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

Re: RFS: nettee



Hi Joel,

sorry for answering myself so late, but I have been busy these days (and
I am still but I'm trying to keep up today).

On Wed, Mar 05, 2008 at 12:02:10AM -0300, Joel Franco wrote:
> However, if you look ate the Debian available packages today, you will
> see that the most do not follow that recommendation.

Thats true, but not a charter for upcoming packages. Its just a sign
that not everybody cares or cared about providing good descriptions.

> Well, i have changed the nettee short description.

Lets not focus too much on the short desc. Its more the long description
that I care about. Something like "a network tee program" as a short
description is okay, because tee is both a common tool and an English
word. But the long description should tell the reader what exactly is a
network tee program. I'm not very satisfied with that, because it uses
an engineer language, but a program to clone computers over the LAN
isn't obligatory used by an engineer. So the description must not be to
complicated (and even for technical packages I pledge for use of normal
language instead of technical terms).
Here is a proposal:

Description: a network tee program
nettee is a program that can be used to transfer data from one computer
to a number of computer nodes simultaneously at nearly full speed of the
network it is connected to.
.
A common use-case for this application is for cloning computer
partitions and disks or moving large database files.
.
Its advantage over netcat+tee is, that it is more simple and can
survive to error conditions like computer nodes dead and transfer
courruption.


> >Please move it to the source package part of the package, for example
> >after the maintainer line.
> 
> ok

Oh, got quiet high. Well, thats okay, while personally I would have
preferred to move it somewhat lower in the source package part (for
example below the maintainer line or so).

> :) now i understand. i made it.

Good.

> Sorry, but it isn't still very clear to me. I understand that the
> copyright file must refer to the Debian license files in a generic way
> and not in a particular way to this package.

Hu? Now I don't understand you.

> Right. That's fine and now i understand why it's useful.
> I have corrected it now :)

Good.

> >- debian/changelog: Needs some work. Changelog entries are not as they
> >  should be. See [1] for some instructions.
> 
> i'd read that, i'm more conscious about that and have changed somethings.
> However, i have to maintain the minimal changes mentionated because it's one of
> my first packages.

The last entry is _very_ confusing. You describe about 6 changes in
_one_ changelog entry and no changelog describes _why_ something has
been changed. But thats bad. After all the sense of the changelog
is for someone else then you (and you, too, in the future) to understand
what has been changed and why it has been changed (which affect does the
change have?).

> >- debian/README.Debian is still in the package. Remember that I and Paul told
> >  you, that its content is not really what the README.Debian is for.
> 
> i don't know which is the better way to fix this issue: i should send it to the
> upstream author or I should rename it to something reflecting the my particular
> use?

I'd suggest you to send it upstream. But it is not suited for
README.Debian, because this file is for Debian specific notes, which
this certainly is not.

Best Regards,
Patrick


Reply to: