OwnCloud Dev Requests Removal From Ubuntu Repos Over Security Holes
operator_error notes a report that ownCloud developer Lukas Reschke has emailed the Ubuntu Devel mailing list to request that ownCloud (server) be removed from the Ubuntu repositories because it contains "multiple critical security bugs for which no fixes have been backported," through which an attacker could "gain complete control [of] the web server process." From the article: However, packages can't be removed from the Ubuntu repositories for an Ubuntu version that was already released, that's why the package was removed from Ubuntu 14.10 (2 days before its release) but it's still available in the Ubuntu 14.04 and 12.04 repositories (ownCloud 6.0.1 for Ubuntu 14.04 and ownCloud 5.0.4 for Ubuntu 12.04, while the latest ownCloud version is 7.0.2). Furthermore, the ownCloud package is in the universe repository and software in this repository "WILL NOT receive any review or updates from the Ubuntu security team" (you should see this if you take a look at your /etc/apt/sources.list file) so it's up to someone from the Ubuntu community to step up and fix it. "If nobody does that, then it unfortunately stays the way it is", says Marc Deslauriers, Security Tech Lead at Canonical. You can follow the discussion @ Ubuntu Devel mailing list. So, until (if) someone fixes this, if you're using ownCloud from the Ubuntu repositories, you should either remove it or upgrade to the latest ownCloud from its official repository, hosted by the openSUSE Build Service."
This makes no sense. Why can't they be removed? Is no one responsible for administrating the server that the Ubuntu repositories are served from? The article doesn't say.
That seems like a lot of dick-measuring on the part of developers. Why wouldn't Canonical simply update the repository with patches that address known security vulnerabilities? Where is the years of support? When you update your package list, the developers of those packages should be able to post updates...
This is why Linux is not desktop ready... to many stubborn minds pushing their way.
I've been away from slashdot for awhile.. who is this Bennett Haselton guy? I read his wiki page, so I know he runs some anti netfilter websites.
On slashdot, is the the new Jon Katz, Michael Sims, or more like a new Timothy, or something else entirely? Is he an 'editor' or just a frequent summary submitter?
Open source is Free as in nobody is paid to fix the security holes!
Someone had to do it.
He's a frequent contributor.
Frequent summary submitter describes it very good:
http://slashdot.org/story/13/1...
http://tech.slashdot.org/story...
http://yro.slashdot.org/story/...
http://news.slashdot.org/story...
He just makes very long submissions. And since this week, a troll has been very busy, submitting stories:
http://slashdot.org/submission...
http://slashdot.org/submission...
and writing But-what-does-frequent-contributor-Bennett-Haselton-think-about-this posts into stories. The term "Frequent contributor" has been used in a summary by an editor, and is already associated with him by /. users:
http://hardware.slashdot.org/c...
The general issue with Bennett Haselton is simple.
Everyone else in the world submits articles, slashdot summarizes them, links back to the full article, and the comments here ensue.
In some cases the article links are just a link back to the article submitters own blog (and this is gently mocked but usually tolerated), in other cases the links are broken (also mocked), in some cases they are linked to an unrelated article (you bet we mock this too), and very occasionally for those people who enjoy the thrill of the hunt, they do go back to an original article in some legitimate or quasi-legitimate source of news. (Hooray!) (In which case we can mock everyone who didn't read TFA.)
Bennett however, as if you've read any of his articles you will know, is special. He read about the virtues of conciseness, efficiency, brevity and then wrote a short epic about how why they really shouldn't apply to him.
When he looked at what it would take to get his very own blog up and running he quickly realized that it was a pretty serious undertaking. He'd have to register somewhere, choose a password, maybe even pick a theme. Do you know how much that would cut into his actual writing time? Several minutes, at least, and he really just doesn't have that kind of time to spare, what with already being slammed just keeping up with writing down every thought that pops into his brain.
So, long story slightly less long, he decided why not just use slashdot itself as his very own personal blog? It saves him having to sign up for one, and better still he argues, saves us a mouse click by eliminating that superfluous step of having to click through to get to the full article.
After having this explained to him, Bennett rejected the argument and suggested we should be delighted at being able to reach his thoughts without having to make that one extra click to an external source.
