Y2K TCL style

[Rob Seaman]

Comp.risks (RISKS DIGEST 19.35) had an interesting Y2K message today:

    Date: Tue, 26 Aug 1997 21:09:43 +0100 (BST)
    From: Lloyd Wood <L.Wood at surrey.ac.uk>
    Subject: Tcl 8.0 Y2K Risk

    Sun's scripting division has released Tcl 8.0, a fashionable bytecode
    compiler reworking of the Tcl interpreter to help Tcl scale better.

    Making it compilable required some changes in semantics, but there
    were some other changes as well. In particular,

    http://sunscript.sun.com/TclTkCore/8.0.html

    Notes:

      There are also a few other minor incompatibilities in Tcl 8.0 and
Tk 8.0:
      [..] 2.2-digit years are now parsed differently by the clock
command to
      handle year 2000 issues better (years 00-38 are treated as 2000-2038
      instead of 1900-1938).

    Supporting two-digit years in the first place was risky enough, but
this
    change is bound to catch a lot of people out.

    It looks like the millennium problem may have come early for
cutting-edge
    Tcl scripters with legacy code.

    <L.Wood at surrey.ac.uk>PGP<http://www.sat-net.com/L.Wood/>

It's good to see that other folks are paying attention to the problem.
Note, however, the differences between the FITS solution and this:

    1) FITS has resolved the existing ambiguity of two-digit years
       by clarifying them as explicitly twentieth century.

    2) FITS has actually fixed the problem (at least until Y10K).

(Comparing FITS and TCL is otherwise comparing apples and oranges, of
course.)

The moderated, high S/N Risks Digest is available via mailing list or as
the comp.risks newsgroup.  Back issues are available via ftp or from:

    http://catless.ncl.ac.uk/Risks

General information is available from:

    http://www.CSL.sri.com/risksinfo.html

Rob Seaman
---
seaman at noao.edu, http://iraf.noao.edu/~seaman
NOAO, 950 N Cherry Ave, Tucson AZ 85719, 520-318-8248
PGP: 98 8D 8B 49 74 9A 41 88  3A 43 87 54 51 BF 30 4B