Shareaza

I enjoy tracking/testing the code updates provided by the Shareaza programmers but miss using the script for semi-automated Daily Build updates available a while back. Using the principle of "release early/ release often" here is my attempt to update that script for the recent build and file locations.

Shareaza Daily Updater v0.18

Modified from Shareaza Daily Updater v0.14 by ???
Caution: This is a very rough attempt to get this started again. It didn't break anything on my machine. I make no guarantees about yours.

Pointed at files generated for the Shareaza sourceforge.net Daily Builds location
Added very rough checkpoints via popups to indicate download, shutdown, reinstall, and restart phases


FractalMode

good to see the script back again

i just took a few minutes to (partially) fix 2 old bugs.
1. The script did only check for today's and yesterday's installer, now it looks months back. (that code section is still missing some stuff, like checks for years, endless loop prevention and extended http response code handling)
2. The script didn't support reading Shareaza registry values on 64bit Windows, now it does.

on a side note... please get rid of all the popup windows soon

Thank you for the code contributions and corrections. I am excited that this seems to be working on someone else's machine besides mine. V0.20 should reproduce quite closely the behaviour of v0.14, the pre-Sourceforge version of SDU.

Shareaza Daily Updater v0.20

1. SDU now defaults to silence per request Set VerboseMode = 1to re-enable display of status messages.

Up next...

SDU should download file and install only if newer than current Shareaza. Suggestions for how to retrieve the daily release number for the local Shareaza install would be greatly appreciated.

changes in 0.21:
* fixed updater failing because no build of current svn revision is available. script now checks past three revisions by default
+ script now retains closed state if Shareaza wasn't running before the script killed it

please note that the current url-guessing method is flawed, slow, ineffective and the script will possibly make dozens of requests to the web server before it hits an existing file. it should probably be replaced by a method using rss/atom feeds.