[postgis-devel] Proposal to drop support for PostgreSQL 8.3 in PostGIS 2.0

Paragon Corporation lr at pcorp.us
Thu Oct 14 10:05:20 PDT 2010


>  From my point of view it's dependent upon the overhead placed on
developers, i.e. by supporting all these extra versions are we making more
work for
 > ourselves? In terms of function signatures, 99% of the time is spent
writing the function while 1% is pasting the prototypes so for me it's a
reasonably
> insignificant part of the process.

I'm saying there is a lot of overhead supporting 4 versions regardless of
how little change there is.  Its not 3 we will be supporting.

It's FOUR versions.  (8.3,8.4,9.0, and 9.1).  We just don't have the man
power to test FOUR versions and all OS permutations of that.  And if we
can't test it, I just assume not claim to support it.

> In terms of supporting versions, I'm generally quite happy to keep
compatibility up to the point where we need a specific minimum version
unless it causes a giant fork in the code. Then again, I appreciate that
building Win32 installers can take a lot of time.

Supporting 8.3 is holding us back.  When ST_DumpPoints came out we had to
change it to work in 8.3 (that wasn't a biggie).  We can't use Window
functions or CTEs in any of our functions.  We have to watch our backs when
doing aliases on columns because 

8.3 doesn't support the field a  format -- granted I think that's a bad
habit anyway.

We can't support RETURNING QUERY (with dynamic), because 8.3 doesn't support
that.

What I'm saying is we might not have functionality that uses these things
yet.  But we CAN:T EVEN DREAM of having such functionality because 8.3
doesn't support that.  We've (I can only guess have ugly aggregate condition
8.3 code in our agg functions because 8.3 doesn't support WINDOW in
aggregates.


> My gut feeling is that we should leave the supported PostgreSQL versions
where they are for the moment
So my gut feeling is -- WE SHOULD NEVER SUPPORT MORE THAN 3 VERSIONS in any
release (particularly a MAJOR one)

I>  I'd have no objection if you only built Win32  binaries for the latest 2
releases. -- WELL Its up to 5 now (even if we don't support 8.3), as we will
have to support 8.4, 9.0 (32, 64-bit) and upcoming 9.1 (32, 64-bit)  :)
But seriously that's the least of my concerns.  Just want to have a clean
code base with as minimal conditional code balancing the painfulness to
users vs. our ability to make changes and test them as quickly as possible.

Thanks,
Regina 






More information about the postgis-devel mailing list