| Option to _not_ cache credentials in the keyring? [message #21801] |
Tue, 14 August 2007 00:20  |
Eclipse User |
|
|
|
Originally posted by: awsaba.gmail.com
Hello,
We have been looking at using Mylyn at work, but the fact that there is
no option to _not_ cache credentials to disk and prompt when needed
(like the CVS plugin) is a real problem for us.
Since I haven't really used Mylyn because of this, can someone say if is
there some fundamental design/workflow reason that it would not be
possible to only cache a password in memory? I spent about 10 minutes
browsing the Trac and base connector sources, and it looks most of the
work to store and read the credentials is done by the Mylyn core, so
this problem would be common even if we were using a different
connector. Is there any objection to creating an enhancement request in
bugzilla for this?
The last time I really checked out Mylar (probably 6-12 months ago), the
only discussion I could find had to do with configuring a repo with
anonymous access for reading, but then prompting for creds when making
changes. I think it was on the old mylar newsgroup, which no longer
appears to be available, so I don't know how/if that was ever resolved.
-Tony
|
|
|
| Re: Option to _not_ cache credentials in the keyring? [message #22387 is a reply to message #21801] |
Fri, 17 August 2007 16:07  |
| Eclipse User |
|
|
|
Originally posted by: rob.elves.eclipse.org
This is a reasonable request so yes, please create a new bug report and
we can discuss design considerations there.
-Rob
Antony Saba wrote:
> Hello,
>
> We have been looking at using Mylyn at work, but the fact that there is
> no option to _not_ cache credentials to disk and prompt when needed
> (like the CVS plugin) is a real problem for us.
>
> Since I haven't really used Mylyn because of this, can someone say if is
> there some fundamental design/workflow reason that it would not be
> possible to only cache a password in memory? I spent about 10 minutes
> browsing the Trac and base connector sources, and it looks most of the
> work to store and read the credentials is done by the Mylyn core, so
> this problem would be common even if we were using a different
> connector. Is there any objection to creating an enhancement request in
> bugzilla for this?
>
> The last time I really checked out Mylar (probably 6-12 months ago), the
> only discussion I could find had to do with configuring a repo with
> anonymous access for reading, but then prompting for creds when making
> changes. I think it was on the old mylar newsgroup, which no longer
> appears to be available, so I don't know how/if that was ever resolved.
>
> -Tony
|
|
|
| Re: Option to _not_ cache credentials in the keyring? [message #23310 is a reply to message #22387] |
Mon, 20 August 2007 23:18 |
| Eclipse User |
|
|
|
Originally posted by: awsaba.gmail.com
Robert Elves wrote:
> This is a reasonable request so yes, please create a new bug report and
> we can discuss design considerations there.
> -Rob
>
Done, it is bug 200634.
https://bugs.eclipse.org/bugs/show_bug.cgi?id=200634
>
> Antony Saba wrote:
>> Hello,
>>
>> We have been looking at using Mylyn at work, but the fact that there
>> is no option to _not_ cache credentials to disk and prompt when needed
>> (like the CVS plugin) is a real problem for us.
>>
>> Since I haven't really used Mylyn because of this, can someone say if
>> is there some fundamental design/workflow reason that it would not be
>> possible to only cache a password in memory? I spent about 10 minutes
>> browsing the Trac and base connector sources, and it looks most of the
>> work to store and read the credentials is done by the Mylyn core, so
>> this problem would be common even if we were using a different
>> connector. Is there any objection to creating an enhancement request
>> in bugzilla for this?
<snip>
-Tony
|
|
|
| Re: Option to _not_ cache credentials in the keyring? [message #577326 is a reply to message #21801] |
Fri, 17 August 2007 16:07 |
| Eclipse User |
|
|
|
This is a reasonable request so yes, please create a new bug report and
we can discuss design considerations there.
-Rob
Antony Saba wrote:
> Hello,
>
> We have been looking at using Mylyn at work, but the fact that there is
> no option to _not_ cache credentials to disk and prompt when needed
> (like the CVS plugin) is a real problem for us.
>
> Since I haven't really used Mylyn because of this, can someone say if is
> there some fundamental design/workflow reason that it would not be
> possible to only cache a password in memory? I spent about 10 minutes
> browsing the Trac and base connector sources, and it looks most of the
> work to store and read the credentials is done by the Mylyn core, so
> this problem would be common even if we were using a different
> connector. Is there any objection to creating an enhancement request in
> bugzilla for this?
>
> The last time I really checked out Mylar (probably 6-12 months ago), the
> only discussion I could find had to do with configuring a repo with
> anonymous access for reading, but then prompting for creds when making
> changes. I think it was on the old mylar newsgroup, which no longer
> appears to be available, so I don't know how/if that was ever resolved.
>
> -Tony
|
|
|
| Re: Option to _not_ cache credentials in the keyring? [message #578009 is a reply to message #22387] |
Mon, 20 August 2007 23:18 |
| Eclipse User |
|
|
|
Robert Elves wrote:
> This is a reasonable request so yes, please create a new bug report and
> we can discuss design considerations there.
> -Rob
>
Done, it is bug 200634.
https://bugs.eclipse.org/bugs/show_bug.cgi?id=200634
>
> Antony Saba wrote:
>> Hello,
>>
>> We have been looking at using Mylyn at work, but the fact that there
>> is no option to _not_ cache credentials to disk and prompt when needed
>> (like the CVS plugin) is a real problem for us.
>>
>> Since I haven't really used Mylyn because of this, can someone say if
>> is there some fundamental design/workflow reason that it would not be
>> possible to only cache a password in memory? I spent about 10 minutes
>> browsing the Trac and base connector sources, and it looks most of the
>> work to store and read the credentials is done by the Mylyn core, so
>> this problem would be common even if we were using a different
>> connector. Is there any objection to creating an enhancement request
>> in bugzilla for this?
<snip>
-Tony
|
|
|
Powered by
FUDForum. Page generated in 0.07317 seconds
Home