not all change is progress
September 15, 2014
Direct download links: MP3 & Ogg
Slackel – ownCloud0:09:17 News
Good News / Bad News
FSF and Debian join forces to help free software users find the hardware they need
Akamai Warns of IptabLes and IptabLex Infection on Linux, DDoS attacks
New Distro Releases; One Ceases
Gentoo Live DVD
“Q4OS is now stable enough to be recommended for everyday use”
Stepping Down from Bodhi Linux Lead (Update: Bodhi Linux is NOT Dead – It is just Changing Hands)
New App Releases
Hands-on: WordPress 4.0 adds useful features to a rich platform
Amazon Brings Prime Instant Video To All Android Phones In US, UK And Germany
You WILL be Running Red Hat GnomeOS
Revisiting How We Put Together Linux Systems (comment threads on LWN and Lennart’s G+ post worth a read)
Sandboxed Gnome apps
Odds ‘n’ Sods
Standard Flavored Markdown (bunfight, and the quick-read version)
Chrome OS can now run Android apps, no porting required
Freexian’s first report about Debian Long Term Support (and Holger Levsen’s thoughts on how the process is working)
DebConf14: QA with Linus Torvalds
With Named Data Networking, a group of researchers promise a future without servers and IP addresses (and what it‘s all about)
Anyone interested in the origins of TCP/IP, and why that alone might not be appropriate to take us forwards, would probably enjoy this presentation by Van Jacobson
0:51:20 First Impressions
Jesse looked at Slackel, whilst Paddy was handed ZevenOS for next time.
A huge thank you to Campbell Barton for his continued support over at Gratipay, and to johanv, DeepGeek and an anonymous donor for their Flattrs. And a special thank you to Charlie Ebert and Clive van Hilten, both of whom became Monthly Supporters and joined Jeroen van Rijn, Peter Kidd, Christopher Atkins, Issac Carter and Brendan MacWade in providing us with a small predictable income stream. Thanks guys – it really is appreciated!
As we mentioned, Jesse has now set up a Luddite G+ Community, so if you use that service we now have another way to get in touch.
We received a large post bag this time, so an up-front thank you to GLaDER, Jezra and Charlie Ebert for your various comments.
Frank Bell, Jonathan Groll and Shay the Daft Punk all got in touch with comments following our look at the *boxen window managers last show.
On the question of usability, stability and backwards compatibility, Rob Landley provided some interesting historical information, and made a point echoed by Torvalds in the video linked above – it’s not the kernel to user-space interface that tends to break things, but items further up the stack. And AdamT questioned whether Linux Mint really would be a good choice for non-technical users looking for a stable platform with up to date apps.
Paddy’s plug for a non-vi/non-Emacs based simple text editor provoked some comments. Ivor O’Connor wrote approvingly that the more traditional editors can act as full-blown development IDEs, Jonathan Groll flagged up the ubiquity of GNU’s Readline, and Russell Dickinson plugged ne, another lightweight editor that unfortunately doesn’t quite tick the boxes Paddy was praising the Sanos editor for.
On systemd, Florian flagged up the GSoC project aimed at producing shims to keep the GNOME desktop working on the BSDs, and Brad Alexander confessed to a growing affection for PC-BSD and the Lumina desktop. Whilst talking about this, Paddy mentioned the recent BSD Now episode on Lumina.
Nathan D. Smith wondered why the bundling of libraries with apps is considered so bad since “disk is cheap”. Rob Landley and Russell Dickinson provided counterpoints to Nathan.
Jonathan Nadeau emailed Paddy to clarify that his motivation for planning to write a new text-to-speech engine for Sonar GNU/Linux was largely just about reducing uncertainty.
Jonathan Groll pointed Joe towards the Zalman HDD enclosures as a possible solution for his ISO woes, and Rufus Rieder linked to a post detailing appropriate chipsets for use when creating your own Hackintosh (also, no doubt, intended for Joe).
Henrik put forward a defence for the use of Wine versus dual booting, and Joel contrasted his picture stitching efforts using Hugin with the results he achieved from Microsoft’s Image Composite Editor. Jesse took a leaf out of Apple’s playbook, telling Joel that “you’re holding it wrong” ;)
Finally, thanks to Morten Juhl-Johansen Zölde-Fejér for pointing us towards mapillary.com, a crowdsourced alternative to Google Street View, and which none of your hosts had come across before.
Many of us are looking for a way to break free of proprietary services, but to still enjoy the benefits that cloud storage and computing seem to offer. Much hyped by credulous bloggers and other podcasts, we thought it time to cast an honest gaze over ownCloud, a dual-headed corporate and community project. Our verdict? Decidedly mixed.
Rolling out of that discussion, and thinking about the realities of the multi-device owning world we now live in, Paddy brought up what he thinks the FOSS community really needs to deliver on in order to free us from the centralised control of our data for once and all.
in your ownCloud discussion you mentioned it using open document formats by default and raised the issue of editing these on mobile devices. Actually, there is an application to do this, called OpenDocument Reader (http://blog.tomtasche.at/p/openoffice-document-reader_9.html). Contrary to the name, it now supports editing as well as viewing documents.
