blr_stall

53 views
Skip to first unread message

Dimitry Sibiryakov

unread,
Apr 29, 2024, 10:57:48 AMApr 29
to firebir...@googlegroups.com
Hello All,

what is the difference in loops with and without blr_stall?

I see that backup created BLR for reading data without it while DSQL and DFW
generates this verb. And both approaches work which is surprising because I see
EXE_receive to call execute_looper before copying of output data into buffer
only for StallNodes.
How does it work?

--
WBR, SD.

Dimitry Sibiryakov

unread,
May 19, 2024, 7:10:57 AMMay 19
to firebir...@googlegroups.com
Dimitry Sibiryakov wrote 29.04.2024 16:57:
>   How does it work?

Just to answer myself: blr_stall is an abomination of initially simple design
of looper that exited only in two cases:

1) end of program
2) client interaction is needed for blr_send/blr_receive/blr_select.

blr_stall is used to force stop looper with ability to continue without need
for client interaction. Perhaps it was a workaround for some problem with
runahead execution. As a side-effect it created uncertainty in looper state
which in turn used to cause request synchronization errors if used in a wrong way.

--
WBR, SD.

James Starkey

unread,
May 19, 2024, 7:31:13 AMMay 19
to firebir...@googlegroups.com
No, you idiot, that’s not what it’s about.

GDML was designed so a host program count fetch a record and decide which among several things it might do.  I could simply go on to the next record, update the current record (in one or more ways), or delete the record.  Which of these was determined by the message number.  Blr_stall says wait for further instructions.

I don’t care about your foolish opinions of things you don’t understand, but I strongly object to you making stuff up.  There are impressionable people on this list who might assume you have some clue of what you are talking about.


Jim Starkey


--
You received this message because you are subscribed to the Google Groups "firebird-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to firebird-deve...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/firebird-devel/831295bd-9d46-4c19-9095-e6eb7746f44d%40ibphoenix.com.

Dimitry Sibiryakov

unread,
May 19, 2024, 7:35:20 AMMay 19
to firebir...@googlegroups.com
James Starkey wrote 19.05.2024 13:31:
> GDML was designed so a host program count fetch a record and decide which among
> several things it might do.  I could simply go on to the next record, update the
> current record (in one or more ways), or delete the record.  Which of these was
> determined by the message number.  Blr_stall says wait for further instructions.

Waiting for further instruction to decide what to do by message number is
what blr_select does, not blr_stall. These are different verbs.

--
WBR, SD.

Jim Starkey

unread,
May 19, 2024, 11:21:42 AMMay 19
to firebir...@googlegroups.com
You may have noticed the blr_receive, blr_select, and blr_send are 12,
13, and 14, respectively, while blr_stall is 155, putting it near the
last of the blr extensions, putting it late in the Interbase era.  I
can't tell you by whom or why it got added, though one could argue that
separating stall and message select operations makes a certain amount of
logical sense, at least enough that I wouldn't argue against it.

I hope everyone on this list is aware that attacking the original
designer, founder, and implementer is profoundly offensive.  You guys
didn't design the system and if it hadn't been for my work, you wouldn't
have a project.
Jim Starkey, AmorphousDB, LLC

Dimitry Sibiryakov

unread,
May 19, 2024, 11:28:02 AMMay 19
to firebir...@googlegroups.com
Jim Starkey wrote 19.05.2024 17:21:
>
> I hope everyone on this list is aware that attacking the original designer,
> founder, and implementer is profoundly offensive.  You guys didn't design the
> system and if it hadn't been for my work, you wouldn't have a project.

Sure. That's why I explicitly said that your original design was simple (and
thus effective).

--
WBR, SD.

Mark Rotteveel

unread,
May 21, 2024, 5:56:25 AMMay 21
to firebir...@googlegroups.com
On 19/05/2024 17:21, Jim Starkey wrote:
> I hope everyone on this list is aware that attacking the original
> designer, founder, and implementer is profoundly offensive.  You guys
> didn't design the system and if it hadn't been for my work, you wouldn't
> have a project.

I think you are reading a little bit too much in Dimitry's poor choice
of the word "abomination". I don't see an attack in it, nor do I think
one was intended.

Please keep in mind that most of us here are not native English
speakers, and sometimes miss the nuance in the meaning of words.

Mark
--
Mark Rotteveel

Reply all
Reply to author
Forward
0 new messages