| 08:15 |
awt |
!!ticker btc usd |
| 08:15 |
deedbot |
$28,028 |
| |
~ 4 hours 44 minutes ~ |
| 13:00 |
asciilifeform |
http://logs.bitdash.io/pest/2023-05-29#1026457 << very nifty. this is the py gui client ? |
| 13:00 |
bitbot |
Logged on 2023-05-29 23:43:50 awt: Got nick tab completion working. |
| 13:05 |
awt |
asciilifeform: yes |
| |
~ 1 hours 32 minutes ~ |
| 14:37 |
awt |
asciilifeform: For out of order packets, are you planning to send them directly to the gui/client after issuing a getdata and just let the gui decide what to do with it? |
| 14:38 |
asciilifeform |
awt: correct ( tho ftr implies a bit moar separation b/w guiism & 'engine' than asciilifeform is having ) |
| 14:39 |
awt |
I believe this explains the presence of the timestamp ordered tab in your station impl.? |
| 14:40 |
asciilifeform |
awt: aha. needs moar thought tho |
| 14:46 |
awt |
One thought I had for handling broadcasts from foreign nets (i.e. no ancestor found) was to pop open a new tab and tack on any descendants as they arrive). Could insert into the main tab later if somehow a common ancestor appears. |
| 14:49 |
asciilifeform |
awt: imho sounds rather ugh ( new tab, and now can have over9000 tabs? and search box nao worx only in a given tab? ) |
| 14:52 |
awt |
right? but the alternative is yet another "order buffer" afaics |
| 14:55 |
asciilifeform |
the alternative, afaik, is to permit reordering of lines in the gui (if a missing predecessor arrives, or 'sort by...' button is hit) |
| 14:55 |
asciilifeform |
(if in mode 'sort by time', no need for any reordering/redraw) |
| 14:59 |
awt |
In that case the message would not be visible in the 'sort by message hash' view until a predecessor arrives? |
| 14:59 |
asciilifeform |
awt: does your gui actually try to store ~all~ lines scrollable in the window, in a dedicated buffer ? ( asciilifeform's wx attempt -- does not ) |
| 15:00 |
awt |
asciilifeform: it does, yes. |
| 15:00 |
asciilifeform |
awt: in asciilifeform's scheme, if a precedent is null/missing, msg is displayed 'by time' |
| 15:01 |
asciilifeform |
awt: ideally you dun want to store all msgs (from entire history of universe !) in a 2nd buffer. this'll be terribly gnarly/slow |
| 15:02 |
asciilifeform |
only store the # that fit on screenfull (and from period calculated from scrollbar pos) |
| |
↖ |
| 15:02 |
awt |
asciilifeform: yes. saving that fun for later though. |
| 15:02 |
asciilifeform |
a ok |
| 15:03 |
awt |
I think phf mentioned he loaded in a novel's worth without issue so I think I have some time. |
| 15:05 |
asciilifeform |
prolly not catastrophic issue ~until~ you potentially need to reload entire universe on acct of a gap fill |
| 15:07 |
awt |
I believe it's fairly straigh forward to insert lines anywhere in a Text. Could be some issues retagging regions though. |
| 15:08 |
asciilifeform |
while it's a txt (rather than complex soup w/ inline imgs etc) |
| 15:10 |
awt |
right |
| 15:10 |
awt |
Still not entirely sure how to keep track of inline images in the tree. |
| 15:24 |
asciilifeform |
afaik the only pill there is some variant of this |
| 15:24 |
bitbot |
Logged on 2023-05-30 15:02:05 asciilifeform[jonsykkel|deedbot]: only store the # that fit on screenfull (and from period calculated from scrollbar pos) |