Maybe you can give it a try…
Hi Daniel – yes, I mentioned the lack of ODT support on Android (other than AndrOpen Office, which I also mischaracterised as a LibreOffice rather than an OpenOffice rebuild). Thanks for the link. I’ll give OpenDocument Reader a spin before the next show, and will report back. I’d previously overlooked it because – as you suggest – of the name.
Yet another fantastic show, so thank you.
Paddy – Thank you for your reply regarding ‘ne’ as opposed to Sanos. I am *still* yet to try Sanos. :O I love to hear about and try alternatives to software I am already using, which leads me down almost endless rabbit holes.
I have downloaded PC-BSD 10 (IIRC) but have not yet worked up the courage to try it. It would be installed on a laptop which is already quad-booting between Windows 8 and several Linux distributions.
I ought to credit your show with having me try Debian for the first time in my more than 10 years with Linux. I am not using it full-time but was glad I at least tried it out.
I was curious about Paddy’s mini coreutils rant. My understanding was that GNU Coreutils was the heart of “gnu’s not unix” – just a one for one port of all the classic UNIX utilities. When you say it is bloated, do you mean additional programs have been added, or that the programs themselves are getting bloated?
Hi Nathan – rant? That wasn’t a rant, just a gentle grumble :P
To address your actual question, whilst the GNU Coreutils have enthusiastically embraced a tendency towards feature-creep, they were hardly the first to do so. It might be pretty old now, but I’m still happy to point people towards Kernighan and Pike’s 1984 article [27KB PDF] as a starting place when thinking about this.
It’s also ironic that whilst platforms have increasingly tended towards the tightly-coupled/bloated/monolithic, there has continued to be strong support for loose-coupling and componentisation within application development. And whilst Pike himself declared the old *nix philosophy outdated a few years back, Go, the development of which he is closely associated with, is itself now providing many ‘net-focused companies the agility that is a side-effect of this sort of development architecture. Funny old world.
Paddy – I have finally compiled the Sanos text editor and have to say it looks great. I typically use a console editor for small editing tasks, mainly on configuration files. The Sanos editor does not have as many features as ‘ne’ (“the nice editor”) but it is definitely smaller and stil provides key sequences that I am familiar with. I had to look at the source code to discover that [Ctrl][g] listed the help available as using [F1] launched GNOME’s Help application.
I hope you didn’t mind my question via Twitter: “Why do you use Chromium?” As I tried to indicate, I was genuinely curious, not looking to start a flame war. Your reply was very reasonable and mentioned some good points. I too remember the days of Firefox’s incredibly memory requirements, which drove many people to Chrome. I am a FLOSS fan and so have tried Chromium. My main problem is that as there are not official releases of Chromium, the version available and potential bugs, can be different across Linux distributions and indeed platforms. The person who packages it for Fedora does so reluctantly because of the Chromium’s dependencies and other factors. Is there any way of identifying what build of Chromium is considered stable? I often wonder how packagers choose a specific point in time to compile it. I have successfully run nightly builds of Chromium but unfortunately encountered an issue with a web-based system I use on a daily basis. For now I continue to run Firefox, which has improved markedly in its memory requirements. I find Firefox to be amazingly stable, even having run nightly and beta builds for months at a time.
Hi Russell – you’re right; the Sanos editor is pretty light on features. But, as I said on the show, it’s cleanly coded, so is pretty straightforward to extend as you wish (if you have a basic grasp of C).
I only tend to look at Twitter very periodically, I’m afraid, and removed my personal account a good few months back; here, or a mail to show@ or paddy@ – is by far the best way to get hold of me. Having said that, I will be dipping my toes back into the social media scene by joining the G+ community Jesse has set up for us.
And to address your actual point – I’ll take another look at Firefox, and also some other browsers. Chrome really is getting to the point where it’s damn near an OS platform on its own, and Firefox does seem to be off down the same road. Come to think about it, a round-up of some of the lesser used browsers – and those that are, still, just browsers – might make for an interesting show segment…
Paddy – Thanks for your reply. I was just listening again to episode 15 where you mentioned that you had “escaped the evil empire”, if my memory is correct. I first listened to your latest shows, then went through the back catalogue and as a result got shows out of chronological order.
On the topic of Twitter, I don’t know why but I have never quite “got it”. I am much more comfortable in G+ though there are some features I would love to see added there.
I think I’ll now use Sanos editor in preference to ne, given it’s so small and light. Thank you again for mentioning it on the show. This afternoon, while trying Sanos editor, I looked again at ‘mc’, the Midnight Commander, an ncurses dual-pane file manager. I recalled that it included an editor, but in fact found it has ‘mcedit’ and ‘mcview’. I quickly discounted them as mc is a massive 1Mb, compared with the tiny Sanos editor.
