Main Area and Open Discussion > General Software Discussion
What the hell is OpenCandy?
40hz:
Yes, indeed it is a functional difference. Several things run on your machine without being installed
-wraith808 (March 31, 2011, 08:55 PM)
--- End quote ---
Just a minor niggle... ;)
I think you might have missed what I was trying to say there.
Based on what I learned from my professors, the fact that a piece of code is capable of being run is proof positive an installation occurred. There is absolutely no "functional difference" between installing to RAM or HD, because in either case it accomplishes the exact same function - getting a piece of binary code into a place where it can be executed by the target system.
To paraphrase so there's no confusion: If it's runnable - it's been installed.
-40hz (March 31, 2011, 11:39 PM)
--- End quote ---
You have a very ... unorthodox view of being installed. :huh:
The act of installation requires some very discreet actions that would seem to put lie to your explanation, and this is coming from someone who did installers for several years many moons ago on many varied OSes. -wraith808 (April 01, 2011, 07:40 AM)
--- End quote ---
If it's an unorthodox view, it's one shared by many. :P
I started referring to the act of loading an executable into working memory as an "install" after hearing Marvin Minsky refer to it that way during a lecture in 1977. Back then, we didn't differentiate between the act of "installing" in your sense (i.e. on a drive or other storage media) and loading it into RAM for execution. It was the act of loading and running that constituted the "install."
And since we're trotting out our CVs, my view is from the perspective of someone who (like yourself) has programmed and installed software on various OSs for...well, let's just say a bit more than several years, starting with IBM's System/360 on mainframes, and Digital Research's CP/M on personal computers. :)
For one thing, installation by any practical definition implies that the software in question can be run again without any need for further installation (leaving out the situation of a software upgrade, since the act of upgrading is changing the delta of the installed software.
--- End quote ---
That's a new one for me. Bounced it off a couple of system admin cohorts of mine. Both felt your qualification that it "can be run again without any need for further installation" is not a valid criteria because they routinely install and run software utilities (not updates) that execute once and then remove themselves from the system. :tellme:
We could continue this endlessly, presenting different examples and exceptions. But why bother? If you don't agree with my definition of what constitutes a software "installation," that's fine too.
Sometimes "agreeing to disagree" is educational in its own right. 8) :Thmbsup:
f0dder:
40hz: FWIW, I've never head anybody use "your" definition either. For me, and people I know, "installing" generally means "putting something on the system (more or less) permanently". Running an executable definitely doesn't count as "installing" in my book.
In previous versions, OpenCandy did get it's DLL installed on the system... but (according to the OC guys) it was just lying dormant there until (possibly) used during the uninstall process.
JavaJones:
It seems like there is a lot of getting hung up on terminology here. Which is ironic because one of the most important points made here - by 40hz - is that the meaning of words *is* important and is being potentially subverted here by OC. On that point I have some concern myself. Nonetheless I think whether something is "installed" or not is ultimately tangential to what is at issue here and of actual concern. After all, I'm sure Wraith would not argue that a virus that simply loads itself into memory and formats your hard drive without ever "installing" anything is ok simply because it's not being "installed" (and neither would anyone else I'd wager :D). So using "installed" as a measure of trustworthiness, safety, or anything else that is really of concern here is not really useful.
So what's really the concern here for those who are uncomfortable with OC? Well, there's certainly the association with adware and spyware of the past. But I happily ran CrashPlan in ad-supported mode for some time, and if I hadn't needed backup sets and wanted to use their online storage, I probably would have continued to do that. I've also used several other ad-supported apps. I don't mind them in principle. Do others here who are objecting to OC's system fundamentally reject all adware? If not, it's an interesting and important distinction.
Spyware, on the other hand, I do broadly reject, at least where I'm aware of it. So what exactly is spyware? The common understanding is it's software that collects information about you and sends it back to a controller, presumably a central server somewhere, for some unknown and unstated purpose. I am *still* unclear on whether OC is doing this, but my understanding from reading this thread is that *yes*, they *are* doing this, in that information is going back to OC. The problem then, at least in my view, is that data is being collected from a position of potentially elevated permissions vs. a web browser (in most cases a browser is not able to see what software you have installed, for example) and *no* disclosure is being made of that. Yes, websites do this all the time, and I'm not happy about that either, yet I continue to browse the web. But for those that have concerns about this sort of thing, there are also common tools available to block this kind of behavior. A good incoming *and* outgoing firewall will catch what OC is doing, as NOD apparently did, and warn the user, allowing them to block it. So tools are available to handle this situation is well. Nonetheless the elevated position that OC is in as far as access to my system in my opinion demands an elevated level of communication regarding its activities.
