transcript for db<>fiddle Feedback
2019-11-14 12:30:58  Paul White
Why is this Q & A on Databases rather than meta?
2019-11-14 15:09:33  Jack Douglas  replying to  Paul White
it is a 'meta' post on databases — have you noticed the 'question type' drop-down on the 'ask question' page? So that makes it equivalent to a post on rather than in old money.
2019-11-14 15:10:08  Jack Douglas
You can ping the OP by clicking on the identicon on the question itself (or on an answer) even if they haven't yet chatted in a room.
2019-11-14 20:26:05  Paul White  replying to  Jack Douglas
Oh. I see. Hm. That difference seems rather small when there's only one "site". Ok, anyway, I still think it looks a bit odd.
2019-11-14 20:27:17  Paul White  replying to  Jack Douglas
Perhaps an initial message from the OP in the chat window would be more intuitive.  
Like "this chat room is for discussion of the question..." found on chat.SE when comments are moved to chat.
2019-11-14 20:41:19  Jack Douglas  replying to  Paul White
that's an interesting thought but to do when someone answer, auto-add a message from them too at that point? That would get a bit strange I think
2019-11-14 20:41:39  Jack Douglas
'but to do' = 'but what to do'
2019-11-14 20:41:54  Paul White
I don't understand.
2019-11-14 20:42:01  Paul White
Oh now I do.
2019-11-14 20:42:30  Paul White
Well the original message could cover that scenario. Just add the answer avatar to the right bar.
2019-11-14 20:43:21  Paul White
@Jack Something like that anyway.
2019-11-14 20:44:11  Jack Douglas
thanks Paul I think you are right we would be better of doing something along those lines…
2019-11-14 20:45:23  Jack Douglas
we will probably still *allow* clicking on identicons on the question/answer because finding the right person in the identicon stack is extra work if you have just read an answer and want to ask for clarification
2019-11-14 20:46:03  Jack Douglas
hmm, we could even have a link next to the 'edit' link something like 'ask for clarification'
2019-11-21 13:52:44  Paul White
@Jack What's the status of the db<>fiddle integration?
2019-11-21 13:55:17  Jack Douglas  replying to  Paul White
2019-11-21 13:55:25  Jack Douglas  replying to  Paul White
sort-of working…
2019-11-21 13:56:52  Jack Douglas
in chat, we currently only expand the latest fiddle so if you refresh, earlier fiddles posted revert to `<><>` — obviously in q&a we won't do that but I think it would slow things down in chat too much if we had more than a few
2019-11-21 13:57:48  Jack Douglas
do please feel free to post fiddles in here to test them out. the 'batch n' text is just a placeholder for now, and I think we'll just remove it entirely
2019-11-21 14:10:04  Paul White  replying to  Jack Douglas
I'm more likely to use one in a question.  
Is the markdown syntax documented somewhere?
2019-11-21 14:11:08  Paul White
Looks like it's just the url surrounded by `<>`
2019-11-21 14:12:09  Jack Douglas
yes, but I'm still not sure that is the best syntax — any thoughts?
2019-11-21 14:13:05  Jack Douglas
once I'm sure we've settled on a particular syntax, I'll add it to [the meta post](/meta?q=72).
2019-11-21 15:01:42  Paul White  replying to  Jack Douglas
I had a quick look at other extensions when the issue was first raised, but I have no idea what is or isn't allowed or best practice.
2019-11-21 15:02:26  Paul White
I'm still hoping the question/answer UI will add shortcut keys or a formatting toolbar at some stage.
2019-11-21 15:02:56  Paul White
The `<>` symbol doesn't seem unreasonable, but does it need to be at both ends?
2019-11-21 15:03:20  Paul White
Oh I suppose it is, for consistency if nothing else
2019-11-21 15:06:09  Paul White
I wonder how often people would want to post a raw db<>fiddle link without it becoming integrated?
2019-11-21 16:54:00  Jack Douglas  replying to  Paul White
6 to 8 minutes for a first cut of that — James is working on it now
2019-11-21 16:55:01  Jack Douglas  replying to  Paul White
possibly not if it is like the header markdown with '#' — in which case it would need to be at the beginning of a line, which seems like a reasonable restriction
2019-11-21 16:55:44  Paul White
Yikes what happened to the formatting
2019-11-21 16:55:53  Jack Douglas
needs refresh
2019-11-21 16:56:02  Paul White
of course it does
2019-11-21 16:56:19  Paul White
confirmed fixed
2019-11-21 16:56:20  Jack Douglas
chat editing has landed
2019-11-21 16:56:22  Paul White
oh wow
2019-11-21 16:56:38  Jack Douglas
:) :)
2019-11-21 16:57:11  Paul White
clicking the dots next to the star does something odd
2019-11-21 16:57:36  Jack Douglas
should cycle to the less often used buttons
2019-11-21 16:57:41  Jack Douglas
does it not?
2019-11-21 16:58:00  Paul White
I can't really tell
2019-11-21 16:58:07  Paul White
It does change to *something*
2019-11-21 16:58:15  Jack Douglas
permalink icon?
2019-11-21 16:58:23  Jack Douglas
edit history should be in there too
2019-11-21 16:58:25  Paul White
It could well be that
2019-11-21 16:58:30  Jack Douglas
(but only if the message was edited)
2019-11-21 16:58:59  Jack Douglas
will add 'up arrow edits recent messages' I just didn't want to overload that commit any further
2019-11-21 16:59:09  Paul White
that's pretty cool, but very fiddly
2019-11-21 16:59:59  Jack Douglas
too small?
2019-11-21 17:00:18  Paul White
good job I have a "precision" touchpad
2019-11-21 17:00:50  Paul White
2019-11-21 17:01:02  Paul White
just testing
2019-11-21 17:02:03  Paul White
Anyway I answered a question on MSDN today that I think could be useful here and I was planning a db<>fiddle link to be part of it
2019-11-21 17:02:25  Paul White
Lurching back on topic
2019-11-21 17:08:35  Jack Douglas  replying to  Paul White
ok that's arrived now — only a couple of buttons and only on questions right now but the rest will be quick I think
2019-11-21 17:09:04  Jack Douglas  replying to  Paul White
that is a great incentive for me to get that working asap :)
2019-11-21 17:10:03  Paul White  replying to  Jack Douglas
6-8 seconds too late for Erik I think
2019-11-21 17:24:37  Jack Douglas
notice that :|
2019-11-21 22:23:44  meme
2019-11-21 22:57:34  Jack Douglas  replying to  Paul White
it's not perfect but it's now working. At some point I will deal with the screen refreshing issue which degrades the question/answer editing experience (but doesn't affect you when just viewing the post).
2019-11-22 05:15:12  Paul White
2019-11-22 05:15:46  Paul White
thanks @Jack!
2019-11-22 09:43:01  PeterVandivier
is it practical to use the same renderer for the embed results as on dbfiddle? just wondering about the line-feeds getting stripped from the Naughts & Crosses board
2019-11-22 09:43:35  PeterVandivier
unless that's a "just my browser" thing?
2019-11-22 09:44:06  PeterVandivier
![Screenshot 2019-11-22 at 09.43.49.png](/image?hash=fab9ca6f41af91e00a59d2052b9c62d76b980bbfecbfcea120c1cb8617a800b7)
2019-11-22 09:47:28  Paul White
Looks the same for me on Chrome.  
I hadn't realized the board representation ought to be displayed 3 x 3 but that does make sense.
2019-11-22 09:51:01  PeterVandivier
yea, i noticed it from the screencap
2019-11-22 09:51:17  PeterVandivier
plus side, i got to play with running the embedded fiddle :-) 
2019-11-22 09:52:23  PeterVandivier
i thought it might be an LF / CRLF thing so i tacked in `|| chr(13)` to the final `array_to_string()` call
2019-11-22 09:52:34  PeterVandivier
which gave me the idea for the post! :-D
2019-11-22 09:52:59  Paul White
Is there a missing word in your opening paragraph?  
> Currently it looks like editing running the fiddle...
2019-11-22 09:53:54  PeterVandivier
two, in fact  
> ... editing and/or running  ...
2019-11-22 10:03:34  Jack Douglas
@PeterVandivier I think an option to update the post when you change the fiddle would be excellent, thanks for the suggestion. The other suggestions I need to think about — db<>fiddle itself has no concept of 'lineage' (though ctrl-Z in the TopAnswers editor does 'undo' which might be helpful in conjunction with your 'update this post' button — ie undo until you can copy the fiddle you want, then redo all the way and paste back in)
2019-11-22 10:15:15  PeterVandivier
yea, i figured lineage might be a complicated thing that doesn't have tooling to support it as yet, but once i got the idea in my head of displaying the diff, it seems like a prerequisite - or at least like a datastore that wants to exist 
2019-11-22 10:21:38  PeterVandivier
also - back when i knew less about how the dbfiddle backend worked, i think i assumed you were already tracking the lineage - and thereby storing the diffs between related fiddles, instead of the full text of each fiddle each time
2019-11-22 10:22:28  PeterVandivier
i know i start to feel guilty after 4 or 5 or 10 edits where it isn't quite like i want - knowing you're now storing 10 garbage fiddles for my one example
2019-11-22 10:24:10  PeterVandivier
lol - for my last post actually, i hammered through a bunch of iterations on my localhost to make sure i had the schema setup correctly before creating the fiddle so i wouldn't spam you with pointless garbage
2019-11-22 10:29:46  Jack Douglas
postgres=# select count(*), sum(pg_column_size(fiddle_input))/1024/1024, sum(pg_column_size(fiddle_output))/1024/1024 from fiddle;

count | column | column
--- | --- | ---
605742 | 516 | 1217
2019-11-22 10:32:16  Jack Douglas  replying to  PeterVandivier
total storage for 600k fiddles is about 2GB, so not something I worry about at all
2019-11-22 10:32:31  PeterVandivier
2019-11-22 10:32:58  PeterVandivier
still, y'know... signal v noise n'at
2019-11-22 10:33:18  PeterVandivier
i'm the guy who tries not to push commits until i know the build is gonna pass
2019-11-22 10:33:34  PeterVandivier
which is actually something i've been trying to change about myself
2019-11-22 23:59:21  Jack Douglas  replying to  PeterVandivier
thanks for mentioning that Peter, for some reason I didn't notice it at all, and I've seen that fiddle dozens of times the way it is supposed to look. I've fixed it to render more like it does on db<>fiddle proper.
2019-11-23 09:37:27  PeterVandivier
2019-12-11 14:26:53  PeterVandivier
@Jack can I safely assume you'd prefer db<>fiddle issues / features / etc go here rather than going forward?
2019-12-11 14:28:09  PeterVandivier
was about to post one on the OG post, but then I thought about adding a topanswers specific suggestion and figured it might look untoward
2019-12-11 14:38:29  Jack Douglas  replying to  PeterVandivier
yes please :)
2019-12-11 14:42:31  PeterVandivier
2019-12-11 14:43:28  PeterVandivier
ruh ruh, do gifs auto play infinitely? 😬😅
2019-12-11 14:59:45  Jack Douglas  replying to  PeterVandivier
no easy way to stop that… except not to post them ;)
2019-12-11 15:00:36  Jack Douglas
it seems to me that's something the browser should have a setting for
2019-12-11 15:05:16  PeterVandivier  replying to  Jack Douglas
i _think_ slack does a 30-second play then pause until click thing? 
2019-12-11 15:05:27  PeterVandivier
just what i'm conditioned to expect
2019-12-11 15:06:51  PeterVandivier
ah... it's webex that does no-play until click and then you can click out of the overlay
2019-12-11 15:08:14  Jack Douglas
they must be very clever!
2019-12-11 15:08:57  PeterVandivier
yea, slack has the `/collapse` command which minimizes all one-boxed media fo the user who subs it
2019-12-11 15:16:36  PeterVandivier  replying to  Jack Douglas
seems like you are too, it's what you're already doing for static images 😜
2019-12-11 15:17:22  PeterVandivier
2019-12-11 15:17:23  Jack Douglas  replying to  PeterVandivier
oh I forgot about that — James did that :)
2019-12-11 15:17:52  PeterVandivier
well, smart by proxy then ;p
2019-12-11 15:27:14  PeterVandivier
actually, looks like it's already set up for gifs too - just needs the thumbnail to display the static image (and possibly a `GIF` watermark/overlay like webex does)
2019-12-11 15:33:21  Jack Douglas
yippee found image.animation_mode in Firefox. only helps me of course!
2019-12-11 15:35:19  PeterVandivier
2019-12-11 15:35:48  PeterVandivier
offhand - do you know if  `\set` options are available in the postgres dbfiddle?
2019-12-11 15:36:30  PeterVandivier
rubber-ducking [this nugget]( and i can't replicate in psql locally
2019-12-11 15:37:00  Jack Douglas  replying to  PeterVandivier
I don't think so
2019-12-11 15:37:06  PeterVandivier
trying to evangelize the good fiddle to the 👋New Contributor
2019-12-11 15:39:29  Jack Douglas  replying to  PeterVandivier
you can use `set` just not `\set`
2019-12-11 15:39:30  Jack Douglas
2019-12-11 15:40:58  PeterVandivier
weird - thought i tried that syntax variant, guess not 🤷‍♂️
2019-12-12 07:50:19  PeterVandivier  replying to  Jack Douglas
re: loading 60k rows in a fiddle - i'm guessing it wouldn't be a welcome feature request to upload files to the fiddle rather than copy-paste? seems like one could start to balloon up that "space used" metric pretty darn quick
2019-12-12 09:40:26  Jack Douglas  replying to  PeterVandivier
there are plenty of fiddles that generate lots more than 60k rows — and I do think we need an 'import csv' feature. otoh I'm not sure the UI will handle 60k actual inserts!
2019-12-12 09:40:48  Jack Douglas
All I can say for sure is that 'space used' won't be the limiting factor
2019-12-12 09:45:07  PeterVandivier
2020-01-05 21:54:08  Jack Douglas
@PeterVandivier re YYYY-MM-DD, how do we actually do this for SQL Server? I thought it might be SET DATEFORMAT but no dice.
2020-01-05 21:59:20  Jack Douglas
we are using if that makes any difference
2020-01-06 11:36:57  PeterVandivier  replying to  Jack Douglas
tbh, i figured it was a quirk in your display layer. i wouldn't _think_ it's a service setting like `DATEFORMAT` - every installation I've ever worked on via SSMS has had ISO as the display default regardless of the client localisation 🤔
2020-01-06 11:38:57  PeterVandivier
i guess i'm also not really familiar with the db<>fiddle architecture that well either. is it something you'd be able / willing to share that i could spin up on my localhost to play with?
2020-01-06 11:40:26  PeterVandivier
(i'm also _very much not_ a web developer so unless the deploy script is pretty tight it may be sort of an academic exercise of me poking at it with a stick to no purpose)
2020-01-06 11:47:42  PeterVandivier
some instructive examples though -
2020-01-06 11:48:49  PeterVandivier
re-reading the docs, i'm sort of remembering that `DATEFORMAT` specifically may only affect date interpolation - but the `date` datatype doesn't / wouldn't / shouldn't get handled any differently by the service based on that setting
2020-01-06 12:18:55  Jack Douglas  replying to  PeterVandivier
if only there was some sort of databases Q&A where we could put questions like this to the world!
2020-01-06 12:21:21  PeterVandivier
2020-01-06 12:57:17  Jack Douglas  replying to  PeterVandivier
I found the answer on a Russian clone of SO:
2020-01-06 13:09:06  PeterVandivier
2020-01-06 13:10:40  PeterVandivier
2020-01-06 14:24:16  Andriy M  replying to  Jack Douglas
Out of curiosity, which answer did you use? They are all in Russian, except the one that's in (I think) C#
2020-01-06 14:41:17  Jack Douglas  replying to  Andriy M
I didn't notice they are all in Russian — the top one seemed like it might be saying that the `value` attribute is of type `datetime` ("SqlDateTimeКласс имеет свойство , Valueкоторое уже типа DateTime"), which was quick to look up and verify :)
2020-01-06 14:42:11  Jack Douglas
not to mention a relief because I was tying myself in knots Googling and reading the docs
2020-01-06 14:44:08  Jack Douglas
now I look a bit harder of course the original SO answer is in English — I don't think I scrolled down past the accepted answer:
2020-01-06 15:50:37  Andriy M  replying to  Jack Douglas
'The SqlDateTime class has a property, Value, which is already of the DateTime type' is what is says. Your Russian is very good! ; )
2020-01-06 15:55:57  Andriy M
Oh, I answered your reply but only now I saw that there was an original in English. Interesting, so the Russian answers in the thread were translated...
2020-01-06 16:07:07  Jack Douglas  replying to  Andriy M
by a machine or is it good Russian?
2020-01-06 16:23:39  Andriy M  replying to  Jack Douglas
Now that I know it's a translation, it actually does look like machine-translated. Initially it simply looked as if written in Russian by someone thinking in English, but correct grammatically nevertheless. I've now looked closer at the other answers, and they are substantially worse in terms of grammar (and sometimes sense too).
2020-01-07 16:47:23  PeterVandivier
@Jack is it a major structural change to support schema creation for postgres in db<>fiddle?
2020-01-07 16:48:09  PeterVandivier
seems like it might be non-trivial
2020-01-07 16:48:39  PeterVandivier
happy to put it up here as a feature request though if it's something you're intrigued by
2020-01-07 18:03:04  Jack Douglas  replying to  PeterVandivier
it's trivial but I want to be sure it isn't a security risk
2020-01-07 18:47:53  PeterVandivier
2020-01-07 20:17:30  Jack Douglas  replying to  PeterVandivier
can you tell from the docs whether `grant create on database database_name to "Peter";` allows anything that could be dangerous?
2020-01-08 09:21:25  PeterVandivier
as with everything i sort of assume the answer is "_it depends_" but i'm reading up on it now and i'll let you know if i get any productive insight
2020-01-08 09:21:34  PeterVandivier  replying to  Jack Douglas
as with everything i sort of assume the answer is "_it depends_" but i'm reading up on it now and i'll let you know if i get any productive insight
2020-01-08 09:22:35  PeterVandivier
ah, edit message behaviour is mildly different from SE
2020-01-08 10:09:34  Jack Douglas  replying to  PeterVandivier
yes we are planning to fix some aspects of that
2020-01-14 08:49:20  PeterVandivier  replying to  Jack Douglas
still around for extended types :P
2020-01-14 08:55:19  Andriy M  replying to  PeterVandivier
Why is the datetime type called 'raw'? : )
2020-01-14 09:00:40  PeterVandivier  replying to  Andriy M
updated all the fiddles in the post
2020-01-14 09:01:01  PeterVandivier
raw as in "unchanged"
2020-01-14 09:01:13  PeterVandivier
not using a cast or function
2020-01-14 09:01:32  PeterVandivier
just the word i thought of in the moment i guess 🤷‍♂️
2020-01-14 09:03:27  PeterVandivier
for context i originally noticed it when reading [this bug report over on]( i saw it again today just by chance
2020-01-14 09:10:09  PeterVandivier
unrelated: i'm not sure if i like or dislike the fact that you can't do 2-letter domain names for the .xyz TLD. just tried to check on as a shortened URL but it looks like it's not valid :-/ 
2020-01-14 09:12:30  PeterVandivier
also @Jack i hope you'll forgive that cheeky last edit :-P
2020-01-14 11:03:09  Jack Douglas  replying to  PeterVandivier
I've fixed this so it shows the iso format for all types, but do you think we should show the whole precision as John Eisbrener seems to be requestion?
2020-01-14 11:03:28  Jack Douglas  replying to  PeterVandivier
not cheeky at all :)
2020-01-14 11:30:53  PeterVandivier  replying to  Jack Douglas
yea, i think it's totally reasonable. pretty sure he's asking for dbfiddle to have similar behaviour to ssms for date formatting. e.g. dt(0) show seconds precision dt(1) shows tenths of a second, dt(2) shows hundreds, etc...
2020-01-14 11:32:09  PeterVandivier
![Screenshot 2020-01-14 at 11.31.05.png](/image?hash=4c5f5bb2e64f69ef41166e644fa8d24cd88814fe2e1e71ce922ed190010c5140)
2020-01-14 11:33:04  PeterVandivier
that being said - while it may be a reasonable ask, i wouldn't be all that surprised if you told me it's surprisingly complicated to generalise that kind of formatting behind the scenes
2020-01-14 11:39:49  PeterVandivier
although it's in place intuitively for numerics so maybe not 🤷‍♂️
2020-01-14 11:39:50  PeterVandivier
2020-01-14 11:40:01  PeterVandivier
![Screenshot 2020-01-14 at 11.38.51.png](/image?hash=eb39c1b918298a79a4fce9b54543efe163ec3939f4dcbcba1a004436db33865b)
2020-01-14 11:43:30  PeterVandivier
i would also like to take this opportunity to complain that switching between a macbook UK keyboard and a windows US keyboard inside a VM is the _worrrrrrrrst_
2020-01-14 11:45:01  PeterVandivier
also apparently my windows VM is on a DataCenter trial license so i get to see what it looks like when the license expires next week. #ForScience !
2020-01-14 11:52:21  Jack Douglas  replying to  PeterVandivier
I can't figure out a way to get the precision form the .net driver — only the data type:
2020-01-14 11:52:23  Jack Douglas
2020-01-14 11:52:41  Jack Douglas
apart from that, does this 2014 fiddle look right to you?
2020-01-14 12:00:59  PeterVandivier  replying to  Jack Douglas
it does 👍
2020-01-14 12:01:59  PeterVandivier  replying to  Jack Douglas
reminds me of [this old tumbleweed]( i never deleted b/c i'm still a little proud of the question
2020-01-14 12:07:49  PeterVandivier
like - i don't think there _is_ a good date "precision" that's communicated from the service to the client. i wouldn't be hugely surprised if it's a string parsing UI shim in SSMS / Visual Studio
2020-01-14 12:20:32  Jack Douglas  replying to  PeterVandivier
I've fixed the others up!
2020-01-14 12:30:06  PeterVandivier
2020-01-14 14:17:27  Andriy M
@Jack Speaking about accuracy/precision, how do you handle `smalldatetime` in dbfiddle? [It's shown with seconds]( but [its official accuracy is 1 minute](
2020-01-14 14:20:10  PeterVandivier  replying to  Andriy M
[are you sure about that? :-P](
2020-01-14 14:21:03  PeterVandivier
i think that's the same as how it's displayed in SSMS (even though it's misleading there too)
2020-01-14 14:21:18  PeterVandivier
but no one cares b/c SSMS is the standard
2020-01-14 14:23:23  Jack Douglas  replying to  Andriy M
I saw that in the docs too but also noticed that despite that it seems to show the seconds correctly (try running that fiddle several times and you see the seconds update). Would be an excellent question for TA if you care to post it! I bet Sean knows the answer…
2020-01-14 14:26:15  PeterVandivier
2020-01-14 14:26:32  PeterVandivier
i didn't even look at your fiddle before pulling up my old one
2020-01-14 14:28:44  PeterVandivier
oh snap! ~~difference from [2017]( to [2019]( -  except for i just re-ran the 2017 fiddle and it updated to show seconds?!
2020-01-14 14:28:49  PeterVandivier
2020-01-14 14:28:56  PeterVandivier
shouldn't have re-run it!
2020-01-14 14:29:30  PeterVandivier  replying to  Jack Douglas
there's no way you can access the old cached results, right?
2020-01-14 14:49:21  PeterVandivier
![Screenshot 2020-01-14 at 14.49.07.png](/image?hash=692d1ca9e1fdea017bcdedfc45e94181809afd3f9fca11fb6db90b55cdff86d2)
2020-01-14 14:49:28  PeterVandivier
☝️ fwiw
2020-01-14 14:51:58  PeterVandivier
i'll test on my updated laptop at home and if it's different i'll post it up (unless someone beats me to it :-P)
2020-01-14 15:32:48  Andriy M  replying to  PeterVandivier
I think I am:  
2020-01-14 15:33:38  Andriy M
When you add 5 seconds incrementally, a `smalldatetime` value doesn't change.
2020-01-14 15:34:43  Andriy M
Adding 30+ seconds increments the minutes because the result is rounded up
2020-01-14 15:42:52  PeterVandivier  replying to  Andriy M
sorry for the confusion - i was interpretting the db<>fiddle handling as some crazy patch i hadn't heard about where `smalldatetime` is now storing seconds
2020-01-14 15:43:34  PeterVandivier
the link to the 2017 fiddle i posted _would have shown_ the behaviour you and i both expect (and that we both screencapped) - except for i re-ran it and blew out the cached results
2020-01-14 15:44:32  PeterVandivier
so either the db<>fiddle backend is doing something very funny to `smalldatetime` all of a sudden or there's an even more shocking unannounced & undocumented change to that datatype
2020-01-14 15:45:36  PeterVandivier  replying to  Andriy M
i got the same results as you in my local test screencap ![](
2020-01-14 17:09:11  Jack Douglas  replying to  PeterVandivier
no, 'fraid not
2020-01-14 17:11:25  PeterVandivier  replying to  Jack Douglas
well if you'll take my word for it, the 2017 fiddle link used to show the same as both mine & @Andriy 's screencaps
2020-01-14 17:11:36  PeterVandivier
so... possibly a regression?
2020-01-14 17:12:07  PeterVandivier
can't imagine what would cause that though 🤷‍♂️
2020-01-14 17:31:53  Jack Douglas
2020-01-14 17:31:59  Jack Douglas
could that explain the difference?
2020-01-14 17:34:06  Jack Douglas
or one of:
2020-01-14 17:34:08  Jack Douglas
2020-01-14 17:34:19  Jack Douglas
2020-01-14 17:35:14  Jack Douglas  replying to  PeterVandivier
sure it wasn't 2012?
2020-01-14 17:49:14  PeterVandivier  replying to  Jack Douglas
as sure as i can be, that's what i linked to on [that comment](
2020-01-14 17:57:37  PeterVandivier  replying to  Jack Douglas
i can't see any of those possibly being the cause. but who knows
2020-01-14 17:58:31  PeterVandivier
like i said though - i'll try and patch & check again on my other laptop (tomorrow probably). if i can't repro there, then i'm guess it's a bizarre internal regression
2020-01-14 17:59:04  PeterVandivier
if i can't then somehow MS snuck in bonus data density in a CU and didn't tell anyone
2020-01-14 17:59:13  PeterVandivier
but that seems unlikely :-p
2020-01-14 18:33:35  Andriy M  replying to  Jack Douglas
Actually... there's [something severely wrong]( with dbfiddle's datetime handling at the moment. Why does it show the current time in each case?
2020-01-14 18:38:15  PeterVandivier
2020-01-14 18:44:49  PeterVandivier
i wonder if the change on the dotnet typing is assigning the current time from the app driver instead of bringing back the value from the DB
2020-01-14 22:25:37  Jack Douglas  replying to  Andriy M
you are a genius @Andriy 
2020-01-14 22:25:56  Jack Douglas
I think it is all fixed up and working the way you expect now
2020-01-14 22:26:39  Jack Douglas
I've truncated the seconds from `smalldatetime` as they are now correctly always 0
2020-01-14 22:26:43  Jack Douglas
cc @PeterVandivier 
2020-01-15 00:06:53  fds
Hi Jack! All that's left is to say thank you again, I believe. Came here to repeat my issue with Firebird LIST() as instructed, but then I see it's already fixed. :) Haha, and just yesterday I spent a good chunk of time registering on, then building up the sufficient "reputation" only so I could post that bug report. It was worth it, of course, in the end, but maybe this is then my new bug report: the suggestion/reports link should be updated on the dbfiddle site to point here.
2020-01-15 00:11:01  Jack Douglas  replying to  fds
Hi @fds and thanks for dropping in — yes I managed to fix the bug thanks to [this faq page]( I found!
2020-01-15 00:11:41  Jack Douglas
Also thanks for letting me know there is a db<>fiddle link still pointing to — did you get there on mobile by any chance? The desktop link already points here!
2020-01-15 00:14:21  fds
Not on mobile, Mac laptop Apple Safari, but it very well might be mistaking that for mobile. Now I'm wondering what the "real" site is supposed to look like. Everything still functioned fine. I see like a play button / hamburger menu, click on it, and up pops a menu containing "comments/suggestions/bugs." That's what lead me to
2020-01-15 00:15:36  fds
Okay, I see it's not about mobile after all, simply based on window size.
2020-01-15 00:15:58  Jack Douglas
ah yes, I remember now
2020-01-15 00:16:00  fds
I don't keep my browser windows terribly wide.
2020-01-15 00:16:25  Jack Douglas
I've fixed the link in the hamburger now too :)
2020-01-15 00:16:43  Jack Douglas
sorry you had to go to all that trouble to report the issue!
2020-01-15 00:17:59  fds
It's all right-- only shame it probably would have benefitted you more if I spent that time over here rather than there. Thank you again! The speed at which you fix things is scary.
2020-01-15 00:18:52  Jack Douglas
heh, not everything, but thanks for the encouragement on that too
2020-01-15 00:19:12  Jack Douglas
I'm very pleased we have Firebird on db<>fiddle, it is a nice little database
2020-01-15 00:22:13  fds
Got to be honest I only play with Firebird for fun & fascination myself, but I also keep it running on my home server. I was amusing myself with this when I noticed the bug.
2020-01-15 00:22:47  fds
A fun silly recursive CTE example from the SQLite docs that as an experience I adjusted to work in any dbms I could find. :)
2020-01-15 00:24:19  fds
Still got to try and make it work in Db2 thank to dbfiddle, never had that installed myself.
2020-01-15 00:24:39  Jack Douglas  replying to  fds
that's brilliant
2020-01-15 00:25:15  Jack Douglas
on a par with this:
2020-01-15 00:25:18  Jack Douglas
2020-01-15 00:25:46  fds
Haha, here are the other versions that are 100%:
2020-01-15 00:26:49  fds
But again, I take no credit, it's simply from the SQLite manual,
2020-01-15 00:27:01  fds
All I'm doing is entertaining myself by porting it over to everything else.
2020-01-15 00:29:34  Jack Douglas  replying to  fds
2020-01-15 00:29:45  Jack Douglas
in case you haven't done postgres yet :)
2020-01-15 00:30:40  Jack Douglas
have you got a link to the sqlite version, I'd like to tweet it…
2020-01-15 00:31:15  fds
Nah, the PG version was one of the first and most amenable. :) I posted all the links, including the SQLite one above, that's verbatim from the manual
2020-01-15 00:31:52  fds
I'm yet to tackle Db2, and I'm fighting some kind of string truncation issue with MySQL 8, at least on dbfiddle.
2020-01-15 00:33:16  fds
Note in PG you don't need that ugly double SELECT because it allows you to ORDER BY on string_agg(). That hack was solely for the Firebird version.
2020-01-15 00:36:26  Jack Douglas
ah, didn't notice that message scrolled!
2020-01-15 00:37:32  fds
Yeah, I'm probably doing something wrong, and not properly marking what I'm typing as being "replies to you," but unsure how to do that.
2020-01-15 00:38:05  fds  replying to  Jack Douglas
let's see
2020-01-15 00:38:12  fds
got it :)
2020-01-15 01:25:29  fds
Figured out what was wrong with the MySQL 8 version, happy to say it wasn't a dbfiddle issue. There's simply a default 1K limit on group_concat(), but you can override.
2020-01-15 02:04:01  fds
Finally, Db2: This was fun :)
2020-01-15 07:09:11  Andriy M  replying to  Jack Douglas
Hey Jack, very nice, it [does appear]( to be working correctly now. However, when I tried to re-run [the same test script](, it failed for me. As far as I've managed to establish, it's `datetimeoffset` that breaks it. That is to say, any script with a `datetimeoffset` in it seems to fail now, so I can't give you a link for testing, but you can just try running `SELECT SYSDATETIMEOFFSET();` in any version of SQL Server to see for yourself.
2020-01-15 09:02:00  Jack Douglas  replying to  Andriy M
ooh yes that's not good — I think I worked it out now though thanks again for helping me with all this (also @PeterVandivier)
2020-01-15 09:15:36  Andriy M  replying to  Jack Douglas
Confirmed, things are back to normal in this department, great job!
2020-01-15 09:18:21  Jack Douglas
2020-01-15 09:18:55  PeterVandivier
2020-01-15 09:24:27  Andriy M
I wish could suggest something on the presentation part, I mean in particular reflecting the `n` of `datetime2(n)` correctly in the output
2020-01-15 09:27:04  Andriy M
However, I did notice something peculiar. You can also specify an arbitrary precision for the `time` type as well. And I've noticed that `time(0)` at least shows without the fractional part, while `datetime2(0)` still shows the `.0000000`, [demo](
2020-01-15 09:28:31  Andriy M
All of `time(1)` through `time(7)` show 7 digits in the fractional part, though
2020-01-15 09:31:16  Jack Douglas  replying to  Andriy M
I looked into this a bit, but couldn't work out how to get the precision info in .net
2020-01-15 09:31:37  Jack Douglas
there were hints it might be possible, but it may well be beyond me
2020-01-15 09:31:57  Jack Douglas
as soon as we have a community here I'll post a question :)
2020-01-15 09:33:04  Andriy M  replying to  Jack Douglas
Oh, will it be a dedicated community for VB.NET? Are you intending to keep those people apart from everyone else?
