CNET 'Reviews' Process Lasso

On Dec 22nd 2009, one of CNET's download.com 'editors' reviewed Process Lasso v3.70.6. There was some good and bad in the review, but the end conclusion was great.

Unfortunately, the graph legend apparently was not shown on to the reviewer due to the initial window size. They must have been using a small laptop, netbook, or virtual machine. Sure, they could have resized the graph or main window to have the legend shown, but they didn't know that. This caused them to be confused about what the graph metrics meant.

They also neglected to ever mention ProBalance, arguably the primary function of Process Lasso. They did, however, highly praise the many of the other features, such as Default Priorities, Default CPU Affinities, High Performance Power Scheme processes, and Gaming processes.

The reviewer correctly asserted the software was too technical, and he or she had to refer to the documentation multiple times. This is something I'm working on improving.

They concluded that the product really did make a difference in their PC's performance, and ended up recommending the software.

Overall, good news. Their rating was 3.5 stars. Average user rating is currently 4 stars. I honestly think it deserved at least 4 stars, especially since they rated a much older version with 4 stars (see article update below). It all depends on which reviewer checks your software I guess. For this reviewer, I think it being so technical threw them off. Still, I can't complain that they concluded it did make a difference - unlike so much other optimization software.

They concluded that the product really did make a difference in their PC's performance, and ended up recommending the software.

UPDATE: I noticed that Download.com still lists the old Process Supervisor (the original name of Process Lasso). Ironically, they reviewed it and gave it 4 stars years ago, lol. The original reviewer focused on its out-of-control process restraint (now ProBalance). Its only gotten better since then. I think I will request a re-review by v4, as its not good that the primary feature wasn't even mentioned in CNet's newer review.

[ Link to Process Lasso x32 on CNET ]

Read More

The hazards of reading user reviews

Every once in a while, I go to check user reviews on popular download sites. I want to see how users are reacting to my software. In some ways, this is a necessary evil. I need feedback. For instance, I know my GUI is apparently very ugly, and that's my top priority now - as you can see in v3.80 (and more changes planned for v3.90).

Fortunately, 95% of Process Lasso reviews are VERY POSITIVE. That's great, and wonderfully inspiring.

However, checking user reviews isn't without its hazards. I take the few negative comments personally, and I've created enemies, somehow. I suspect its because I've negatively reviewed software in the past, under my own name. Who knows though.. For whatever reason, I do have people who come and negatively review every almost every release of my software. They either hate me, hate life, or like being trolls -- or maybe they just are convinced of their position and just aggressively asserting their belief. I don't know. UPDATE: I am happy to report that one of these people changed his tune, giving me renewed faith in humanity ;)

The other problem is negative reviews based on misunderstandings. I can't tell you how often I read a review where the person says something along the lines of "I didn't download this, but you can do that with the Windows task manager. What a rip off!". Yes, they obviously missed the point of Process Lasso, and assumed its just another task manager. That is an easy mistake to make, especially if you just take a quick glance at the name and screenshot without reading further. Most people just don't have the attention span to read a paragraph that explains the software's purpose. That's understandable.. there's a lot of text on the net.

Given that I tend to take negative reviews to heart, I have to limit my checking of reviews. I did so this week, so I won't again for at least 6 months. In the meantime, I'll be trying my best to deliver quality software. It sometimes frustrates me that I have to work so hard to create a prettier GUI because I'm not using .NET or some other RAD (Rapid Application Development) language, but its worth it in the long run (less RAM use) -- my core user base appreciates me coding in native C++, at least I think they do ;).

Anyway, to any other software developers out there. Here's my advice (as obvious at may be):

• NEVER post a negative review of someone else's software under your own name. Even if their software is a total gimmick (i.e. snake oil), and you feel compelled to reveal that fact. Don't do it. They will come back and negatively review your stuff ten times worse.
• NEVER create fake profiles (I never have). Also make sure any resellers don't do this. Sure, its tempting to correct misunderstandings, but its just not worth it. It calls into question all the legitimate positive reviews.
• NEVER respond to negative comments, it feeds trolls and makes it look like you are inflating your own rating. I find it hard to follow this advice, hence I don't even look at the reviews but once in a blue moon.
• NEVER listen to anyone complaining because your software isn't more free than it already is.
• NEVER let them get you down. Its easy to be a critic. Don't let the cynics and critics kill your inspiration. Again, I find it hard to follow this advice. I am out of patience for all the critics of the world, who complain about this and that - never creating anything themselves. I am NOT talking about people who make suggestions or CONSTRUCTIVE criticism, that's a different story. We all know the type of person I'm speaking of. As with anything in life, there are negative people out there, and either you tune them out, or they infect you with their negativity.

For whatever its worth... These things seem obvious, but I've had to learn the hard way.

Read More

Running Process Lasso with Highest Rights - new beta improvements

I'm approaching final for v3.80. I've accomplished all but a few items remaining on my list, then I'll enter the testing (and regression testing) stage.

Slowly over the course of the next few days, I'll introduce a few of the new features in v3.80. Today's feature is better automatic startup of Process Lasso with Highest Rights for Vista+ systems with UAC turned on.

As many of us know, Windows Defender blocks automatic startup of all applications that want to run with Highest Rights. You can't even create an ignore list. I reckon this is a good thing, from a big picture perspective.

If you are running Windows XP or below, this article doesn't apply to you. So, you can stop reading now unless you are curious.

Since many people need to run Process Lasso with Highest Rights, even though it is not normally recommended, we have further accommodated that capability. Remember, Highest Rights is the mode that allows any single instance of Process Lasso to view and manage the processes of all users (including system processes). In some cases, this is certainly desired. In most cases, its not.

Originally, I had planned to work around Windows Defender by adding Process Lasso to the Windows Task Scheduler's list of application to launch when a user logs in. However, I changed my mind and tried a different approach. I call it the 'proxied launch'.

While Windows Defender doesn't allow launching of High Rights processes, it does allow launching of low-rights processes, who can then launch a High Rights process. Of course, when done this way, the UAC elevation dialog is shown to the user. In current beta, its actually shown twice - once for the GUI and once for the core engine. I plan to lower that to once before final.

Whether or not this is the way I stay with, I don't know. As a side note, the user has the choice of using the proxied launch or not. Still, the official recommendation is to just let Process Lasso run with normal rights. The only exclusions to this rule are single-user systems where a system service needs managing, or servers where the processes of multiple users need managing by a single instance of Process Lasso running as a service.

Read More

WARNING: Beware of pirated copies of Process Lasso Pro

Some users may be tempted to save money by downloading illegal copies of Process Lasso Pro. I highly recommend against doing that. Illegal copies of Process Lasso downloaded from the internet could be harboring malware, or be modified in such a way that they cause other problems. I am not going to claim that every illegally hosted copy of Process Lasso Pro has malware in it, but do you really want to take that chance? As a general rule, NEVER install software that you don't get from a trusted source (such as Bitsum Technologies).

Process Lasso Pro is very affordable. Why not just buy a license and be safe? If you truly can not afford one, email me and I'll work with you.

I work tirelessly on Process Lasso. I know we all have come to expect free software, but nothing in this world is free. The reason so many free projects are forced to turn into commercial projects is simply because the world doesn't work on charity. That's our society, and its beyond our control. If someone would feed and house my family, I would happily give all my software away for free. Until that day, please understand why I charge for Process Lasso Pro, and why I encourage you to purchase a license if you are able to afford one.

Thank you for your patronage.

Read More

New low cost license experiment

I've long thought software costs too much. It seems that every time I go to buy software, I get depressed. It is just so darn expense. That is why I've always tried to provide affordable licenses for my software, and I think I've done a good job of that.

