Hell Oh World!

I read this comment on Slashdot on one of the reasons why Linux adoption is slow. Thought it was quite interesting:

When someone gets Windows, he installs it, starts it up and starts clicking around. Some things will work, some won't, but those that won't don't discourage him. After all, everyone says Windows is so easy to use, every dumbo can work it out. So they try. I mean, who wants to be dumber than... And they try. And putz around and finally (maybe after reinstalling, when they managed to click somewhere they really, really shouldn't) it works. When someone gets Linux, he installs is, starts it up and starts clicking around. Some things will work, some won't, but those that won't discourage him. After all, everyone says Linux is a geeks-only system, nobody but a true blooded geek can figure it out. So they don't even try and toss it as "too complicated".

I recently got an email from Nikhil Sinha who’s interested in writing some code for antidialer. He’s looking fix some stuff and also make a tray icon plugin which, I guess, is one of the must-have features that I was too lazy to write. It would be quite cool if he does get around to doing it.

Finally got back to working on ayttm after a, well, work break ;)

I spent the last 3 hours fighting with gpgme-1.x and have finally managed to get some things working. The strangest problem (and it took me quite a while to catch hold of) was that the deprecated gpgme_data_rewind(foo) was working but its non-deprecated replacement gpgme_data_seek(foo, 0, SEEK_SET) wouldn't work at all, unlike what the gpgme info docs said. Copy+paste it from the info docs and it still wouldn't work. I'll update if I get it working.

I’ll be at work this weekend too, and next week I’ll be practically living in office. Production releases are fun. Would have been even more fun if I was reimbursed for it ;)

Instead of GPGME API reference., I went for GPGME Reference and I found the GPGME API Reference Manual.

… apparently does not exist. Michael Belanger found out that Ayttm does not compile with GPGME v1.x, after which I found out that we’re compiling against GPGME v0.3.x. A lot seems to have changed since and I managed to get a list of what changed after a google search. If the 6 year old post on the gnupg mailing list still holds true, I’ll have to generate API docs using gnome doc tools on the gpgme source.

I had forgotten to test keyboard shortcuts for the ayttm chat window and main window. The chat window shortcuts were associated with a menu which was not associated with any window at all, hence the accelerators never got activated. Slept for most of the day and got it working at night. *Yawn* a very eventful day ;)

I also noticed something funny in the morning. Ayttm signs on with a blast of static noise, more static when you send/receive messages and a long static blast when it signs out. Ran through the code and found out that the .au file is cat'ed to /dev/audio. Initially I thought this was wrong and tried to find alternates. Turns out that is how you do sound with the .au format, so I've got to think of something else.

Today the ATM machine of my bank doled out two 50 Rupee notes for me when I asked for 100 Rs. This is good. Now I won’t have to run around asking for change :)

Just put up a couple of userpics… that’s me :D

I've finally got almost exactly what I wanted out of the combo of MTNL and Sify. I now use MTNL as my primary connection and whenever the router overheats and bails out on me, sify kicks in. There's some minor race condition-like situations where if a site is requested while the transition is happening from sify to mtnl, all requests to that site are routed through sify. I'll probably have to strictly alternate the interfaces to resolve this -- eth2 up would lead to eth1 down and eth2 down would lead to eth1 up.

I was thinking about redundant gateways as well, but feeling too lazy to go for it...