So now we just mock Bennett.
I think that sums it up fairly concisely, at least relative to what Bennett would have said. ;)
He also contributes frequently.
Lukas from ownCloud here (the one mentioned in that article). I have to say, that this quickly escalated in a way that I did certainly not intend to. However, I'd like to clarify one thing.
The article states "for which no fixes have been backported". With that I meant to refer to the Ubuntu packages and not Version 5 or 6. We still support ownCloud 5 for security patches and critical bugfixes and ownCloud 6 for bugfixes and security patches. This might have been unclear.
I sent this request to Ubuntu because we're very much concerned about our users. While some of us might know that using the "Universe" repository is not a that great idea for internet facing software, most people don't. Furthermore, I don't believe it's the responsibility of the developer to update packages in every single distribution out there. Especially with distributions such as Ubuntu you have to follow quite complex processes such as SRU which consumes a lot of time.
Additionally, some people in the comments seem to claim that "one developer of ownCloud is noted as maintainer for the Debian package". This entry is a legacy entry and as you can see in the changelog at http://metadata.ftp-master.deb... Thomas did last modify the packages at 11 Oct 2012.
We're always recommending to our users to use one of the supported installation methods such as owncloud.org/install where we even provide our own repositories for most distributions.
(Disclaimer: Opinions expressed in this post are solely my own and do not necessarily also express the views of the ownCloud project or my employer)
If you wonder about the typos in above comment, that is the result of typing in the /. comment box using a tablet device. The /. comment function is another technology scratch that piece of s* that ought to be outlawed.
PHP and most (all?)PHP-based software have been shitpiles of exploitable bugs for years now. The core reason is that the "convenience" features of PHP, including the type-laziness, directly contribute to a high defect rate. Ready to be exploited.
So the Israelis created something to make the Internet more "transparent".
Dont touch anything PHP-based.
So now we just mock Bennett.
I'm not sure if he deserves all the mockery. Maybe some people think that he has posted a couple of silly opinion pieces, but that does not make him a malicious monster.
Welcome to the Western World Jungle. You can essentially buy every insane drug you can think of if you really want to poison yourself. Same with PHP.
Just avoid it personally and tell other people. That's all you can do.
PHP: meh
insecure programming practices: Building SQL-Statements from string concatenation (no format strings for example) and so on
rather slow
NO INCREMENTAL SYNC!
only pro: Server runs on a cheap webspace.
And now go and have a look at seafile.com
The editors that permit this do.
I'm the one who started with the all the posts ending in "He's a frequent contributor" yesterday. Well said. I'd like to add is that Dice dishes out his opinion and he never comes to comments. I feel that if he should get special treatment, he might as well be a commenter. I think the feeling of being shoved his dumb opinions is what creates the equal reaction of repeated mocking. The exchange is a one way transmission of blabber so this is the result.
He's definitely not a malicious monster. However, Slashdot (or DiceDot) basically endorses his opinion pieces - which then get intelligently torn apart in the comments. And there's is no accounting for this from the other end. It's different if a story gets posted and we all know it's a stupid idea. It's another dynamic when DiceDot endorses a retarded opinion and we completely call them out, but then the cycle just repeats. The equal reaction is the a cycle of mocking, because people blatantly point out why his opinion fails and there is no accountability on his part or DiceDot's... just another repetition of the cycle.
Bennett has been posting these long ramblings since a very long time before Dice bought Slashdot. Unfortunately, I think that your complaints are not likely to be heard because Slashdot seems to have had a policy for a long time of not recruiting editors from people who regularly read the site...
I am TheRaven on Soylent News
Theres also the fact that his ideas are often enough repellent, such as when he explains how we dont really need the double jeopardy or self incrimination protections of the 5th amendment, or how Computer Acceptable Use Policies and the corresponding network IDS and filtering systems are literally Hitler.
His ideas are very often absurd, and appear very much as if he recently learned about (or began thinking on) a topic, and immediately crafted an opinion on how everyone else who is an expert in said field is wrong.
Notable entries in this category:
* Why the 5th amendment is totally unnecessary.