Let's keep some perspective here though. This does not need to be stated in scary terms in order to avoid being seen in a bad light. I believe OC could require better communication of its activities through its partners and their installers while not necessarily reducing opt-in significantly. Here's an example:
"This installer is powered by OpenCandy! As part of a free service, OpenCandy will check your system for potential software upgrades to improve performance and capabilities. This check will collect basic non-personal information about your system and store it securely on our servers. If you'd like to decline this service, simply uncheck the box below." That's 30 seconds of thought put into the wording; a good marketer could do a lot better, keeping the important information will making it more appealing. That's what marketers do, and that's ok.
They could even make it a bit more controllable and potentially get more customers by doing something like this: "...This check will collect basic non-personal information about your system and store it securely on our servers. If you'd prefer not to have your information sent to our servers, we can still perform a local check and offer some recommendations if you select the "local check" radio button below. You can also choose to decline this service by selecting the "do not check my system" radio button." In a situation like that I might still opt-in to the local check.
Adding a sentence of info about OC's service helping to support developers might also be a good thing, depending.
In the end I think the problem 40hz has, and which I share, is that OC is not very open about what it's doing, and in some/many cases even seems to be completely unmentioned in the equation (e.g. the offer appears to be coming from Microsoft for installing IE9, with no mention of the fact that OC brokered the deal). Disclosure is a big deal to many people.
All this being said the average person doesn't give a crap. :P
- Oshyan
Renegade:
Anyone using Wakoopa? :D
wraith808:
I started referring to the act of loading an executable into working memory as an "install" after hearing Marvin Minsky refer to it that way during a lecture in 1977. Back then, we didn't differentiate between the act of "installing" in your sense (i.e. on a drive or other storage media) and loading it into RAM for execution. It was the act of loading and running that constituted the "install."
-40hz (April 01, 2011, 12:15 PM)
--- End quote ---
If we're going back *that* far, then at that point there wasn't really anything that even resembles installations of today, especially given the change in media and storage technology, so I think that given the change in the English language and technology, to carry over a term from that time to apply to the methodologies of today is a pretty big stretch...
And since we're trotting out our CVs, my view is from the perspective of someone who (like yourself) has programmed and installed software on various OSs for...well, let's just say a bit more than several years, starting with IBM's System/360 on mainframes, and Digital Research's CP/M on personal computers. :)
-40hz (April 01, 2011, 12:15 PM)
--- End quote ---
I'd guess a bit more than several years would depend on what the definition of several years is. ;)
For one thing, installation by any practical definition implies that the software in question can be run again without any need for further installation (leaving out the situation of a software upgrade, since the act of upgrading is changing the delta of the installed software.
--- End quote ---
That's a new one for me. Bounced it off a couple of system admin cohorts of mine. Both felt your qualification that it "can be run again without any need for further installation" is not a valid criteria because they routinely install and run software utilities (not updates) that execute once and then remove themselves from the system. :tellme:
-40hz (April 01, 2011, 12:15 PM)
--- End quote ---
Personally, I wouldn't consider that an installation any more than I would consider copying a file to your computer or executing said file from a usb drive an installation. /me shrugs
But I think that with that out of the way, i.e. you've defined what you refer to as installation as anything that runs on your computer, then that seems to erode your position even more (see arguments by renegade and f0dder) ;)
It seems like there is a lot of getting hung up on terminology here. Which is ironic because one of the most important points made here - by 40hz - is that the meaning of words *is* important and is being potentially subverted here by OC. On that point I have some concern myself. Nonetheless I think whether something is "installed" or not is ultimately tangential to what is at issue here and of actual concern. After all, I'm sure Wraith would not argue that a virus that simply loads itself into memory and formats your hard drive without ever "installing" anything is ok simply because it's not being "installed" (and neither would anyone else I'd wager :D). So using "installed" as a measure of trustworthiness, safety, or anything else that is really of concern here is not really useful.
-JavaJones (April 01, 2011, 02:49 PM)
--- End quote ---
First, installation was brought up by the nay sayers, not me. It was in their own words that the problem with OC was that it installed without their knowledge.
Second it is the words that are my entire point of this conversation. OC is *not* adware, nor spyware, nor malware, but it is being categorized as such. Currently adware, spyware, and malware are all negative terms, and the threat of such labels is enough to force action in one way or another. But, if you dilute it using it in edge cases, or cases that have *nothing* to do with the terms in question, you begin to erode the power of the term (see rape, domestic abuse, and racism for examples of such). And that would be a real shame to see happen.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version