You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by "C. Michael Pilato" <cm...@collab.net> on 2007/02/16 09:08:22 UTC
svnsync and non-interactive authentication. bug?
While documenting svnsync in The Book, a question occurred to me. I see
that svnsync has --username and --password command-line options, as well as
the related --no-auth-cache and --non-interactive options.
But what happens if both the source and target repositories of your
mirroring process require authentication, and as different users? Are you
just out of luck in terms of getting non-interactive functionality? Is this
a shortcoming of svnsync, or am I overlooking something?
(I admit it -- I haven't tested this to see what happens. It's 4am, and I'm
going cross-eyed, so this mail is as much a TODO that will wind up back in
my Inbox as it is a question for someone else to answer.)
--
C. Michael Pilato <cm...@collab.net>
CollabNet <> www.collab.net <> Distributed Development On Demand
Re: svnsync and non-interactive authentication. bug?
Posted by "C. Michael Pilato" <cm...@collab.net>.
C. Michael Pilato wrote:
> Daniel Rall wrote:
>> On Fri, 16 Feb 2007, Garrett Rooney wrote:
>>
>>> On 2/16/07, C. Michael Pilato <cm...@collab.net> wrote:
>>>> While documenting svnsync in The Book, a question occurred to me. I see
>>>> that svnsync has --username and --password command-line options, as well as
>>>> the related --no-auth-cache and --non-interactive options.
>>>>
>>>> But what happens if both the source and target repositories of your
>>>> mirroring process require authentication, and as different users? Are you
>>>> just out of luck in terms of getting non-interactive functionality? Is
>>>> this
>>>> a shortcoming of svnsync, or am I overlooking something?
>>>>
>>> If I had to guess it won't work, but like you I haven't actually tried it.
>> Sounds like a bug to me. We'd need a second set of
>> --username/--password command-line options to correct this problem.
>
> I was thinking the same thing.
By the way, I tracked this in issue #2717, which I just closed as FIXED (per
the commit in r23433).
--
C. Michael Pilato <cm...@collab.net>
CollabNet <> www.collab.net <> Distributed Development On Demand
Re: svnsync and non-interactive authentication. bug?
Posted by "C. Michael Pilato" <cm...@collab.net>.
Daniel Rall wrote:
> On Fri, 16 Feb 2007, Garrett Rooney wrote:
>
>> On 2/16/07, C. Michael Pilato <cm...@collab.net> wrote:
>>> While documenting svnsync in The Book, a question occurred to me. I see
>>> that svnsync has --username and --password command-line options, as well as
>>> the related --no-auth-cache and --non-interactive options.
>>>
>>> But what happens if both the source and target repositories of your
>>> mirroring process require authentication, and as different users? Are you
>>> just out of luck in terms of getting non-interactive functionality? Is
>>> this
>>> a shortcoming of svnsync, or am I overlooking something?
>>>
>>> (I admit it -- I haven't tested this to see what happens. It's 4am, and
>>> I'm
>>> going cross-eyed, so this mail is as much a TODO that will wind up back in
>>> my Inbox as it is a question for someone else to answer.)
>> If I had to guess it won't work, but like you I haven't actually tried it.
>
> Sounds like a bug to me. We'd need a second set of
> --username/--password command-line options to correct this problem.
I was thinking the same thing.
Now, how do we do that compatibly? I mean, ideally, the option names would
all carry the "from/to" or "source"/"target" or "source"/"sync" (haha!)
naming. But we already have flat --username/--password.
The best thing I can think of which let's us look forward without changing
the meanings of things today is:
* introduce the four new options
* make "--username FOO" mean "--from-username FOO --to-username FOO";
same for the password args
* only allow --username if neither --from-username nor --to-username
are present (to discourage less-than-clear meanings); same for the
password args
Sound sane?
Anybody partial to a particular pair of new option prefixes? I kinda like
"source" and "sync" myself (so --source-username, --sync-username, etc.
It's cute, yet clear.
--
C. Michael Pilato <cm...@collab.net>
CollabNet <> www.collab.net <> Distributed Development On Demand
Re: svnsync and non-interactive authentication. bug?
Posted by Daniel Rall <dl...@collab.net>.
On Fri, 16 Feb 2007, Garrett Rooney wrote:
> On 2/16/07, C. Michael Pilato <cm...@collab.net> wrote:
> >While documenting svnsync in The Book, a question occurred to me. I see
> >that svnsync has --username and --password command-line options, as well as
> >the related --no-auth-cache and --non-interactive options.
> >
> >But what happens if both the source and target repositories of your
> >mirroring process require authentication, and as different users? Are you
> >just out of luck in terms of getting non-interactive functionality? Is
> >this
> >a shortcoming of svnsync, or am I overlooking something?
> >
> >(I admit it -- I haven't tested this to see what happens. It's 4am, and
> >I'm
> >going cross-eyed, so this mail is as much a TODO that will wind up back in
> >my Inbox as it is a question for someone else to answer.)
>
> If I had to guess it won't work, but like you I haven't actually tried it.
Sounds like a bug to me. We'd need a second set of
--username/--password command-line options to correct this problem.
Re: svnsync and non-interactive authentication. bug?
Posted by Garrett Rooney <ro...@electricjellyfish.net>.
On 2/16/07, C. Michael Pilato <cm...@collab.net> wrote:
> While documenting svnsync in The Book, a question occurred to me. I see
> that svnsync has --username and --password command-line options, as well as
> the related --no-auth-cache and --non-interactive options.
>
> But what happens if both the source and target repositories of your
> mirroring process require authentication, and as different users? Are you
> just out of luck in terms of getting non-interactive functionality? Is this
> a shortcoming of svnsync, or am I overlooking something?
>
> (I admit it -- I haven't tested this to see what happens. It's 4am, and I'm
> going cross-eyed, so this mail is as much a TODO that will wind up back in
> my Inbox as it is a question for someone else to answer.)
If I had to guess it won't work, but like you I haven't actually tried it.
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org