.fsf@knowledgetools.de> <86zl3wc3im.fsf@knowledgetools.de> Message-ID: <86vddcgq3i.fsf@knowledgetools.de> Hi Levente, >> I guess you mean calling it persistent-slot-definition-class? > Yes. renamed. Thanks, Tomas From attila.lendvai at gmail.com Thu Mar 4 14:08:05 2010 From: attila.lendvai at gmail.com (Attila Lendvai) Date: Thu, 4 Mar 2010 15:08:05 +0100 Subject: [Cl-perec-devel] allegro port In-Reply-To: <86zl2ogtv4.fsf@knowledgetools.de> References: <86sk9zsdc9.fsf@knowledgetools.de> <86fx5zs95j.fsf@knowledgetools.de> <86vdeuqoak.fsf@knowledgetools.de> <86zl2ogtv4.fsf@knowledgetools.de> Message-ID: hello Tomas, > I have updated the new version with the portability patches. ?Please see > http://www.knowledgetools.de/tmp/temporaer/tomas/ for more details. i've pushed most of your changes to our repos (HEAD for now) with a few fixes here-and-there... the rest waits for Levy to review when he gets to it. note: judging from the patches, the type handling in perec is causing headaches on allegro... it is on our TODO to refactor it to use the type stuff from hu.dwim.util. wui already uses it for the component factories... thanks for your efforts! -- attila ps: kami (on channel #dwim.hu on irc.freenode.net) is working on oracle support... sync up with him if needed... From tomas.hlavaty at knowledgetools.de Tue Mar 9 10:55:22 2010 From: tomas.hlavaty at knowledgetools.de (Tomas Hlavaty) Date: Tue, 09 Mar 2010 11:55:22 +0100 Subject: [Cl-perec-devel] allegro port In-Reply-To: (Attila Lendvai's message of "Thu, 4 Mar 2010 15:08:05 +0100") References: <86sk9zsdc9.fsf@knowledgetools.de> <86fx5zs95j.fsf@knowledgetools.de> <86vdeuqoak.fsf@knowledgetools.de> <86zl2ogtv4.fsf@knowledgetools.de> Message-ID: <866355zrvp.fsf@knowledgetools.de> Hi Attila, > i've pushed most of your changes to our repos (HEAD for now) with a > few fixes here-and-there... the rest waits for Levy to review when he > gets to it. thanks a lot! > note: judging from the patches, the type handling in perec is causing > headaches on allegro... it is on our TODO to refactor it to use the > type stuff from hu.dwim.util. wui already uses it for the component > factories... Yes, there are some tricky bits around subtypep but it seems to be more or less working or allegro. Do you have any plans for a Clozure port? I have briefly looked into it and hit a problem with +unbound-slot-value+ in hu.dwim.computed-class. Basically, perec uses a trick which is not portable (works on sbcl and allegro but not clozure and clisp for example). Not sure how much work would this be to make this portable? BTW, I have commited more patches and the Allegro port should now be working with the new hu.dwim.perec. Cheers, Tomas From tomas.hlavaty at knowledgetools.de Tue Mar 9 11:09:10 2010 From: tomas.hlavaty at knowledgetools.de (Tomas Hlavaty) Date: Tue, 09 Mar 2010 12:09:10 +0100 Subject: [Cl-perec-devel] QA Message-ID: <86sk89yco9.fsf@knowledgetools.de> Hi all, I would like to ask what is your policy regarding quality of the live repository? There are quite a few tests failing there so I wonder what is the difference between live and head and also how do I know that the code is stable enough for merging it into my own repository for further work? Thank you, Tomas From attila.lendvai at gmail.com Tue Mar 9 11:36:06 2010 From: attila.lendvai at gmail.com (Attila Lendvai) Date: Tue, 9 Mar 2010 12:36:06 +0100 Subject: [Cl-perec-devel] QA In-Reply-To: <86sk89yco9.fsf@knowledgetools.de> References: <86sk89yco9.fsf@knowledgetools.de> Message-ID: > I would like to ask what is your policy regarding quality of the live > repository? ?There are quite a few tests failing there so I wonder what > is the difference between live and head and also how do I know that the > code is stable enough for merging