That is disturbing. I've had *.google-analytics.com nullrouted ever since I heard of GA.
> SHA-1 of the machine's MAC address.
There's only 2^48 MAC addresses, there are 2^160 SHA-1 hashes. With computing power at where it is today, a bruteforce over the MAC space wouldn't take long.
On the other hand, spoofing the data is also an alternative way of opposition...
> a bruteforce over the MAC space wouldn't take long
Indeed, best supercomputers are running today around 10 petaFLOPS (= 10^16 FLoating-point Operations Per Second) and there are 2^48 ~= 10^15 possible MAC addresses.
A FLOP is quite a lot simpler than a SHA-1 hash, however.
The Bitcoin network might be a better comparison since it uses doubled SHA-256; currently it's at around 3e16 hashes per second, or 6e16 SHA-256 operations per second. That's enough to hash all possible MACs in less than a second.
It doesn't matter whether it is free or not. My only aim was to provide information that might be relevant for some people interested in trying out Atom (admittedly I mixed this information with my personal position towards Google and the ever-present tracking aka surveillance) so that they know the implicit price of its usage.
You are free to use Atom or anything else, but you should always have as much information available as possible to correctly judge what you pay for any given thing.
Except you aren't providing information in a vacuum - you are also providing a negative personal opinion. I would rather see responses which purport to educate include a more balanced perspective such as the potential benefits such analytics might confer to end-users in the long run.
I fully respect peoples right to privacy. I just wish those who frequent Hacker news were more accepting of the fact that privacy is a trade off; one which often comes at the expense of the benefits of sophisticated analytics.
Are you serious? You expect me to provide a Fox News fair and balanced response in any of my comments on Hackernews? I could understand your point if I would have tried to be polemic, but I surely wasn't.
The other position was already provided for by github and numerous other posters, while I pointed out a fact that was not yet mentioned.
By the way, "benefits of sophisticated analysis" for whom? Google's ad revenues? I actually agree that github could provide a better product with the information gathered, but I really don't think that google should be in between me and github.
isn't that the purpose of the free software community? you find a malicious feature then you modify the software source code and release a clean version for interested people?
Sorry to tell you, it is not free.
https://atom.io/faq
"We haven't settled on pricing yet, but you can expect it to be competitively priced compared to similar editors."
Executing analytics code consumes CPU and network resources that I pay for so I do think people should have a say in what a third party does with them.
I believe that the final product isn't going to be free - so that changes the landscape of the argument. Then again, maybe they won't have analytics in that version.
Note: I'm not making a comment on it being a paid for app - they've released bucket loads of open source code they wrote to create the editor. For that they should be praised.
Works, but the gist is now gone. However, it looks so similar to sublime, I can't believe it. In case anyone is looking for it: the command is
curl -L https://www.atom.io/api/updates/download -A "Atom/0.1 CFNetwork/1.5" > Atom.zip
So there's only one comment that gives the impression that there's some kind of troll going on. Just to let you know: It actually works, nothing fishy just someone posting the direct link to the app that's exposed via their update Api.
There's no activation or anything, it will just take you straight into the app.
Their clear wish is for this to be invite only. I agree that their 'security' should probably better if they want to protec themselves against certain types of people, but your argument is akin to justifying robbing a house of its contents because it was only locked up with no alarm and you had to break one window to get in.
No, my argument is not akin to that because piracy is not akin to theft. Downloading the beta from a public URL is not depriving Github of anything or breaking anything of theirs, whereas if I broke a window to steal something I've left the owner with a broken window and deprived them of the contents.
Yet on HN there have been numerous 'security breaches' defended solely due to the fact that the data was accessible via a public URL and if they wanted it secured they should have done better.
So, now that the zip file is missing from GitHub, some people are starting to upload it elsewhere. Can the people who got it from GitHub earlier post the result of running
md5 Atom.zip
? That might help mollify the paranoid conscience.
(EDIT: or any other hash function for that matter.)
I think the atom.zip is being updated continuously. Today I finally got my invitation and it gives a different shasum too. File diff shows the Info.plist of my newly downloaded version as of 2014-03-07 is:
No. Of course not. Seems like the easiest decision they made all week.
Imagine it was your software. And that somebody had pirated your beta and actually posted instructions on how to do so on the forums of your own website. What would you do?
Naturally, you'd delete the comment the second you saw it, then start working on patching the hole.
