<fossy>Keep sending critisims of the code my way if you have more :) <OriansJ>odds are if you need to put prepend_string or postpend_string in a loop, there is probably a more efficient way of doing it by just having index variables and walking the strings <OriansJ>I would recommend breaking the variable functions into a seperate .c file and creating a .h which only has the struct definition and a few commonly shared functions' prototype <fossy>I was thinking about that, splitting into multiple files <fossy>I think its a good idea - ill go with it <fossy>4. Split variable stuff out of kaem.c <OriansJ>looking at how ->pos is being used; odds are you probably don't even need it in the struct if you just did proper local index variables and the passing of ints where appropriate <fossy>After this refactor I plan to take a bit of a break from kaem, to be honest I dont enjoy working on the same piece of software every day <OriansJ>possibly add foo=1 ./bin/bar sort of environment variables to the future TODO list <fossy>Not a blocker for pushing though? <fossy>But the multiple variable substitution is a blocker right <fossy>I do not enjoy breaking things <fossy>OriansJ: how do you plan to integrate kaem with mescc tools? As a submodule? <OriansJ>fossy: yes it will become a proper submodule <OriansJ>with updates when I feel the current kaem commit is good enough <fossy>Take your time, but any further things to make the current commit good enough than those 4? <OriansJ>create feature branches in kaem's repo when you want your work reviewed before squash and merge <fossy>Quite reasonable and very normal <OriansJ>The master branch is your responsibility to keep it clean and working on as many commits as you can <fossy>From this refactor on, 100% of commits on master will build and past tests <OriansJ>and yes, getting those 4 done will make me very happy <fossy>Once those 4 are done, would you like push, or another reivew? <OriansJ>and then we can cascade it down into mescc-tools, mescc-tools-seed and finally stage0 if everything is nice and happy <fossy>dddddd: if you have more comments I would like to hear them very soon, those "finicky details" would be good to hear <dddddd>sure, about EjN or the next one? <dddddd>I can anotate EjN and send you back and/or do it with the next one (with the 4 points in place) <fossy>unless you would prefer the next. <dddddd>In fine doing both. I'll do EjN now. <fossy>^-^ thank you both, my coding is getting better slowly I think ***roptat_ is now known as roptat
<fossy>dddddd: FYI I broke the builtins. <fossy>Please ignore the return TRUE stuff, thats broken, im fixing it <dddddd>return TRUE wrt builtins, ok... I'll skip that. ***ChanServ sets mode: +o rekado_
<fossy>OriansJ: how do you feel the best way to pass both index_input and index_target in and out of a function is? I was thinking a struct <OriansJ>fossy: you might not need to pass both in and out of a function. <OriansJ>as you can simply know that index_target will be index_target+bytes_processed; which would make those sort of functions a single integer return. <fossy>OriansJ: I think I have found a M2-Planet bug <fossy>lemme get you a minimal test case <fossy>can you redeclare the same variable inside a goto, after it has been run once, in normal C? <fossy>I don't use gotos much, but they were by far the cleanest way to add multiple variable support to kaem, only 5 extra LOC <fossy>way more trivial than i thought :D <fossy>wow, kaem is so much faster without postpend chat