12:00 PM
Mr_Monkey
That's good to know.
2019-03-10 06933, 2019
12:00 PM
samj1912
well - here's the thing
2019-03-10 06951, 2019
12:00 PM
samj1912
shifting to Sir would mean you don't have to worry about scaling the indexer later
2019-03-10 06902, 2019
12:01 PM
samj1912
but it would mean having a python sqla orm
2019-03-10 06916, 2019
12:01 PM
samj1912
if you want - you could use your existing JS orm to shift to solr
2019-03-10 06923, 2019
12:01 PM
samj1912
I would leave it to you to prioritize
2019-03-10 06937, 2019
12:01 PM
samj1912
the current schema is simple enough that the JS one can be used
2019-03-10 06959, 2019
12:01 PM
Mr_Monkey
That could be a good stepping stone.
2019-03-10 06907, 2019
12:02 PM
samj1912
but in the future if you want to scale to a complex schema/faster indexer you can shift to sir
2019-03-10 06943, 2019
12:02 PM
samj1912
anyway - you would first want to prioritize this
2019-03-10 06950, 2019
12:02 PM
Mr_Monkey
I think that's a good way forward, that doesn't seem to me to hinder later needs
2019-03-10 06905, 2019
12:03 PM
samj1912
so you want to move to sir?
2019-03-10 06930, 2019
12:03 PM
Mr_Monkey
No, I think using the existing ORM might be more manageable
2019-03-10 06935, 2019
12:03 PM
samj1912
ah okay
2019-03-10 06936, 2019
12:03 PM
samj1912
cool
2019-03-10 06943, 2019
12:03 PM
samj1912
so I would start with this
2019-03-10 06957, 2019
12:03 PM
samj1912
look up existing Solr JS api helpers
2019-03-10 06905, 2019
12:04 PM
samj1912
figure one you'd like to use
2019-03-10 06913, 2019
12:04 PM
samj1912
and make a simple solr schema for now
2019-03-10 06949, 2019
12:04 PM
samj1912
2019-03-10 06903, 2019
12:05 PM
samj1912
instead of tag you will have name
2019-03-10 06911, 2019
12:05 PM
samj1912
everything else I guess remains the same
2019-03-10 06918, 2019
12:05 PM
samj1912
id = bbid
2019-03-10 06929, 2019
12:05 PM
samj1912
you have a ngram for autocomplete
2019-03-10 06945, 2019
12:05 PM
samj1912
and maybe throw in a "type" field
2019-03-10 06913, 2019
12:06 PM
Mr_Monkey
OK, indeed that's simple enough. Would that be all that is needed for a simple schema then?
2019-03-10 06918, 2019
12:06 PM
samj1912
yup
2019-03-10 06947, 2019
12:06 PM
samj1912
and since you seem to have just 1 core and all your entities are indexed there you can also have a unified search like you currently do
2019-03-10 06907, 2019
12:07 PM
samj1912
so all you would be indexing is the name of the entity, its bbid and its type
2019-03-10 06925, 2019
12:07 PM
samj1912
all 3 fields would be stored and indexed
2019-03-10 06938, 2019
12:07 PM
samj1912
keep the ngram field indexed but not stored
2019-03-10 06941, 2019
12:07 PM
Mr_Monkey
I suppose aliases would be good as well, but maybe not a lot more complicated ?
2019-03-10 06914, 2019
12:08 PM
samj1912
2019-03-10 06920, 2019
12:08 PM
samj1912
should be just that
2019-03-10 06929, 2019
12:08 PM
Mr_Monkey
>keep the ngram field indexed but not stored
2019-03-10 06929, 2019
12:08 PM
Mr_Monkey
/me is a noob, didn't understand
2019-03-10 06945, 2019
12:08 PM
samj1912
ngram is used for partial searches
2019-03-10 06906, 2019
12:09 PM
samj1912
like if you search for "Sam" "samj1912" will show up in the results
2019-03-10 06919, 2019
12:09 PM
samj1912
solr stored fields are any fields that you want solr to return
2019-03-10 06926, 2019
12:09 PM
samj1912
indexed fields are any fields you want to search
2019-03-10 06939, 2019
12:09 PM
Mr_Monkey
Ah, I see.
2019-03-10 06920, 2019
12:10 PM
samj1912
since ngrams basically break up something like samj1912 into - "sam", "amj", "mj1" etc. you do not want those things to be returned by solr
2019-03-10 06924, 2019
12:10 PM
samj1912
just searchable
2019-03-10 06935, 2019
12:10 PM
samj1912
(I am assuming trigrams here)
2019-03-10 06944, 2019
12:10 PM
Mr_Monkey
Got it. I see how that's configured in the schema
2019-03-10 06917, 2019
12:11 PM
samj1912
so yeah - this should be very simple to do all this
2019-03-10 06918, 2019
12:11 PM
dns[m] has quit
2019-03-10 06933, 2019
12:11 PM
samj1912
maybe you can piggy back on you existing solr servers hosted for mb
2019-03-10 06935, 2019
12:11 PM
dns[m] joined the channel
2019-03-10 06940, 2019
12:11 PM
samj1912
since it is just 1 more core
2019-03-10 06947, 2019
12:11 PM
samj1912
I would leave it to zas to decide
2019-03-10 06923, 2019
12:13 PM
Mr_Monkey
Yes, I'll see with him about hosting
2019-03-10 06950, 2019
12:13 PM
samj1912
but yeah all of this should be simple enough
2019-03-10 06901, 2019
12:14 PM
Mr_Monkey
So that path wouldn't prevent us from later on creating more complex schemas down the line?
