On 09/28/2015 08:22 AM, Nio Wiklund wrote:
Den 2015-09-28 kl. 13:56, skrev Marc Deslauriers:
On 2015-09-28 07:48 AM, Nio Wiklund wrote:
Den 2015-09-21 kl. 18:03, skrev Nicholas Skaggs:
On 09/19/2015 07:29 AM, Marc Deslauriers wrote:
On 2015-09-19 07:04 AM, Nio Wiklund wrote:
Hi Marc,
My standard ISP's mail server is down. So I tried to send this mail
via Google's
web mail interface. In my saved version of the mail, there is an
attachment
"SCC-test.ods", size 17692 bytes. I'll try again ...
Best regards
Nio
That's a very interesting spreadsheet, thanks for doing that. It seems
most
issues are:
- A known bug in udisks that prevents the erase function from working
reliably
(LP: #1460602)
- A mismatch between syslinux versions between the host system
creating the usb
key and the version being installed (LP: #1325801)
- A limitation of the fix for the syslinux issue that prevents amd64
images from
being created on i386 systems (LP: #1446646)
I believe a simple way to eliminate every one of these issues at once
in UDC
would be to get rid of the persistence functionality. All of the
currently
supported images can now be written directly to a usb device as-is.
The syslinux and boot loader mangling that is currently required for
persistence
and used to be required for older end of life releases is a constant
source of
problems each time a new release come out and is difficult to get working
properly on mismatched architectures.
Removing persistence would also simplify the user interface and would
remove the
need of having the "Erase Disk" button, therefore eliminating the
known udisks
issue as a side effect.
Advanced users who require generating a usb key with persistence can
simply use
one of the other tools that are available. Perhaps in the future
persistence
could be added back by creating a partition in the free space left
over after
copying over the image.
Changing UDC to remove persistence and to copy images as-is is a
trivial change,
and I am willing to volunteer to do it.
Marc.
Marc, this sounds like a great way to accomplishes two things at once.
It both simplifies the application for users, while making the process
more robust. I'm in support.
Nicholas
Hi Nicholas and Marc,
I have added two test cases to the spreadsheet, now updated to
'SDC-test1.ods'. You find it via the following link
http://ubuntuforums.org/showthread.php?t=2291946&page=2&p=13358864#post13358864
Would it be possible to remove persistence and to copy images as-is
already in Wily? Or will Wily be released with the most acute bugfix but
with some remaining serious bugs in the SDC, or maybe even without any
bugfix, if the limited time will only allow for more important
bug-fixes? Anyway, I can volunteer to help testing, what you develop, Marc.
I have assumed that we are talking about 'removing persistence' in the
next release, 16.04 LTS, because there must be a formal decision. Is it
like this, Nicholas, or can we 'save the SDC is Wily' with Marc's
suggestion?
Hi Nio,
I'm currently working on SDC in my free time. You can track progress here:
https://code.launchpad.net/~mdeslaur/usb-creator/imaging-rework
I still have a few things to fix, and then I'll build test packages to try out.
It's likely too late to have any rework in Wily, as we're already past feature
freeze and UI freeze. That being said, I'd definitely like it to get SRUed into
all releases when I'm done.
Marc.
Hi Marc,
That's great news, that you have already started :-)
I did not know how it works. Please send a mail directly to me, when you
have something to test, and I will test it and give you feedback as soon
as possible.
It is as I feared, too late for the Wily release. But the next LTS will
be the important one, and SRU should be accepted. I'm thinking of the
previous LTS versions. to provide a reliable built-in tool to install
the next [LTS] version.
Best regards
Nio
I'm sure Marc will have everyone on this thread give feedback as soon as
something is ready :-) As Marc said, since this will look to be SRU'd
(which will also need everyone's help with getting accepted!), all the
stable versions of ubuntu will benefit. So we needn't think of being
"late" for wily as a large issue. The next LTS as well as the previous
will get the update.
Thanks again everyone. I can't wait to see the results of this effort.
Nicholas
--
Ubuntu-quality mailing list
[email protected]
Modify settings or unsubscribe at:
https://lists.ubuntu.com/mailman/listinfo/ubuntu-quality
