| 5 Jan 2024 |
 champignoom | or otherwise, the future is still distant, and a dirty workaround might become the workhorse | 15:27:12 |
| champignoom | ___ | 15:28:32 |
| champignoom | another thing | 15:28:44 |
| champignoom | the parser might get upgraded, but the syntax won't change much, right? | 15:29:29 |
| champignoom | so can we already use v1 parser to do /something/ ? | 15:30:19 |
| champignoom | like, we might not be able to use tags/macros/... but plain text description is already available, right? | 15:30:48 |
| champignoom | i mean in terms of gtd | 15:31:53 |
 boltless | - free-form might has some huge breaking change
- parsed output CST would be completely different e.g. heading won’t contain it’s level on it’s node name as we can now nest it endlessly
- have you heard about inline block syntax? (neither do I) | 15:32:13 |
| boltless | so yes, in terms of GTD, I think you can use v1 for now | 15:32:30 |
| champignoom | very nice | 15:33:01 |
 max397 | .vhyrro | 15:34:17 |
 .vhyrro | The good news is:
- I recently finished the client
- I am working on the backend GTD system
- I recently finished the db system | 15:34:22 |
| .vhyrro | and the parser... yeah the parser is what it is | 15:34:27 |
| max397 | .vhyrro just don't forget my request | 15:35:47 |
| .vhyrro | do you think i remember your request | 15:35:57 |
| .vhyrro | what was it | 15:35:58 |
| max397 | I had one request
when you started with the gtd system
the biggest thing I was missing from the old apis | 15:36:05 |
| max397 |
an api to modify task state | 15:36:36 |
| .vhyrro | brother with this api your ass is going to be able to do even more than that
I'm hoping to make the API cover a lot of the most common uses | 15:37:20 |
| max397 | stfu | 15:37:23 |
| max397 | well I just need that
and ofc a way to get the tasks
but if I can modify task state I'll do epic things (neorg-kanban my beloved) | 15:37:51 |
| champignoom | do we have an interface of the server/client, on which we can already build the UI? | 15:43:31 |
| champignoom | for instance if the client returns Vec<Tasks> in 10ms, we can start working in parallel on the UI with a simple lua function that visit all .norg files in 1s. | 15:48:08 |
| champignoom | and once available, just replace get_tasks_slowly_in_lua() with get_tasks_quickly_with_norgopolis() and all other parts won't change | 15:50:44 |
| champignoom | The idea is that, if norgopolis and the new parsers are performance improvements of some existing functionalities, they shouldn't be blocking. | 15:53:02 |
| max397 | well but that's not it | 15:53:15 |
| max397 | data will be stored in a db | 15:53:21 |
| max397 | it won't be possible to just replace a function | 15:53:33 |
| champignoom | yeah and db is improvements of filesystem | 15:53:42 |
| champignoom | querying from filesystem is slower but does the same job as querying from db | 15:54:00 |
