Zhang Yunjun

After the above checking is done. The implementation of `timeseriesApp` should be simple, take a look at the `TimeSeriesAnalysis` class at [L206](https://github.com/insarlab/MintPy/blob/master/mintpy/smallbaselineApp.py#L206). The new application could call this TimeSeriesAnalysis object...

Hi @falkamelung, all GPS related features should have major changes in the coming version. I will keep this feature in mind and check the detailed implementation strategy when we started...

That's good to know, @falkamelung. Is there a publicly available implementation of MIDAS code for reference?

Thank you @ehavazli.

Using `lscpu` sounds fine with me, since it only affect the translation of `all`. Of course, we should explain it well in the comments of the function in `cluster.py` and...

Hi @falkamelung and @Ovec8hkin, I have implemented `num_core = os.cpu_count() / num_cpu_threads` as the translation for `numWorker = all`, as Falk suggested. Cheers.

Thank you Falk and Eric for the info. This means we still have not located the cause yet. The "conflict during multiple HDF5 writing processes" as described in #692 still...

All the other errors, e.g. `no access to file`, `negative dimension size`, have been fixed by the PR above. The remaining ​re-occurring error is `OSError: Timed out during handshake while...

Hi @andretheronsa, I agree that better handling of the error message is desired. I implemented it to not raise Error message in order to support the time-series with the latest...

We could add this in [tropo_pyaps3.py#L544](https://github.com/insarlab/MintPy/blob/83a2fb187699879dc208ab6ef3fa2f11ef274e1c/mintpy/tropo_pyaps3.py#L544). It should at least address one of the two FAQs. Could you issue a PR for it @pbrotoisworo?