MariaDB-5.5-57 for OpenVMS

[Neil Rieck]

This just in from Mark Berryman

Mark Berryman informed me that he has finally found the time to continue releasing newer ports of "MariaDB for OpenVMS". (yay!)

Now that his toolset is in place we might see newer versions of MariaDB sooner on OpenVMS. Congratulations Mark.

Anyway, you can download a BETA of MariaDB-5.5-57 (Alpha and Itanium) from here:
http://www.theberrymans.com/php_kits/

Caveats:
1) This is a BETA
2) Not supported, and no warranty of any kind (like MySQL and MariaDB)
3) Mark still sounds overloaded so please to not bug him
4) For people who compile OpenVMS programs against Mark's libraries (like me), changes have been made which you will need to test on an alternate system. For example, his previous product (MariaDB-5.5-25) had separate directories for Alpha and Itanium libraries. This release changes that to a single common directory. Carefully read all of the file "Readme.VMS" before you begin.

Neil Rieck
Waterloo, Ontario, Canada.
http://neilrieck.net/OpenVMS-Programmers-Corner.html
http://neilrieck.net/docs/openvms_notes_mysql_mariardb.html

[Bill Gunshannon]

On 10/11/2017 08:24 AM, Neil Rieck wrote:
> This just in from Mark Berryman
>
> Mark Berryman informed me that he has finally found the time to continue releasing newer ports of "MariaDB for OpenVMS". (yay!)
>
> Now that his toolset is in place we might see newer versions of MariaDB sooner on OpenVMS. Congratulations Mark.
>
> Anyway, you can download a BETA of MariaDB-5.5-57 (Alpha and Itanium) from here:
> http://www.theberrymans.com/php_kits/

Which might sound impressive until you see that MariaDB is already
up to 10.2 in the stable tree and 10.3 Alpha.

>
> Caveats:
> 1) This is a BETA
> 2) Not supported, and no warranty of any kind (like MySQL and MariaDB)
> 3) Mark still sounds overloaded so please to not bug him
> 4) For people who compile OpenVMS programs against Mark's libraries (like me), changes have been made which you will need to test on an alternate system. For example, his previous product (MariaDB-5.5-25) had separate directories for Alpha and Itanium libraries. This release changes that to a single common directory. Carefully read all of the file "Readme.VMS" before you begin.

bill

[Jan-Erik Soderholm]

Den 2017-10-11 kl. 17:01, skrev Bill Gunshannon:
> On 10/11/2017 08:24 AM, Neil Rieck wrote:
>> This just in from Mark Berryman
>> Mark Berryman informed me that he has finally found the time to continue
>> releasing newer ports of "MariaDB for OpenVMS". (yay!)
>> Now that his toolset is in place we might see newer versions of MariaDB
>> sooner on OpenVMS. Congratulations Mark.
>> Anyway, you can download a BETA of MariaDB-5.5-57 (Alpha and Itanium)
>> from here:
>> http://www.theberrymans.com/php_kits/
>
> Which might sound impressive until you see that MariaDB is already
> up to 10.2 in the stable tree and 10.3 Alpha.
>

You have probably not looked very closely at the MariaDB home page.
Much easier to just throw in a comment, I guess...

5.5.57 is the last/latest 5.5 release and it is still in "GA" with
support and patches to 2020. The *next* major release was 10.x.

Both are actively supported and such things and security CVE's
fixes goes into both releases.

5.5.57 was released 19 Jul 2017, not *that* old, I'd say.

So exactly what do you mean with "already up to 10.2"?

Not that I like MariaDB/MySQL anyway, but I do not like
when someone is careless with the truth either.

[Bill Gunshannon]

On 10/11/2017 12:05 PM, Jan-Erik Soderholm wrote:
> Den 2017-10-11 kl. 17:01, skrev Bill Gunshannon:
>> On 10/11/2017 08:24 AM, Neil Rieck wrote:
>>> This just in from Mark Berryman
>>> Mark Berryman informed me that he has finally found the time to
>>> continue releasing newer ports of "MariaDB for OpenVMS". (yay!)
>>> Now that his toolset is in place we might see newer versions of
>>> MariaDB sooner on OpenVMS. Congratulations Mark.
>>> Anyway, you can download a BETA of MariaDB-5.5-57 (Alpha and Itanium)
>>> from here:
>>> http://www.theberrymans.com/php_kits/
>>
>> Which might sound impressive until you see that MariaDB is already
>> up to 10.2 in the stable tree and 10.3 Alpha.
>>
>
> You have probably not looked very closely at the MariaDB home page.
> Much easier to just throw in a comment, I guess...
>
> 5.5.57 is the last/latest 5.5 release and it is still in "GA" with
> support and patches to 2020. The *next* major release was 10.x.
>
> Both are actively supported and such things and security CVE's
> fixes goes into both releases.
>
> 5.5.57 was released 19 Jul 2017, not *that* old, I'd say.
>
> So exactly what do you mean with "already up to 10.2"?
>
> Not that I like MariaDB/MySQL anyway, but I do not like
> when someone is careless with the truth either.
>
>

Guess it depends on which MariaDB your looking at. :-)
mariadb.com is pushing MariaDB Server 10.2.9.

bill

[Jan-Erik Soderholm]

Maybe they do, so what? I have not looked at the "news" in 10.x, but
since 5.5.57 is still activelly developed and supported, I guess there
is a reason for that. Maybe some incompatibility against 10.x, too few
new features (or new features not needed by evereyone).

Anyway, I do not think that the port of 5.5.57 is that bad that your
post made it sound like.

