IRC log for #stepcode on 20130618

00:37.54*** join/#stepcode kesha_ (~kesha@
00:58.32*** join/#stepcode kesha_ (~kesha@
01:05.49*** join/#stepcode kesha_ (~kesha@
01:55.18*** join/#stepcode kesha_ (~kesha@
02:14.47*** join/#stepcode kesha_ (~kesha@
13:06.03*** join/#stepcode kesha (~kesha@
13:13.33*** join/#stepcode kesha_ (~kesha@
17:26.04*** join/#stepcode kesha_ (~kesha@
18:52.32*** join/#stepcode kesha_ (~kesha@
18:57.31*** join/#stepcode kesha_ (~kesha@
19:51.47*** join/#stepcode mpictor (~mark@2601:d:b280:9:d63d:7eff:fe2d:2505)
19:56.01kesha_Hi mpictor, just sent u a mail. Check inbox and guide me..
20:06.16kesha_Using the previous build/ again, make builds successfully.
20:12.21mpictorhi kesha
20:12.56mpictorcould you push a branch to github?
20:15.03mpictorwithout seeing the code, the only thing I can think of is that some braces are no longer printed in the generated code
20:17.47mpictorkesha_: 150 is a good start! :)
20:30.55kesha_Made a branch ks-issue21 on my system, then "git remote set-url origin". Now how to push? "git push origin ?"
20:31.36mpictorthat should work. I usually do 'git push origin mybranch:remotebranch'
20:31.57mpictoroh, I misread your first line
20:32.27mpictor'git push origin' probably wouldn't work, but the one I suggested should
20:33.53kesha_What shall I fill up mybranch and remotebranch ?
20:34.26mpictorks-issue21 for mybranch, and remotebranch could be the same or something else
20:35.01mpictorthe 'git remote set-url' command shouldn't be necessary if you are using the same computer as before
20:35.36mpictorif a different computer, you'll need your ssh key
20:38.55kesha_crap, the command prompt is engaged since a long time after entering my github username and password.
20:40.29kesha_can you walk me again through pushing it onto a new branch ?
20:42.24mpictorwhat error did you get?
20:42.42kesha_no error !
20:42.58kesha_I killed the process
20:44.40mpictor'git push origin ks-issue21:ks/issue21' ought to work
20:44.56mpictoroh! the url you used may not work with ssh
20:45.15mpictortry 'git remote set-url origin'
20:47.50mpictorwell, makes it sound like the https url would work
20:48.46mpictorI didn't know you could have write access except with git@
20:57.34kesha_Its being pushed :)
21:06.46mpictoris about to get a thunderstorm...
21:09.33kesha_in balcony and its raining heavily out...
21:09.54kesha_btw, you can see that branch now. :)
21:10.27mpictorI don't see it at
21:11.52kesha_Can you see here ? -
21:11.59mpictorI can
21:12.39mpictoroh, it says they're still calculating info about this branch
21:12.52mpictormaybe that is why it doesn't show up in the list
21:13.25kesha_oh, ya. Shortly it will be there
21:32.53mpictordid you download an archive of the code and then do 'git init'? it looks like most of the history is gone in your branch :/
21:33.03mpictorno wonder it took so long to push the branch
21:33.35mpictorfortunately, it should be pretty easy to fix
21:35.49kesha_I cloned the repo from github and did git init.
21:36.20mpictoroh you don't need to do git init after git clone
21:36.31mpictorI guess it wiped the history out
21:36.49kesha_oh, I saw .. 1337 behind :/
21:37.00mpictorcreate a backup of your repo
21:37.14mpictorcreate a new dir, and clone stepcode into that
21:37.37mpictorin the repo with the changes, delete the .git dir
21:38.01mpictorcopy the .git dir from the freshly cloned repo into your old repo
21:38.15mpictorthen create a new branch and add your changes
21:38.47mpictorI found the cause of the error
21:38.53kesha_what ?
21:39.24mpictorI guess one of the warnings you fixed was 'unused variable first'
21:39.49mpictoryou moved it from outside a loop to inside the loop, so it's always set to 1
21:40.12mpictorthe loop doesn't actually show up on that page
21:40.44mpictorthe loop starts on line 1812, hidden in the LISTdo macro
21:40.46mpictorLISTdo( attr_list, a, Variable )
21:41.27mpictorI don't know why it thought the variable wasn't used, but it is incorrect
21:42.23mpictorif we can't find a way to silence the warning, we'll have to leave it
21:43.22kesha_previously also I had encountered with such instances where cppcheck claims variable to be not used, but they were actually used.
21:43.42kesha_I might have been a little careless with this one :(
21:44.07mpictorthose macros don't help to make the code easier to read!
21:44.29mpictorthey reduce the number of lines, but that's about it
21:45.36mpictorif you haven't seen the definition of LISTdo, it's at
21:47.46kesha_I reinserted "int first=1", but still the same error ! :/
21:50.19kesha_oh, no. It vanished off :)
21:51.53mpictorwhen you redo your commit after replacing .git, please break it up into smaller chunks
21:52.12kesha_regarding that transferring of .git/ ,if I delete .git/ , wont it remove all commits also ?
