Re: [sqlite] [SPAM?] Re: Explicit "read transaction" with journal_mode=WAL.

2019-07-31 Thread Richard Damon
If, and it is possible for there to be sequencing to enforce it, You KNOW that step 1, get a read transaction (which the OP presumes implies includes getting the lock) occurs before step 2, and thus step 4 seeing changes from step 3 says something is wrong. Yes, if you only can use the

Re: [sqlite] [SPAM?] Re: Explicit "read transaction" with journal_mode=WAL.

2019-07-31 Thread Thomas Kurz
Would it be possible for you to give some feedback (just an estimation) whether or not a suggestion might be considered? - Original Message - From: Richard Hipp To: SQLite mailing list Sent: Wednesday, July 31, 2019, 16:10:13 Subject: [sqlite] [SPAM?] Re: Explicit "read transa

Re: [sqlite] [SPAM?] Re: Explicit "read transaction" with journal_mode=WAL.

2019-07-31 Thread Richard Hipp
On 7/31/19, Simon Slavin wrote: > On 31 Jul 2019, at 12:57pm, test user wrote: > >> Is there a standard place where people can request features to be added to >> SQLite? > > Here. You've already done it. The developers of SQLite read this list and > will consider the things you wanted to do

Re: [sqlite] [SPAM?] Re: Explicit "read transaction" with journal_mode=WAL.

2019-07-31 Thread Simon Slavin
On 31 Jul 2019, at 12:57pm, test user wrote: > Is there a standard place where people can request features to be added to > SQLite? Here. You've already done it. The developers of SQLite read this list and will consider the things you wanted to do and whether it's worth providing a better

Re: [sqlite] [SPAM?] Re: Explicit "read transaction" with journal_mode=WAL.

2019-07-31 Thread test user
Yeh I think this is the way to go currently. I just wanted to be sure I understood it correctly and that it doesn’t exist already. Is there a standard place where people can request features to be added to SQLite? The Fossil repo perhaps? On Wed, 31 Jul 2019 at 12:53, Richard Damon wrote: >

Re: [sqlite] [SPAM?] Re: Explicit "read transaction" with journal_mode=WAL.

2019-07-31 Thread Richard Damon
As has been pointed out, your function that is called can do the BEGIN and then a SELECT that hits the database to force the obtaining of the read lock. The fact that the BEGIN didn't get the lock is then not seen by the users of your API. IF at some point a new BEGIN SHARED IMMEDIATE operation

Re: [sqlite] [SPAM?] Re: Explicit "read transaction" with journal_mode=WAL.

2019-07-30 Thread Richard Damon
On 7/30/19 5:33 PM, Olivier Mascia wrote: >> Le 30 juil. 2019 à 22:39, test user a écrit : >> >> What I would like is something like BEGIN READ, which will not block >> writers for its duration. >> >> This "read transaction" can see all committed transactions that happened >> before it, but none