Bugilla search interface changing soon
Myk Melez
> As Myk announced a week ago
> <http://groups.google.com/groups?selm=3CF80456.6030300%40mozilla.org>,
> bugzilla.mozilla.org is due to be upgraded to Bugzilla 2.16 at June 10
> 0800 UTC.
>
> As I announced last December
> <http://groups.google.com/groups?selm=3C189E24.7CEE92F9%40mailandnews.com>,
> this upgrade will introduce the redesign of Bugzilla's search page.
That announcement was premature, as is this one. When I upgrade
bugzilla.mozilla.org on Monday I will *not* be upgrading to the new
version of the search page. Instead, I'm going to leave the current
version in place.
Then, at some point after the upgrade I'm going to make the new version
available at a different URL to give users a chance to familiarize
themselves with it before I make it the default interface for
bugzilla.mozilla.org.
-myk
David Tenser
Why not forcing the users (developers) to use the new interface
directly? After all, we're application developers and we know how to
adopt new interfaces/methods. Get rid of the old, deprecated interface
right away!
Christian Biesinger
> When I upgrade bugzilla.mozilla.org on Monday I will *not* be upgrading
> to the new version of the search page.
Eh, why not? The new one is _so_ much better looking and easier to use.
The old one, on the other hand, is difficult, intimidating and long.
--
"They that can give up essential liberty to obtain a little temporary
safety deserve neither liberty nor safety."
-- Benjamin Franklin
Gervase Markham
> bugzilla.mozilla.org on Monday I will *not* be upgrading to the new
> version of the search page. Instead, I'm going to leave the current
> version in place.
>
> Then, at some point after the upgrade I'm going to make the new version
> available at a different URL to give users a chance to familiarize
> themselves with it before I make it the default interface for
> bugzilla.mozilla.org.
The upgrade to Bugscape (Netscape's internal Bugzilla) which gave them
the new query page happened just after it was checked in, and when it
still had several bugs. This caused Netscape engineers to become annoyed
with the new interface because:
a) It didn't work properly (e.g. keyword searching was broken), and
b) It was different to b.m.o, so they had to use two interfaces at once.
Both of these reactions were understandable at the time. However, both
of these issues are now resolved, but Myk is still under pressure from
some Netscape engineers to make the old interface available even after
the upgrade. He has written a template to do so, and (as stated above)
plans to make it the b.m.o default even after 10 June. This is going on in
http://bugzilla.mozilla.org/show_bug.cgi?id=137855
I suggest that if you believe the new interface is an improvement on the
old and would like it to be the default, you should mail him politely to
say so. (This obviously does not preclude making the old interface
available.) He can then weigh up the level of support for each version
when making the decision on the day.
Gerv
Myk Melez
> Myk is still under pressure from
> some Netscape engineers to make the old interface available even after
> the upgrade.
No one has pressured me, but some users (from engineering, QA, and
management) have complained about the change. Other users (including
myself) have been eagerly awaiting the new version.
My plan is to balance these concerns by making the new interface
available as soon as possible after the upgrade but leaving the old
interface the default for a period of time to give users time to make
the transition (probably between one and two months, but it could be
sooner if a user preference for the interface version gets implemented).
That's a reasonable approach and treats current users better than the
alternative, which is to force them to learn the new form the moment I
upgrade rather than when they have time for it. For more information on
why this is a better approach, see the discussion in bug 137855.
> I suggest that if you believe the new interface is an improvement on the
> old and would like it to be the default, you should mail him politely to
> say so. (This obviously does not preclude making the old interface
> available.) He can then weigh up the level of support for each version
> when making the decision on the day.
That would be pointless, as I've already decided to retain the old
interface for the reasons stated above and in the bug report.
Setting follow-ups once again to netscape.public.mozilla.webtools.
Those of you who are interested can follow the discussion there. Let's
not continue to spam multiple newsgroups with it.
-myk
Gervase Markham
> available as soon as possible after the upgrade but leaving the old
> interface the default for a period of time to give users time to make
> the transition (probably between one and two months, but it could be
> sooner if a user preference for the interface version gets implemented).
User preferences for interface versions are a bad idea, because they
force you to continue to support n releases of a user interface. mpt
expressed it very well in the bug when he explained why users complain
when an interface is made more usable.
If you plan to do this, for how long are you going to support the old
interface? Indefinitely? If not, you are just giving people more time to
get used to the old interface and so they'll be more annoyed when you
finally remove it.
> That's a reasonable approach and treats current users better than the
> alternative, which is to force them to learn the new form the moment I
> upgrade rather than when they have time for it.
