Targetlynx execute "keep manual changes" not working?

per_oxid

If I wish to update the RT for one compound but keep the manual changes made for the other compound integrations then I
1) save the *qld file
2) Go to Processing->Execute and select "Integrate" and "Keep Manual Changes".

This keeps some of my manual changes but also removes a lot of my manual changes!

I really don understand why targetlynx removes my manual integrations in such a random manner when I have selected "Keep Manual Changes". It does keep some manual changes but remove most of them. When you have spent hours controlling and manually correcting a large dataset it is very frustrating that you can not alter integration method for one compound without losing all the manual corrections made for the other compounds.

per_oxid

I also frequently get crash
"Quantify MFC Application has stopped"
Processing on Waters workstation using Masslynx 4.2 SCN1045
Very frustrating when I have 40 000 MRMs to review

per_oxid

The issue with MFC crash occurs when I integrate all 200 samples and 90 MRMs in targetlynx (18 000 chromatograms).  If I select a lower nr of samples the crash does not occur. The loss of manually integrated peaks despite selecting keep manual changes seems to be related to peak detection/compound detection. Even manually integrated peaks are removed if they fall outside the compound identification parameters in the method.

I would like to see the following changes in Masslynx/Targetlynx SCN 1045

1)
I would like an option to run the integration method for just one compound at the time without reprocessing all the compounds. If I have reviewed 90 compound and made manual changes this is my final result, i don't wish any change to these reviewd/corrected compounds. But I may still want to change integration method for a compound nr 91 during the review and reprocess this one compound with an integration method. 

2)
If I select keep manual changes from the execute command I wish that the integration method maintain all of my manual changes.

3) 
I wish to be able to use a fixed retention time for compound identification, set a start an stop time for compound identification integration and that integration is total signal with baseline at zero between these two time points. This is useful for compunds where some samples are low and where a baseline is not perfect (often the case for very narrow MRM windows). Fixed retention time is available in SCN1045 but the baseline detection does not work for me I wish it to go down to zero.

4)
Possibility to build MRM method in excel as table with variables (with only one variable per column) that can be pasted directly into an acquisition method (experiment file *.EXP).
(MRM can be imported from excel using QunaPedia but I cannot include start and stop time for MRM acquisition in the import. Skyline is a very good option but the export of MRM windows from skyline to *exp file does not work for integration start/stop and polarity )