2020-01-15 09:39:22  Andriy M  replying to  Jack Douglas
What about the time zone info for `datetimeoffset`? Dbfiddle currently [doesn't seem to show the time zone]( Is that a problem too at the moment?
2020-01-15 09:54:32  Jack Douglas  replying to  Andriy M
that shouldn't be a problem — it just needs to be added to the format string
2020-01-15 09:54:40  Jack Douglas
I will add that at some point
2020-01-15 09:55:17  Jack Douglas  replying to  Andriy M
I'd like to split the SO scope into smaller communities here if we can get people interested
2020-01-15 10:27:51  Andriy M  replying to  Jack Douglas
2020-01-15 10:28:31  Andriy M  replying to  Jack Douglas
Makes total sense to me, though at the same time, it may not be always obvious where to draw the boundaries
2020-01-15 10:29:21  Andriy M
I guess it's a discussion worth its own meta question
2020-01-15 10:30:00  Andriy M
Or a few
2020-01-15 10:31:10  Jack Douglas
all very true!
2020-01-15 10:33:39  Jack Douglas  replying to  Andriy M
what programming languages do you use, presumably not still delphi?
2020-01-15 10:58:50  Andriy M  replying to  Jack Douglas
Actually I do still work with Delphi and almost nothing else, I mean as a major part of my work and not counting SQL Server. I've had to deal with an ASP script or two a few times, and also do some Windows batch scripting to solve some tasks and as a helping tool but that's just my choice. I know PoSh would be much superior most times, but it requires learning and I've been increasingly lazy learning new stuff lately
2020-01-15 11:01:29  Andriy M
I know some other things as well, a little, as I've never used them professionally and haven't created anything major using them to claim I *actually* know them
2020-01-15 11:06:57  Jack Douglas  replying to  Andriy M
if we had a Delphi community here would you be interested? I'm asking as an academic question — we'd probably see if things like C# would fly first, but I'd be interested to know nonetheless.
2020-01-15 11:24:07  Andriy M  replying to  Jack Douglas
Not sure. When I first joined Stack Overflow, I thought Delphi would be speciality but I quickly found how much less I know about it than some other people. Our company used a rather outdated (pre .NET) version at the time, even though at least Delphi XE2 was already around then. We *have* switched to XE8 since then, but I'm still not very versed in the newer features, so I wouldn't be able to provide much help.
2020-01-15 11:44:57  Jack Douglas  replying to  Andriy M
interesting, thanks :)
2020-01-16 10:06:58  PeterVandivier
@Jack is it worth trying to close / lock the [ post]( ?
2020-01-16 10:08:55  PeterVandivier
give it a "historical significance" banner and stop new answers?
2020-01-16 11:24:34  PeterVandivier
@Jack i know you're suuuuper busy with TA atm, but just to bump the issue :), [this seems like a freebie plug for cockroachdb on db<>fiddle](
2020-01-16 11:28:14  Jack Douglas  replying to  PeterVandivier
heh that does show it would be useful :)
2020-01-16 11:28:56  Jack Douglas
we are moving DC over the next few weeks/months so if it does happen it'll be after that
2020-01-16 11:29:19  Jack Douglas
probably worth moving any active feature requests you have here though?
2020-01-16 11:31:31  Jack Douglas  replying to  PeterVandivier
I don't know if it's even necessary — the real problem is that the link in the hamburger was still pointing at the post!
2020-01-16 11:31:55  Jack Douglas
fds — [a day ago](  
>Not on mobile, Mac laptop Apple Safari, but it very well might be mistaking that for mobile. Now I'm wondering what the "real" site is supposed to look like. Everything still functioned fine. I see like a play button / hamburger menu, click on it, and up pops a menu containing "comments/suggestions/bugs." That's what lead me to
2020-01-16 11:39:46  PeterVandivier  replying to  Jack Douglas
~~whaaaaaaat?! 😱 we're not gonna be neighbors anymore?? 😢~~  
misread as "_moving_ **to** _DC_", plz disregard
2020-01-16 11:41:47  Jack Douglas
moving to DC might be nice, idk :) 
2020-01-16 11:42:32  PeterVandivier
2020-01-19 07:44:38  fds
MySQL 5.7 appears to be out of disk space at the moment: (Errcode: 28 - No space left on device)
2020-01-19 11:50:35  Jack Douglas  replying to  fds
thank you, that's all fixed now!
2020-01-20 18:48:35  Andriy M
@Ami This reminds me of a similar issue I used to have when searching with Google. Basically, I would first enter one request, say 'foo', click 'Search', get the results, then enter 'bar' into the search box, 'Enter', get new results, then click 'Back', and I would get the previous search results, for 'foo', but the search box would show 'bar'. The last time I experienced that was quite some time ago, and I never found out for sure if it was an issue with the site or my browser, but I believed it to be with the browser (or browser + OS maybe). For the record, I don't experience the issue you've described in my Firefox on Win 7.
2020-01-21 13:59:24  GeorgePalacios  replying to  Andriy M
That still happens to me all the time lol!
2020-01-21 14:01:44  Andriy M  replying to  GeorgePalacios
In Firefox?
2020-01-21 14:29:18  GeorgePalacios  replying to  Andriy M
I get it in Chrome
2020-01-21 15:31:12  Andriy M
Based on my past experience, I'm guessing Chrome would do this on any page that has an input form. And of course would be a typical example, which is why the issue has been raised. It used to work like that for me in Firefox, but apparently they've since changed/fixed it
2020-01-22 15:23:12  Jack Douglas
@John are you really asking if we can have some generic mechanism for executing batches multiple times? That sounds like something that might be possible. What would we do with the output from the batches though — presumably not just display it all (like this:
2020-01-22 15:23:56  Jack Douglas
I guess we could compare the output and if they are all the same then print just once (or should the 'repeat' count)
2020-01-22 16:02:51  Andriy M
@John @Jack `GO` is a client-side feature. In addition to having to figure out how to handle multiple outputs it can produce, implementing it may also pose a question whether (or which) other client-side features should be implemented as well (including those of other platforms). Or does dbfiddle have any implemented already?
2020-01-22 16:04:16  Andriy M
It's like the MySQL `DELIMITER` instruction. Does dbfiddle support MySQL's `DELIMITER`?
2020-01-22 16:47:47  PeterVandivier  replying to  Andriy M
[deja vu 😜](
2020-01-22 16:59:41  Jack Douglas  replying to  Andriy M
I don't think it does
2020-01-23 07:41:39  Andriy M  replying to  PeterVandivier
Ah yes, we've already had that discussion
2020-01-28 11:47:36  John aka hot2use  replying to  Jack Douglas
I'm not sure about the "generic" bit, but a mechanism to execute something x number of times would be cool.
2020-01-28 11:49:28  John aka hot2use  replying to  Andriy M
Seeing as `GO` seems to be a client-side "operation" as per Andriy's feedback, I could retract my suggestion altogether. 
2020-01-28 11:49:36  John aka hot2use
Opinions? Anybody?
2020-01-28 12:21:50  PeterVandivier
i've been bit by it and it would be nice, but imo it's not the biggest deal
2020-01-28 12:23:18  PeterVandivier
if & when db<>fiddle is open-sourced a la TA, it'd be a fun "up for grabs" type issue, but i don't think it'd important enough to divert Jack or James from other work atm
2020-01-28 12:24:27  PeterVandivier
esp. b/c it seems like the kind of thing that _seems_ simple enough, but might actually require some non-trivial work on the backend to support it
2020-01-28 12:48:02  Andriy M
I'm only slightly worried by the fact that it would created a precedent for others to request all sorts of client side features to be implemented for other DBs as well, and some of them might be either too difficult or too rubbish but Jack would be too nice to refuse and try to implement them anyway, and dbfiddle might end up unnecessarily complicated, in terms of both maintenance and usage
2020-01-28 12:49:23  Andriy M
`GO number` itself already raises questions, for instance how to handle multiple result sets, especially if someone accidentally runs it with a big number
2020-01-29 08:12:09  John aka hot2use
I have retracted the feature request, edited it and left it here for historical reasons.
