| 1 |
On 2/14/2018 9:35 AM, Francesco Riosa wrote: |
| 2 |
> |
| 3 |
> On 14/02/2018 05:37, Robin H. Johnson wrote: |
| 4 |
>> On Tue, Feb 13, 2018 at 09:32:32PM -0500, Brian Evans wrote: |
| 5 |
>>> I have a plan I would like some eyes on... |
| 6 |
>>> |
| 7 |
>>> I want to gradually *BAN* the use of virtual/mysql and |
| 8 |
>>> virtual/libmysqlclient as dependencies. |
| 9 |
>> Overall I agree, but there's some slight concerns I have. |
| 10 |
>> |
| 11 |
>>> To accomplish this, force dev-db/mysql-connector-c to be the only souce |
| 12 |
>>> of libmysqlclient.so. |
| 13 |
>>> |
| 14 |
>>> Packages that choose to support libmariadb.so instead can include a |
| 15 |
>>> libmariadb USE to hook up to dev-db/mariadb-connector-c that will be |
| 16 |
>>> introduced (and they can live side-by-side). The motivation for this |
| 17 |
>>> could be licensing with libmariadb being LGPL instead of GPL. This is |
| 18 |
>>> similar to ffmpeg/libav, except the libraries can co-exist. |
| 19 |
>> Have all the concerns about using slightly different libmysqlclient.so |
| 20 |
>> builds been resolved? Esp for pre-built binaries (I don't know if there |
| 21 |
>> are any left in the tree). |
| 22 |
>> |
| 23 |
>>> The current providers of virtual/mysql would get a new USE flag that is |
| 24 |
>>> MASKED for all users for the transition period and pull in the lib |
| 25 |
>>> package(s) when that USE is disabled. |
| 26 |
>>> |
| 27 |
>>> virtual/mysql would become a server reference for USERS only. It would |
| 28 |
>>> be a QA warning violation to depend directly on virtual/mysql as it can |
| 29 |
>>> live anywhere. |
| 30 |
>> This part worries me slightly. I do understand that mysql-embedded is |
| 31 |
>> retired entirely, but apps that spun up their own local mysqld instance |
| 32 |
>> would still be affected this this change. |
| 33 |
> |
| 34 |
> Checked a random desktop install and found these packages depending on |
| 35 |
> virtual/mysql: |
| 36 |
> |
| 37 |
> vivo@Monfi ~ $ equery d virtual/mysql |
| 38 |
> * These packages depend on virtual/mysql: |
| 39 |
> dev-db/mariadb-10.2.12 (server ? ~virtual/mysql-5.6[embedded=,static=]) |
| 40 |
> dev-libs/apr-util-1.6.1 (mysql ? =virtual/mysql-5*) |
| 41 |
> dev-libs/cyrus-sasl-2.1.26-r11 (mysql ? virtual/mysql) |
| 42 |
> dev-libs/redland-1.0.17-r1 (mysql ? virtual/mysql) |
| 43 |
> kde-apps/akonadi-17.12.2 (mysql ? virtual/mysql) |
| 44 |
> net-analyzer/net-snmp-5.7.3_p3 (mysql ? virtual/mysql) |
| 45 |
> net-mail/mailutils-3.4 (mysql ? virtual/mysql) |
| 46 |
> sci-mathematics/glpk-4.63 (mysql ? virtual/mysql) |
| 47 |
dev-libs/apr-util-1.6.1 (mysql ? dev-db/mysql-connector-c:=) |
| 48 |
dev-libs/cyrus-sasl-2.1.26-r11 (mysql ? dev-db/mysql-connector-c:=) |
| 49 |
dev-libs/redland-1.0.17-r1 (mysql ? dev-db/mysql-connector-c:=) |
| 50 |
kde-apps/akonadi-17.12.2 (mysql ? dev-db/mysql-connector-c:=) |
| 51 |
net-analyzer/net-snmp-5.7.3_p3 (mysql ? dev-db/mysql-connector-c:=) |
| 52 |
net-mail/mailutils-3.4 (mysql ? dev-db/mysql-connector-c:=) |
| 53 |
sci-mathematics/glpk-4.63 (mysql ? dev-db/mysql-connector-c:=) |
| 54 |
|
| 55 |
|
| 56 |
> |
| 57 |
> It would be interesting to know how their ${*DEPEND} should look after |
| 58 |
> the change. |
| 59 |
> |
| 60 |
> Honestly I'd also like to understand the rationale better, why this |
| 61 |
> change is needed? |
| 62 |
> Are the implementations of Oracle mysql and Mariadb so different that |
| 63 |
> it's not possible anymore to consider those equivalent (for what |
| 64 |
> packages see)? |
| 65 |
|
| 66 |
This is needed for a few reasons: |
| 67 |
|
| 68 |
1) Virtuals cannot rebuild dependencies when there are subslot changes |
| 69 |
2) dev-db/mysql-5.7 and dev-db/percona-server-5.7 are prevented from |
| 70 |
entering the tree simply due to the client library SOVERSION change. It |
| 71 |
is difficult for the user to swap between server versions. |
| 72 |
3) libmariadb.so is a pthread_once library which many other packages are |
| 73 |
not expecting. So they may have to write things two different ways |
Archives