It doesn't treat well the current users who like the new interface but
don't want to have to remember to do something different (click a
different link, edit the URL) in order to use it. My assertion is that
the number of users who like the new interface and want to use it far
outweighs the number who want to stick with the old one - because
everyone agrees the new interface is more usable.
As I've said in the bug, a reasonable approach would be to make the more
usable interface the default, and provide a link in the footer to the
other interface, with a note at the top saying "this interface is going
away in six months - please learn the new one" or however long you choose.
Gerv
Myk Melez
> User preferences for interface versions are a bad idea, because they
> force you to continue to support n releases of a user interface.
Preferences force nothing, especially not supporting "n" (where "n" is
presumably any number), they merely add the option to support more than
one interface.
> mpt
> expressed it very well in the bug when he explained why users complain
> when an interface is made more usable.
mpt's dismissive attitude towards user complaints in the face of clearly
superior design is a mistake not only because user complaints are a
priori relevant but also because interface designs are to some extent
arbitrary and trade off some forms of usability for others.
> If you plan to do this, for how long are you going to support the old
> interface? Indefinitely?
No, I plan to support the old interface for some time, probably between
one and two months.
> If not, you are just giving people more time to
> get used to the old interface and so they'll be more annoyed when you
> finally remove it.
That's just silly; a few more months won't have an appreciable effect on
how used to the old interface people are.
> It doesn't treat well the current users who like the new interface but
> don't want to have to remember to do something different (click a
> different link, edit the URL) in order to use it.
The upgrade tonight is going to improve the situation for fans of the
new interface (because it will make it available to them), pave the way
for further improvements for those fans (by taking the first step
towards making the new interface the default), and not screw fans of the
old interface or those who could care less, just want to get their work
done, and don't want to have to learn a new interface right now. What
you want will happen, just not as fast as you like. Hold your horses.
> My assertion is that
> the number of users who like the new interface and want to use it far
> outweighs the number who want to stick with the old one - because
> everyone agrees the new interface is more usable.
Actually, no, not everyone agrees the new interface is more usable, but
even if they did, they still have to learn how to use it, and they
should have time to do so.
-myk
Jehan
>> interface and so they'll be more annoyed when you finally remove it.
>
>
> That's just silly; a few more months won't have an appreciable effect on
> how used to the old interface people are.
Except that if the old one is the default, new users will learn the old
interface.
And for people who prefers the old one, they won't have any incentive to
use the new one: it will be easier for them to use the old than to use
the new one (one click farther or new bookmark to set).
If the idea in the long run is to use only the new interface, you should
make the old interface less reachable than the new one and not the other
way around.
Moreover, my guess is that with the release of Mozilla 1.0, you'll get a
bunch of newbies who are not technical users and stayed away from the
beta releases. And those users are the more likely to get confused by
the change of interface.
Jehan
Gervase Markham
> presumably any number), they merely add the option to support more than
> one interface.
If you provide e.g. three interfaces to the same CGI, and then want to
change how things work internally, you have to change all three. In
addition, if any one of them breaks for some reason, you have to fix it.
You also need to determine, when problems are reported, which interface
the reporter is using.
All of these are an additional burden on programmers and support staff.
> mpt's dismissive attitude towards user complaints in the face of clearly
> superior design is a mistake not only because user complaints are a
> priori relevant but also because interface designs are to some extent
> arbitrary and trade off some forms of usability for others.
User complaints are relevant; however, most complaints about the new
design came from people exposed to a version which had bugs in (because
Netscape decided to upgrade their internal bug tracker to the Bugzilla tip.)
Yes, UI design is a trade-off. From a usability perspective, what do you
think is worse about the new design?
Gerv
Lisa Chiang
> User complaints are relevant; however, most complaints about the new
> design came from people exposed to a version which had bugs in
> (because Netscape decided to upgrade their internal bug tracker to the
> Bugzilla tip.)
>
> Yes, UI design is a trade-off. From a usability perspective, what do
> you think is worse about the new design?
>
> Gerv
>
I think that the user complaints are relevant, even from users who used
the earlier version of the new query screen from Netscape's internal bug
database (which got upgraded early to the changes). In looking at
Bugzilla's new query page, there are only two UI visible changes that I
see between Bugzilla and Netscape's internal bug database:
1. The keyword field was added. (I had filed the bug originally when
this field was missing which ended up being fixed in the bugzilla codebase).
2. The default for the "email and numbering" section displays
"contains" instead of "is".
I don't believe that it can be dismissed that the user complaints are
due to users being exposed to the query screen before all the bugs were
fixed. In trying to use the new interface to do our jobs, we provided
feedback to Myk and filed bugs in the proper webtools components as
appropriate.
Lisa