Their terms of service aren't exactly a grand stand for freedom of expression:
We may, but have no obligation to, remove Content and Accounts containing Content that we determine in our sole discretion are unlawful, offensive, threatening, libelous, defamatory, pornographic, obscene or otherwise objectionable or violates any party's intellectual property or these Terms of Service.
While I understand that Github is a private body and that their site is, well, their site ... forcibly deleting someone's gist simply because they don't like that that someone has found a way around their invite process smacks of underhandedness.
Will they suspend his account permanently as a result of this leak? Will he have his account limited in some other way, preventing him from having private repos, for example?
Will they lash out against anyone sharing Atom.zip that they can trace back to a github account?
It's just all very "our show, our rules" of them to delete the gist.
And yes, I understand that it is their show. And yes, their ruleset too.
Doesn't make me feel any comfier in the knowledge that they'll simply delete anything that they don't approve of.
Imagine someone posting the url to access your commercial product because your dev didn't secure the download. And you have control over the post. What would you do then? And Github can delete anything they want. If anyone has problem with that, they sure can sue Github.
Imo its perfectly reasonable to not want your new product leak out before official release for numerous reasons, like it not being finished/ready for example.
This beta invite program is designed to let us carefully control how many new people get access to Atom each day so we can provide timely bug fixes and support responses.
isn't that the same message if the user took it down? I.e. how do you know GitHub took it down themselves? btw, I also feel it's pretty wrong to exploit a workaround like this to obtain the software against the authors' wishes.
I think the editor uses the same underlying technology that Chrome uses to display web pages. That's a lot of overhead for a text editor, but it makes it really easy to extend and do crazy things, basically anything that is possible in a browser. I honestly don't know what would make it faster.
maybe not in the sense that the GitHub owners are constantly refreshing the GitHub home page, but they could certainly have notifications set up for posts that mention GitHub. plus, anything as relevant as this is, is going to reach them pretty quickly by many different routes.
It is rather disappointing that github heads down the Mac OS path while they base their company around a concept which is rather strange to most things Apple - openness and sharing. Writing for Mac OS first also makes it harder to port to other platforms as it is a rather idiosyncratic platform.
If they insist on pushing Mac OS AND want to be seen as a good free software citizen they might want to consider putting some effort in the GNUstep project. Once GNUstep is up to par they should be able to build for that on Linux and Windows with minimal porting efforts.
Still, I'm disappointed in the lack of vision many of these so-called 'open source' companies exhibit. Don't they realise that by pushing these closed, proprietary platforms they undermine the very base of their own existence? The software and community which github depends on was not the result of the generous sponsorship from Microsoft and Apple after all...
I'm making an assumption that the majority (if not entirety) of GitHub employees use a Mac, so why wouldn't they make an editor for themselves on a platform that they use? I doubt they're doing it to bolster the OSX ecosystem as much as meet their own needs.
Because the product they provide is tailored to open source/free software. No matter how much you might be enamoured by Apple and its polished products it is hard to get around the fact that these are the opposite of open and free.
Remember what happened when Linus decided he wanted use BitKeeper to maintain Linux? Also remember how that worked out in the end? Free and/or open source software and closed, proprietary companies often make odd bedfellows.
Wtf. Why would one start with the software-patent and lawsuit-based platform?
Does nobody at github have any sense of ethics? Github was built on open source software and Linux. Why not give something back? Why support those who spend national budgets on anti open-source lawsuits?
Err... because you want to try out an experimental text editor that's rumoured to be quite nice and you run Windows? seriously, what on earth is the point of your post? Are you trying to make a snarky comment about something? if so, at least flesh it out with some actual humour or context.
> 4. To help us improve the editor, Atom sends usage information to Google Analytics. See [atom/metrics](https://github.com/atom/metrics) for details.
That is quite horrible in my opinion. Do I want an application on my computer that constantly sends information to Google and Github?
What kind of information is it sending? (from the link)
Collected Data
+ A unique identifier that is generated by computing the SHA-1 of the machine's MAC address.
+ The screen width and height
+ The version of Atom being used
+ The name of each item opened in a pane such as EditorView, SettingsView, and MarkdownPreviewView
+ The amount of time the current window was open for
+ The amount of time the current window took to load
+ The amount of time the app took to launch
Item no. 4 is the critical one. I really don't like this. However, you can apparently disable this:
> If you do not want this information reported, disable this package from the Metrics section of the Settings view `(cmd-,)`.
(edit: removed typo)