Re: [arch-general] [arch-dev-public] removed mariadb 10.1.27 from [testing]
On Tue, 2017-09-26 at 21:28 +0200, Christian Hesse wrote:
Hello everybody,
after the release of 10.1.27 upstream was made aware of a regression, so the release has been pulled from the downloads system. The fix will be in 10.1.28. I removed mariadb 10.1.27 packages from [testing].
Thank you Christian - related note - can you share thoughts around 10.2? thanks. -- Gene lists@sapience.com
Genes Lists via arch-general <arch-general@archlinux.org> on Tue, 2017/09/26 16:41:
On Tue, 2017-09-26 at 21:28 +0200, Christian Hesse wrote:
Hello everybody,
after the release of 10.1.27 upstream was made aware of a regression, so the release has been pulled from the downloads system. The fix will be in 10.1.28. I removed mariadb 10.1.27 packages from [testing].
Thank you Christian - related note - can you share thoughts around 10.2?
We tried to upgrade to MariaDB 10.2.6, which ended in a disaster. The client library has been renamed from libmysqlclient to libmariadb and it has seen structural changes for data types. Programs fail to link or - if they do link - crash or misbehave, including data corruption. I do not see any way to upgrade any time soon. IMHO data structures have to be made opaque in favor of accessing data via functions. OpenSSL suffered a similar problem, which has been solved with version 1.1.0. MariaDB upstream would have to do something similar. -- main(a){char*c=/* Schoene Gruesse */"B?IJj;MEH" "CX:;",b;for(a/* Best regards my address: */=0;b=c[a++];) putchar(b-1/(/* Chris cc -ox -xc - && ./x */b/42*2-3)*42);}
On Tue, 2017-09-26 at 23:06 +0200, Christian Hesse wrote:
... We tried to upgrade to MariaDB 10.2.6, which ended in a disaster. The client library has been renamed from libmysqlclient to libmariadb and it has seen structural changes for data types. Programs fail to link or - if they do link - crash or misbehave, including data corruption.
I do not see any way to upgrade any time soon. IMHO data structures have to be made opaque in favor of accessing data via functions.
Thanks - so it's clear - are you saying that the recommended method of running[1] mysql_upgrade does not in fact work properly? [1] https://mariadb.com/kb/en/library/upgrading-from-mariadb-101-to-mar iadb-102/ -- Gene lists@sapience.com
Genes Lists via arch-general <arch-general@archlinux.org> on Wed, 2017/09/27 09:31:
On Tue, 2017-09-26 at 23:06 +0200, Christian Hesse wrote:
... We tried to upgrade to MariaDB 10.2.6, which ended in a disaster. The client library has been renamed from libmysqlclient to libmariadb and it has seen structural changes for data types. Programs fail to link or - if they do link - crash or misbehave, including data corruption.
I do not see any way to upgrade any time soon. IMHO data structures have to be made opaque in favor of accessing data via functions.
Thanks - so it's clear - are you saying that the recommended method of running[1] mysql_upgrade does not in fact work properly?
[1] https://mariadb.com/kb/en/library/upgrading-from-mariadb-101-to-mariadb-102/
No, mysql_upgrade updates the table data (schema 'mysql'), that's not a problem. The server itself works just fine. But API/ABI in library /usr/lib/libmariadb.so and header files in /usr/include/mysql/ changed. Lots of programs break because of this. -- main(a){char*c=/* Schoene Gruesse */"B?IJj;MEH" "CX:;",b;for(a/* Best regards my address: */=0;b=c[a++];) putchar(b-1/(/* Chris cc -ox -xc - && ./x */b/42*2-3)*42);}
participants (2)
-
Christian Hesse
-
Genes Lists