Posts Tagged ‘geek’

My kingdom for a VGA cable

2010.11.26 22:44 by Leo Antunes - 1 Comment

So you have two geeks in a university room after a relatively late and (at least for one of them) unproductive learning session. It’s just natural that they decide to kick back and watch some mind-numbingly stupid geek series, which in this case happened to be Stargate SG-1 (so absurdly shitty its actually very entertaining).
The first lazy geek instinct is to just watch it on the laptop that has the file, which – with its 11″ display and shitty speakers – doesn’t turn out to be a great idea. The next try involves the other laptop, but a 13″ screen isn’t that big of an improvement.
Since the room our intrepid heroes are in happens to have a pretty decent built-in projector and a couple of small – but still a lot better than a laptop’s – Bose speakers, the obvious next step would be using it. The only problem is the lack of a VGA cable.

Inspired by the brief sight of MacGyver on the 11″ screen, one particularly enterprising geek comes up with the challenge of making a VGA cable out of the only material available at the time: one horribly yellow cat5 ethernet cable.
Being the helpful little extra-dimensional entity that it is, the internet happily provided all the needed information and after some slight problems trying to appropriately deprive the cat5 of its connectors (no scissors and no blades of any kind in sight) and some annoying and manual sticking-cable-to-socket action…

what did you mother tell you about sticking things in sockets?

our reluctant hackers get it right:

it's ALIIIIIVEEEE!!!

The final solution looked like this:

nothing like cable salad for dinner

And if you’re wondering where those white wires came from, one final touch of über-hackerdom:

notepads have never been more useful

This might seem like overkill, but after a nice nice 4 hour movies and series marathon, we can safely say it was totally worth it (but no, we didn’t stand 4 hours of Stargate; even geeks have their limits).

 

Just in case the Instructables page gets hosed at some point, here’s the invaluable connection diagram, originally scraped off of a since dead Geocities page.

Portal was great, this is just wow.

2009.07.07 21:16 by Leo Antunes - 0 Comment

I liked Portal just like probably the rest of the geeks out there, but then comes a guy and makes an ASCII re-imagining of it. I found it just so amazingly clever and ingenious that I couldn’t help posting about it. It felt like a simple tweet (I’m still trying) wouldn’t do it justice.

 

(via Offworld, via BoingBoing)

Just rinding the Wave

2009.05.29 22:01 by Leo Antunes - 0 Comment

I know everyone’s been talking about it, and that I’m not supposed to be a news blog or anything like that, but I’m pretty excited about Google Wave after watching the presentation.

I try really hard not to be a Google fanboy, but I must admit I sympathize with the company and the way they work. The “do no evil” stuff might not be entirely true, but it’s corporate behavior is certainly well within what I consider a reasonable trade-off between the modern needs of a huge company and the lofty ideals of social responsibility.
It could all be a ruse to get big enough to rule the world and the paranoid side of me isn’t completely dead, but I give them the benefit of the doubt for now.
Anyway, I digress in my justification of admiration for Google.

I haven’t yet read the protocol specs, but I’m curious to see how they deal with connection problems between federating servers and the conflicting edits that might arise from such a situation. Also curious about possible desktop client integration and offline use.
I’ll hopefully have time to browse through it in the next few days.

Twhat?

2009.05.07 11:55 by Leo Antunes - 0 Comment

I give up. I was gonna complain about how impossible it has been these last weeks to get some geek news without having to read through thousands of headlines about how Twitter is gonna save the world and cure cancer, but now I’m gonna try and use it for real – belatedly, of course – to see if it’s really as useless for me as I’ve always thought.

Note that I won’t go as far saying it’s useless for everyone – specially since it’s more dangerous to mess with twitoholic fanboys by cursing Twitter online than cursing Mohamed inside a Mosque – but I just don’t think it matches my usual web habits.

