[query/service] make available driver progress information

[danking]

It now looks like this:

In [2]: import hail as hl
   ...: t = hl.utils.range_table(100)
   ...: t._force_count()
   ...: t.show()
Submitted batch 1774275, see https://batch.hail.is/batches/1774275/jobs/1
Batch 1774275: completed jobs: 100%|████████████████████████████████████████████████████████████████████| 1/1 [00:14<00:00, 14.76s/it]
Submitted batch 1774315, see https://batch.hail.is/batches/1774315/jobs/1
Batch 1774315: completed jobs: 100%|████████████████████████████████████████████████████████████████████| 1/1 [01:03<00:00, 63.06s/it]
+-------+
|   idx |
+-------+
| int32 |
+-------+
|     0 |
|     1 |
|     2 |
|     3 |
|     4 |
|     5 |
|     6 |
|     7 |
|     8 |
|     9 |
|    10 |
|    11 |
|    12 |
|    13 |
+-------+
showing top 14 rows

[danking]

stacked on #11745

[danking]

@tpoterba I have two concrete questions for you here:

1. What do you think of this (very minimal) amount of information about the driver batch?
2. What should we do about boolean fields going forward? Here I only accept the strings "True" and "true" but I think we've thus far accepted any non-empty string as a true value. I suppose we should probably continue that process, but I'd like to get your thoughts before I go that route.

[tpoterba]

1. I think it's better than nothing! Ship it.

2. This is for env vars, right? I think people often use HAIL_BLAH=1 from the command line to set stuff, so maybe look for that too? Chris probably has more grounded opinions about this.

[tpoterba]

@chrisvittal can you comment on this second point?

[chrisvittal]

For flag style env variables, I think we should use present and non-empty. We could also go with a reasonable set of 'affirmative' values, like say (all case insensitive) true, t, yes, y, 1.

[danking]

I'll go with present and non-empty.

[danking]

OK, ready for review.

[tpoterba]

oops, didn't realize this was hanging on me!