On Sat, 10 Mar 2007 18:15:17 +0100, Thorsten Leemhuis wrote:
Michael Schwendt schrieb:
> On Sat, 10 Mar 2007 16:31:56 +0100, Thorsten Leemhuis wrote:
>> Axel Thimm wrote:
>>> User management is delicate and fedora-usermgmt is not the way to go.
>> It solved a problem afaics, that we have no better solution for in RHEL5.
>
> It doesn't. It adds a feature that can be enabled on demand. The feature
> fixes a problem, but since the tool is not enabled by default, it doesn't
> fix anything for us.
Okay, then let me rephrase: fedora-usermgmt can be used to solve a
problem if the local admin wants to do it; fedora-usermgmt works the
same way as useradd would if not configured and should do no more or
less harm then useradd for users.
Does that match better?
Yes.
> There are other reasons why EPEL might not want the tool. Naming
them is
> left as an exercise once all the critics have understood what the tool does
> and how it works.
/me still would like to see good reasons to ban it...
For instance, but probably not limited to:
- It is a wrapper around a system utility (user{add,del}/group{add,del}).
- It is not particularly convenient to turn it on at install-time.
- Non-standard user/group maintenance tools are not supported in RHEL.
- Two set of tools, different possibly conflicting behaviour (if enabled).
- You need people to maintain the package/tools for many years.
- You need to keep it alive for many years, since offering it implies
that your users may deploy it.
- The relative values in pkg scriptlets are misleading as they look
like constants, e.g. fedora-useradd 42 joe is not like useradd -u 42 joe
- It is not pretty, see e.g. "fedora-useradd --help".
- Has it seen enough QA before it's made available to the target group
of EPEL?
/me sill wants this problems solved in Fedora-land once (e.g. in
both
EPEL and Fedora)
CU
thl