Re: First draft of Beta announcement

On 05/12/2017 10:03 AM, Joshua D. Drake wrote:
> On 05/12/2017 09:40 AM, Josh Berkus wrote:
>> Josh, Justin, Andres, Gunnar, Peter:
>>
>
>>> Most people won't know what that means. I can conclude that it means
>>> that we can now connect to PostgreSQL and say, "please give me a read
>>> only or a read/write host" but I am sure I am not 100% correct. I know
>>> more about PostgreSQL than most who will care about this announcement.
>>
>> The idea is more to get people -- specifically driver and ORM authors --
>> interested enough to bother looking up the feature. Not to describe it
>> in full, which would take a paragraph.
>
> Sure but still, the line doesn't really mean anything. Perhaps:
>
> * Driver API for read only or read/write database routing? (I know
> that's wrong but I think you know what I am getting at.

How about just "New target_session_attrs parameter"? I'm more concerned
with giving users a string which they can search on than a definition.
Anything real is going to be too complex for an announcement.

>>> * WAL support for Hash Indexes
>>>
>>> Crash safe Hash Indexes or ACID compliant Hash Indexes.
>>
>> Crash safe is good.
>>
>>>
>>> WAL is irrelevant in terms of the announcement.
>>>
>>> New "monitoring" roles for permission grants
>>>
>>> Is roles supposed to be plural?
>>
>> Yes.
>
> Then let's list them with context. Something like:
>
> New roles, x,y and z for monitoring purposes

No, because that would extend this to multiple lines. The role names
are long.

>
>>>
>>> Push Down Aggregates to Foreign Servers
>>>
>>> What does this mean? (I know what it means), the majority of our readers
>>> will not.
>>
>> If you can come up with a wording here which is clearer but takes one
>> line, be my guest. I was unable to.
>
> Planner support for aggregates on foreign (linked) tables
>
> I added linked because it is a term that Oracle and MSSQL DBAs will get.

WFM.

>
>>> I don't think we need to say anything more than:
>>>
>>> Version 10 has a high number of backwards-incompatible changes. For a
>>> list of these changes please see the [Release Notes](link to release
>>> notes).
>>
>> I disagree. As a rule, we don't break backwards compatibility so pg 10
>> is going to be a shock to a lot of people. We really haven't seen this
>> quantity of breakage since 8.3, which was released nine years ago, long
>> before the majority of our current users were using Postgres. Given
>> that -- because of partiitoning and logical replication -- many users
>> will want to upgrade to 10 the month it comes out, I think we need to
>> point out *in detail* why they will want to do extra testing. At a
>> minimum, this includes the change in version numbering, the renaming of
>> xlog to wal, and dropping support for FEBE 1.0.
>
> We should point out IN EXCRUCIATING DETAIL, in the release notes.

pgsql-hackers decided that having extra detail in the release notes
beyond a listing of compatibility issues was unwanted (and some of these
compatibility issues aren't even in the compatibility section). So my
plan is to have a list of major items in the announcement, and a later
doc somewhere else with detail and examples, likely my blog + the What's
New page.

--
Josh Berkus
Containers & Databases Oh My!