Since I’m backpedaling I might as well join identi.ca, but nobody outside the geek world seems to know about it, so I’m peer-pressured, so to speak, into Twitter.

Migrating like the ducks

2008.10.05 18:07 by Leo Antunes - 2 Comments

So yeah, apparently I wasn’t as satisfied with the blog as I thought I was.

Just dumped Drupal in favor of WordPress.

The main reasons were:

  • Simple OpenID comment support. I didn’t want to have to code that into Drupal for the reasons explained in the next point. This is related to another small project of mine which will hopefully be in a usable state soon(ish).
  • Code simplicity. I enjoy how Drupal is coded, from a technical standpoint, but it’s just not practical for my humble needs. It’s overkill. WordPress may feel more hackish in some points, but at least I feel relatively comfortable jumping into the code and customizing it to what I want. (although getting this site ready involved just hacking a theme together and installing some plugins)
  • Bloat. Drupal is too resource-intensive. Side-effect of the “everything plus the kitchen sink” complex.

I tried to keep all the permalinks working and hopefully the very few people that follow this through RSS won’t have any problems.

There are a couple of issues I already spotted:

  • WebKit based browsers seem to be running into a javascript error when entering a new comment. The comment gets saved, but the user gets redirected to a 404 page. I suspect the error affects other browsers as well, only the handling is somewhat different, but I haven’t really looked into it yet. If there’s anyone reading this using Google Chrome or Safari I might be forced to debug this further.
  • The theme looks like crap on IE6 and older. I don’t really care for them and I have no idea how many people still use them, so I’m probably just gonna leave it like that. (It’s times like these I really appreciate not working as a web-programmer anymore)

UPDATE: yeah, I screwed up and flooded the RSS aggregators with duped posts. Thankfully only a handful of people read this.

Exatamente como Jesus…

2008.07.13 06:27 by Leo Antunes - 0 Comment

…só que ao invés de 3 dias, demorou quase um ano.

Pra quem não entendeu, isso tem a ver com este post: o GB voltou ao ar. Ressuscitado depois de muita enrolação da minha parte.

Pra quem ainda não sabe do que se trata, não faz a menor diferença na vida de vocês.
Pra quem sabe, é suco de nostalgia!

Guck 0.1 released

2008.07.13 06:20 by Leo Antunes - Comments Off on Guck 0.1 released

Finally the first release of Guck.
Not a very important step for mankind, but it was worth the time fooling around with the code.

The dragon’s alive, but I slayed this gecko instead (twice)… do I still get the ladies?

2008.07.12 07:49 by Leo Antunes - 1 Comment

I’ve been obsessing over the last few hours about a small problem I was sure I could solve with a single RegExp: parsing specific attributes out of HTML tags.
Just when I was about to give up and wrap the whole damn thing inside a loop, I came up with this pearl (no pun intended, I wasn’t even coding Perl):

/(?:[\w-]+(?<!href|src)(?:\s*=\s*(?:('|").*?\1|[^\s><'"]+?(?=\s|>|\/))))(?=[^<]*?>)/gi

This one allows only href and src to pass through, for instance.

During this venture I discovered this very nice online RegExp tester, which helped me get past some mind blocks. It’s got some annoying bugs, but visualizing the matches and doing quick edits prevail over the few small annoyances!

Even though I enjoy coding regexes quite a lot, I can’t help feeling they sometimes look like brainfuck, or some equally bizarre thing.

UPDATE: So yeah, my little monster apparently had some corner-cases uncovered. Fixed (hopefully) all but one: a leftover “>” will trigger the removal of any attribute=value pair before it. Since lookbehinds can’t have varied lengths, we can’t be absolutely sure we’re inside a tag. Since in my case it’s better to remove more then less, I still consider it a good solution to the problem, but certainly not the most readable or maintainable… damn, for a second I really thought I had found an elegant solution!

PS.: when your RegExes create horizontal scrollbars, you know there’s gotta be something wrong with you…