2020-01-29 08:41:51  Jack Douglas  replying to  John aka hot2use
thanks @John but do keep them coming :)
2020-01-29 08:53:56  John aka hot2use  replying to  Jack Douglas
I certainly will
2020-01-30 20:09:41  Jack Douglas
@PeterVandivier that's actually deliberate because when there are several it can slow down rendering unacceptably. We have a plan to fix it though by caching db<>fiddle results on the client. Will update your post when that lands!
2020-01-30 20:39:50  PeterVandivier  replying to  Jack Douglas
cool beans & no worries! glad to have it referenced explicitly though. fwiw - i think it'd be totally acceptable to throttle fiddle one-boxing in chat to one or 2 batches or something sensible like that. tbh, i was actually a bit surprised/impressed that fiddle one-boxing in chat is supported at all so hear it needs to be throttled a bit makes a lot of sense
2020-02-03 14:56:36  PeterVandivier
@Jack should there be a community etiquette around max size of embedded fiddles? i was going to edit those firebird answers to be one-boxed but the (first one at least) was large enough that i paused. 
2020-02-03 17:42:06  Jack Douglas  replying to  PeterVandivier
you can hide some of the batches, though topanswers doesn't honour that yet — it should though and we'll no doubt get round to that because it'd be useful here
2020-02-04 08:52:16  Jack Douglas
@Arioch do you know about the 'split batch' button?   
![Screenshot 2020-02-04 at 08.49.49.png](/image?hash=244b61bb24957a970d9e7d8d73ab3e4af579ea29d9a37b1408e18de089ec60e2)  
 I don't think the PHP Firebird/Interbase driver supports multiple statements per batch like some others do so we can't enable this in db<>fiddle.
2020-02-04 09:28:07  Arioch  replying to  Jack Douglas
@JackDouglas - split batch is not a problem, copy-paste is not hard to do. The problem is visual clutter and noise. Non-grouped statements are harder to read. It is pity DB-Fiddle can not split the script into statements and feed them to server one by one...
2020-02-04 11:04:02  Jack Douglas  replying to  Arioch
I agree on all counts — however db<>fiddle has a golden rule: we do not parse SQL that is being processed. It's a rule that has served us well so far, it saves us from all sorts of complexity down that route.
2020-02-20 10:20:36  PeterVandivier
looks like the cached previous post may be back. just went in to post a new answer and i'm seeing the "brittle fiddle one-boxing" rev 1 looks like
2020-02-20 10:21:18  PeterVandivier
i'll leave it for a few mins if you want me to check anything on my end
2020-02-20 23:29:58  Jack Douglas  replying to  PeterVandivier
I think that might be a hangover from the earlier bug — it caches answers to each question individually so if you answered here while the bug was active, then try and answer after it's fixed, you'll get the old text. Only once though.
2020-02-20 23:30:23  Jack Douglas
this is one of the few places you'd notice because generally we don't post multiple answers to the same question…
2020-03-03 15:51:24  PeterVandivier
i feel like it might be too cluttered, but just because the thought crossed my mind... how do you feel about using "gitmoji" style tags for this post?
2020-03-03 15:52:11  PeterVandivier
e.g. - instead of `bug`, you have `🐛bug`, instead of `declined`, you have `🚫declined`
2020-03-03 15:53:00  PeterVandivier
just thinking that as the number of answers grows ever _huger_, the tags in aggregate are sort of making my eyes glaze over when i expand the whole list
2020-03-03 15:53:20  PeterVandivier
wondering if a visual "helper" indicator is prudent
2020-03-03 15:54:24  PeterVandivier
2020-03-03 15:54:31  PeterVandivier
by way of further reading
2020-03-03 15:55:21  PeterVandivier by way of constructive real-world example
2020-03-03 19:39:57  Jack Douglas  replying to  PeterVandivier
we are still thinking about answer tags — not really specifically for this post or dba.ta at all, but for Code Golf
2020-03-03 19:40:27  Jack Douglas
at some point we'll move db<>fiddle issues entirely to GitHub I think so it won't be an issue here
2020-03-03 19:41:09  Jack Douglas
I'm not sure answer tags would be useful anywhere other than CG or possibly TeX, but they'd be available
2020-03-03 20:47:06  PeterVandivier
Oh cool, I wasn’t even thinking proper tags, tbh. Just of using emoji in the preformatted text. I’m all for solving the general problem though rather than the local one 🙂
2020-03-09 04:37:12  Bob Zeller
This is great! The functionality is all there unlike sqlfiddle that coughs violently. My only suggestion is to secure a graphic designer to make it look less pre-911. The graphics and layout have that UNIX macho aesthetic designed to intimidate the unseasoned and terrify the novice. Lot's of good work going on here.
2020-03-09 11:37:23  Jack Douglas  replying to  Bob Zeller
thanks Bob, will bear in mind the UX tips :)
2020-03-09 12:25:13  PeterVandivier
re: null-row collapse. i assume you've also previously thought about adding the "_null_" text placeholder to differentiate empty-string cells from true `null` cells. was there a compelling reason to leave it out? lots of PITA config between drivers for the different engines or something?
2020-03-09 12:25:37  PeterVandivier
or has it just not come up before?
2020-03-18 14:31:31  PeterVandivier
just noticed that it's only integration here - web portal has the null placeholder :-P
2020-03-18 14:32:39  PeterVandivier
related: is something up with MySQL 8.0 on db<>fiddle atm? can't seem to run it @Jack 
2020-03-18 14:32:42  PeterVandivier
2020-03-18 14:53:31  Jack Douglas  replying to  PeterVandivier
yes that's right
2020-03-18 14:53:59  Jack Douglas
there are a few similar issues
2020-03-18 14:54:02  Jack Douglas
::: quote 37 17227 ba1d6fd8881a87067fe28d71a5ddc594f0146228d32fe77a6399565d3b928bd8 212,223,236 62,103,153
Jack Douglas replying to PeterVandivier[ *— a month ago*](#c17227)  
>you can hide some of the batches, though topanswers doesn't honour that yet — it should though and we'll no doubt get round to that because it'd be useful here
2020-03-18 14:54:18  Jack Douglas
^^^ apart from that one I mentioned earlier
2020-03-18 14:58:39  Jack Douglas  replying to  PeterVandivier
fixed, thanks — some log files had expanded to fill the hard drive
2020-03-18 15:13:06  PeterVandivier
friggin log files always up to no good
2020-03-31 07:11:12  PeterVandivier
::: quote 4 32429 70ac1a052b69788b840dc7af631544ba35aa71997240f3a4b5b0b7f57bc166a4 212,223,236 62,103,153
Josh Darnell *— 11 hours ago* *in [stop the merry-go-round](/databases?room=4#c32429)*  
>You know what would be cool?  If @Jack had popular sample databases up on db<>fiddle.
2020-03-31 07:14:05  PeterVandivier
i was thinking maybe a "featured fiddle" could be nice. idk if there's anything like a "popular" fiddle or something that gets a lot of hits, but it might be fun to tweet out interesting fiddles once a week or something
2020-03-31 07:14:14  Jack Douglas  replying to  PeterVandivier
Yes, and it's been requested before (by @Ami)
2020-03-31 07:14:59  Jack Douglas
I'm doing a whole bunch of work on db<>fiddle right now and this is something I'm keen to add if possible
2020-03-31 07:15:55  PeterVandivier
i wonder if there's a sensible way to parse out at some step which objects are used in the fiddle and drop all others so you don't burn through disk 
2020-03-31 07:16:36  PeterVandivier
anyway, i'd be happy to volunteer with porting the candidate databases between platforms assuming that's helpful
2020-03-31 07:17:33  Jack Douglas  replying to  PeterVandivier
that's a very kind offer. I suspect each RDBMS will want it's own idiomatic sample DB though
2020-03-31 07:17:58  PeterVandivier
and assuming that's something you were thinking you want (AdventureWorks on SQLite, DVDRentals on SQL Server, etc)
2020-03-31 07:18:00  Jack Douglas
so "Sakila" for MySQL, "AdventureWorks" for SQL Server
2020-03-31 07:18:31  PeterVandivier
lol, might make sense to just use the popular ones where they'll be used, eh?
2020-03-31 07:19:07  Jack Douglas  replying to  PeterVandivier
how do you mean?
2020-03-31 07:20:01  Jack Douglas
2020-03-31 07:21:08  Jack Douglas
2020-03-31 07:21:23  Jack Douglas  replying to  PeterVandivier
Is DVDRentals 'the' sample db for Postgres?
2020-03-31 07:21:24  PeterVandivier
yea, i was thinking there'd be cross-platform support for each database rather than limiting AdventureWorks to just sql server for example
2020-03-31 07:21:40  PeterVandivier  replying to  Jack Douglas
2020-03-31 07:21:54  PeterVandivier
but i suppose thats a bit silly
2020-03-31 07:22:25  Jack Douglas  replying to  PeterVandivier
it has the one big advantage of being useful when comparing different rdbms's
2020-03-31 07:23:04  Jack Douglas
but perhaps there only needs to be one sample database that works across dbs
2020-03-31 07:23:14  Jack Douglas
plus the idiomatic one for that platform
2020-03-31 07:23:36  PeterVandivier  replying to  Jack Douglas
that's rather what i was thinking, yea
2020-03-31 07:24:08  PeterVandivier
the current experience being that you can click between platforms with the same hash
2020-03-31 07:24:26  PeterVandivier
but idk how many people besides me do that frequently
2020-03-31 07:24:26  Jack Douglas
I'd quite like to make the topanswers data dump a regularly updated sample db in postgres — that kind of kills the SEDE bird very simply
2020-03-31 07:25:02  PeterVandivier
lol, yea that'd be a very pleasing symmetry
2020-04-07 09:21:46  Jack Douglas
@fiatjaf thanks — this is a bug with the Ace editor. It's also visible on some of the db<>fiddle help pages such as [this one]( where we've used enclosed numerics. CodeMirror (which we use here on TopAnswers instead of Ace) does not seem to have the same issues — and also plays much better on mobile. I think on balance we should switch to CodeMirror for db<>fiddle too, but my hesitation is down to the syntax highlighting, which is quite different :  
what do you think about the CodeMirror highlighting? (the first one above — you can test [here](
2020-04-16 13:05:00  Andriy M
@fiatjaf I'm guessing this ligature thing (and probably the issue you're experiencing) is browser-specific, because my browser (FF, ironically enough) doesn't replace two consecutive fs with the ligature character.
2020-04-16 13:11:11  Andriy M  replying to  Jack Douglas
To be honest, I like db<>fiddle's current colour scheme slightly better, although it's a very minor issue for me, perhaps I'm just more used to those colours (because they are very similar to what SSMS uses). On the other hand, db<>fiddle's (Ace's?) syntax highlighter doesn't seem to recognise some keywords (e.g. `with`, as can be seen from your screenshots).
2020-04-16 15:51:43  Jack Douglas  replying to  Andriy M
thanks for the input @Andriy, I don't use SSMS much so that's good to know. It does fell like on balance a switch would be better. I might even be able to tweak the colour scheme!
2020-04-16 19:53:04  Jack Douglas
::: quote 2 37572 ba1d6fd8881a87067fe28d71a5ddc594f0146228d32fe77a6399565d3b928bd8 212,223,236 62,103,153
Jack Douglas *— a few seconds ago* *in [The Heap](/databases?room=2#c37572)*  
>@PeterVandivier do you know you can link your account from your profile?
>@@@ answer 1058
2020-04-16 21:30:17  PeterVandivier
Bit sleepy but so I don’t forget - can’t seem to paste a fiddle from my iPhone 
2020-04-16 21:31:18  PeterVandivier
2020-04-17 08:03:01  PeterVandivier  replying to  Jack Douglas
> 21000: ERROR: more than one row returned by a subquery used as an expression CONTEXT: SQL function "set_se_user_id" statement 5
2020-04-17 11:19:56  Jack Douglas  replying to  PeterVandivier
hmm, that looks like it was broken for a while. I think I have fixed the issue and along the way finished off the feature that allows you to link to multiple SE sites if necessary. Would you mind trying again?
2020-04-17 12:08:19  PeterVandivier  replying to  Jack Douglas
2020-04-17 12:16:59  Jack Douglas  replying to  PeterVandivier
nice one ;)
2020-04-24 08:59:00  PeterVandivier
have we discussed "pinned" fiddles here before? like, "best of" or really good ones getting named handles? 
2020-04-24 09:12:19  Jack Douglas  replying to  PeterVandivier
how do we find them?
2020-04-24 09:12:25  Jack Douglas
you mean have a contest or something?
2020-04-24 09:34:09  PeterVandivier
no, not a contest, just like vanity URLs. so like, would be the postgres one, maybe /help would be the fiddle at the help link, then you could add others as and when you come across really well structured or intriguing ones
2020-04-24 09:34:32  PeterVandivier
or sorry, /noughts-and-crosses rather than /tic-tac-toe :-P
2020-04-24 09:34:46  PeterVandivier
or both 🤷‍♂️
2020-04-24 09:37:20  PeterVandivier
maybe build out like a "help section" index -   
* `/rdbms=sqlserver_2019?howto=pivot`  
* `/rdbms=sqlserver_2019?howto=unpivot` 
* `/rdbms=sqlserver_2019?howto=alter-column`
2020-04-24 09:37:37  PeterVandivier
stuff like that
2020-04-24 09:38:11  PeterVandivier
possibly even farm examples from the official documentation
2020-04-24 09:38:40  PeterVandivier
might be a nice synergy with support for AdventureWorks if that happens
2020-04-24 09:53:03  PeterVandivier
and since i'm making grandiose plans for you... :-P could consider automagically embedding featured fiddles into "How To/FAQ" questions on TA to make them searchable here too, which might help segregate "beginner" tier questions out of the more 3l337 feed currently present on dba.ta
2020-04-24 10:15:20  Jack Douglas  replying to  PeterVandivier
that is definitely on the cards
2020-04-24 10:18:37  Jack Douglas  replying to  PeterVandivier
howtos are better suited to some sort of post here that on db<>fiddle
2020-04-24 10:19:18  Jack Douglas
now I think of it the 'help' link should definitely be a link to a post here :)
2020-04-24 10:19:51  Jack Douglas
they are a bit odd currently with the blurb as comments in the batches
2020-04-24 10:20:25  Jack Douglas  replying to  PeterVandivier
> as and when you come across really well structured or intriguing ones      

I don't see any more than you do!
2020-04-24 10:20:41  Jack Douglas
there is no way of discovering interesting fiddles in the 800k that exist
2020-04-24 10:21:09  PeterVandivier
lol, not like you ought to go through the database, but in posts are the interwebs. 
2020-04-24 10:21:38  Jack Douglas
what, search all the posts on the internet?
2020-04-24 10:21:43  PeterVandivier
when you're a-scrollin' through or dba.ta and see a fiddle you like that you think is worth pinning
2020-04-24 10:22:12  Jack Douglas
they aren't usually that valuable out of context
2020-04-24 10:22:50  Jack Douglas
the noughts-and-crosses one is a bit of an exception
2020-04-24 10:23:20  PeterVandivier
🤷‍♂️ i was just thinking about it because i saw a perennial "_how do i pivot data_" question on this morning
2020-04-24 10:23:37  PeterVandivier
felt like that ought to be an FAQ
2020-04-24 10:24:02  Jack Douglas  replying to  PeterVandivier
we need canonical questions and answers here I think — they can then be linked in other answers
2020-04-24 10:24:08  PeterVandivier
perhaps better suited to dbfiddle than either ta or se since the concept is the same, just paginated by platform (& version)
2020-04-24 10:25:09  Jack Douglas
you mean examples?