2019-03-10 06943, 2019
12:14 PM
samj1912
well - I would suggest moving to something like Sir once you have complex relationships between your searchable objects
2019-03-10 06905, 2019
12:15 PM
samj1912
but as of now - this should be good to go
2019-03-10 06929, 2019
12:15 PM
Mr_Monkey
Right, that'll be a chunk of work, but that means GSOC aspirants don't need to know JS+Python at least
2019-03-10 06942, 2019
12:15 PM
samj1912
2019-03-10 06949, 2019
12:15 PM
samj1912
and trying it out
2019-03-10 06911, 2019
12:16 PM
Mr_Monkey
For sure
2019-03-10 06905, 2019
12:17 PM
samj1912
2019-03-10 06924, 2019
12:17 PM
Mr_Monkey
Hah, I was about to ask about continuous indexing/reindexing
2019-03-10 06946, 2019
12:17 PM
samj1912
well I don't know how BB currently handles that
2019-03-10 06916, 2019
12:18 PM
Mr_Monkey
2019-03-10 06934, 2019
12:18 PM
Mr_Monkey
2019-03-10 06900, 2019
12:19 PM
samj1912
but that basically deletes the index and reindexes everything right
2019-03-10 06901, 2019
12:19 PM
Mr_Monkey
Although that's all ES specific
2019-03-10 06904, 2019
12:19 PM
samj1912
what about updates?
2019-03-10 06918, 2019
12:19 PM
Mr_Monkey
Hang on, I'll have to dig it up
2019-03-10 06919, 2019
12:21 PM
Mr_Monkey
2019-03-10 06900, 2019
12:22 PM
samj1912
cool - btw what is body?
2019-03-10 06912, 2019
12:22 PM
samj1912
can I see an example of what it looks like?
2019-03-10 06924, 2019
12:22 PM
Mr_Monkey
The content of the http request body
2019-03-10 06931, 2019
12:22 PM
Mr_Monkey
Hang on
2019-03-10 06926, 2019
12:24 PM
Mr_Monkey
I don't have an example at hand for the body, but it will be a json object representing the entity from the ORM object
2019-03-10 06943, 2019
12:24 PM
samj1912
okay
2019-03-10 06944, 2019
12:25 PM
samj1912
2019-03-10 06905, 2019
12:26 PM
pac23 has quit
2019-03-10 06911, 2019
12:26 PM
samj1912
hmm not updated in 2 yrs
2019-03-10 06905, 2019
12:27 PM
samj1912
2019-03-10 06907, 2019
12:27 PM
Mr_Monkey
I figure I'll find something suitable
2019-03-10 06920, 2019
12:27 PM
Mr_Monkey
I was reading that earlier, looks like there's some alternatives
2019-03-10 06941, 2019
12:27 PM
samj1912
well solr recommends using plain old xmlhr
2019-03-10 06915, 2019
12:28 PM
Mr_Monkey
I'm happy with that.
2019-03-10 06941, 2019
12:28 PM
Mr_Monkey
I mostly need to read up some docs of the API
2019-03-10 06917, 2019
12:29 PM
samj1912
yeah
2019-03-10 06921, 2019
12:29 PM
samj1912
well thats that
2019-03-10 06940, 2019
12:29 PM
Mr_Monkey
2019-03-10 06953, 2019
12:29 PM
samj1912
yeah
2019-03-10 06908, 2019
12:30 PM
Mr_Monkey
JS/Expressjs and HTTP requests will be a breeze
2019-03-10 06915, 2019
12:30 PM
samj1912
let's catch up in 2 weeks around the same time on progress?
2019-03-10 06919, 2019
12:30 PM
samj1912
im off for lunch
2019-03-10 06933, 2019
12:30 PM
Mr_Monkey
OK, that gives me a much clearer picture, and it seems like an easier project than I anticipated.
2019-03-10 06939, 2019
12:30 PM
Mr_Monkey
Thanks a lot for your help samj1912 !
2019-03-10 06941, 2019
12:30 PM
samj1912
yeah
2019-03-10 06943, 2019
12:30 PM
samj1912
no worries
2019-03-10 06950, 2019
12:30 PM
Mr_Monkey
Much appreciated
2019-03-10 06911, 2019
12:31 PM
Mr_Monkey
And bon appétit :)
2019-03-10 06929, 2019
12:39 PM
pac23 joined the channel
2019-03-10 06936, 2019
12:57 PM
pac23 has quit
2019-03-10 06955, 2019
13:12 PM
michelv has quit
2019-03-10 06931, 2019
14:01 PM
Cyna
reosarevok: Can you help me test changes made to the .tt and React files ?
