Forgot to say… I've submitted this for inclusion in NeuroDebian. Hopefully it (fslview 4.0.1) will be available in updates shortly.
To non-Ubuntu users: the port to Qt4 has not been straightforward and there are still significant issues... especially on Mac where the port is badly broken. This is why Ubuntu/Debian is the only platform for which we are currently distributing version 4.x. Non Ubuntu users will continue to get the v3 stream and, until we can get v4 working satisfactorily cross-platform, any new features will continue to be developed for both versions.
Dave
On 4 Mar 2013, at 15:01, David Flitney <[log in to unmask]> wrote:
> On 3 Mar 2013, at 16:20, Jiří Keller, M.D. <[log in to unmask]> wrote:
>
>> Hi all,
>> I upgraded to fslview 4.0.0 (beta) on my boxes and I have to report
>> few annoying bugs and regressions (maybe some are already reported
>> …):
>
> Could you confirm the platform... Ubuntu?
>
>> 1) after launching fslview on fullscreen widow with ortographic view
>> is not fullscreen as it was in fslview 3. Subwindow witdth is 980 px.
>> Workaround: clicking to maximize button. Bug level: annoying
>> 2) after resizing, minimal width of ortographic view increases (!) to
>> 1070 px, which means that two ortogonal views do not fit even on
>> full-HD screen ! On previous versions, I was able to use fslview on my
>> n900 or EEE 701 with 800x600 resolution without problem. Workaround:
>> use fslview3 or buy a new CRT ;-). Bug level: blocker (at least for
>> me)
>
> Hiding the AtlasDock and giving better size hints seems to have rectified these two issues, at least for me.
>
>> 3) atlas tools shown by default, which only occupies space as majority
>> of my images are not in MNI space. Workaround: switch off manually.
>> Bug level: slightly annoying
>
> Now hidden by default. Introduces a new minor bug in that turning on the atlas tools doesn't initially generate the correct resize event (I have no idea why).
>
>> 4) gray crosshair is way less visible than previous green one.
>> Workaround: none. Bug level: slightly annoying
>
> Made them bright green. Perhaps we'll have this user selectable in a future version.
>
>> 5) bug in multi-volume image (zdim>1): in the beginning, volume 0 is
>> in all 3 planes. After selecting z=1, nothing happens. After clicking
>> into one plane, in other two, volume 1 is selected correctly, in this
>> window volume z=0 remains. After clicking in any other plain, z=1 is
>> in all three planes. Same with going back to z=0. But it is even
>> worse: if you restart fslview and change volume z=0->z=1, nothing
>> happens, but after going back to z=0 it shows z=1 in all three planes
>> ! And z=0 for z=1. Again, consecutive clicking into two images solves
>> the bug.
>> Workaround: just click in two images ... Bug level: severe (I
>> discovered it in dual-echo T2 sequence, which is easy to distinguish,
>> but …)
>
> Yes, I have a fix which I'll be releasing asap.
>
>> 6) after undocking of Cursor tools/overlay settings, they become
>> completely separate windows (contrary to fslview3), so one has to
>> handle more entities in task-switching. Workaround: not to use this
>> feature or getting use to new behaviour. Bug level: just annoying
>
> I switched to using the QDockWidget. This gave better support cross platform but, it seems, at
>
>> I know that fslteam is doing really good job, but new fslview is
>> really anoying …
>
> ...I find nothing quite so motivating as finding out how really annoying I've been.
>
>> thank you for solving these problems
>> George
>>
>
> --
> Dave Flitney
> Oxford Centre for Functional MRI of the Brain
> E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717
> URL: http://www.fmrib.ox.ac.uk/~flitney
>
--
Dave Flitney
Oxford Centre for Functional MRI of the Brain
E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717
URL: http://www.fmrib.ox.ac.uk/~flitney
|