From levente.meszaros at gmail.com Thu Feb 21 14:41:59 2008 From: levente.meszaros at gmail.com (=?ISO-8859-1?Q?Levente_M=E9sz=E1ros?=) Date: Thu, 21 Feb 2008 15:41:59 +0100 Subject: [Cl-perec-devel] Announce new cl-perec Message-ID: FYI... just in case anybody other than us is reading this list... The old cl-perec branch was renamed to cl-perec-hydrogen and a new branch under cl-perec-helium was started. The official branch is still the old one that is cl-perec links to cl-perec-hydrogen. The new branch is incompatible with the previous one because we have changed the way hash code is calculated for class names so that 32/64 systems can use the same database. It also uses cl-rdbms-helium which has also been branched. There are several internal cleanups regarding the type system and some in the MOP stuff for better extensibility, but the most important new feature will be the so called tesites. This introduces time parameters into the persistency layer where you can set slot values in terms of timestamps and/or time intervalls. See the tests for more details. And as usual: there is no documentation available. Let us know if you want to write some... :-) levy -- There's no perfectoin From levente.meszaros at gmail.com Fri Feb 22 08:59:39 2008 From: levente.meszaros at gmail.com (=?ISO-8859-1?Q?Levente_M=E9sz=E1ros?=) Date: Fri, 22 Feb 2008 09:59:39 +0100 Subject: [Cl-perec-devel] Announce new cl-perec In-Reply-To: <794f042d0802211115o48582814u5859010b095a3d9@mail.gmail.com> References: <794f042d0802211115o48582814u5859010b095a3d9@mail.gmail.com> Message-ID: On Thu, Feb 21, 2008 at 8:15 PM, Saurabh Nanda wrote: > This may be off-topic, but I'm writing in to confirm whether cl-perec > will be able to support more than Postgres now? I had tried > installing cl-perec some time back, but gave up due to multiple errors > and non-standard dependencies. > > Anyways, this seems like an interesting and promising project, and I'm > still waiting :-) This is not at all off topic. To tell you the truth it is very unlikely that we are going to implement other SQL backends or port cl-perec to other CL implementations. We are using SBCL and postgresql at the moment and as long as they fit our needs we will not spend time on supporting others. On the other hand we are willing to support anybody who wants to port cl-perec to other CL implementations and/or use different SQL backends. cl-perec was implemented to keep protability in mind and cl-rdbms already supports Oracle and sqlite3 to some level but certainly not enough for all cl-perec tests to pass. (and there's quite a few) The patched or non standard libraries as you call them help us to go on without reimplementing them or waiting for the owner to commit our patches or negotiate a solution which satisfies both us and them. Long live the distributed version control... levy -- There's no perfectoin From attila.lendvai at gmail.com Fri Feb 22 09:45:03 2008 From: attila.lendvai at gmail.com (Attila Lendvai) Date: Fri, 22 Feb 2008 10:45:03 +0100 Subject: [Cl-perec-devel] Announce new cl-perec In-Reply-To: References: <794f042d0802211115o48582814u5859010b095a3d9@mail.gmail.com> Message-ID: > The patched or non standard libraries as you call them help us to go > on without reimplementing them or waiting for the owner to commit our > patches or negotiate a solution which satisfies both us and them. and of course all changes are being discussed on the official lists and most of the changes are merged into the official tree as-is within days. at the time of writing there are two branches which ran ahead of their official counterparts: 1) local-time, where i do have the commit bit, but i usuallt wait for Daniel's amen 2) postmodern, where we optimized the login to use ironclad but Marijn prefers not to have an external dependency (and also optimized it meanwhile, so we may roll it back). and we moved it to use local-time for time/date related things instead of an own solution in postmodern. -- attila