02:48 <_whitenotifier-5> [starshipraider] azonenberg pushed 5 commits to master [+15/-75/±2] https://git.io/JtyGR

02:49 <_whitenotifier-5> [starshipraider] azonenberg 04db645 - Removed a bunch of really old sims that don't apply to current hardware

02:49 <_whitenotifier-5> [starshipraider] azonenberg 71ad8ef - Removed more obsolete sims

02:49 <_whitenotifier-5> [starshipraider] azonenberg 3de0a55 - Added more filter experiments. Removed additional obsolete simulations

02:49 <_whitenotifier-5> [starshipraider] ... and 2 more commits.

02:51 <_whitenotifier-5> [starshipraider] azonenberg pushed 1 commit to master [+10/-4/±2] https://git.io/JtyGE

02:51 <_whitenotifier-5> [starshipraider] azonenberg f3b05b5 - Continued push-pull port sims of left side area with ENIG

12:36 <_whitenotifier-5> [scopehal-apps] sam210723 forked the repository - https://git.io/JtMUE

12:39 <_whitenotifier-5> [scopehal] sam210723 forked the repository - https://git.io/JtMUE

13:21 <_whitenotifier-5> [scopehal-apps] sam210723 opened pull request #307: Add *.bin filter to Import dialog - https://git.io/JtyXO

13:29 <_whitenotifier-5> [scopehal] sam210723 opened pull request #396: Add binary capture file parser - https://git.io/JtyXM

13:29 <_whitenotifier-5> [scopehal] sam210723 edited pull request #396: Add binary capture file parser - https://git.io/JtyXM

13:29 <_whitenotifier-5> [scopehal-apps] sam210723 edited pull request #307: Add *.bin filter to Import dialog - https://git.io/JtyXO

13:43 <_whitenotifier-5> [scopehal] sam210723 opened issue #397: 100BASE-FX Decoder - https://git.io/Jty11

13:45 <sam210723> might call it a night there

13:45 <sam210723> all the importer code is in those PRs

13:46 <sam210723> and i added a 100BASE-FX sample as a scope session file

16:15 <azonenberg> Great

16:15 <azonenberg> I'll have a look in a few

16:48 <azonenberg> sam210723: have a look at lib/scopehal/scopehal.cpp#604, ReadFile

16:48 <azonenberg> that basically does what your read function does, no sense adding a new one

16:50 <azonenberg> your BytesToInt/BytesToFloat are pretty redundant, since you should be able to just reinterpret_cast the char* to float* or uint32_t* directly

16:51 <azonenberg> swapping endianness if needed separately

16:52 <azonenberg> for that matter the entire file header should be possible to create as a packed struct

16:52 <azonenberg> then cast

16:53 <azonenberg> Rather than vendor being "Agilent/Keysight/Rigol" can you use the file magic number to set to "Agilent/Keysight" or "Rigol"?

16:56 <azonenberg> Re packed structs you can either use pragmas https://gcc.gnu.org/onlinedocs/gcc-5.2.0/gcc/Structure-Packing-Pragmas.html or https://gcc.gnu.org/onlinedocs/gcc-4.7.0/gcc/Type-Attributes.html

16:56 <azonenberg> This is safe as long as your target architecture allows unaligned memory accesses (performance isnt a big deal for file loading compared to other stuff, so being slow is OK)

16:56 <miek> all the ReadFromFile calls leak memory?

16:57 <azonenberg> miek: i didnt even look for that because i assumed we were gonna get rid of them anyway

17:43 <azonenberg> Also, AKL-PT1 update... doing a bunch of larger scale simulations, including one of the left half of the probe with different levels of fidelity for the ENIG loss model

17:44 <azonenberg> One simulation has been running for 8.5 hours and is still going

17:44 <azonenberg> On a single core

17:44 <azonenberg> I really need to just buckle down and buy sonnet pro so i can multithread the solver lol