2020-04-24 10:27:13  Jack Douglas
I am working on a 'documentation' post type in the PHP community (not live yet) that is structured like the PHP docs — the 'answers' are examples and warnings etc (and 'tagged' accordingly)
2020-04-24 10:27:42  Jack Douglas
this doesn't sound like quite the same — more 'concepts' than 'documentation'
2020-04-24 10:28:30  PeterVandivier
yea, very specifically "bite-size" stuff. like snippets but online
2020-04-24 10:29:14  PeterVandivier
* `/rdbms=sqlserver_2019?howto=rowcount-all-tables`  
* `/rdbms=postgres_12?howto=rowcount-all-tables`
2020-04-24 10:29:53  Jack Douglas
if it's here, the examples could be sorted by how useful people actually found them
2020-04-24 10:30:00  PeterVandivier
with a few of the syntax options for each, and maybe the information_schema example as the last batch 
2020-04-24 10:30:58  Jack Douglas
I will give some thought to a 'concepts' post type on databases.ta I think that could fly
2020-04-24 10:31:15  PeterVandivier  replying to  Jack Douglas
true, but then it's less canonical. and if a post changes over time, it still retains all the stars it ever earned prior to the most recent edit
2020-05-11 12:56:55  PeterVandivier
Odd behavior in fiddle code blocks on iOS at present - render as blank then fill on-click
2020-05-13 08:07:18  PeterVandivier
@Solomon seems like this might be a two-parter.   
1. please publish lifecycle SLAs for dbfiddle  
2. `/v2` endpoint supporting `?rdbms` and `?version` as separate inputs  
2020-05-13 08:09:06  PeterVandivier
for 1. ( and Jack can correct me if I'm lying), I believe that even if a version "ages out" of dbfiddle, the fiddle link itself should remain live because the results are cached until it can be re-run. and if it can't be re-run, then the cache can't expire (although it might be worth testing this is the case for this scenario)
2020-05-13 08:11:38  PeterVandivier
for 2. i like it! :D but i might suggest "`latest_supported`" and "`earliest_supported`" as the args instead of trying to get fancy with numbers. the only problem i can see now is that "`earliest_supported`"  begs for a "`not_after`" companion arg if (for example) you know a demo demonstrating breakage in an early version will "start working" at some point
2020-05-13 09:52:03  Jack Douglas  replying to  PeterVandivier
this is correct — all existing fiddles should *display* — if they don't it's a bug
2020-05-13 09:52:22  Jack Douglas
but if the back-end is retired, they can't be *run* (or edited and run)
2020-05-14 08:52:40  Jack Douglas
@Solomon your post here is worthy of getting a proper answer — would you mind reposting as a general 'meta' question? I think we might close this thread to new answers and do it that way instead going forward as it will be more manageable
2020-05-14 08:53:33  Jack Douglas
the short answer however is that all valid fiddle links (that have not been url edited) *should* work forever
2020-05-14 08:54:24  Jack Douglas
'work' in the sense that they should be viewable — if the back-end has been retired, and the person looking at the fiddle wants to change it and run it, they have to switch to an available backend
2020-05-14 09:18:52  PeterVandivier
@Jack in the `hide=N` arg, is that a bitmask?
2020-05-14 09:59:31  Jack Douglas  replying to  PeterVandivier
2020-05-14 09:59:54  Jack Douglas
if it is shorter that the number of batches the extras are assumed visible
2020-05-14 10:16:20  PeterVandivier
re: the "fun fiddles" talk from awhile back, found this chestnut again just now :-)   
2020-05-14 10:35:37  Jack Douglas  replying to  PeterVandivier
How about a 'poll' question on databases.ta: 'whats your favourite db<>fiddle?'
2020-05-14 10:38:14  PeterVandivier  replying to  Jack Douglas
haha, perhaps. it's not something i _actually_ feel strongly about. just found that fiddle re-reading through an old thread and it sparked joy
2020-05-14 10:38:24  PeterVandivier
felt compelled to share
2020-05-14 10:49:26  Jack Douglas
now it's shared more widely :)
2020-05-14 10:49:40  Jack Douglas
I might link back to that post from db<>fiddle at some point
2020-05-14 11:15:44  PeterVandivier
2020-05-14 15:41:13  Solomon Rutzky  replying to  Jack Douglas
Even though the question is about db<>fiddle irrespective of this site? Just makin sure...
2020-05-14 18:06:25  Jack Douglas  replying to  Solomon Rutzky
yes please — we can say that 'meta' questions can be about TA or about the tools like db<>fiddle and tio, thanks @Solomon
2020-05-19 04:37:41  Solomon Rutzky  replying to  Jack Douglas
Ok. Should it be a "meta.ta" question, or a "databases.ta" question tagged as [meta]?
2020-05-19 05:54:34  Jack Douglas  replying to  Solomon Rutzky
databases.ta tagged as Meta please :)
2020-05-21 14:02:56  PeterVandivier
@Solomon if you wanted to add some of the fiddles from that blog to the "Fiddle Hall of Fame", i think they're worth of some praise :-)  
@@@ question 1017
2020-05-21 14:04:16  PeterVandivier
@Jack is there anything already on your radar for letting vanilla users post as "Community"? or something like that? I'd be happy to put those fiddles up, but i wouldn't want to poach the ⭐️s
2020-05-21 14:14:19  Jack Douglas  replying to  PeterVandivier
that's automatic for polls
2020-05-21 14:14:54  Jack Douglas
similar to wiki questions (which are credited to community automatically whoever posts them)
2020-05-21 14:21:16  PeterVandivier
oh nice!
2020-05-21 14:25:13  PeterVandivier
now i just need to pick the "best" fiddles from @Solomon's blog :-P 
2020-05-21 14:28:10  PeterVandivier
fwiw though actually @Solomon - re-reading your post, i think the issue you're hitting is b/c the fiddles were never run on the deprecated platforms before deprecation. i don't think we addressed it _directly_, but that's something i've run into before as well - when you want to farm a demo across a bunch of different platforms - you have to hit the "`Run`" button manually for each one in order for the resultset to get generated
2020-05-21 14:28:52  PeterVandivier
the `Run` button does a POST call that is needed in order for the subsequent GET to be honored
2020-05-21 14:29:21  PeterVandivier
but once it's posted, then your cache is safe "forever"
2020-05-21 14:30:31  PeterVandivier
you just can't get results for existing fiddles on deprecated platforms _where they were never run_ because there isn't a cached set _for that backend_ and the backend isn't there anymore to make one
2020-05-21 14:31:58  PeterVandivier
but you can repro that same breakage between supported backends too if you skip the "`Run`" button for one of them
2020-06-09 07:41:57  Anonymous 1318
Hi, I was wondering whether db<>fiddle  still supports PostGIS for the Postgres instances? Currently I have some trouble with data type geometry. However, I believe the PostGIS support once existed, I am right?
2020-06-09 07:44:05  Jack Douglas  replying to  Anonymous 1318
that's correct, it was supported at one point
2020-06-09 07:44:16  Jack Douglas
let me take a look and see what happened
2020-06-09 07:46:34  Anonymous 1318
2020-06-09 07:51:14  Jack Douglas  replying to  Anonymous 1318
what pg version are you using or does it not matter?
2020-06-09 07:55:14  Anonymous 1318
generally it doesn't matter. However, I would like to use your version 12 instance for that
2020-06-09 08:06:30  Jack Douglas  replying to  Anonymous 1318
2020-06-09 08:06:55  Jack Douglas
not sure if that means it is definitely working, please let me know if it isn't
2020-06-09 08:09:53  Anonymous 1318
I'll have a look, one moment please :)
2020-06-09 08:14:23  Anonymous 1318
Looks good! Thank you so much for your incredible fast response!
2020-06-09 08:24:15  Jack Douglas  replying to  Anonymous 1318
you are welcome :)
2020-06-09 08:34:29  PeterVandivier
2020-06-15 16:57:59  Solomon Rutzky  replying to  PeterVandivier
Thanks for all of that info. Sorry to not be more responsive, but all has been chaos since mid March as lockdown has meant 3 kids, too young to be self-sufficient, have been home 24/7 and my wife and I both work (well, I try to, but 😿). Anywho, my question was equally about deprecated as well as versions that have never been supported, or entirely invalid "versions". I will try to post this in meta soon. Part of why I haven't is that your info re: answered the deprecated version part, but then I started re-thinking the actual question so I'm trying to get clear on that first.
2020-06-15 18:23:05  Solomon Rutzky  replying to  PeterVandivier
Thanks 😺. As I mentioned, things have been a bit chaotic lately. I have a little bit of cleanup to do on some of them, I think. Also, not sure what to submit as they are all essentially the same thing, just for different RDBMSs.
2020-06-16 08:12:18  PeterVandivier  replying to  Solomon Rutzky
no worries at all, i've just got the one toddler (and my wife is on reduced working hours, too) and we were still feeling a little overwhelmed, so i can only imagine how you must be feeling.