Table of Contents

* Logseq [[logseq]]

Logseq is a #localfirst browser interface for #org (mode). I'm using it for my #exobrain

search could be wider and include more stuff? [[exobrain]]


[2020-12-30] display 'current' graph without clicking on 'page graph' all over again? [[outbox]]

[2020-12-30] hmmm… I'd like to see 'related' pages, e.g. when certain tags are siblings?

[2020-12-30] eh. singe page (toexobrain export) didn't quite work well

might be nice to have several 'layers' [[exobrain]]

e.g. people could be on a different layer
it could be achieved by having them not participate in constraints and just layed out on top with different color?

[2020-12-29] Publishing [[logseq]] [[exobrain]]

ok, nice, this worked (need to run http server though, doesn't work as a fully static page?)

dev mode suggesion: reindex on refreshed code (or force reindex) [[outbox]]

node size proportional to the number of links? [[exobrain]] [[logseq]]

hmm, if title is missing, might be better to use filename (now it's picking the first heading) [[exobrain]]

if it used custom ids would be awesome..

need a mode to exclude headings from notes graph? although not sure [[exobrain]]

e.g. I want two tags to be connected if they are siblings – basically only want connections between tags? I guess it's kinda subjective

colour differently pages with the corresponding file and without (i.e. tags) [[logseq]] [[exobrain]]

[2021-01-07] logseq/graph.cljs at 14e4af295781a00aacdb9104bda9c8121000b2cb · logseq/logseq [[logseq]]

(map (fn [[p aliases]]

aliases for graph view

auto push could push on a separate branch? and the actual normal push could be manual? [[logseq]]

[2021-01-08] right, it treats # as tags? [[org]]

need to highlight connections when I drag the enode [[logseq]] [[exobrain]]

ignore tags: read watch inspiration motivation buy outbox ? maybe even study? they result in too many connections

maybe they are more like attributes.. or tiny dots around tag, each of dots represents a task. dunno

[2021-01-18] logseq/config.cljs at master · logseq/logseq

[2020-12-30] Directory handle not exists (local filesystem) · Issue #918 · logseq/logseq

do not parse tags in quoted text [[logseq]]

[2020-12-30] using heading as title [[logseq]]

Yeah, there're some users who rely on the first heading as the page title.
If you prefer to use the file name as the page title instead of the first heading's title, (TITLE -> filename -> first heading), add :page-name-order "file" to the file logseq/config.edn.

wtf?? [[logseq]]

only seems to happen in publishing mode?

Lookup ref should contain 2 elements: ({:page/name "pkm"})

ask for publishing from console [[logseq]]

would also help for testing.. in a way?

* Orgzly [[orgzly]]

Awesome android app for #orgmode

so, every notebok has a root node
unspecified: inserts at the end of root?

note level; level 0 is root

debugging SQLiteDatabase: /data/user/0/com.orgzly/databases/orgzly_test.db

feat: interface for fs browsing

The main issue is multiple repositories
Imagine we have
we'll have to somehow merge them. Quite annoying!

feat: open org files in orgzly [[org]]

karlicoss/orgzly branch view-org-files

right, so apparently, notes are getting rebuilt every time from database
we could have a temporary database which gets flushed occasionally and never gets displayed?

Wonder if i could implement simple org viewer with org-java? [[org]] [[android]] [[spacedrep]]

feat: capture templates? With screen buttons

[2019-02-21] shit. gradle sync didn't work. tried updating Kotlin but that didn't help…

[2019-02-21] basic idea: first approximation – reuse widget for notebook, just trigger note creation on opening

we want
name, template, target (could be file in the first approximation, but think how we would address subheadings)


Orgzly notification high priority only? Or even a search query

you know if anyone started progress on that? If not I might give it a try.

I'm not aware of any. That would be great, thanks.
I guess it would be a new preference in Settings / Reminders that can contain a search query?

bug: cut note and then create new in an empty notebook – the note you cut disappears

Orgzly quick search widget?

todo state changes

make sure orglzy can handle that??

Orgzly archive

Would require support in notebook interface

[2019-02-10] there were some suggestions already, for tagging in particular

[2019-02-10] Implement auto synchronization · Issue · orgzly/orgzly-android [[orgzly]]

Post about orgzly on fossdroid?

[2019-12-15] Detect org files in subfolders · Issue #225 · orgzly/orgzly-android

ctory tree with bunch of org files around it, and right now orgzly only detects the org files on top level. Would

[2019-12-15] Detect org files in subfolders · Issue #225 · orgzly/orgzly-android

Just checked out attern is having a directorytop level.

[2020-07-11] [orgzly/orgzly-android] Detect org files in subfolders

I switched from orgzly to markor exactly because of the lack of this feature. Yes, I know, these apps differs a lot, but I use them simply as a text editor. I'm still waiting when this will become available in orgzly. Why it's so hard to implement it?

* syncthing [[syncthing]]

suggest heartbeats for syncthing? [[cloud]]

[2019-09-06] dschrempf/syncthing-resolve-conflicts: A small bash script that handles synchronization conflicts with Syncthing. Inspired by 'pacdiff' from Arch Linux.

[2019-09-06] jjkramok/syncthing-conflict-notifier: Detects conflict files in a single syncthing shared folder and sends a push notification if conflicts are found.

[2020-04-12] issue with Out of sync items/ sync stuck at 99%

not sure what exactly caused it, but basically was constantly showing local/global state mismatch
Rest api is on GUI port, key is in advanced settings
use the post request to reset the index (apparently no way from GUI)

commit conflict detector? show alter in the inteface

[2020-12-19] eh? apparently it's already presetn?


ok, syncthing paranoid should simply reuse logic in cloud-heartbeat [[cloud]]

also probably belongs to the same repository altogether

had a bug: in pkm dir [[cloud]]

TODO would be interesting to do fuzzing

  1. before pkm (real dir) pkm/search (symlink) pkm/swoop (symlink)


  1. after pkm: symlink with the same files inside

sync was just stuck

Better indicator for syncthing? In mc and in gui?

Gui – available, mc – not sure how..
mm, doesn't look like there is a nice terminal interface?

run git gc occasionally? can remove quite a lot of objects… [[cloud]]

[2020-06-07] yeah, it's a really good idea.. wonder how to run it more often

[2019-04-17] implement more regular checks for syncthing?

[2019-04-17] also allow selective syncing on mobile/wifi

[2019-06-05] shit, weird. syncthing has got some sort of mismatch between GUI and web file/folder count; borg was showing differnt counts as well (script in dropbox that I used to check for differences) [[cloud]]

* wildcard [[wildcard]]
related: #interop #malleable #spreadsheet

[2020-05-13] Ask HN: Anyone else upvote HN comments to track comments they have already read? | Hacker News [[pkm]]

[2020-05-05] SelectorGadget: point and click CSS selectors [[css]]

[2020-05-11] pull tweet">(10) jestem króliczkiem on Twitter: "@jtraub аа, понял. Крутая идея, да! это было бы оч легко, даже не модифицируя экстеншн если бы браузеры разрешали экстеншнам друг на друге оперировать, а так можно будет добавить в Wildcard" / Twitter

аа, понял. Крутая идея, да!
это было бы оч легко, даже не модифицируя экстеншн если бы браузеры разрешали экстеншнам друг на друге оперировать, а так можно будет добавить в Wildcard

ok, what if we combine them? [[promnesia]] [[worldbrain]] [[wildcard]]

e.g. intercept AJAX request and store them in the database on your disk. Then, add some code to filter it and keep as the browser history?

firefox logging

Another thing I noticed was this bit logging errors all the time for adapters that don't have `scrapeAjax` (I'm on Firefox). I wonder if it's better to check for the att

wip on DSL issue

> Usually I prefer DSLs embedded in a Turing-complete language to provide the TC escape hatch if needed, but here that's precisely what we don't want.

TODO why no escape hatch

I'm (as a somewhat experienced programmer) biased towards having a real programming language, because, because I've struggled with mediocre yaml/json DSL that end up reimplementeing half of the programming language in an incomrehensible ways. E.g. if you already know javascript, you end up frustrated with expressing in DSL what you'd otherwise have don in few lines of code. And the opposite: if you don't know how to program in the first place I'd be much rather

debugging somethine like

>  "xpath": "//*[@class='blog-item']/div[0]"

is kind of a nightmare for anyone, and you need the devtools and JS console for that

TODO something about ids

>  "querySelector": "._i24ijs"

Often (citation needed though) websites have completely dynamic ids, so you would need to do some sort of completely dynamic query to determine the content and the items

[2020-05-07] Restrict expressiveness of site adapters · Issue #17 · geoffreylitt/wildcard

contribute to the discussion here

Generic scraping with chrome inspector?

'exploring queue', mark explored elements and never show them again [[jonbo]] [[wildcard]]

[2020-05-11] bug: Vrroom

[2020-05-12] Fraidycat

[2020-05-16] Generate RSS feeds for any web page | PolitePol

ok, nice so it allows you to choose elements dynamically, could be useful

would be nice to keep images outside of the git repository… perhaps even prune

separate repository for binary assets is better

[2020-05-26] pull tweet">(2) Jared Palmer on Twitter: "PSA: You can usually crank up Webpack rebuild/HMR speed by ~7x on TypeScript projects by doing this to your Webpack config in development. Diff to create-react-app:" / Twitter [[wildcard]]

PSA: You can usually crank up Webpack rebuild/HMR speed by ~7x on TypeScript projects by doing this to your Webpack config in development.

* sympy [[sympy]]

related #sim #math

Add.matches??? [[sympy]]

use wild to deconstruct expressions? [[sympy]]

started solving that system of DEs [[sympy]]

try dsolveset?? [[sympy]]

[2018-12-24] Added a few rules for imageset simplification by hargup · Pull Request 7625 · sympy/sympy

dsolve bug [[sympy]]

add wave eq to sympy?? [[sympy]]

* materialistic app for #hackernews

[2019-12-21] [#C] set link as a title when sharing??

Hn materialistic: change view on press

Hn Materialistic: comment icon is way too large for the number of comments

* git-bug [[gitbug]]

related #git #github #decentralize

[2020-11-02] git-bug isn't syncing PRs?

[2020-11-21] git-bug/import.go at b3318335986618388637a9d35d68b39633e4548a · MichaelMure/git-bug

ugh fuck, this should go to stderr?

* Hypothesis [[hypothesis]]

need to color my and other people highlights in diff colors

[2018-06-11] just be a bit more explicit about not logged in

ok, what I actually want is the extension to be more vocal about logouts


[2019-06-27] ok, it was partially resolved, at least it expands the sidebar now. Swill major problem for usability though, can easily disapppoint new users.

[2019-02-20] hypothesis/client: The Hypothesis web-based annotation client.

Join us on Slack - request an invite or log in once you've already created an account - or in on freenode for discussion.
If you'd like to contribute to the project, you should consider subscribing to the development mailing list, where we can help you plan your contributions.

doesn't work on hackernews, some content-security policy thing [[hypothesis]]

[2020-05-05] history.txt [[hypothesis]] [[annotation]]

doesn't work on github?? ugh.

[2019-07-18] John Maynard Keynes: "Newton, the Man"

hmm, hypothesis doesn't work with Firefox reading mode?

bug: ugh, doesn't work with Linkedin

ugh, seems that pseudoelements like in ampie and promnesia mess with hypothesis highlight [[hypothesis]]

e.g. try to highlight over "eg with blue-blocker sunglasses or f.lux) after sunset" on this page

[2021-03-06] Shirky: Situated Software [[hypothesis]]

doesn't work on pages??

* tantivy [[tantivy]]

Search indexer on Rust
related #search

Try search as you type on Wikipedia [[tantivy]]

host wikipedia backend for a bit and give a frontend link to github issue [[tantivy]]

tantivy-py: need to bump version to -dev or something? [[tantivy]]

* twint [[twint]]

related: #twitter

not sure about –profile-full?? [[twint]]

twint -u karlicoss --database tweets3.db --profile-full

hmm. it looks like it can't parse tweet ids?? [[twint]]

CRITICAL:root:twint.get:Tweet:time data '' does not match format '%Y-%m-%d %H:%M:%S'
1249982874150932480 2020-04-14 09:48:35 BST <karlicoss> I love it, one of the best physics textbooks I know ! Will probably take  years for me to fully grasp, but it's very pleasant to read!  …
1249856617031630848 2020-04-14 01:26:53 BST <karlicoss> I'm almost sure you've seen it, but this is a fascinating read: … (comments too!)
1249820865593847814 2020-04-13 23:04:49 BST <karlicoss> It's probably for the first time in years when I actually liked some user interface. Wow.
1249820017576890368 2020-04-13 23:01:27 BST <karlicoss> Wow, I've just used @liberapay for the first time, and it's been *increadibly* pleasant.  Clean, functional and text-centered interface. No bullshit, you can clearly see who you pay, how and when.
CRITICAL:root:twint.get:Tweet:time data '' does not match format '%Y-%m-%d %H:%M:%S'
CRITICAL:root:twint.get:Tweet:time data '' does not match format '%Y-%m-%d %H:%M:%S'
1249586565417963520 2020-04-13 07:33:48 BST <karlicoss> this is, in part, why I stopped listening podcasts

[2020-04-13] Show HN: Twint – Twitter scraping tool written in Python | Hacker News [[twint]] [[promnesia]] [[hpi]]

hmm maybe use it real quick?

[2020-04-14] Home · twintproject/twint Wiki [[hpi]]

maybe use github wiki?

* org-sync
syncing #github issues with #org mode
using #gitbug instead now

[2019-12-27] not sure if I like sorting by modified time… OPEN and id is better [[orgsync]]

[2020-04-12] for org-sync thing – could keep private notes in a separate file, but link into ids? [[github]]


share link and content [[tusky]]


validate systemd? [[pyinfra]]

ncmpcpp file name too long [[mpd]]

годы пройдут часть 1

update miniflux docs, mention that CLEANUP_ARCHIVE_READ_DAYS docs

PR in toastify for innerHTML

[2019-09-08] Validate inputs before saving · Issue #2 · fregante/webext-options-sync

problem with escaping table name [[visidata]]


 def reload(self):
    tblname = self.tableName
    self.columns = self.getColumns(tblname)
    r = self.conn.execute('SELECT COUNT(*) FROM "%s"' % tblname).fetchall()
    rowcount = r[0][0]
    self.rows = []
    for row in Progress(self.conn.execute('SELECT * FROM "%s"' % tblname), total=rowcount-1):

def getColumns(self, tableName):
    cols = []
    for i, r in enumerate(self.conn.execute('PRAGMA TABLE_INFO("%s")' % tableName)):

Allow users to mark posts as searchable/not searchable? [[mastodon]]

[2020-12-11] Export annotations - ⭐️Feature Requests - Memex Community [[worldbrain]]

Here is a guide on how to use the StorexHub API to get your Memex annotations.
It’s still in developer alpha, so it will require Javascript knowledge in building something useful: 12

sigh.. it's a bit overcomplicated
really need a simple json export..