Hello, I sent some emails to the list, but due to some magic, they never reached the list. Pancake suggested, that it might be interesting to forward our conversation to the list
---------- Forwarded message ---------- From: mobi phil <[email protected]> Date: Tue, May 11, 2010 at 1:09 PM Subject: Re: Fwd: The results of your email commands To: pancake <[email protected]> > Hehe this means hyou send the mail with unsubscribed from address. sorry.. you mean from unsubscribed address... no, I am subscribed... > Are you receiving mails from the list? I thought you were already subscibred Yes, I am receiving emails from the list... >> - Unprocessed: >> some potential bugs, or "for me does not works'" running radare: >> 1. pdf @ sym.main ; i get Too many length > you can change this by removing the screen flood protection e scr.floodprot > or something like this. thanks, will try.. >> 2. if I want to use dwarf: >> addr2line: '$FILE': No such file >> Usage: rsc list <file>:<line> >> 3. doing F6 crashes my X windows session hm.. okidoki ... >> 4. mapping with o, does not accept addresses > will check that later.. But must be working. Can you provide an example? > Which command did you use? >> one day will hack for sure source code of radare.. but till that... >> maybe some hint? > > i should make a new minor release :) please do that... Q: why don't you use git? It is your freedom to use hg, but more people are going for git... would it be huge overhead to add git to your repos? >> Did not get any feedback on my prev email, probably because the most >> important question was at the end: >> there is a considerable difference in speed when using search command >> in debug and non debug mode. Should I understand that when in non >> debug mode, you search through ptrace? Maybe the image of the >> debugged >> process could be copied on demand, and search there... >> (and from my prev email... please comment my suggestion about insert >> mode under 'c' mode) > > You can do !dump or !restore to get a memory snapshot of the process in > disk. And use it to search over there. Yes, it is an alternative... but adding it an option to the search command would save time... > In r2 there are some caching mechanisms, but uses ptrace at the end. So > dumping to disk and searching there will be faster. the same as above > Doing non-requested caches can result on invalid memory contents.. Read that > if you search over a memory block it should keep as is.. But not if you > propagate such cache after the end of the search.. > Looks like an interesting caching optimization for r2. Will keep it mind. Do that :) > Never expect this to happen in r1. You have ways to do it right there. Sorry ... what were you reffering here? >> And the first one, that is not so relevant: >> 1. I understand that you try to follow some vi principles, keys and modal >> editing etc. My very first remark is that when you go to "c" mode >> the key >> to go into "overwrite" mode should maybe not "i" but the eq. in vim >> "gR". > > call it cursor mode :) > > When working in binary files takes not much sense to use non-overwrite > insertion modes.. I just find 'i' more intuitive than gR.. You have an eval > var to enable insert mode without overwriting. > >> The "i" seems counter-intuitive for a vi user, as i would go into >> insert >> mode, whereas you go in radare into "overwrite" mode, so it seems >> like a >> little trap... > > What solution do you propose? as mentioned gR would be the equivalent in vim, but you would be not that wrong with simple "r". As you may know "r" replaces is vim one letter without advance. > You have keybindings in normal letters for all the things you have in Fx. > Just check visual help V? I was not aware... however checked visual help.. > If there's one missing report it and i will add it. But you can create your > own keybindings too. That is what I did not find yet... I think you should create a wiki site... then people might help you with documentation :) > If you think on any change of keys or command names, please propose it :) i > think usability bugs are as important as segfaults. That for sure.. I am happy that you see that like this... > I will forward this mail to nibble. Feel ffree to send it to the mailing > list if you find it interesting :) No problem, will forward it to the list once more, then... -- rgrds, mobi phil being mobile, but including technology http://mobiphil.com -- rgrds, mobi phil being mobile, but including technology http://mobiphil.com _______________________________________________ radare mailing list [email protected] http://lists.nopcode.org/listinfo.cgi/radare-nopcode.org