However, as an experiment in further decreasing the cost of my software, I am now offering 1 and 3 month update subscriptions for $7.95 and $9.95, respectively. Want to buy one? Click here . Of course, given the extremely active development on our software, we must admit that a better value is to get a yearly or lifetime update subscription.

We also have a new official policy regarding subscription renewals. For Process Lasso, we had not clearly articulated it before, though it had been articulated for PECompact. Subscription renewals will cost only 50% of whatever the original cost of the license was. The renewal will be for the term you originally purchased (i.e. 12 months). So, if you purchased a yearly license for $16.95, at the end of 12 months you could renew the update subscription for $8.47. This renewal option is provided in our Support area.

Read More

The many instances of Chrome

As many readers of this blog may know, Google Chrome is different from other Windows web browsers. It isolates each of its tabs into an individual process, instead of containing them all in a single multi-threaded process. Google Chrome essentially forks itself each time you create a new tab. This can easily result in 10-20 chrome.exe processes running on your system. But, don't worry -- that doesn't mean more memory or CPU use. Read on.

Fork? That's when a process creates a new running copy of itself, identical to the first. As always (fork or not), the virtual memory between process instances is mostly shared. Therefore, additional RAM usage is negligible. From a RAM use perspective, there is no big difference between forking a process and spawning a new thread.

How does memory/RAM get shared between processes? Well, that's a different subject, but its really simple - its all in the virtual to physical mapping. Physical memory is the actual RAM, where-as virtual memory is an abstraction of that RAM that can exist in RAM, on disk (i.e. a page file), or anywhere else.

As a sharing example, a page of memory may exist at location X in physical RAM. Well, if two processes reference that same physical page of memory, they both can have different virtual pointers to it. If one process writes to that shared memory, Windows will apply a principle known as copy-on-write. This is where the process that wrote to the memory gets its very own copy of the shared page, because its now uniquely modified for that process. Read more on virtual memory if you want to know more.. back to Google Chrome.

Forking is very common for *nix environments, where applications fork themselves all the time. Windows applications traditionally spawn new threads instead of forking the entire process. Google decided to take a different approach, using forks. Note that the technical aspects of the Chrome fork may differ from a traditional *nix fork, but its conceptually the same.

The biggest advantage of isolating each tab into its own process space is that a crash in any one tab can't (or shouldn't) bring down the entire application (all the tabs). Again, since identical virtual memory is shared between processes, there isn't much extra memory overhead associated with all these Chrome processes. Each process will only allocate memory exclusively applicable to the tab it owns. Further, since the Windows CPU scheduler only sees threads, this forking design doesn't affect how Chrome performs during periods of high CPU contention.

One interesting performance note is that Chrome dynamically lowers the priority class of processes hosting background tabs, similar to what Process Lasso's ProBalance does. So, when you switch tabs, Google Chrome lowers the priority class of the old tab's process and raises the priority class of the new tab's process (to below normal and normal, respectively). I would hope that traditional single-process browsers do the same with individual thread priorities, but I wouldn't bet on it.

This lets the Windows CPU scheduler know which process's threads are most important during periods where there is heavy contention for CPU time slices. Of course, the most important threads will be those in the process thats holding the tab you're currently using. The others are important as well, but don't need quite the same level of real-time responsiveness as the foreground process's threads will. Although Windows has built in mechanisms to boost foreground thread performance, giving the Windows CPU scheduler 'hints' as to what should take precedence is always a good idea - after all, its not psychic.

That's what Process Lasso aims to do -- help give the Windows CPU scheduler a few more hints as to what should get precedence, and to rescue the user in the case of catastrophic CPU monopolization (as demonstrated by our CPU Eater demo). If more applications cared as much about their impact on the rest of the system as Chrome does, then perhaps there wouldn't be as much of a need for Process Lasso's ProBalance.

Well, I've spent enough time rambling on.. back to real work. I hope this helps explain to anyone who is concerned by the 20 different instances of chrome.exe they may see running on their PC ;).

Read More