[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#720327: ITP: esu -- It allows to copy files with different checksums on the fly.

Package name should be "ecp", sorry misspelled that.

I fixed the README and improved some things.

Thanks for your feedback!

2013/8/20 Lars Wirzenius <liw@liw.fi>
On Tue, Aug 20, 2013 at 06:41:22PM +0400, Konstantin Khomoutov wrote:
> On Tue, 20 Aug 2013 10:21:05 -0400
> Ryan Kavanagh <rak@debian.org> wrote:
> > On Tue, Aug 20, 2013 at 03:33:06PM +0200, root wrote:
> > > * Package name    : esu
> > >   Description     : It allows to copy files with different checksums
> > >                     on the fly.
> > >
> > > Basicly a replacement for cp with additional checksum on the fly
> > > support.  It allows MD5, SHA1, SHA224, SHA265, SHA384, SHA512
> > > algorithm to be used.
> >
> > How is this different from rsync? Quoting rsync(1):
> [discussion of meaning of rsync option]

It would be helpful if the upstream README.md and the package description
explain what the checksums are for. Is this a replacement for cp+sha1sum
(or sha256sum or whatever algorithm is used), or is the checksum used
for verifying that the resulting file is copied correctly and has not
become corrupted during the copy? If the latter, does the program do
anything else to ensure a safe copy, such as fsync to make sure the
target file is committed to disk, or flushing kernel buffer caches so
that checksumming the target file happens on data that is read from the
target disk, and not from cache memory? Also an explanation of why this
is useful and why (and when) the kernel's usual mechanisms aren't enough
would be a good idea.

"cp, but with checksums" isn't a useful description of a program. Unless
the program's output includes the checksums (perhaps for later
verification), the checksums don't seem interesting to me as a user. They
seem like an implementation detail rather than an essential feature of
the program.

http://www.cafepress.com/trunktees -- geeky funny T-shirts
http://gtdfh.branchable.com/ -- GTD for hackers

Reply to: