Richard Choi

Established Members
  • Content Count

  • Joined

  • Last visited

Everything posted by Richard Choi

  1. emc: I'm looking at your issues now I'll try and get the other issues to the appropriate developers.
  2. th3ron and I tried to reproduce the crash on create torrent in Windows 7 x64, and could not reproduce it. If someone can post a small file for which the crash reproduces onto rapidshare, along with any settings used to create the torrent, and settings files (if they are different from the default), we will try to reproduce it on our machines. Thanks!
  3. PIxelated: we'll look into that one, we're getting a Vista 64 environment set up for testing
  4. Is this problem specific to Windows 7 64? Has anyone gotten the installer to work properly using regular Windows 7?
  5. One of the changes was that we updated the blacklist of bad programs to include another NVIDIA filter that was showing up quite a bit in crash logs... no other really significant changes that I recall.
  6. No, there was a bug in the past where uTorrent would pop up the Windows UAC if you chose to install the client. If you cancelled out of the UAC dialog, uTorrent would not handle that case properly and end up in a partially-installed state.
  7. wolf550e: the toolbar is not mandatory... what do you mean that you are unable to install?
  8. AJZ: thanks for catching that issue, looking into it now.
  9. Thanks for the help... I will fix this. Edit: @rafi: I used your .dat and did not have filter/label problems... it seemed to properly apply the labels specified by the favorites for any downloaded torrents. @5618: When 1.8 was released, we made it default to use the feed name for pre-1.8 feeds instead of using the feed's metadata, because users may have expected the alias to be the URL. The change in 1.8.3 to automatically apply the feed's name as a label was requested by the users, so it was added. Perhaps it makes sense to add an advanced option for this behavior to turn off...
  10. @5618: so the issue is fixed for you now? I am unclear as to the specific issue you are/were encountering...
  11. rafi: do you mean the speed was smoother and less erratic?
  12. fowl: for multiple reasons, we don't want to sign the .exe after bundling, and after a little investigation, we could not find a simple way to bundle without corrupting the signature's integrity. Any useful suggestions to doing this would be appreciated.
  13. aids10: as soon as the bug reports start dying down
  14. There aren't any significant differences beyond offering browser toolbars for a small % of new installs upon install. More here: The revision needed to be increased to properly parse crash dumps. Outside of the install, nothing is different in the update.
  15. The hash bug has been fixed for some time now, although the *latest* build of 1.9 alpha has not been released yet. It should be released very shortly.
  16. kokobaroko: there are some significant uTP optimizations in the next alpha (coming soon)
  17. Fadeout: were you having upload issues in the previous betas?
  18. We have identified the memory leak (and a few others that existed previously), and should have a new build shortly.
  19. No more testing necessary: memory leak fixed
  20. bt.transp_disposition = 5 should disable uTP
  21. Has anyone else seen ax89's error #1? Also, we are trying to isolate the memory leak. Does it still occur if uTP is disabled? Or if no torrents are active?
  22. We are looking into a solution to get old skins working with a mixture of defaults for new images. Did this problem also occur in the 1.7x-1.8 update for people using skins? New images were added in that update as well. For people getting high memory usage, can you provide us with simple steps to duplicate it?
  23. No, this release does not support DNA functionality.
  24. 1.8.3 and up brought back old RSS status icons (new within 24 hrs, previously downloaded) into the main window. There is also a different RSS icon for feeds that are not functional.
  25. hunt3r: thanks for the link... I am not sure if other devs already investigated this, but I'll take a look. Some of these options may affect performance or code size, but we'll run some experiments.