[virt-tools-list] [PATCH 4/6] IPv6 support: update the create network wizard

[Gene Czarcinski]

On 03/28/2013 12:55 PM, Cole Robinson wrote:
> On 03/27/2013 05:37 PM, Gene Czarcinski wrote:
>> On 03/26/2013 06:51 PM, Cole Robinson wrote:
>>> On 03/26/2013 07:05 AM, Gene Czarcinski wrote:
>>>> This update includes changes to both the glade-3 ui and
>>>> the releated python code (createnet.py).  There have
>>>> been some significant changes made to both files for
>>>> IPv6 support and some additional capabilities.
>>>>
>>>> This update includes the fix so color will work in the
>>>> network creation wizard.
>>>>
>>>> The objective was to get the functionality implemented and
>>>> some future effort may be needed to "improve" the
>>>> python code.
>>>> Signed-off-by: Gene Czarcinski <gene at czarc.net>
>>>> ---
>>>>    ui/vmm-create-net.ui     | 1518
>>>> ++++++++++++++++++++++++++++++++--------------
>>>>    virtManager/createnet.py |  612 +++++++++++++++----
>>>>    2 files changed, 1583 insertions(+), 547 deletions(-)
>>>>
>>> A couple bits here:
>>>
>>> - The color of the createnet header changed: it used to be black.
>> What's wrong with blue? ... OK, it is back to black again.
> Blue is the color of the other wizards, but the createnet wizard needs to be
> really redesigned to have a more consistent look before switching the color.

I have tried to be consistent but that does not mean that I am.  If you 
can point me to some good examples and/or guidelines, I will give it a shot.

For now, I have completed redoing all of the gui and wizard stuff 
incorporating your comments.  Just before wrpaping it up and packaging 
the patches for submit, I did one more test: create, start, stop, and 
then delete/undefine a network ... everything worked fine until the 
delete ... the network was deleted but I got an exception when 
net_select() tried to access the just deleted network???  This happens 
with some net-definitions but not with others??

>
>>> - There's horizontal separators added everywhere, they are spaced weirdly and
>>> don't add anything. Please remove.
>> Just following what was there already.  Trying to get things to look good and
>> not spread out.  glade-3 is helpful at times and a real PITA at other times.
> Hmm, I didn't think there were pre-existing separators, sorry.

There were some ... most have been deleted.  There are still some that 
make sense (at least to me).

>
>>> - Many vbox, like on the 'Name' screen, have a ton of empty entries now. This
>>> doesn't affect the visible UI but is clutter when working in glade. Please
>>> shrink them down to only the required sizes.
>> OK, I got rid of the extras here and elsewhere.
>>> - Check the spacing here:
>>> http://fedorapeople.org/~crobinso/virt-manager/create-net-spacing.png  . The
>>> pre-existing elements don't expand, the added bits do. Also the DNS name field
>>> extends across the whole UI. It should probably only be 20 chars long max.
>> Twenty characters sounds reasonable but in real life it can be longer.  I am
>> not sure how to constrain the gtkentry to just 20 characters or to limit the
>> display from consuming the entire line. I will keep "testing" to see if I can
>> find a way.
> Text entries have a 'width in chars' field which can be in glade. Even if it's
> capped at 20, a user can still enter longer text, it will just scroll to the
> right. 20 might be too small, so feel free to play with it.

Did not know that.  Thank you.  Somehow I got the gtkentry boxes to a 
small, reasonable size without using this feature.  I will need to go 
back and make sure they are setup correctly.


>
>>> Glade can be tricky for some of the subtle things, so make a best effort and
>>> if there are still some things you can't get right I'll fix them up before
>>> committing.
>> I would really like to know/understand how to do some of this with glade-3 and
>> the gui.  Even with google being your friend, available documentation is
>> limited or does not address what I am trying to do.  The last thing I want to
>> do is to dive into the implementing glade/gtk/gdk code to see what can be done
>> and how to do it.
> Yes this stuff is tricky. Here's the jist
>
> - For that particular case, you likely need to play with the horizontal and
> vertical fill/expand settings for every element in the table. For example on
> that forwarding page, try and make new elements use the same set of
> expand/fill values. Generally you don't want anything to vertically expand.
> Also check xalign and yalign

Indeed ... that has been most of today and a bit of yesterday.

