Please draft us a Scheme API Lassi Kortela (17 Sep 2019 18:54 UTC)
Re: Please draft us a Scheme API John Cowan (17 Sep 2019 19:08 UTC)
What should Hello World be like? Lassi Kortela (18 Sep 2019 08:41 UTC)
Re: What should Hello World be like? Lassi Kortela (18 Sep 2019 08:54 UTC)
Re: What should Hello World be like? John Cowan (19 Sep 2019 01:14 UTC)
Re: What should Hello World be like? hga@xxxxxx (19 Sep 2019 02:31 UTC)
Re: What should Hello World be like? Lassi Kortela (19 Sep 2019 15:35 UTC)
Re: What should Hello World be like? Peter Bex (19 Sep 2019 18:02 UTC)
Re: What should Hello World be like? hga@xxxxxx (19 Sep 2019 22:50 UTC)
Re: What should Hello World be like? John Cowan (19 Sep 2019 19:46 UTC)
Re: What should Hello World be like? hga@xxxxxx (19 Sep 2019 20:50 UTC)
Re: What should Hello World be like? Lassi Kortela (19 Sep 2019 21:14 UTC)
Re: What should Hello World be like? Peter Bex (20 Sep 2019 07:49 UTC)
Re: What should Hello World be like? Alaric Snell-Pym (20 Sep 2019 10:55 UTC)
Re: What should Hello World be like? John Cowan (20 Sep 2019 13:25 UTC)
Re: What should Hello World be like? John Cowan (20 Sep 2019 14:04 UTC)
Re: What should Hello World be like? Lassi Kortela (23 Sep 2019 10:54 UTC)
Connection strings, and representing them as lists Lassi Kortela (23 Sep 2019 10:57 UTC)
John's Simple SQL API hga@xxxxxx (19 Sep 2019 11:38 UTC)
Re: John's Simple SQL API John Cowan (19 Sep 2019 20:28 UTC)
Re: Please draft us a Scheme API hga@xxxxxx (17 Sep 2019 19:11 UTC)
Re: Please draft us a Scheme API John Cowan (17 Sep 2019 19:12 UTC)
Re: Please draft us a Scheme API Lassi Kortela (17 Sep 2019 19:32 UTC)
Re: Please draft us a Scheme API John Cowan (17 Sep 2019 19:35 UTC)
Re: Please draft us a Scheme API Lassi Kortela (17 Sep 2019 19:27 UTC)

Re: What should Hello World be like? Lassi Kortela 23 Sep 2019 10:54 UTC

> Given that SQL calls are basically
> imperative, it makes sense not to attempt any more of them after an abort.

+1

>> Just continuing with the regular flow seems unwise to me, as those
>> statements would then be taking place in the parent transaction.
>
> Or would be auto-committed, if there is no parent (as would typically be
> the case), which is the Worst Possible Thing.

Strongly agreed.