Discussion in 'other anti-virus software' started by Templar, Dec 29, 2009.
Could not figure out any productive way to add to the conversation huh ?
Why are you even reading the thread if you don't care?!? Move along.
I have noticed this "hidden" ram usage a long time ago. IMO, the easiest and truest way to see how much ram an app is using is to start up the system and measure free available ram before installing the app, then install the app, and check free available ram again. The difference is how much your app is using. And many times this is NOT what you see in task manager or other programs, it is much more.
That said though, I would agree with some others about ram not being what really matters. Most of us nowadays have plenty of ram. The real thing that matters is system responsiveness and performance.
People are aloud to give their input on forums, that's the whole point of it. If Fajo didnt care, he would not have posted. He also is not attacking any individuals opinions, just giving his point of view - unlike your last 2 posts here, sure they're also opinions, but we can see they are aggressively directed, to someone who did not provoke you in any way.
May I also add, you have no right to tell others on this forum what to do.
Any time a thread comes up on memory use, there are people stating this isnt 1990, and all PCs have 4-8gb of Ram, and no one should care about memory use. Fact is, not everyone had 4gb of ram, and people do care about efficient programs. Otherwise threads like this would not start to begin with.
Tell me exactly how using 50Megs vs 100Megs of ram shows any different in efficient programing. You have no idea what a program does behind the scenes nor do you have any idea how it is coded or coded in a way to use that memory. People wanting Anti-Virus to shove all there usage into the pagefile is just idiotic. The memory that a program uses is there because that's what it needs not there to make you feel happy it uses 5 megs of memory. You want a AV that uses no resources go use placebo and let us know how it goes, it's memory usage should be right up your ally.
No problem happy to be of service. Then again you did not answer any of the questions above lol.
There was not any "?" is your statement above. How is a memory resident program using 100meg better than one using 15meg, all other things being equal?
If someone has 1gb of ram, that is about 10% vs 1.5%. And don't tell me the person should buy more ram.
My question was how can you tell it's more efficient for the program that uses 50megs vs 100megs. How it looks to me and I could be wrong so correct me if I am. But it seems you think programers are slacking on there job, if program A can use 50 megs to do its job But yet program B needs more. Even tho the programs and how they work maybe totally different from one and other even tho they come from the same sector as one and other.
Also don't tell people to buy more ram. Why not ? You don't have to be rich to get ram and it would boost overall system performance not just a AV. People that want more demanding programs should not demand the software vendor sacrifice quality just because they are to stubborn to fork out $30 USD for a 2gig stick. Even less if you ebay it.
The point of this thread is how AV programs hide memory usage. Not to buy more ram, or RAM usage doesn't matter. I have had this RAM debate with you before, and we are not going to changes each others minds on the topic.
This was the thing I responded to and that's the question I asked. As for buying ram yes it's alternative and fix's 90% of all issues people have with a slugish computer.
As for RAM USEAGE and Hiding the ram. There is no real way to hide ram unless you want to shove a crap load of stuff into the pagefile to make it look like your using less resources. The sad part is people seem to think programs should use 5-10 megs of Memory or else its not as efficient or is poorly programed. The fact is as tech moves forward so will programs people that think AV's should use this much ram or that much ram or even run on there 10 year old computer as fast as it does on a new computer are in for a rude awaking.
Now agreed we probably would not change one and other minds. Now does this make ether of us wrong of course not. That's what I like about this board the most is the chance to debate my views and listen to yours.
And it's a good time waster why at work. Well at least until something happens to the Network I maintain.
Fajo, you're missing the point.
1) This is about "hiding" resource usage. There are ways to do such things through APIs and built in OS stacks.
Just because there's a new generation of RAM geeks out there that don't care if their AV use a GIG of ram doesn't mean that some of us still consider a piece of software something that should be streamlined at the same time provide a good function of its intented use.
Very good point was raised by one poster about CPU time and I/O. All you do is to tell us how you don't care, please ignore this thread if you really don't care and thus this thread quality is dropping fast.
Please try to be professional and understand the issues even if you consider them not something relevant to your usage.
System responsiveness is of course the main thing and that goes without saying.
Having said that no one has mentioned bandwidth usage of cloud based solutions, it would be interesting to know just how much traffick/packets are being swapped on for example PrevX, Panda Cloud and even Norton which uses Reputation. A lot of countries now days are introducing bandwidth cost models (Those in a country where this is not something on the horizon don't need to post to tell us how they don't care - Thanks).
On a lighter note I'm happy to see how the performance views have started to emerge.
Happy new year every one..
Oh Fajo, u don't get it do ya? . Its not about the RAM!, ofcourse we are concerned about our old pc's. If u have 1TB RAM its good for you that doesn't mean everyone would have 4GB's. Plus this thread is about antivirus hiding there resource usage, Many websites claim that NIS is the lightest av for the year 2010, which is absolutely wrong. Yes! it does not slow your pc down, but that doesn't mean its light. No Offense
Ok then show me proof it is eating the ram. And I don't mean a screenshot of Physical ram in task manager. As anything can reserve ram then give it up as soon as its needed. I want to see some proof here that it is "Hiding its usage" otherwise this is all just speculation.
This seems to be just semantics to me. Some would consider this a contradictory statement. As a customer do you care if its the lightest or not even though you know it does not slow your PC down. I think the only reason anyone would care is if they have a bone to pick with Norton or they are a competitive ghostwriter
@ fajo: i can't give u proof for now as i'm not a coder/serious hacker. But someone here will be able to do it.. Time will tell
@Zombini: I'm using another SS now. I'm very happy with it. I would not have noticed this memory consumption if i was not using any main memory monitoring tools.
first off im not at all saying one is better than the other here. i like both eset and nis just my views from actually running both.... and i own BOTH ess and nis2010. i can 100% say they are almost equal speed and resource wise. i have tested both on various machine configs from old to new just to see the difference and tbh most of the time nis runs faster. there was next to no boot time difference at all except the first reboot with nortons because it does the insight checks and does a scan, the second reboot after scanning is always much faster yet. i do not see this "extra" usage with nis2010 even with procc explorer. i see nothing out of the usual. nis2010 does use a tad bit more during scans than some others but i see 0 slowdowns even on a older celeron d machine with only 1.5gb of ram. i had to remove kaspersky from that system because it just dragged so much (previous version of kaspersky not 2010). that is my mil's and she even saw the difference and is about as computer illiterate as they get.
i like eset and ess do not get me wrong but i had way to many overall issues with the newest version with the firewall and downloads i had to look elsewhere. and if you check the eset forums i was def not the only one. ever since ess rolled out imo they need a few revisions and updates to fix things each time. for me i dont have the patience for that. again its a great product i had issues and on more than one machine.
im very picky about usage and overall speed and it took me a LOT to try the nis2009 last year and i have to say im still glad i did. i may still switch to something else but thats mainly due to me never being settled on one thing for to long.. so far nis is the second longest one that has been on my systems with the first being nod32 back in the v2 days along side outpost before outpost got crazy bloated...
This argument implies that tools such as Task Manager or Process Explorer inaccurately report memory usage. I am very incredulous that such is the case -- but, I am also more than willing to be proved wrong.
If the memory is used in a different process (such as a service running in svchost.exe), Process Explorer will show you - but you have to know that you should look there.
If the memory is allocated in a driver (some AVs used to do it, don't know what's the situation today), Process Explorer won't show you anything, because it doesn't belong to any process.
Anyway, this whole discussion on memory usage misses an important point (well, I'm sure there are quite a few of them)... why the hell would you want to have free RAM in your computer? Did you buy it just to sit there, maybe eat some power and produce some heat... and be useless otherwise?
I mean - yes, it's bad if an application uses too much memory all the time, but imagine the following strategy of an antivirus:
- suppose the scanning process can be made faster by using specific lookup tables
- check the amount of free RAM on the system at the particular moment
- if there's (more than) enough free RAM, allocate bigger tables used for scanning - i.e. use more memory, but scan faster
- if there isn't enough free memory, use a more conservative allocation strategy - i.e. use less memory, but scan slower
- periodically check the status and switch between those two, if necessary
Now, do you really prefer an antivirus that uses as little memory as possible all the time, no matter what - or would you rather have a more advanced piece of code that can use the free memory (if you have it) for something useful?
Can you really say anything reasonable by the memory usage numbers you see in Task Manager / Process Explorer / ... ?
Very good point! What is so bad about a program USING Ram? It could very well make the program run lighter...
I vote 4 u man This is what i'm looking forward to.I don't wan't av's to hog resources while gaming/ running heavy apps. Many here believes increasing memory increases system speed which is proved wrong [there's a limit, around 4 GB i think, TIP: use same speed&size Dimms on both sockets ie ]
I believe you're missing the whole point yourself.
This is about having resource use openly viewable and not stealthed, it's about knowing how much drag a piece of software is creating on your computer.
You and Fajo can start another thread about how you don't care how much RAM, IO or CPU time is being consumed by your fav AV. This thread isn't about if it's BAD to use RAM or anything in that regard, it's about knowing how much of your computers resources are being consumed.
Please stay on topic.
The problem is that even if you know the "exact" numbers, you cannot draw any reasonable conclusion from them (even if you think you can).
If one program has higher numbers shown in Task Manager / Process Explorer / whatever other tool), you cannot say that it uses more memory. And now I'm not talking about "hiding" the usage anyhow, but simply about the fact that you have no idea how the program works internally.
Maybe it shares memory blocks between multiple processes (thus if you sum the usage of those several processes, you basically get the number N-times bigger than reality). Maybe it maps files (or memory blocks) multiple times within a process - thus raising the numbers you see again, but not really using more memory.
Maybe it lowers the memory usage actificially by swapping out its whole memory every few seconds - then the numbers you see will be very low, but the operating system will have to page the data in (from disk) very often, making your system slower.
Maybe the program performs "big" operations in driver - so it allocates the memory in driver, and you won't see these numbers anywhere (but it's no hiding, it's simpy the way it works if you do things in driver).
Maybe there is a reason for running a service inside of svchost, thus shifting a part of memory usage into a system process (and I'm not saying NIS does that, I think it's probably a side effect of something, but I don't know that).
Maybe the program uses more memory only at the moment a lot of free memory is available.
So, great, be happy knowing the exact numbers of your resources being used - but face the fact that they are basically random numbers for you - you have no idea what they mean. Therefore, whether you know the exact numbers, or if something is "hidden" for you (either on purpose, or just as a side-effect of the particular implementation), it doesn't change anything - you'll know the same in both cases.
As already written, the only thing that matters is the (subjective) response of your system.
Read the posts I said prove that they are doing this. Otherwise this whole thread is nothing but speculation.
I never said I did not care infact IO reads and CPU time are important. but wanting all programs to use the same amount of ram is pointless they are all coded different and have different functions people are comparing apples to oranges here.
That whole sentence about made me fall out of my chair laughing.
Separate names with a comma.