>
> - For getting spacing right and for everything to fit nicely, really just
> focus on spacing of containers like boxes, tables, etc. The main content of a
> page should be a vbox, every chunk of UI will probably be its own box _within_
> that top level vbox. If you add any boxes or tables, make sure their 'spacing'
> value is consistent with the rest of the layout, usually 3 pixels for table
> column/row spacing, 6 for hbox/vbox, and 12 for the global vbox.
>
> - If you want to indent an entire chunk of UI like a table, don't tweak
> spacing/alignment on individual elements, stick the whole thing in an
> 'alignment' widget, and use that to indent it all with one setting.
>
> - Sometimes there doesn't seem to be any way to get an element to NOT expand,
> like a small combo box in a table with a longer widget. You can shrink it, by
> sticking the widget in a 2 item box, set it to not expand, stick an
> 'alignment' in the other entry, and have _that_ expand.
>
> But as I said, don't drive yourself crazy with it, just make a best attempt
> and I can try and fix up some of the weird bits as an add on patch. It can be
> quite frustrating getting this stuff right.

While I might have not done things the "right way" given your comments 
above (and  I will go back and check what I have done), most of it 
"looks good" at least to me.  There are a few things you might not be 
aware of such as the Route-Via and also that a valid definition can have 
neither IPv4 or IPv6 and thus both are hidden.

Pending "issues" --

1. The bug I found when some networks are deleted (this is a problem 
somewhere in host.py).

2. It is not clear what the settings are for isolated versus routed.  I 
believe I understand it for IPv6 but need to verify some stuff for 
IPv4.  When no addresses are specified, then it is truly isolated.  But, 
if an address is specified but no <forward> is in the definition, what 
does than mean?  At the very least it means that dnsmasq will be started 
as a DNS on that network.

3.  Under 0.9.4, the "Hint" message/label under the IP address entry-box 
was fixed in place.  Under the gtk3.2 branch is moves if you stretch the 
window.  This may be a feature an

d not correctable.

4.  I really wanted to put the Domain Name entry on the Name Page along 
with the Network (interface) Name entry.  IMO, the Network Name entry 
MUST have the focus when the page comes up.  With the second entry, I 
could not get this to happen ... the focus was always the Domain Name 
... at least the first time the wizard was run during a wirt-manager 
execution.  Additional running of the wizard resulted in the focus being 
whatever was the focus the last time the wizard was run.  Since this 
"hurt", my solution was to move it to what is now the Miscellaneous 
Settings Page (formally Forwarding).

Although I am much interested in getting the IPv6 support into 
virt-manager, this has also been a learning experience for me with 
respect to python and glade-3 (the last time I ran glade was when it was 
just plain, old glade).  So far I have been positively impressed with both.

Gene
>
> Thanks,
> Cole
>
>>> Please squash this patch together with patch #1, the UI changes aren't much
>>> useful in isolation.
>>>
>>> Also, for testing UI it's quite useful to use the libvirt test driver if you
>>> haven't:
>>>
>>> ./virt-manager --connect test:///`pwd`/tests/testdriver.xml
>>>
>>> The UI has a couple quirks:
>>> - There's a separator between the IPv4 config and IPv6 config, I'd remove it.
>>> - Please hide the entire IPv6 frame if the network doesn't have any IPv6 bits.
>>> This is similar to how the interface details panel works. The testdriver
>>> should help show all the interface details bits
>>> - If you have a suite of networks you are using to test the UI, please also
>>> add them to tests/testdriver.xml so I can use them as well. It's as simple as
>>> just copy+pasting the entire <network> XML into that file, in the same section
>>> as the rest of the <network> instances.
>>>
>>> Thanks,
>>> Cole
>>>
>> I have all the changes done but I want to try an polish this a bit so it will
>> be at least a day before I resubmit.
>>
>> BTW, I figured out host to run the tests ... it was obvious once I understood
>> what was happening.
>>
>> Gene
>>
>