21:53.24mpictorthe changes will remain in the files, as long as you haven't checked out a different branch or used 'git reset <...> --hard'
21:53.43mpictoralso, please make a backup just in case
21:54.47kesha_How do I go about fixing- "C-style pointer casting" ?
21:55.02mpictorcan you give an example
21:55.50kesha_for example, src/fedex_plus/
22:02.12mpictorI guess use static_cast or reinterpret_cast
22:04.19mpictoroops, that particular example should use dynamic_cast since EntList inherits from SimpleList
22:05.06mpictorthat line would become         nm = ( dynamic_cast<SimpleList *>(ent) )->name;
22:07.33kesha_okay :) And last one, there cppcheck says many functions are not used. What shall I do to those functions ?
22:07.47kesha_I dont think commenting out all of them a nice idea !
22:08.10mpictorit depends where they are
22:08.30mpictorin fedex_plus, unused functions will probably remain unused
22:08.41mpictorso it's ok to comment them out
22:09.31mpictorin one of the libraries - express, exppp, cl* - they might be useful
22:10.16kesha_heh ,, most of them are from those libraries only !
22:10.30mpictorI'd leave them alone
22:11.33mpictordon't bother fixing any warnings in src/clprobe-ui
22:11.51mpictormany of the others are definitely used by the generated code
22:12.04kesha_Its just the remanant of prev ui :)
22:12.38mpictorI haven't deleted it because it could be useful as a reference on how to do something
22:12.57mpictor(and how not to do many other things)
22:13.40kesha_so, let those all warnings keep screaming ?
22:14.13mpictorI tried to fix it... many horrible macros in the InterViews widget libraries, and I got many compiler errors that I couldn't figure out
22:14.45mpictorI guess so
22:15.33mpictorwe can use a code coverage tool like gcov to get a better idea of what's actually used
22:15.54mpictorit takes a very long time to run, though
22:17.26mpictoroops, I misspoke
22:17.42mpictorgcov / lcov tells you what functions are called in a given run
22:18.07mpictorwhile a static tool like cppcheck warns you if it doesn't find function calls
22:18.26mpictorif *both* say a function is unused, it probably isn't used
22:20.24kesha_ just shows the stats of *how many* functions - where does it show *which* all?
22:21.20mpictoryou have to click on a dir, then a file
22:22.12mpictorthe numbers there are very low
22:23.38mpictorhmmm... maybe they are, maybe not... I don't remember if I ran any of the p21read tests
22:24.10mpictorI only provided the link as an example; we would need to re-run it because a lot has changed since then
22:24.59kesha_ ??
22:26.13mpictormain() has the number 40 by it; I think that is the number of times the line was executed
22:26.32mpictorif all of the tests were done, I think that would be 150 or so
22:27.35mpictorthere is a function summary for each file, but for c++ it shows the mangled names
22:30.33kesha_hmm..anyways, by tomorrow or max. day after that all the genuine warnings reported by cppcheck will be cleared off. :)
22:31.26kesha_Then, what will we go for next ?
22:31.52mpictorif you need to find where something is printed in fedex_plus, there is a way to find it fairly easily - but you'll need a commit that isn't in master yet
22:31.58mpictor(one sec)
22:33.27mpictor'git cherry-pick 0eb4f8ba547' , then 'cmake .. -DSC_TRACE_FPRINTF=ON'
22:34.14mpictorthe generated code won't compile, but each line will have a comment listing a file and line in fedex_plus
22:34.16kesha_did u code it now ?
22:34.45mpictorI did it a while back, but it got broken when I renamed some stuff from SCL* to SC* - I missed that one
22:35.20mpictorI would have told you earlier if I knew you would need to debug the generated code - sorry!
22:36.59mpictorafter you work through the cppcheck warnings, what about looking at the gcc warnings for the generated code?
22:37.34mpictorthose will be more difficult to trace since they will mostly originate in fedex_plus
22:39.51kesha_oh, fear and they are large in numbers too !
22:40.41mpictorfix one and chances are you'll fix a hundred other identical warnings :)
22:44.03kesha_How about going to some other issue. I think compiler warnings can be handled when I am sort of little bit busy. I have plenty time now, so I can take up a fresh new thing to work upon and have enough time to read and understand all related code.
22:44.36kesha_This is just what I feel, you can even say no! :D
22:45.14mpictorwhat about #200 then?
22:45.47mpictoror for a change of pace, #235
22:52.47mpictoronce you think you're done with #21, enable testing and run 'make test' to run all of the tests
22:53.44mpictorbrlcad, starseeker: any point in keeping the 203wseds schema around? all of the files we have for it don't match it and show up on cdash as failed tests
22:53.57kesha_#235 sounds interesting. I haven't done such a thing, so any pre-requistes knowledge or links you could guide me with to get a headstart ?
22:54.24kesha_Sure, I will run 'make test'
22:56.43mpictorI'd look at the cmake documentation -
22:57.16mpictortheir wiki has some useful stuff on it, but some of it is outdated
23:02.00kesha_cool. Get back to you soon. My bed is calling me ;)
23:02.09mpictorhah I bet

Generated by Modified by Tim Riker to work with infobot.