LATEX-L Archives

Mailing list for the LaTeX3 project

LATEX-L@LISTSERV.UNI-HEIDELBERG.DE

Options: Use Classic View

Use Monospaced Font
Show Text Part by Default
Condense Mail Headers

Topic: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Sender: Mailing list for the LaTeX3 project <[log in to unmask]>
From: Barbara Beeton <[log in to unmask]>
Date: Mon, 1 Jul 2002 13:02:51 -0400
Content-Type: TEXT/PLAIN; charset=US-ASCII
MIME-Version: 1.0
Reply-To: Mailing list for the LaTeX3 project <[log in to unmask]>
Parts/Attachments: TEXT/PLAIN (83 lines)
claire connelly suggests, with respect to the lppl permitting
substitution of a new maintainer for one who has "disappeared",

    If what we're concerned about here is someone ``hijacking'' a
    particular package, then it might make more sense to define some
    restrictions on uploading to CTAN and leaving the license such
    that anyone can modify the package and make it available somewhere
    else.

no, please, no.
anyone *can* now modify a package as long as they change the
name.  the purpose of the restriction is to guarantee
reproducibility.

the ams already struggles with material submitted by authors
that uses old versions of "standard" packages.  it adds a great
deal of time and cost to producing our publications.  but at
least we are generally certain that no one has made unexpected
changes, and can usually trace through the history to see what
might be causing problems.

if we can't be sure that authors are using standard (even if
obsolete) versions of packages, that will only increase our
cost and headaches.

if file names aren't changed when a modified file is installed
on a shared system (as are the norm at many universities), most
users won't even know they're not using the standard version.

    CTAN could be (and is, I thought) regarded as the ``official
    repository'' of TeX/LaTeX packages, and might require a higher
    standard of proof before allowing someone other than the original
    author to upload a changed version of a package.  [...]

yes, we regard ctan as the trusted source for packages, but if
someone already has a particular package available locally,
that person isn't going to go to ctan to get a new copy.

    That way the LPPL could be a very free license while still
    preserving an official TeX source tree with a significant level of
    sanity checking.  ...

not enough sanity, unfortunately.

    ...  If the changes were too great to allow complete
    backwards compatibility, the CTAN maintainers could enforce a name
    change on the package (while at the same time recommending the
    newer package with a note in the CTAN Catalogue or similar).

the ctan maintainers don't have time to do this level of
chacking.  i suspect they do make a cursory check on the
"ownership" of a package before posting, but actually testing
it is out of the question.  the problem of "disappearing"
authors is real (though not always their fault), and i'd love
to see something in the license to get around the problem.
but opening things up so that changes can be made to copies
not on ctan isn't the answer.

    I don't think that the license has to assume that anyone making
    changes is up to no good and restrict people's ability to make
    those changes or to make those changes available in some form.  At
    the moment, the LPPL doesn't prevent an original author from
    making significant changes to their package that break backwards
    compatibility or even completely change its functionality.

i don't think there *is* any assumption that someone making
changes is up to no good.  i think that reproducibility is
the only important assumption in this regard.  if the original
author does make incompatible changes, then s/he will be faced
with deserved slings and arrows.

    On a related note, if CTAN, the LaTeX Project, TUG, or one of the
    European user groups could provide the resources for a centralized
    bug-tracking system that all CTAN authors could use, such a system
    could be a very valuable way to keep track of problems, fixes, and
    even the activity and availability of authors.  (I'm thinking, of
    course, of Debian's BTS (bugs.debian.org), which tracks an
    enormous amount of information without being especially
    complicated.)

this is a good idea.
                                                        -- bb

ATOM RSS1 RSS2