OMNeT++/OMNEST Bug Tracker - OMNeT++
View Issue Details
0001046OMNeT++runtime / Qtenvpublic2018-07-26 05:392018-08-01 17:08
ManiAm 
 
normalminoralways
newopen 
 
 
0001046: sort event pane by columns
[feature] ability to sort the events by clicking on each column.
No tags attached.
png 2018-07-25_22-34-44.png (241,768) 2018-07-26 05:39
https://dev.omnetpp.org/bugs/file_download.php?file_id=213&type=bug
png
Issue History
2018-07-26 05:39ManiAmNew Issue
2018-07-26 05:39ManiAmFile Added: 2018-07-25_22-34-44.png
2018-07-31 11:43attilaNote Added: 0001366
2018-08-01 17:08ManiAmNote Added: 0001367

Notes
(0001366)
attila   
2018-07-31 11:43   
Could you please elaborate a bit on what exactly you would use this for?
Is it possible that some filtering would be enough instead?

I'm asking because the current implementation is heavily geared towards scalability in the number of log entries in the buffer. For this, and some other reasons, adding something like sorting to this inspector would require significant effort. Filtering functionality is already in place, so extending that to be more useful should be less painful.

On top of this, once the order of the entries can be different than in which the events occurred, the display might become confusing to the user - for example if the sorting is stored across executions, but the user forgets that they changed it sometime in the past, and now wonders why the entries are all mixed up.
(0001367)
ManiAm   
2018-08-01 17:08   
>> Could you please elaborate a bit on what exactly you would use this for?

To group related entries together. For example, if I am interested in OSPF hello messages, then I click the 'Name' to group all hello message exchanges between all nodes. Can we do this using filtering? I don't think so.

>> but the user forgets that they changed it sometime in the past ...

The Qtenv can reset this feature to default upon each execution. If the user needs to sort entries based on a column, then he needs to activate it again.

>> I'm asking because the current implementation is heavily geared towards scalability in the number of log entries in the buffer.

I understand. This is a (nice to have) feature, but if it hurts the performance under the hood, then you can just ignore it :)