Archive for December, 2005

Hackfest 10000 Mortem

Friday, December 16th, 2005

Since last week:

Vinny:

  • Managed to get PyOgre to compile on the Mac.
    • One downside though: it doesn’t fully work with OpenSDL
  • PyOgre runs and a has started to tie into Commotion.
  • Scenes in Ogre can be created through PyOgre in Commotion.

Graham:

  • Finals… *sigh*

Luke:

  • Linux on kitchen sink.
    • Loaded up OpenWRT onto my Linksys WRT54G.
    • Loaded up Linux onto my Treo 650.
  • Did some MythTV work especially to fix my settings that allow me to have the MythTV box automagically shutdown when idle and startup when it needs to record.

Tonight:

Vinny:

  • Messed with commotion.
  • Downloaded a version of emacs thats not from the 70’s.
  • Spent some time on real-time document collaboration.
    • Real-time in this case means 2+ people are editing at the same time and all other editors see their changes in real time.
    • SubEthaEdit is an excellent example of this principle.
    • DocSynch markets itself as being exactly what Vinny wanted, but turned out that almost everything is still "planned".
    • Considering JEdit as a potential base to implement a cross-platform version.  Hopefully the protocol will be generic enough that others will implement it for other editors (emacs, eclipse, etc…).
    • BEEP for Java would be great for this if Vinny can find a open version that supports peer features.

Graham:

  • Played with Emacs syntax tables and got the syntax system to tell
    him what was punctuation.
  • Contimplating the next steps for his prototype.
  • Potential next step: rewrite the code to do character level timing so that its easy to manage timing across edits.  And, crack the problem of serializing the data across sessions.

Luke:

  • Did some reading of arm/linux documents and source code.
  • Thought about trust systems and personal data.  Can one create services in which the user doesn’t need to trust you?
  • Discussed rich instant messanger services.

I’ll linkify things later.

Atlhack Podcasts

Monday, December 12th, 2005

Links to Atlhack podcasts go under this page.

Podcast 1111

Monday, December 12th, 2005

The audio is alive!

Hackfest 1111 Mortem

Thursday, December 8th, 2005

Since Last Week:

New cool posters from pushmepullyou and methanestudios.
Podcast goes here.

Vinny-
Running in circles.
Graham-
Doing homework and running in circles.
Ben-
Nothing, except a birthday.

Today’s Plans:

Vinny-
PyOgre should be running.
Graham-
Need to fix bug where text properties spread. Need them to be confined to periods.
Ben-
Figure out how to make JUnit classes that wrap regular classes.

Today’s Reality:

Vinny-
Installed newer version of Python (2.4) and scons (python make). Pyst is not working.
Graham-
Made clock properties non-sticky, which fixed the issues. Working on rendering the invisible clocks.
Ben-
Will autogenerate testclass. You manually wrote a test class so he can design the code generator.
Will need to fix grading threads, each in their own thread-group. So when a student starts a JFrame in their program, you can shut down the JFrame as well.

atlhack art: sticker

Sunday, December 4th, 2005

If we find a way to produce static stickers (pdf):

sticker

Hackfest Podcast

Friday, December 2nd, 2005

This week’s hackfest was brought to you by the letters M and P and the number 3!  Our first "podcast".

Hackfest 1111

Thursday, December 1st, 2005

Hackfest 1111, it’s on for next week.
See previous hacks for description.

Hackfest 1110 Mortem

Thursday, December 1st, 2005

Today’s notes are informal, but action-packed!

Luke commits to starting part-time technology business.
Vinny is doing an audio podcast. He will put the link here.
J knits a pink hat.
Graham still trying to fix his clocks code. Invisible clocks don’t work because somehow text properties are off.

Vinny is working on Gmaps schools application – running live on rails.

Luke helps Graham with various emacs snags.
Helpful emacs commands: f10 gets you to the menu.

Luke found the page tables for his handheld coding.
He wants to get an idea for the segment sizes and how it shows up in virtual address space.
He wants to put the kernel in a well-understood place so he is just interacting with linear memory.