2019-03-10 06902, 2019
14:03 PM
Cyna
I've tried recompiling the resources and running the redis-server again but it didnt work
2019-03-10 06942, 2019
14:04 PM
reosarevok
Cyna: are the components listed in server/components?
2019-03-10 06952, 2019
14:04 PM
reosarevok
(that's a new thing that might not be documented properly yet?
2019-03-10 06955, 2019
14:04 PM
reosarevok
)
2019-03-10 06945, 2019
14:05 PM
Cyna
I want to test change to the existing .tt file as of now
2019-03-10 06951, 2019
14:05 PM
reosarevok
Oh
2019-03-10 06914, 2019
14:06 PM
Cyna
I am trying to understand catalyst
2019-03-10 06926, 2019
14:06 PM
reosarevok
Those should change automatically as far as I know, unless you're embedding a React component in which case you'll still need to list it in root/server/components
2019-03-10 06939, 2019
14:06 PM
Cyna
Hmm...
2019-03-10 06947, 2019
14:06 PM
reosarevok
I'm at the shop on mobile now, but if you push and link me to your branch I can take a quick look
2019-03-10 06914, 2019
14:08 PM
Cyna
Im changed index.tt
2019-03-10 06918, 2019
14:08 PM
Cyna
2019-03-10 06906, 2019
14:09 PM
reosarevok
What are you changing there? I don't know it by heart and can't find a different properly on mobile :)
2019-03-10 06930, 2019
14:09 PM
Cyna
I've changed the href
2019-03-10 06906, 2019
14:10 PM
Cyna
<li><a href="[% c.uri_for_action('/admin/attributes/attribute_index', [ m ]) %]">[% m %] hell</a></li>
2019-03-10 06918, 2019
14:10 PM
Cyna
But Im getting the output
2019-03-10 06937, 2019
14:10 PM
Cyna
2019-03-10 06938, 2019
14:11 PM
Cyna
I must've got *attribute_name + hell*
2019-03-10 06951, 2019
14:12 PM
reosarevok
Is models defined? That's inside an if models, you might actually be rendering the else
2019-03-10 06907, 2019
14:13 PM
Cyna
Yea the models is defined in the controller
2019-03-10 06912, 2019
14:15 PM
reosarevok
Ok. I'll take a look when I get home if you haven't figured out in the meantime
2019-03-10 06917, 2019
14:15 PM
Cyna
Sorry reosarevok
2019-03-10 06922, 2019
14:15 PM
reosarevok
Might take like 30 min
2019-03-10 06933, 2019
14:15 PM
Cyna
I just found out that I had edited the wrong copy
2019-03-10 06955, 2019
14:15 PM
Cyna
It worked after changing the right copy which my container is binded to
2019-03-10 06901, 2019
14:16 PM
Cyna
Thanks and sorry for the trouble
2019-03-10 06949, 2019
14:16 PM
reosarevok
Haha
2019-03-10 06955, 2019
14:16 PM
reosarevok
Well, that's good to hear!
2019-03-10 06911, 2019
14:17 PM
reosarevok
Do let me know if you need help with something :)
2019-03-10 06933, 2019
14:17 PM
Cyna
Yea sure... Im figuring out perl right now
2019-03-10 06939, 2019
14:21 PM
CallerNo6 joined the channel
2019-03-10 06939, 2019
14:21 PM
CallerNo6 has quit
2019-03-10 06939, 2019
14:21 PM
CallerNo6 joined the channel
2019-03-10 06953, 2019
14:30 PM
Zastai joined the channel
2019-03-10 06940, 2019
14:46 PM
ahmedkrmn joined the channel
2019-03-10 06901, 2019
14:48 PM
Zastai
2019-03-10 06940, 2019
14:49 PM
Zastai
And the GitHub repo is no immediate help since it's .rst files, with the bulk of the API documented at document expansion time
2019-03-10 06959, 2019
14:55 PM
CatQuest
usb sticks are ancient? 🤔
2019-03-10 06921, 2019
14:56 PM
CatQuest
[13:27] <yvanzo> good instant
2019-03-10 06922, 2019
14:56 PM
CatQuest
hah! that one. 👍
2019-03-10 06942, 2019
14:57 PM
outsidecontext joined the channel
2019-03-10 06935, 2019
15:23 PM
Zastai
(and it looks like the mbids in listens are now deprecated in favour of msids, which I assume are MessyBrainz IDs - but that's not documented, and
messybrainz.org does not seem to document an API either)
2019-03-10 06907, 2019
15:33 PM
amCap1712 has quit
2019-03-10 06928, 2019
15:40 PM
Freso
Zastai: Fetching or submitting listens?
2019-03-10 06943, 2019
15:40 PM
Zastai
Currently fetching
2019-03-10 06948, 2019
15:40 PM
Freso
For submitting, you shouldn't submit MSIDs, always MBIDs.
2019-03-10 06935, 2019
15:41 PM
Freso
iliekcomputers is likely the one that can best answer your questions.
2019-03-10 06941, 2019
15:41 PM
Zastai
I'm fetching, but getting null for all _documented_ fields; only MSIDs are filled. (Imported scrobbles)
