Well, I spent a bunch of time on triggers last night. They seem to be working pretty well at home. There is one issue remaining for single line triggers that I need to work out (I think my indices are off.) I also think I know how to do multiline triggers.
If the sy != ey then I have multiline. So, I can then query the line numbers from text block. I can then get the geometry for each line from start_idx to end_idx. each of these line geometries will form an area that I can use as a trigger block. Should make it easy to setup the triggers I think.
There seems to be an issue on this amd64 box here, the first trigger isn't appearing. So, either I broke something last night without realizing it, or something very wacky is going on.
So, we went down to check out the Toronto Young Street festival a couple of times this weekend. (One of the venues was 2 blocks from my house). Saw Great Big Sea playing down at Young and Dundas Square. Was a pretty good concert, with some freaky huge giraffes in the background. Seems that Toronto has a new branding scheme: Toronto Unlimited it has a pretty good video associated with it that they showed.
In other news, I had to go pick up a cable modem, which apparently dosen't work in my house as the signal strength isn't strong enough. At which point Rogers said my cable would be down for a week. We then proceeded to argue with them for about an hour until they give us the old modem back until they fix the new one.
I sat down and wrote the trigger code for Ewl on the weekend, twice. The first time it wasn't going as I wanted, and was a bit sloppy so I saved it away and started again. The re-write went pretty quick, I like the way it came together. Now, if only it worked at all, heh. I'll have to keep playing with that a bit I think.
So, I've been spending some time looking at security stuff the last few days. I've let my knowledge slip a bit over the ages. Started looking at it again after discovering the Toronto Area Security Klatch. So, I printed off a copy of Smashing the stack for fun and profit which is a really interesting read and have been heading on from there.
Spent some time puttering around shellcode.org which has some pretty good info, and read through some of the shellcode community postings on livejournal.
It brings up some interesting sounding programs and possibilities. The one that caught my eye was a source code scanner to find bad format string usage in programs (Can't remember the name or url at the moment). It would be an interesting exercise to extend that concept, except write a full on c compiler. After you get the code into your abstract syntax tree start the analysis from there. You can keep track of what is the dirty data and how its used and probably do other bounds checks and that kinda thing. Not sure how well it would work, but it seems like an interesting idea at least.
Anyway, more to look at and more to read. If you've got some interesting links, send em my way. Articles, URLs, tools, whatever.
Well, I've broken down and started a log here at livejournal. I've finally decided I'm just too lazy to write my own.
In other news, I'm working on some updates to the Ewl_Text widget to allow for selections and what I'm calling 'triggers'. A trigger is essentially an HTML style link inside the text. So, you can get mouse over, mouse out, and click events off of the trigger text. It could be interesting getting it to work correctly with Evas_Object_Textblock. Should be an interesting experiment. Luckily a selection is just a specialized type of trigger so I should be able to use the same code to handle both of them.