Process Lasso v5.0.0.20 FinalProcess Lasso v5.0.0.19 FinalProcess Lasso v5.0.0.18 FinalProcess Lasso v5.0.0.17 FinalProcess Lasso v5.0.0.16 Finalhttp://www.bitsum.com/files/prolasso.ziphttp://www.bitsum.com/files/prolasso64.ziphttp://www.bitsum.com/changelog/pl/changes.htmv5.0.0.20
One of the optimizations I employ is reusing memory instead of having to free and reallocate it. Yes, it does make things a bit more complicated, but it improves efficiency considerably. Sadly, while I reviewing the code I saw a logic error that needed repairing. I have no idea if it ever manifested in the real world, I've not seen it. It would have manifested most likely as a crash.
However, I did fix something that I know did manifest in an XP test bed I have with a unique software environment that includes NIS 2011: a very small memory leak in the core engine. I do not know how many systems were affected, I certainly have not seen this on any other test beds. Also, it is a very small and slow leak that would take a good amount of time to build up (and be continually paged out in the meantime). So, it is unlikely anyone who was affected actually noticed. Regardless, I repaired it and am issuing this update.
Thus, we have v5.0.0.20:
[.20]Fix.Core: Fixed a logic error in memory reuse optimization (unknown if manifested in real world)
[.20]Fix.Core: Fixed a small memory leak seen in exhaustive testing under XP /w NIS 2011 present (special handling involved with NIS2011)
[.20]Change.GUI: Updated Serbian Latin
[.20]Change.GUI: Updated Finnish
Please keep in mind this version was a large leap, and I certainly do wish I had polished it more prior to the first final version. However, we are where we are, and I can say that it is now where it should have been to start with. Yes, I am embarrassed with the admission that I discovered more bugs. At least I discovered these before they were even reported, for what it is worth

.
The polishing continues...
» Click to show Spoiler - click again to hide... «
v5.0.0.19
I had a user report that rules weren't being enforced in a timely manner. I partially addressed this in v5.0.0.17, but it turned out there was one more little tweak to do. This tweak was actually rather important, and improves efficiency in addition to fixing the application of No sleep and automatic gaming mode.
I do apologize for overlooking this. I thank all users who report bugs. There are so many features of Process Lasso and it runs in so many different environments that it is imperative users report these bugs. Usually I fix bugs very rapidly once I become aware of them (often within 1 hour I have a new build out). It is exceedingly important that this version excel. Although it is embarrassing that any bugs slipped through testing, I must admit and own up to them - so users know they existed and know a fix is now available.
ProBalance, default priorities, default affinities, and most other rules were never affected. Only No Sleep and Gaming Mode were guaranteed to be affected, with an imperceptible improvement to the performance of Default Power Schemes (though it never malfunctioned).
This build is being created now and should be available in the next few minutes.
v5.0.0.18
During testing a few hours ago I noticed the auto-update was failing in XP. Needless to say, I was a bit bummed and surprised by this. An investigation showed the cause was quite obvious (at least in retrospect). I must have been tired when I wrote the original code. The surprising thing is that it worked elsewhere to be honest, which is why this slipped by my previous testing.
This build fixes the updater, BUT it will not start working until the *next* update after this one because this build contains the fix.
This should be the FINAL fix to the automatic updater. Now that it is sorted out I can begin issuing updates like a crazy person ;p
This build will be available in the next few minutes, I am making this post a bit prematurely.
v5.0.0.17
This build fixes a reported issue with the watchdog rules (not default affinities, priorities, power schemes, but rather the terminate/restart/affinity on X% CPU/MEM). They were not being applied instantaneousness and a fix was applied to the Clear All button in their dialog.
Russian and Italian languages were also updated.
This post has been edited by cybpsych: May 30 2011, 08:03 PM