[Simon Clubley]

On 2017-10-11, Bill Gunshannon <bill.gu...@gmail.com> wrote:
>
> Guess it depends on which MariaDB your looking at. :-)
> mariadb.com is pushing MariaDB Server 10.2.9.
>

This doesn't seem to be any different from the situation with RHEL and
its free rebuilds in that RHEL 7 is the RHEL version currently being
pushed but that doesn't change the fact that CentOS 6 is still supported.

This push current version but also support {major version - 1} mindset
appears to be common (and a sign of growing maturity) in some parts of
the open source world.

Simon.

--
Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world

[Arne Vajhøj]

Jan-Erik already indicated it, but let me elaborate. The version
numbers are misleading as MariaDB's major version number went
from 5 to 10 skipping 6, 7, 8 and 9.

So they have a new major version 10.x but are still supporting
the latest release of the last major version 5.x.

Like VSI may chose to support VMS 8 for a long time
after 9 is out.

Arne

[Michael Moroney]

<ar...@vajhoej.dk> writes:

>> Guess it depends on which MariaDB your looking at. :-)
>> mariadb.com is pushing MariaDB Server 10.2.9.

>Jan-Erik already indicated it, but let me elaborate. The version
>numbers are misleading as MariaDB's major version number went
>from 5 to 10 skipping 6, 7, 8 and 9.

Seems that marketing and other non-technical types have been tinkering
with version numbering of software and tech. Used to be that a bump in the
major number (by 1) indicated either a rewrite or a significant technical
change or addition. For example with VMS, introduction of VMS on Alpha or
Itanium, or clustering. But now Microsoft and apparently Apple have
skipped version 9. Firefox is now at Version 56.0. Similar with Chrome and
Opera. Now MariaDB skipped 4 versions.

[Stephen Hoffman]

On 2017-10-12 01:47:49 +0000, Michael Moroney said:

> Seems that marketing and other non-technical types have been tinkering
> with version numbering of software and tech. Used to be that a bump in
> the major number (by 1) indicated either a rewrite or a significant
> technical change or addition. For example with VMS, introduction of VMS
> on Alpha or Itanium, or clustering. But now Microsoft and apparently
> Apple have skipped version 9. Firefox is now at Version 56.0. Similar
> with Chrome and Opera. Now MariaDB skipped 4 versions.

Ayup.... Major new features of the platform get a patch-level change
(VAX/VMS V5.0-3, with DECwindows), major new version that wasn't
(OpenVMS Alpha V8.2), releases reserved for betas or for specials or
limited distributions (OpenVMS Alpha V7.2-6C1, V7.2-6C2, V8.0, & V8.1),
new features shipped out without any version changes and doc'd solely
via patch notes (V8.4-era UPDATE kits), new features shipped out via
ECO kit and sorta-but-not-really doc'd via config file comments (TCP/IP
V5.7, & ECOs), RTLs that are permanently GSMATCH'd into oblivion with
the same value used for both upward and downward compatibility, the
"VMS" to "OpenVMS" naming, need I continue?

Biggest pragmatic issue with any of the versioning morass has been the
longstanding lack of a generic capability-sensing and patch-sensing API
within OpenVMS. This as the OpenVMS version number has become and
arguably always was a marketing construct, being misused as and twisted
into an exceedingly half-baked capability and patch indicator.

How OpenVMS — which has implemented a plethora of
unfortunately-often-cryptic APIs for access to all manner of arcana —
was never extended to provide introspection into the local and cluster
software configuration? Maybe because the OS folks don't themselves
usually need to do that? Donno. Neither PCSI nor LMF ever got
callable interfaces, for that matter.

Picking version numbers and the associated debates is often fodder for
controversy. Dealing with the version sprawl is never easy, with
mixed-version clusters and mixed installations and kits that'd best
work across ranges of versions. What's even harder for a development
team is addressing the sprawl and reducing the configuration
permutations, once it's become entrenched and familiar. That includes
treating networking and network services as an add-on, and the "fun"
involved in accreting the management of the version matrices onto the
end-users, among other details.

Fun times.





--
Pure Personal Opinion | HoffmanLabs LLC

[David Froble]

Well, at least we didn't get "OpenVMS V1.0" ...

[Stephen Hoffman]

On 2017-10-12 17:01:22 +0000, David Froble said:

> Stephen Hoffman wrote:
>> On 2017-10-12 01:47:49 +0000, Michael Moroney said:
>>
>>> Seems that marketing and other non-technical types have been tinkering
>>> with version numbering of software and tech. Used to be that a bump in
>>> the major number (by 1) indicated either a rewrite or a significant
>>> technical change or addition. For example with VMS, introduction of VMS
>>> on Alpha or Itanium, or clustering. But now Microsoft and apparently
>>> Apple have skipped version 9. Firefox is now at Version 56.0. Similar
>>> with Chrome and Opera. Now MariaDB skipped 4 versions.
>>
>> Ayup.... Major new features of the platform get a patch-level change
>> (VAX/VMS V5.0-3, with DECwindows), major new version that wasn't
>> (OpenVMS Alpha V8.2), releases reserved for betas or for specials or
>> limited distributions (OpenVMS Alpha V7.2-6C1, V7.2-6C2, V8.0, & V8.1),
>> new features shipped out without any version changes and doc'd solely
>> via patch notes (V8.4-era UPDATE kits), new features shipped out via
>> ECO kit and sorta-but-not-really doc'd via config file comments (TCP/IP
>> V5.7, & ECOs), RTLs that are permanently GSMATCH'd into oblivion with
>> the same value used for both upward and downward compatibility, the
>> "VMS" to "OpenVMS" naming, need I continue?

>> ...

>> Fun times.
>
> Well, at least we didn't get "OpenVMS V1.0" ...

Yeah, that would have been weird. Nobody'd ever use V1.0, V1.5,
V1.5-1H1, and then V6.1.

[Craig A. Berry]

Although I have heard of VMS referred to as "NT 1.0" :-).