Adaptive Advertising (again)

Ask questions and give answers about computers, mobile devices, game boxes, PC security and all manner of geeky stuff.
User avatar
yogi
Posts: 5950
Joined: 14 Feb 2015, 15:49

Re: Adaptive Advertising (again)

Post by yogi » 13 Mar 2019, 14:08

Linux server managers would be out of a job if it wasn't for Windows.

My dad's bigger than your dad. nah nah nah nahhh

User avatar
Kellemora
Posts: 3482
Joined: 16 Feb 2015, 11:54

Re: Adaptive Advertising (again)

Post by Kellemora » 14 Mar 2019, 12:18

True that!

How many servers would be in business if they all had to pay the big bucks to Mickey$oft. One out of a thousand?

User avatar
yogi
Posts: 5950
Joined: 14 Feb 2015, 15:49

Re: Adaptive Advertising (again)

Post by yogi » 14 Mar 2019, 13:08

You got to remember that a lot of servers out there belong to big companies which pass the cost of operations onto the consumer. It really doesn't matter if they pay a premium price for their software or not. The bigger issue is finding qualified people to keep the servers online.

User avatar
Kellemora
Posts: 3482
Joined: 16 Feb 2015, 11:54

Re: Adaptive Advertising (again)

Post by Kellemora » 15 Mar 2019, 13:11

I do know, based on the cost to built Titan, and its expected lifetime, it cost 2 thousand dollars an hour to build and run.
They do have access to it at the college but you have to sign up for how much space you need and for how long. Many jobs are piggybacked on with other jobs to save time and costs.

But the Titan is not exactly like a server farm where you can lease space and usage based on things like number of connects at a time and things I know absolutely nothing about.

My little website is running on a massive host, probably with millions of other websites.
In this regard, I'm like Windows users, don't know what's going on under the hood, just that it works, hi hi.
But I do have one step up, I write my own html without using templates or programs to do it, other than a simple text editor.

User avatar
yogi
Posts: 5950
Joined: 14 Feb 2015, 15:49

Re: Adaptive Advertising (again)

Post by yogi » 15 Mar 2019, 16:38

Every once in a while I run into articles that say the client/server model is obsolete and needs to be replaced. Then they go off into a parallel universe and explain how to do it. Same goes for passwords. They are the weakest link to secure communications and need to be replaced. I guess biometrics is one approach but apparently there are other methods to assure only the right person gets access to the right data. It's all geek to me. :lol:

User avatar
Kellemora
Posts: 3482
Joined: 16 Feb 2015, 11:54

Re: Adaptive Advertising (again)

Post by Kellemora » 16 Mar 2019, 12:50

It seems with all the technology we do have in the computing world. Things like malware, trojans, and ransomware would be discovered immediately before they could be transferred on to someones computer.

After all, for the file to get uploaded to the server to a persons file, it has to be read and stored by the computer, so it should check for something malicious and stop it from being uploaded. This way no one would have to worry about downloading it.
Yes I know, it's all 0's and 1's so unless they know what they do when strung together, how can they tell?
Even so, it seems like certain strings of code would be necessary to cause an adverse affect and they could block those or at least sandbox them until a programmer looks to see what it is doing, just in case it is on the up and up.

I had an image file once a long time ago, I built it up pixel by pixel myself as a map for a game I was playing.
Nothing at all strange about it, but after we got new computers, virus checkers would flag it as harmful.
Also, for the gaming machines, we would create the Casino's Logo and make it appear on the screens also.
Every one of these, when opened on a Windows computer would be flagged as a virus. I have no idea why. Most were made using Windows Paint, saved as PCX and converted to the image format used in the game machines.

I use simple passwords on my own machines here at home, but quite a bit more complex on websites where I need a password to get in. I figure if someone wants to single me out to get into my machines, they will do so regardless of what I use as a password, hi hi. To many back door ways to get in based on programs I have installed. Like Flash Player.

User avatar
yogi
Posts: 5950
Joined: 14 Feb 2015, 15:49

Re: Adaptive Advertising (again)

Post by yogi » 17 Mar 2019, 17:19

I think you answered your own question regarding preemptively stopping malware from being distributed. There are too many back doors, thus zero day attacks are common. There is no signature for the average virus until it is created. Think of it as being a checksum. You don't know the value of the checksum until the file is saved. Much of the malware and virus software reaching it's targets is simply authorized by the user. For example, this website can easily be cloned in it's appearance. When you come here the next time and see that everything looks normal, you go ahead and open the web page and start reading my meanderings. It's too late by that time should you have inadvertently logged into a malicious web site. Furthermore, maybe nothing happened that you can see, but a trojan was downloaded along with, say, the logo for our web page. That trojan can sit on your computer for a long time before it is activated remotely to do its dirty work. You may not even know it's involved with a DDoS network. All you will see is a slow response on the network for a short period of time. In reality you have become part of a bot net.

That's just one simple example. The best preventative measures I've seen are what they call heuristic analysis. The functions in your operating system are constantly being monitored in this method. Should anything happen that is out of the ordinary for your system, it gets flagged, quarantined, or deleted until you have an opportunity to review it. While this is a good approach it's success depends on the ability of the software to "learn" about your system to see what exactly is normal. A lot of bad things can happen in that learning process.

Then there is the situation you describe wherein a lot of false positives are generated. That's generally a case of being too paranoid and flagging anything that doesn't fit the pattern. Malicious software is often successful because it imitates valid and normally safe software. It works the other way round too where perfectly good software could do things that only malicious software has done in the past. It's game of hide-and-seek when it comes to malware and viruses. Sometimes the bad guys are ahead of the game, sometimes not.

User avatar
Kellemora
Posts: 3482
Joined: 16 Feb 2015, 11:54

Re: Adaptive Advertising (again)

Post by Kellemora » 18 Mar 2019, 12:34

The frau got a nasty virus on her computer a few months back, and we figured out exactly where it came from too.
It was embedded in an image file she downloaded from a website.
Although she uses a virus detection program, apparently they only work if they know the name of the virus as a download.

By the same token, there were a few tools we used to download on-line for our use on the computers.
And like you said, somebody else put up a website using the exact same name as the original file, only it was slightly larger in size than the original. This is how they got another virus into her computer.

Once the virus program embedded in the image did run, a separate virus scan of her computer found it and deleted it.
But all she had to do was view that photo again, and bang it was back again.
Moved it to trash and then emptied the trash and no more problems from it.

Even in the Linux world, sometime a program will make it as far as a repository before someone catches something it is doing that it shouldn't. Like giving the ability to turn on a camera when the program has nothing to do with a camera.
But with thousands of eyes watching these programs, they soon find it and remove it from the repository. Most often, they never make it onto a repository. I normally don't download a program from a website unless it is a must have program, and then I first find out what the checksum is supposed to be, or the true file size.

User avatar
yogi
Posts: 5950
Joined: 14 Feb 2015, 15:49

Re: Adaptive Advertising (again)

Post by yogi » 18 Mar 2019, 14:34

One reason I don't bother with anti-virus (AV) programs on my Windows computers is the experience Deb had on her system. Not all viruses are caught by a single AV program. When downloading a single file I always run it through the VirusTotal software I have installed or upload it to their web site if the file is small enough. I also have a utility called Process Explorer which can check every process running against 50+ antivirus databases in real time. Thus I can tell if a virus is currently running. If it is dormant I'd have to find it some other way.

Malicious code embedded in images is particularly sinister because it can be executed simply by downloading the picture. No other action is necessary on the part of the target machine.

You are very lucky to have been able to rid the machine of the virus simply by deleting a single file. Many virus makers anticipate this and change registry entries, modify dll files, or embed their code in some hidden part of the OS. In some cases the root cause is stashed in BIOS. Fortunately UEFI has a way of preventing that, but there is still a lot of BIOS out there.

User avatar
Kellemora
Posts: 3482
Joined: 16 Feb 2015, 11:54

Re: Adaptive Advertising (again)

Post by Kellemora » 19 Mar 2019, 12:02

Well, if I recall, after we deleted the image file, we also ran a program that looked for missing files. In other words, files that program may have needed to run. Or files that are not associated with any other program.
Deleting them is sometimes risky, because they are there for games you may load from another drive.

Back in the XP days, it was not uncommon for me to wipe a drive and reinstall windows and the programs I was using.
Most often, this was because I bought a bigger hard drive, and a matching size external drive to use for backups.
Although time consuming, I did have dealer install disks for all of the Windows XP versions I had, so they could be loaded unattended. Not like the headache of installing from the OEM disks with all the crapware on them.
The machines always ran faster using a dealer install disk!

User avatar
Kellemora
Posts: 3482
Joined: 16 Feb 2015, 11:54

Re: Adaptive Advertising (again)

Post by Kellemora » 19 Mar 2019, 12:07

Forgot to mention. I used to have a program on my XP machine that told me WHERE every single file was placed on my computer when I installed a new program. Or even used a test version of the program.
You would not believe the hundreds of places a program writes to when you take one for a trial run with a 5 to 30 day trial usage period. Or how many places they write to if you do a full install from a paid for version.
Very few programs written for Windows will run from their own folder, and then only if you put the dll files in the dll file folder on the machine.
Many of the early simple games, like little brick out, and others would run entirely from the folder you installed them in.
You could simply move the folder to a new computer and it would work without reinstalling the game.
This is the way many things on Linux work, no messing with the computers files, hi hi.

User avatar
yogi
Posts: 5950
Joined: 14 Feb 2015, 15:49

Re: Adaptive Advertising (again)

Post by yogi » 19 Mar 2019, 15:33

When you install a program into Windows it typically will be registered and does poke around to many shared parts of the system. This is normal for Windows. I'm not entirely clear why it's normal, but the architecture has been pretty much the same since DOS days. Now and days you can get many of the popular programs in "portable" versions. These programs are stand alone and have no dependencies. Almost any browsers can be installed as a portable version for example. They run from the one folder (or USB memory stick) onto which they were installed.

I'm not sure what you mean by OEM disks because all the ones I've seen are trash. These are indeed stripped down versions of the full Windows Operating System and will run faster prior to the OEM making their mods. I've never found one of those OEM disks to be free of the crap that the computer manufacturer foists upon unsuspecting users. Same goes for those computers with recovery partitions. The OS in those partitions has all the bloatware that the OEM can fit into it. It is handy at times, but I prefer to buy an unadulterated version of the OS right from the source and then modify it to my own specifications. Guaranteed not to have crapware that way - unless you consider the OS itself to be crap. :lol:

User avatar
Kellemora
Posts: 3482
Joined: 16 Feb 2015, 11:54

Re: Adaptive Advertising (again)

Post by Kellemora » 20 Mar 2019, 10:37

OK, during the time period I'm talking about, you could buy:
From Microsoft the commercial versions of each of their XP OS versions, such as XP Home, XP Pro, XP Pro MCE, etc.
From the computer maker an OEM version to reinstall whatever version was on that model computer.
Not supposed to be sold, but you could get Dealer/Repair Shop totally clean versions, without the OEM crapola.
The thing about the Dealer/Repair Shop version is it was designed for unattended installation. You didn't have to answer the questions until you ran it the first time, same as when you buy a new computer from a store. With one main difference. The Dealer/Repair Shop version already had one user established, usually with the entry point of Owner Password, and a burn in test to run the computer through all kinds of tests before selling it.
The copy of XP Pro I had also could print out a page showing all the components in the computer, CPU, RAM, HD size, etc.
Since I bought so many computers from this one dealer, he always tossed in a Dealer install disk for me, no charge.

User avatar
yogi
Posts: 5950
Joined: 14 Feb 2015, 15:49

Re: Adaptive Advertising (again)

Post by yogi » 20 Mar 2019, 15:35

The way I understood it back in the days when I had a job is that OEM versions of Windows were special licensing agreements. These systems were not the same as those sold to the general public and could only be purchased in bulk quantities. One OEM license was impossible to get, for example. The idea was to encourage manufacturers and corporations to buy huge numbers of Windows OS licenses to install in the products they sold or bare bones computers that corporations purchased for internal use. While the OS was not allowed to be changed, there were any number of additions that could be made, such as the corporate logo during the boot process.

People like Dell Computer would give out what they called "recovery disks" that essentially was the modified version of the OEM plain vanilla system. I don't think you could install from scratch using the recovery disks, but it was like having an image (iso) from which to draw on in case of a fatal crash. The alternative to these recovery disks was a partition that was installed alongside Windows. One recovery disk could be made from that partition just in case your whole computer went up in flames.

Since the OEM license was a lot cheaper than the retail version, Microsoft had a vested interest in not allowing OEM software to be distributed to the public freely. That doesn't mean it could not be done. It just wasn't part of the deal.

User avatar
Kellemora
Posts: 3482
Joined: 16 Feb 2015, 11:54

Re: Adaptive Advertising (again)

Post by Kellemora » 22 Mar 2019, 11:28

Every OEM disk I had that came with a computer, would only work on that exact computer. About the only thing you could change was the HD, so it much have checked the Mobo serial number or something.

The Dealer disks I had could be installed on any computer, without having to go through the install process as you had to from a Real Windows disk. You still had to go through the set-up process, but this was after it was installed, so it went a whole lot faster. Or I should say, you didn't have to sit there for hours to get the OS installed.

I don't remember where we talked about computer games, but I learned something new, although I partially knew some of it. Almost all games load part of the game into your computer, but I thought this was just to access their game servers.

Mainly I was trying to figure out how FarmTown handled so many players, and had all this interactivity between them.
Turns out, my way of thinking was really off the wall, it works nothing at all like I imagined. Although you are playing the game on their servers, sorta, almost everything is done on your own computer, and only the data is exchanged every split second with their databases.
Let's start with something simple, like doing the Cooperative Quests, where you look to see what each player needs.
In this case, you are not really seeing their farm on a Game-Server, you are only seeing an Image on your own computer, no Game-Server involved. Nor or you going to a players farm to do this part of the game.
When you open FarmTown, it loads the game module for your farm onto your computer, then takes you to your own default farm, still basically on your own computer, but is also speaking with YOUR GameRoom on their Game-Server.
This is how the game becomes interactive. If someone clicks to visit YOUR farm, they are going to YOUR GameRoom on the Game-Server. All the game server does is maintain an up to date image of the data on the data servers, nothing more.
When you do work on your farm, all this work is done on your computer and only saved to the database on the data server which is then reread into the gameroom on the gameserver, so other players can see what is going on, and or also working on your farm at the same time.
No actual work takes place on the game server other than image updates shared between those logged into your game room on that server. If no one is logged into your gameroom, the image does not update on the game server.
It is the database that is doing almost all the heavy work of recording what you are doing, and refreshing your screen every split second. You can work faster than the data is going back and forth, but your on-board game program is showing you what you are doing, and you could be ahead of what the database is done. You can sometimes see this when harvesting crops as there will be a shadow of what is not yet sent to the server, and sometimes you can see the buffer as a light square for a second or two as it unloads. Go to fast and you end up with a few extra of these light squares and a warning to slow down, hi hi.
But, the program on your computer is what is doing all the display work, and exchanging data with the database.
This is how they use so few resources at their end, in order to have millions of players all playing the same game.
Technically, each person is playing their own individual game, with the data going into the dataserver.
If someone else joins your game, they are seeing it in the gameroom on the gameserver, but the gameroom is also just a database transferred to their computer which is doing most of the work of parsing the data and creating the display.

I didn't do very good at explaining what was told to me, they did it much more eloquently than I did, hi hi.

It's like our webpages are nothing more than an html file, which is sent to our web browswer which in turn parses the data into the graphical display of the web page. Here too, almost all the work is done by the users computer viewing the web page. All they are doing is sending the text file. Their server is a little busier if it is a php program with a database, but not that much busier. The php file creates the html file on their server, then their server sends the text file to your computer. Seems games work in about the same way.

User avatar
yogi
Posts: 5950
Joined: 14 Feb 2015, 15:49

Re: Adaptive Advertising (again)

Post by yogi » 22 Mar 2019, 14:43

I probably didn't explain it well enough, but I was alluding to the same process that you describe when I fantasized how Farm Town might be played. The playing field is constructed from the client side software you install on your computer. It's built locally and not on any server per se. All the servers involved are in reality databases. Your local computer takes input from those databases to construct your playing field.

I play one game similar to SIMs, and it operates pretty much the same way I described my idea of Farm Town. All the furniture in my SIM room and city is downloaded from the game server in the cloud, but only once. After an item is downloaded, the graphics information is stored on my computer. Revisiting the room causes it to be reconstructed from the cache of items stored in my local game client space, not from the server. One day I noted that there was about 10GB of unexplained file space being taken up on my HDD. After some research I discovered that this game I play was storing everything in my rooms/cities/characters on my computer. There is no limit, apparently. If I play the game long enough (years) my hard drive would eventually be filled up with all the unique items from the game. I solved that problem by simply deleting the inventory cache every so often. That does not disrupt the game, but it does force the rooms to load again from the server. Farm Town must do something similar. The playing field is constructed on your computer and the items you see in the playing field are cached on your hard drive. Of course there is communications with the game server and character (farmer) server to keep track of any changes in your local environment. The changes, however, are not in graphic form but in numerical or character string form. Updating that kind of data is a lot quicker than updating full blown images.

The multi-player concept would seem trivial. All visitors to your farm have a database entry describing the status of their character. Thus any enhancements to the visitor's character (farmer) get sent back to the same server that is keeping track of a farmer's status. It really doesn't matter which farm the visitor is working because only his attributes are stored in the database. It doesn't matter where he acquired those attributes. The same can be said for your farm. It doesn't matter who makes changes to the landscape. The servers simply keep track of what changes are current. The farms are replicated for each player working on your farm by drawing from the two databases; one for your farm and one for the farmer. Using the HTML analogy is close to what is happening in Farm Town. Everybody on the same website sees the same thing because there is a server in the cloud that has all the content information (not a fully rendered web page). Also, when you go back to that website a second or third time, it is rebuilt from your browser's cache and not downloaded from the server unless something changed in the interim. Even then only the changes are downloaded.

I don't know about you, but I think I explained it to myself well enough to understand what is/was happening in Farm Town. :mrgreen:

User avatar
Kellemora
Posts: 3482
Joined: 16 Feb 2015, 11:54

Re: Adaptive Advertising (again)

Post by Kellemora » 23 Mar 2019, 10:22

Yeppers, you do explain things a whole lot better than I do.
It is still a mystery how they store all that information for 6 million players for years without losing any data.
Even if you quit playing, you can go back again years later and start back up where you left off.

User avatar
yogi
Posts: 5950
Joined: 14 Feb 2015, 15:49

Re: Adaptive Advertising (again)

Post by yogi » 23 Mar 2019, 15:30

Think about something you already know and accept. Think Google. Google has records in its database that go back to the beginning of their presence on the Internet. If you use the right search terms you can come up with any article, picture, sound track, video, and the gods only know what else. Not only does Google have this information, it is stored on many computers scattered all over the world and is being accessed by millions of people 24x7. They deliver it to you in milliseconds. I'm certain most front page searches are retrieved in your browser faster than what you can do on your local servers. LOL

How do they do that? Actually, it's a pretty well guarded secret, but the point is that Google serves more people than any game server, or even ten game servers, will ever be capable of doing. Your entire Farm Town game and character consists of two strings of data, one for the farm and one for the farmer. These two strings are used to build the playing field on your local computer and 98% of the time the graphics for that field is stored in your computer's memory. Storing a million strings of data is trivial if you have the right database. Ten million or a hundred million just involves a little more indexing, but it's not that complicated.

User avatar
Kellemora
Posts: 3482
Joined: 16 Feb 2015, 11:54

Re: Adaptive Advertising (again)

Post by Kellemora » 24 Mar 2019, 12:11

Awhile back when we were talking about this same thing. You gave me the link to a website that I spent days reading and studying.
When we save things on our computer, we normally do so in Folders and sub-Folders so we can find the data when we need it, and it's all together in one Folder. Like I have a main folder for Organized Photo's. Inside this is a sub-Folder for each persons individual name, one for Animals, etc. But inside each persons individual folder, I have more sub-folders, one for Group pictures they are in, one for pictures of just them, and one for a close-up of their Face. Plus in many cases a folder marked Misc. Inside the Misc. Folder I may have a folder for the houses, cars, pets, and memorabilia, like a trip to the lake, or a cross country tour, etc.

OK, back to that website you sent me to. Sorry I no longer have the link, it was made on another computer that fried, and this was before the Google Bookmark bar too.
In any case, you didn't have to worry about folder or much of anything else. Everything you want to store is accessible through a Search Engine, almost like using the Web Search Browser. You want to find a photo of Uncle Frank, just type Images containing Frank, Franklin, or Frankie. And everyone on your hard drive with that in the name would pop up.
Looking for a document but don't remember the name you filed it under, just type Document containing the words "and then he jumped over the" and it would bring up documents containing that string.

The thing about it was, it was supposed to be much faster than using your on-board computer search feature, which could take a long time to find something, because things were not stored in folders.
This doesn't sound to smart to me, like everything but everything in a single folder, but it may be what things are coming to in the future, due to better searching capabilities.

I know I have thousands of duplicate files on my hard drives and backup drives due to my redundancy, hi hi.
Trouble is, I can't trust the programs that hunt for duplicates. I have many files by the same name, but in specific folders, and all too often these duplicate checking programs see them all as the same file, when they are nothing alike.

User avatar
yogi
Posts: 5950
Joined: 14 Feb 2015, 15:49

Re: Adaptive Advertising (again)

Post by yogi » 24 Mar 2019, 18:48

Too bad you don't have the link to that alternative search site. I'd be interested in looking it over again. I certainly don't recall now how I came up with it, but I do have a vast collection of bookmarks. Hopefully it's in there somewhere.

As you must know there are no such things as folders when it comes to CPU's. It's all done with memory addresses and blocks of mapped memory. Windows uses a high level search algorithm so that it's fairly easy for us mere humans to understand. The price you pay for that intuitiveness is that it takes a long time to find something. Lunux, and it's related OS's, have something called grep. I had to learn a little about it when I worked for a living but could not tell you much these days. It would be the perfect solution for your file searches in that it works directly on memory and does not have to interpret folder names or human keywords. You do give it string content or file attributes for it to work, but that's not the same as trying to figure out Google search commands, for instance. Unfortunately, grep is so sophisticated that it might require a lot of time studying it in order to get it to do what you need done.

Post Reply