* More questions about the 5th amendment, indicating a lack of understanding of its background and purpose (leading one to question in what way Bennett was qualified to raise objections to it).
* Why corporate network filtering and intrusion prevention are tyranny
* Why you should ignore every lawyer's advice of "dont talk to cops".
There are hundreds more, if you do a search for Bennett Haselton. The guy is well intentioned-- he clearly has a passion for getting rid of censorship and fixing the world-- the trouble is that hes proven massively susceptible to the Dunning-Kruger effect.
Was. His account no longer exists. All prior articles no longer have a link to his profile page, and manually typing it in gets a "not found".
No, because renaming it has the same effects on existing systems. The installed package "ownCloud" is no longer there (by that name) so future usage of apt-get can still break.
Of course it can. The repository maintainer can introduce a new package pwnCloud and turn ownCloud into a metapackage that requires pwnCloud. This "transitional package" pattern happens often in Ubuntu updates.
He *posts* frequently. He contributes very little.
If PHP ought to be banned, then what migration path do you propose for (say) Wikipedia, which runs MediaWiki software, which is written in PHP? This migration path proposal might give ownCloud's developers ideas on how to migrate from PHP.
My tablet has it's own spell checker, why doesn't yours?
"The likes of Facebook and WhatsApp are free to those whose privacy is of zero value."
my favorite is writing a 10 page rant on the lg optimus prime when everyone who comments here pretty much knows low end android phones are junk basically on purpose and will never get updates or bug fixes. all the 10 pages could have been summed up with "i bought a cheap android phone, and i can confirm it sucks"
If you are an decently qualidied Adminsitrator, then you always conciously choose between the following:
a) You customize/install/update/recompile/patch the software you need on your own time. Usually you do thos when the service availability is absolutely critical and at the same time no out of the box solution exists
b) You use an "out of the box" solution. This solution should be supported, and used within its nominal use case.
Ubuntu very clearly states that Universe packages may - at best - only receive a minimal quality check at the distibution release and are patched by maintainers, which are not necessarily authors of the software nor employees of ubunut. As such their time which they may spend to predictably react to problems is limited, and, if anything in their life changes they just have to stop doing anything for the package without further warning - if the packge is important enogh for you, donate money to the maintainer and pay him.
I appreciate that the author loudly raises his concers, but i think anybody running an unsupported port of an program is responsible for himself. Pulling the pckage is not good. I for my part run any service for myself (file sharing etc) on a machine which only shows a single port for a vpn to the outside world. If something other than a security problem in the VPN software apprears, i would prefer to contunue using (and reinstalling) the packages which I chose.
If I run SW which faces the internet, then if fix it myself
Call the Ubuntu specific version PwnCloud...
Thank you, I'll be here all week.
LK
"Hi. This is my friend, Jack Shit, and you don't know him." - Lord Kano
Although I've used Linux as my main OS for many years, the idea that bundling applications locked to version that cannot be update is insane and one of the things that I hate about Linux distros. Ubuntu did the same stupid thing with Firefox and Open Office at one point. Being stuck with outdated and potential insecure software, unless you compile your own or used another unofficial repository, is crazy. This is a great example of a system that is designed to fail and a huge security flaw.
I do often compile and install or directly install debs or add other repos. It isn't difficult but can become a hassle when it expects a base Linux environment that is very different. It is about time for some standardisation in the Linux distros. That would also help with a broader adoption of Linux in a desktop role and attract more commercial software to Linux that is currently Windows only. Commercial devs can chose between developing for a small number of Windows versions or a shitload of constantly changing version of Linux. Learn something from the example of Android as a commercially successful version of Linux...
Locking the core OS and software necessary to provide a common base makes some sense but this is taken too far. Either keep software in repositories updated or don't provide them. Ubuntu don't have to be the ones updating but they can have a policy of removing software that isn't keep up to date and banning it from future versions. Shift it back to the original developers to decide what distros to support and install the software directly, rather than through the broken repository approach.
What I'd like to know is why someone would apt-get a web service from Ubuntu who notoriously lags behind the current version in pretty much everything. Especially one that should have its own internal update feature (like Pydio)
Other great tool that is restricted by "security" (for us o for someone else).