fix: wrong transcription order

[ltorje-8x8]

The receiving timestamp will be sent to skynet to be used as absolute timestamp for the audio chunk - the network lag is presumed to be less than tens of milliseconds (a latency of 20-50 ms is expected and it is considered to be ok as loss of time).

[codecov]

Codecov Report

Attention: 7 lines in your changes are missing coverage. Please review.

Comparison is base (86273ae) 20.86% compared to head (37b3e33) 21.85%.
Report is 6 commits behind head on master.

Additional details and impacted files

@@             Coverage Diff              @@
##             master     #520      +/-   ##
============================================
+ Coverage     20.86%   21.85%   +0.99%     
- Complexity      289      309      +20     
============================================
  Files            74       76       +2     
  Lines          6222     6269      +47     
  Branches        833      838       +5     
============================================
+ Hits           1298     1370      +72     
+ Misses         4699     4671      -28     
- Partials        225      228       +3     

Files	Coverage Δ
...va/org/jitsi/jigasi/transcription/Transcriber.java	0.00% <ø> (ø)
...g/jitsi/jigasi/transcription/WhisperWebsocket.java	28.39% <46.15%> (+28.39%)	⬆️

... and 8 files with indirect coverage changes

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0a047ff...37b3e33. Read the comment docs.

[damencho]

Please fix the imports to be star imports.
Please avoid making style changes in the same commit with some fixes as it make it impossible for review, so can you isolate the fix in separate commit?

About the style of the java code all existing code follow
https://desktop.jitsi.org/Documentation/CodeConvention.html
And we tried to enforce it with the checkstyle.