Power of valgrind and cgdb combined

I just found out how to make valgrind ask if it should invoke cgdb when it finds an error:

valgrind --db-attach=yes --db-command="cgdb -nw %f %p"

This is incredibly useful. For those of you who don't know, cgdb is a curses interface to gdb, which displays a split view between the code and the gdb console. A bonus is that it uses vim-ish keys and commands. Valgrind is a very powerful memory debugger, which can find memory leaks and find illegal writes and reads to/from memory done by an application.

Addicted to vim and screen

I've become addicted to vim and screen. With it, I can get a very powerful development environment over one ssh connection. I switch between windows in screens with ctrl-a [0-9]. I switch between splits in vim with some custom bound key combos, ctrl-j and ctrl-k. I can issue regular expressions and I can use all the powerful and flexible editing commands vim offers. Just try pressing d and then any 'movement'. For example dt;. That one will cut all text before the next ;. Then there's the visual modes, v and V. Try selecting some code with it and pressing =. That will reindent that code. And this is just poking on the surface of what's possible with vim.

Artificial Stupidity

seqseq & int: error (7)
seqseq & float: error (7)
byteseq & seq: error (7)
byteseq & byte: error (7)
byteseq & int: error (7)
byteseq & float: error (7)
intseq & seq: error (7)
intseq & byte: error (7)
intseq & int: error (7)
intseq & float: error (7)
floatseq & seq: error (7)
floatseq & byte: error (7)
floatseq & int: error (7)
floatseq & float: error (7)
seqseq & seq: error (7)
seqseq & byte: error (7)
seqseq & int: error (7)
seqseq & float: error (7)
byteseq & seq: error (7)
byteseq & byte: error (7)
byteseq & int: error (7)
byteseq & float: error (7)
intseq & seq: error (7)
intseq & byte: error (7)
intseq & int: error (7)
intseq & float: error (7)
floatseq & seq: error (7)
floatseq & byte: error (7)
floatseq & int: error (7)
floatseq & float: error (7)
seqseq & seq: error (7)
seqseq & byte: error (7)
seqseq & int: error (7)
seqseq & float: error (7)
byteseq & seq: error (7)
byteseq & byte: error (7)
byteseq & int: error (7)
byteseq & float: error (7)
intseq & seq: error (7)
intseq & byte: error (7)
intseq & int: error (7)
intseq & float: error (7)
floatseq & seq: error (7)
floatseq & byte: error (7)
floatseq & int: error (7)
floatseq & float: error (7)
seqseq & seq: error (7)
seqseq & byte: error (7)
seqseq & int: error (7)
seqseq & float: error (7)
Passed Test!


...

How to get rid of ssh brute force attackers

Simply bind ssh to another port. The bots are programmed for port 22 and won't use any other port.

Note to oneself

What to not do in Nethack: Eat a zombie. It's very likely rotten.

Windows box fixed

Turns it it was ZoneAlarm that was blocking images it thought was ads. Disabling it also fixed the non-functional site.

Browsing in windows fscked

My mom noticed that some sites weren't rendered correctly on the winxp box we've got (that I rarely use, but that's another story) and some were even not viewable at all.

One of the sites that aren't viewed correctly is www.resedagboken.se. The photos in the middle of the page is simply... not there. This happened in both firefox and IE (!!). I examined the page source and compared it to the page source on my linux laptop. The result is: the <img> tags for these images are missing on that computer -- they are replaced with whitespace.

I'm having a hard time thinking of anything else than spyware that could cause this... so I'm currently scanning it with ad-aware. Who knows what that poor box might have gotten on it.

Firefox 1.5 rc2 is cool

SVG support, canvas and css3 columns -- what else do you want?
It's also a tad bit faster than 1.0.x. However it still has quite a bit of bugs.