This repository has been archived by the owner on Mar 6, 2022. It is now read-only.
/
speed_up_ideas
51 lines (40 loc) · 2.16 KB
/
speed_up_ideas
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Hopper uses 10 seconds (9739ms) to load the big binary I am working on, we also need speed. Bottleneck to reach this goal is the loading of the web elements, not backend.
Speed up
[Frontend]
- web worker to analyze instruction highlighting
- challange : clustersize.js will have to search for the object again....
it migth not make preformance any better :/
- Load as you scroll?
- Not sure about this one, it will speed up the loading
however, if I want to do something I migth be stuck at loading
either way.
- migth hurt under a CTF
- maye the terminal interface can come to the rescue, have a core running a socket
you can interface with web / terminal since terminal can connect to the socket and
tell the web to update.
[Backend]
- comrpess the json, use lookup values for instruction
instead of instering all of them
- I compressed data with 34.392134472587784%
- Have the backend find each block, then the frontend
can use this data to easier to sepearte blocks
- less work for javascript
- minimize reflow anywhere I can....
- (write some benchmark code like I did for javascript)
[Both]
- cleaver algorithms and data structures
- Some places I can write better data structures and algorithms because of new knowledge on the way to design the system.
QA:
why not just switch from web to qt?
I can write code in the web way faster than I can with QT. Also
it will look nicer. There are some bottlenecks with writing code
for the web that I just have to be cleaver about. However I'm sure QT can hold
way more objects in frame than a browser can, but program speed is not the reason
why I choose to write the GUI in web. I want a seperate backend and frontend and having
the web as a frontend makes the seperation easily.
Remove json
I kinda want to keep json mainly for readability reasons, makes reading
the code a lot easier. However the fact that it generated ~17MB of data is
way to much. Trafering that over the socket usally makes my session end,
the times the session does not end it takes way to much time for the socket
to recive the data. Not good. (this is when I run the backend not on localhost)