it into my own repository for further > work? the integrated libraries in live are in a state that compile, load and work fine together (which means that we don't know about issues that are the result of integration headaches). no more, no less. dwim.hu runs from the live repos. as of failing tests: they are part of the bug database. i try to put with-expected-failures everywhere where i see expected failures, although it's not always easy due to macrology and i don't run all our tests regularly (e.g. perec). so, we try not to have failing tests, but we are not too nervous if there are some. and we definitely prefer failing tests to unexposed bugs.... -- attila From levente.meszaros at gmail.com Tue Mar 9 13:47:21 2010 From: levente.meszaros at gmail.com (=?ISO-8859-1?Q?Levente_M=E9sz=E1ros?=) Date: Tue, 9 Mar 2010 14:47:21 +0100 Subject: [Cl-perec-devel] allegro port In-Reply-To: <866355zrvp.fsf@knowledgetools.de> References: <86sk9zsdc9.fsf@knowledgetools.de> <86fx5zs95j.fsf@knowledgetools.de> <86vdeuqoak.fsf@knowledgetools.de> <86zl2ogtv4.fsf@knowledgetools.de> <866355zrvp.fsf@knowledgetools.de> Message-ID: On Tue, Mar 9, 2010 at 11:55 AM, Tomas Hlavaty wrote: > Do you have any plans for a Clozure port? ?I have briefly looked into it We are not going to do a Clozure port, nor any other port. On the other hand we would be happy to more patches! > Basically, perec uses a trick which is not portable (works on sbcl and > allegro but not clozure and clisp for example). ?Not sure how much work > would this be to make this portable? In what sense is it not portable? > BTW, I have commited more patches and the Allegro port should now be > working with the new hu.dwim.perec. Thank you very much! Eventually they will be merged in, but I'm fixing the test suite first. Cheers, levy -- There's no perfectoin From levente.meszaros at gmail.com Tue Mar 9 13:49:38 2010 From: levente.meszaros at gmail.com (=?ISO-8859-1?Q?Levente_M=E9sz=E1ros?=) Date: Tue, 9 Mar 2010 14:49:38 +0100 Subject: [Cl-perec-devel] QA In-Reply-To: <86sk89yco9.fsf@knowledgetools.de> References: <86sk89yco9.fsf@knowledgetools.de> Message-ID: On Tue, Mar 9, 2010 at 12:09 PM, Tomas Hlavaty wrote: > I would like to ask what is your policy regarding quality of the live > repository? ?There are quite a few tests failing there so I wonder what > is the difference between live and head and also how do I know that the > code is stable enough for merging it into my own repository for further > work? In addition to what Attila said: I will finish the automated nightly tests, so that it runs for all our systems on dwim.hu. This supposed to test head and live separately and you will be able to compare the results. Hope this helps, levy -- There's no perfectoin From tomas.hlavaty at knowledgetools.de Tue Mar 9 14:38:13 2010 From: tomas.hlavaty at knowledgetools.de (Tomas Hlavaty) Date: Tue, 09 Mar 2010 15:38:13 +0100 Subject: [Cl-perec-devel] allegro port In-Reply-To: ("Levente =?utf-8?B?TcOpc3rDoXJvcyIncw==?= message of "Tue, 9 Mar 2010 14:47:21 +0100") References: <86sk9zsdc9.fsf@knowledgetools.de> <86fx5zs95j.fsf@knowledgetools.de> <86vdeuqoak.fsf@knowledgetools.de> <86zl2ogtv4.fsf@knowledgetools.de> <866355zrvp.fsf@knowledgetools.de> Message-ID: <861vfty2zu.fsf@knowledgetools.de> Hi Levy, >> Basically, perec uses a trick which is not portable (works on sbcl and >> allegro but not clozure and clisp for example). ?Not sure how much work >> would this be to make this portable? > In what sense is it not portable? I thing the problem is that +unbound-slot-value+ cannot be (portably) used as a "value", e.g. in hash-tables, see the attached test case. It seems to have special meaning in some implementations. >> BTW, I have commited more patches and the Allegro port should now be >> working with the new hu.dwim.perec. > Thank you very much! Once more patch, now it should work:-) > Eventually they will be merged in, but I'm fixing the test suite first. No problem & thanks. Tomas -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: standard-instance-access-unbound.lisp URL: From attila.lendvai at gmail.com Tue Mar 9 15:17:35 2010 From: attila.lendvai at gmail.com (Attila Lendvai) Date: Tue, 9 Mar 2010 16:17:35 +0100 Subject: [Cl-perec-devel] allegro port In-Reply-To: <861vfty2zu.fsf@knowledgetools.de> References: <86sk9zsdc9.fsf@knowledgetools.de> <86vdeuqoak.fsf@knowledgetools.de> <86zl2ogtv4.fsf@knowledgetools.de> <866355zrvp.fsf@knowledgetools.de> <861vfty2zu.fsf@knowledgetools.de> Message-ID: > I thing the problem is that +unbound-slot-value+ cannot be (portably) > used as a "value", e.g. in hash-tables, see the attached test case. ?It > seems to have special meaning in some implementations. fix pushed. -- attila From tomas.hlavaty at knowledgetools.de Wed Mar 17 12:10:46 2010 From: tomas.hlavaty at knowledgetools.de (Tomas Hlavaty) Date: Wed, 17 Mar 2010 13:10:46 +0100 Subject: [Cl-perec-devel] oracle port Message-ID: <86ocinkv21.fsf@knowledgetools.de> Hi Tomi, I am starting to work on porting hu.dwim.rdbms to oracle (and hu.dwim.perec later) and I would like to ask what state was the oracle backend last time it was maintained? What version of oracle database did you use? Thank you, Tomas From tomi.borbely at gmail.com Wed Mar 17 13:10:21 2010 From: tomi.borbely at gmail.com (=?ISO-8859-1?Q?Borb=E9ly_Tam=E1s_?=) Date: Wed, 17 Mar 2010 14:10:21 +0100 Subject: [Cl-perec-devel] oracle port In-Reply-To: <86ocinkv21.fsf@knowledgetools.de> References: <86ocinkv21.fsf@knowledgetools.de> Message-ID: <53c1332f1003170610g2233f39bi9c39c45f69ea1996@mail.gmail.com> Hi Tomas, We used the 'Oracle 10g Express Edition' as I remember. The rdms port is almost complete (~90%). It means that SQL commands/transactions should work. The problems we have are related to the mapping between cl-rdbms types and the oracle sql types. The cl-rdbms types are closer to the postgresql. Oracle does not have boolean type, int16,int32,... types, so they are mapped to char(1), and number, which makes the reverse mapping ambigous. I started the perec port to oracle, but that was never finished. I would say it's cca. 50%. Some queries generated by the query compiler are not valid in oracle sql (e.g. there is no 'case' expression). Furthermore we have added functionality to perec after I tried the port, (like using views, dimensional stuff), in which I don't have experience how hard is to port. That's all, I can remember now. If you have some concrete question just drop me a mail. Regards, Tomi 2010/3/17 Tomas Hlavaty > Hi Tomi, > > I am starting to work on porting hu.dwim.rdbms to oracle (and > hu.dwim.perec later) and I would like to ask what state was the oracle > backend last time it was maintained? What version of oracle database > did you use? > > Thank you, > > Tomas > -------------- next part -------------- An HTML attachment was scrubbed... URL: From tomas.hlavaty at knowledgetools.de Thu Mar 18 12:57:44 2010 From: tomas.hlavaty at knowledgetools.de (Tomas Hlavaty) Date: Thu, 18 Mar 2010 13:57:44 +0100 Subject: [Cl-perec-devel] oracle port In-Reply-To: <53c1332f1003170610g2233f39bi9c39c45f69ea1996@mail.gmail.com> (=?utf-8?Q?=22Borb=C3=A9ly_Tam=C3=A1s=22's?= message of "Wed, 17 Mar 2010 14:10:21 +0100") References: <86ocinkv21.fsf@knowledgetools.de> <53c1332f1003170610g2233f39bi9c39c45f69ea1996@mail.gmail.com> Message-ID: <86k4t9lrcn.fsf@knowledgetools.de> Hi Tomi, thanks for your answers. > That's all, I can remember now. If you have some concrete question > just drop me a mail. I'll see when I dive more into it;-) Thank you, Tomas