On 12/17/05, Phil Steitz <[EMAIL PROTECTED]> wrote: > We seem to be converging on a "standard" (mentioned in > /releases/release.html as an ASF standard, but I can't find it > documented anywhere) for md5 checksum files that takes the form > > hash *<name> > > I am curious as to where the extra stuff comes from (i.e., why we > decided to add the * and the file name). Looking around apache, I > find > > httpd uses > hash <name> > (note no "*") > > ant, struts, tomcat, geronimo use > hash > > commons components are split between just having the hash and adding > on the "*name". > > It seems silly to me to add the extra contents, since this is just > duplicating the file name inside the file. Of course, I am sure there > is a good reason for this and I just don't know it, so maybe someone > can clue me in or point me to a link where this is documented? > <snip/>
Don't know if there is a link that documents this in the apache.org domain, but I'd recommend we stick with: hash *<name> since: * This retains the input file information even if the checksum file is purposely or accidently (such as during download) renamed. * Some verification tools may require the filenames since a checksum file can contain entries for more than one input file (one per line). * I once happened across a Windows based tool that distinctly looked for the filename at the 35th character in the line, 32 for md5 checksum, a space, a file type demarker and the filename (though I don't recollect its name, so this may sound like a urban legend ;-). The * before the <name> is a file type marker, and indicates the checksum is of a binary file. This is documented, for example, in the Linux md5sum man page [1]. Thus, adding the extra content, though not strictly necessary, is good practice nevertheless. -Rahul [1] http://www.die.net/doc/linux/man/man1/md5sum.1.html > Thanks! > > Phil > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
