Perl 6 - the future is here, just unevenly distributed

IRC log for #askriba, 2017-11-27

| Channels | #askriba index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
07:58 ashimema joined #askriba
09:31 karjala_ joined #askriba
10:45 ashimema is frew around?  I'm a bit confused.. I appear to suddenly be getting 'DROP TABLE "x" CASCADE;` when I prepare upgrade that include new tables using DeploymentHandler..
10:46 ashimema I'm sure I didn't used to get these lines in the auto sql.. is there a recent change to deploymenthandler that may have introduced these at all?
10:46 ashimema https://github.com/frioux/DBIx-Class-DeploymentHandler/commit/1813e4691aa746336efe798e973cedec8b4d8ea4 perhaps?
10:47 ashimema It makes Postgres spit out nasty warnings unless I go in an add 'IF EXISTS' into the statement (or remove the DROP before the CREATE line entirely)
10:48 ribasushi ashimema: almost certainly the problem is where you identified it yes
10:48 ribasushi frew wakes up ~4 hours from now usually, he'll likely respond then
10:48 ashimema I'll dig back and see what the default value of that variable is.. I'm sure I've never changed it.
10:49 ashimema thanks ribasushi..
10:49 ashimema how was the weekend btw.. sorry I missed you on friday :(
10:49 ashimema bloomin kids
10:50 ribasushi it was superb - I got to drink beer with folks, and spent the daytime in museums in stuff - best perl conference I attended in a while ;P
10:50 ribasushi s/in stuff/and stuff/
10:51 ribasushi there were some good talks from what I hear too, you should def. try it next time
10:58 karjala_ I seem to have a little problem with $row->id
10:59 karjala_ when I print it in my code as warn Dumper({id => $row->id}) I get id => 1, but when I do id => "".$row->id, I get 13 (which is the correct number)
10:59 karjala_ I'll check again, one moment
11:00 ribasushi karjala_: it's not a problem as such, it is a shitty design (again)
11:00 ribasushi ( sec grabbing link )
11:00 karjala_ whose design? where?
11:00 karjala_ ok
11:02 ribasushi karjala_: go ahead and `warn Dumper [ $row->id ]`
11:02 karjala_ I'll do some more tests, right now.
11:02 karjala_ Just a sec
11:02 ribasushi see what that returns, it should be several values
11:03 karjala_ no actually my question here was not accurate
11:03 karjala_ i did different things in fact
11:03 karjala_ like: $id = int $row->id and then print $id (and got 1)
11:03 karjala_ and then print "$id" and got 13
11:04 karjala_ So I don't know what I should test now
11:04 karjala_ no
11:04 ribasushi karjala_: I told you *specifically* what to test ;)
11:04 karjala_ ok
11:06 karjala_ that printed [19] (the correct answer) - I think need to reproduce the error I got originally, and paste the exact code in pastebun
11:06 karjala_ pastebin
11:08 karjala_ just a sec, I'm trying
11:09 ribasushi yeah it's not what I originally thought, you'll need more details
11:11 karjala_ Oh found it
11:12 karjala_ I had my $id = int $file->id, instead of ->id;
11:12 karjala_ XD
11:12 karjala_ comma instead of ;
11:13 karjala_ so the next line was using the previous $id from the higher scope
11:13 karjala_ mystery solved
11:13 karjala_ thanks
11:13 ribasushi :)
11:23 ribasushi karjala_: did you figure out your reconnection stuff from Fri?
11:30 karjala_ I didn't try mstrout's new dev release, no. Should I do it? I don't intend to use a dev release.
11:30 karjala_ in production
11:30 karjala_ But I did this solution:
11:31 karjala_ whenever an exception occurred at the web request handler, I check whether the connection is on, and if not I connect
11:32 ribasushi this is a pretty problematic "solution" as a lot of state within DBIC is likely being incorrectly left over from this
11:33 ribasushi this is in fact the very bug I was suggesting you are hitting, which is fixed by myself within the "dev release", which is not an official release for reasons unrelated to technology
11:33 karjala_ ok when I say reconnect, I mean create a new schema object
11:33 karjala_ does that sound better?
11:33 ribasushi that would work, yes, albeit being a really crappy workaround
11:33 karjala_ lol
11:35 karjala_ because it's slow, I guess, eh? the other solution is to use the _dev version which I'm afraid of doing
11:35 karjala_ is this dev release production ready you trust?
11:36 ribasushi karjala_: as far as I can tell - yes ( but I am biased, as I essentially wrote all of it )
11:37 ribasushi karjala_: frew has been doing large scale testing at their site recently, perhaps they discovered some issues that I've not heard of yet
11:39 ribasushi karjala_: in any case: before deciding whether to upgrade or not
11:39 ribasushi you should try your test script with it
11:39 ribasushi to see if the problem is actually solved
11:39 ribasushi because if it isn't - then this becomes a different conversation entirely
17:41 karjala_ joined #askriba

| Channels | #askriba index | Today | | Search | Google Search | Plain-Text | summary