Hi, Here's a quick update. V 3.0 M1 (alpha) will be released in a few days with some important notes. It's not feature complete yet, but it will include (among other things) two important improvements: - the new SQL engine: we changed a little bit the query APIs and completely rewrote the internal mechanisms of execution planning and query execution. The query parser is the same as the strict parser of v 2.2, so the language is the same, apart from some small changes and additions (eg. now we have a DISTINCT keyword for projections) - remote transactions: until v 2.2 all the transactions were executed on the client and transferred to the server only at commit time, so any SQL statement during a transaction was executed on the server and outside of the transaction itself. In V 3.0 we added management of remote transactions: tx are still executed on the client, but as soon as an SQL statement is executed, the transaction is flushed to the server (not yet committed of course), so the SQL statement will run inside current transaction.
So the answer to your original question is finally YES, in 3.0 we will have remote transactions, SQL inside transactions and a consistent behavior of remote and plocal instances. About details and docs, many things will be updated in next few days and released with the M1 Thanks! Luigi 2017-01-21 3:58 GMT+01:00 odbuser <[email protected]>: > Can you provide an update on this feature (particular SQL participating in > transactions)? > > I can't figure out by the roadmap, issues for 3.0, or the 3.0 > documentation whether SQL can be participate in a transaction along side > the graph api. > > On Thursday, July 14, 2016 at 2:22:17 AM UTC-4, Luigi Dell'Aquila wrote: >> >> Hi, >> >> This is our intention, so the answer is "it's very likely, but not yet >> implemented, even in alpha" >> >> Thanks >> >> Luigi >> >> >> 2016-07-14 0:41 GMT+02:00 odbuser <[email protected]>: >> >>> Bump... this one is very important for anyone using transactions. Any >>> comments? >>> >>> >>> On Tuesday, June 21, 2016 at 10:05:47 PM UTC-4, odbuser wrote: >>>> >>>> Are the SQL and JS query/commands really going to participate in >>>> transactions in 3.0? According the docs, this is the case but is it so? >>>> Can you provide the links to the issues in git that will resolve this >>>> issue? I'm interested in how this will work... especially select >>>> statements executed during a transaction. >>>> >>>> http://orientdb.com/docs/last/Transactions.html >>>> >>>> "In this scenario you can have different isolation levels with >>>> commands. This issue will be solved with OrientDB v3.0 where the >>>> transaction will be flushed to the server before to execute the command." >>>> >>> -- >>> >>> --- >>> You received this message because you are subscribed to the Google >>> Groups "OrientDB" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected]. >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> -- > > --- > You received this message because you are subscribed to the Google Groups > "OrientDB" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- --- You received this message because you are subscribed to the Google Groups "OrientDB" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
