While we a absolutely commited to delivering the high quality applications it is a known fact that "there always is one more bug". Especially releases which refrect a moving target or unfinished snapshot, there can be some problems which the users (or beta testers) should be aware of.
All listed entries are sorted by severity and group by topic as far as possible. A bug reference number is provided in case you want to provide additional information to one of the items and as far as
Some problems are so seriouse that they might lead to a programm crash which in the worst case can result in corrupted or lost data. Be sure to keep them in mind.
The following problems are known but not dangerous anh mainly focus around ugly user interface behavior, complicated program handling or limited usability.
-
Wrong Cachefile permissions. (MB000010)
The temporary folders and cache files have the wrong permssions. We have to set them more explicitly so that we can predict the behavior. Currently we use the default permissions which will cause problems with the /tmp/MassBoy folder.
Workaround: none
-
Zoom buttons not validated. (MB000010)
We should disable the zoom buttons if a certain zoom action makes no sense. This include disabling the zoom panel and its "Apply" button (zoom panel handling might be the most tricky candidate in this case).
-
Zoom to selction has no autoscroll. (MB000010)
While autoscroll behavior is already inside the current implementation it is not yet activate since the sync with the scroller and upper views is not available yet. This means that current the zoom to selection can only perform "zoom in" operation and not "zoom out". The final release should support both.
-
Zoom popup command keys don't work. (MB000010)
The command keys only work with the new main menu items. Additionally command keys won't work if they are only reachable with a CMD-ALT-[key] combination. Remapping the keyboard can fix this (both problems are NeXTs bug. Ref. Nr.: [107879])
-
Printing not perfect. (MB000010)
The code currently does not remember the default paper settings. By default we will use landscape printing...this is currently hardwired. If NSPageLayout won't fix this problem in 4.2 we should do our own default handling.
Additionally we do not really take care of the printable area. We use a default margin which might not be correct depending on the available printer.
-
Dumb redraws. (MB000010)
We often redraw the entire views which makes redraws slower then necessary. E.g. setting the selection only requires a redraw in the selection lines area and not the entire view. With large documents this can be nasty.
-
Y-percent scale shows too few ticks. (MB000010)
The ticks on the vertical axis do not get drawn at this time. This only gives you the two extreme values 0 and 100. Future versions will included a smarter y scale (similar to the current x axis) which will automatically show the most useful ticks
-
Scale view ticks not perfect. (MB000010)
The tick lines of the scale view do not really reach up to the axis base line...we need to fix that at some point.
-
Scans numbered "0, ... , n-1" not "1, ... , n". (MB000010)
We currently use the cumper index numbering and not the more common mathematical style of scan numbers. A future version should fix this at least for the UI so that numbers get adjusted for graphical output.
-
Time scale not displayed immediately. (MB000010)
When you activate the time information from the Preferences panel it will not show up right away. You have to scroll, zoom or force any other redraw of the chromatogram window in order to get them displayed. This will be fixed automatically in future releases as we switch to real "per window" settings and a working inspector panel.
-
Spectra List Panel does not resort columns. (MB000010)
The original idea was to have the table resorted based on the first column...this is not implemented yet but should be availabel soon.
-
Spectra List Panel does not obey threshhold. (MB000010)
Even peaks with very low intensities are displayed.
-
Spectra Panel needs better "empty" view. (MB000010)
One startup the view looks ugly...we need some nicer displa behavior when no scan is selected.
-
Spectra Panel shows wrong results. (MB000010)
We currently snap peaks to integers but don't add same values. This will cause wrong displays as same masses will not be visible. With 0.3 the entrie peak handling will be updated and this problem will just go away.
-
Spectra Panel has strange scale. (MB000010)
The spectra panel uses a scale which starts at zero and not at the lowest mass scanned. This will be fixed in a future release and is currently solved by presetting the visible range to the real data range which has been recorded.
-
Ugly keyboard UI does not work. (MB000010)
We need to support the regular TAB, DEL and other keys for keyboard control. Currently we only use a dirty hack.
-
Filehandling a little rough. (MB000011)
There are many places where we store file which could stand some extended exception and read/write success checks then we have in place up to now. This affects primerily the library handling.
-
Search and Save Lib are identical
Currently the default search library is identical to the library where a user did sove his spectra to. This needs to be seperated in the future.
-
User can switch to unwriteable library
When trying to add a spectruma user has the chance to switch the library. This however does not prevent him from switching to a library which he can not write to.
-
Adding spectrum with existing name will overwrite
If a spectrum is saved to a library and a name conflict occures it is possible for an admin to ignor the worning an save the file anyway. This currently will replace the existing file instead of adding a different representation of the same spectrum.
-
Chemcal elemennt table incomplete (MB000020)
Currently only a very limited set of atoms is known to the system. This is mainly because we have no full periodic system right here on the beach...where version 0.5 did evolve.
-
Inspectors holds document reference too long. (MB000010)
The Inspector panel retains the document for too long. This causes a delayed free or "no" free at all. We need to look into that. (E.g. post a "focusChanged" notification on resigning main status?)
-
Windows cascade strangely. (MB000010)
While the cascading of opened windows works it openes the first window at a strange position.
If you had a problem which is not yet mentioned here or if you can provide further details which would make it easier to track down a specific problem we would be very glad to hear from you!
Until we get our interactive bug report application up and running you can simply sent us an email.