Re: version numbering: was there a netcdf 3 in 1992?

Glenn,

Though we've already discussed this, I'm sending this anyway to get it
into the archives, so I can just look it up when the question comes up
again.

> I was digging through some netcdf archives, and
> discovered reference to "netcdf v3.0" in correspondence
> I had with Mark Rivers in October 1992!
> 
> Was this just my mistake at the time, or was the a version
> discontinuty at some point?

Here's the chronology, as best I can reconstruct:

  Version       Date of announcement

  1.05          1990-01
  1.06          1990-07
  1.07          1990-07
  1.08          1990-08
  1.09          1990-10
  1.10          1990-10
  1.11          1991-03
  2.0 beta      1991-10
  2.0           1991-11
  2.01          1991-11
  2.02          1991-12
  2.3 beta      1993-03
  2.3           1993-04
  2.3.1         1993-05
  2.3.2         1993-06
  2.3.2pl1      1993-06
  2.3.2pl2      1993-06
  2.3.2pl3      1994-07
  2.3.2pl4      1995-01-31
  2.4 beta1     1995-10-11
  2.4 beta2     1995-10-13
  2.4 beta3     1995-11-13
  2.4 beta5     1995-12-06
  2.4 beta6     1996-01-08
  2.4           1996-02-28
  2.4.1         1996-03-19
  2.4.2         1996-05-02
  2.4.3         1996-08-12
  3.1a          1996-08-30

In February 1993, we were discussing whether to call the next version
2.3 or 3.0 (I believe you favored 3.0), but we ultimately decided on
2.3.  I've appended a note sent to the netcdfgroup mailing list about
this time that confirms this.  There was never a version 3.0 announced
or released, as far as I can determine.

--Russ

 From address@hidden Mon Feb  1 07:10:27 1993
 Received: by unidata.ucar.edu id AA28658
   (5.65c/IDA-1.4.4 for netcdfgroup-send); Mon, 1 Feb 1993 14:10:30 -0700
 Received: from buddy.unidata.ucar.edu by unidata.ucar.edu with SMTP id AA28654
   (5.65c/IDA-1.4.4 for <address@hidden>); Mon, 1 Feb 1993 14:10:28 -0700
 Organization: .
 Keywords: 199302012110.AA28654
 Received: by buddy.unidata.ucar.edu id AA21638
   (5.65c/IDA-1.4.3 for address@hidden); Mon, 1 Feb 1993 14:10:27 -0700
 Date: Mon, 1 Feb 1993 14:10:27 -0700
 From: Russ Rew <address@hidden>
 Message-Id: <address@hidden>
 To: address@hidden
 Subject: new release delayed a bit

 Hi,

 First a reminder: administrative requests to be added to or deleted from
 this mailing list should be sent to address@hidden, *not*
 address@hidden.

 Question 4 in the NetCDF Frequently Asked Questions list is:

 4:  When will the next version be available?  What changes are included?

 The previous answer was that we would have the next version available in
 January 1993, but January is over and the new version is not quite ready
 yet.  We hope to have a beta release available in another week or two.  I'm
 just calling it "the new version" because we're still discussing whether to
 call it netCDF 2.3 or netCDF 3.0.  We are making no incompatible changes in
 the interface, but the additions of some new functions may justify a new
 release number.

 Here's the updated answer to Question 4 in the FAQ list (pub/netcdf/FAQ on
 unidata.ucar.edu): 

 A:  We hope to make the next version available in February 1993.  It contains
     bug fixes, portability enhancements, performance enhancements, and new
     capabilities.

     We implemented some new optimizations for the library that result in
     significant speedups for accessing cross-sections involving
     non-contiguous data.

     New capabilities include some additional interfaces that provide a more
     general form of hyperslab access.  This supports sub-sampling along
     specified dimensions and a mapping between the points of the hyperslab
     and the memory locations of the corresponding values.  In a generalized
     hyperslab, an index mapping vector is used to define the mapping between
     points in the generalized hyperslab and the memory locations of the
     corresponding values, so data values that are written or read need no
     longer be contiguous in memory.

     There are also some new interfaces that can be used to write, read, and
     inquire about records, where a record may contain multiple variables of
     different types and shapes.  Where before you had to access a record's
     worth of data using multiple calls, now you will be able to use a single
     call.

     The ncdump utility supports several new command-line options including
     the ability to specify for which variables data values will be output,
     to provide brief annotations in the form of CDL comments to identify
     data values for large multidimensional variables, or to provide full
     annotations in the form of trailing CDL comments for every data value.

     The new release also includes a prototype implementation of a C++
     interface for the netCDF data access library.  It provides all the
     functionality of the C interface, improves type safety by eliminating
     all use of void* pointers, and is somewhat simpler to use than the C
     interface.  With the C++ interface no IDs are needed for netCDF
     components, there is no need to specify types when creating attributes,
     and less indirection is required for dealing with dimensions.  However,
     since this is a prototype interface and implementation, it may be
     changed before a supported version is released.

 ________________________________________________________________________
 Russ Rew                                       Unidata Program Center
 address@hidden                         UCAR, PO Box 3000
                                                Boulder, CO 80307-3000