Regarding Firefox versus Chromium, I am really torn between the two. I love features of both, and that leads me to switch between them too often. I have also been testing Opera Developer, the return of Opera to Linux. It seems very much like Chrome but there are definitely some differences in its UI. I especially like the tab live preview which appears when you hover over a tab. I’m having to settle with Firefox because I simply can’t find current, stable releases of Chromium for my main Linux distribution – Fedora. I would love to hear a review of other web browsers, though I believe the field is quite small.
As always, thanks to yourself, Jesse and Joe provide. Sometimes I find myself at almost polar opposites to your views, but it’s great because it has me thinking about entirely different views. Please remain true to yourselves and I’m sure the show will have a great future.
Oops… that sentence should have read “As always, thanks to yourself, Jesse and Joe for the entertainment you provide.” I paused while writing that to check the spelling of Jesse’s name and forgot to finish it. :O
Paddy – When discussing the Sanos editor you mentioned you had made some changes or additions to it. Could you please share those? I would be interested to see what they are.
Considering the deserved stick I’ve given devs on previous shows about code quality, I’m not about to let my own hackishness into the light of day. Maybe in the future, if I ever get time to tidy it up.
I like the idea you posed about SystemD, that being if people really do hate it so much, they’ll go their own way and develop one or more Linux distributions that don’t use it.
Given all the rancor over the init system that’s more than an init system, I fully expect this to happen. And the controversy could give a boost to the BSDs. Joe — you always say Fedora means pain, but have you ever tried running OpenBSD on the desktop?
Steven I hope that systemd cold have a proper alternative feature wise (Solaris for instance have a nice service management and sandboxing but the new owners at the moment are making sure that all nails on that coffin are properly secured :( ) but the problem is that there is much hate involved and to little of actual coding.
Now that I’ve heard the whole show, including the part about you guys giving BSD a go, I hope you do it. It’ll be entertaining for us, if painful for you.
I spent 6 months at one point running a laptop with OpenBSD — The optical drive was dead, it didn’t boot off USB, and OpenBSD still had (and probably still does have) a floppy installer. Yep, a 3.5-inch floppy image.
I was able to get everything working well enough, first with the stock FVWM window manager and later Xfce, but doing anything complicated in OpenBSD, or any other BSD for that matter, can really eat into your time.
Just getting Java to run was a weeklong odyssey. And the software-update mechanisms and philosophies really make you appreciate Linux. Unless you run from ports and love compiling, you often get NO updates between releases. So all the security gains you theoretically get from OpenBSD are somewhat muted by the fact that you’r e running the same versions of most or all apps for a full six months. And then you have to go through a painful update process, or the infamous “nuke and pave,” as you put it.
I had worse luck in FreeBSD. It’s not clear which combination of update commands you need to do to update the base system, the binary packages and the ports. More than once I started an update that began building ports for the entire system for DAYS and DAYS, and that doesn’t take into account build failures. I ended up with a non-bootable system more than once.
If you’re obsessed with the idea of OpenBSD, FreeBSD, NetBSD or DragonFlyBSD, you can probably make it work, but it’s a lot of pain. Joe, it makes Fedora look like cake.
I love the idea of BSD, and I have a lot of (perhaps misplaced) affection especially for OpenBSD, but running it on the desktop is just too much work.
I’ve always said that BSD needs what Linux has in the dozens. Namely distros. They are few and far between, but I think that PC-BSD is pretty much the only hope that BSD has on the desktop for even semi-regular people. So I’d focus on that and maybe try to get Kris Moore — PC-BSD Founder and rival (Jupiter Broadcasting!!) podcaster on the show to give you the PC-BSD overview.
And you should all install and run PC-BSD for awhile — forget those obscure Linuxes for a few weeks.
Thanks Steven for sharing your BSD experiences. You recommended giving PC-BSD a try, but I couldn’t tell for sure if PC-BSD was included in your BSD experiences. If so, could I ask which release?
I’ve been curious about using PC-BSD on the desktop for myself, but haven’t had the time to give it a proper evaluation. I like your suggestion that one of the hosts give it a more lengthy trial (assuming they can afford the time). Given the radical difference in eco-systems between the BSDs & Linuxes, it could really take some time to get a feel for the BSD experience.
I’m looking forward to hear about the host’s experiences, but willing to wait until they’ve had sufficient time.
I haven’t done so for quite awhile, but I have installed PC-BSD in the past. This was a few years ago when they offered KDE and Fluxbox, I think, and the whole thing was very KDE-based. I’m not at all KDE-based, so I didn’t stick with it.
Now that PC-BSD is desktop-agnostic, and you can install just about anything, it seems to be a better choice. The one difference is that years ago they had a live image. Now I don’t think they offer a live DVD, and not being able to see how my hardware reacts to the system until I do a full install — it’s just too backward for me.
Excellent point Steve about the lack of a live DVD image. Thanks.
I would like just to raise a point of how in “Revisited
Linux System” some parts could be immensely useful both
to users and developers/packagers/maintainers.
Particularly to the less technically skilled users (and
if we want to “give Linus that desktop” OS parts needs to
be as painless as possible to them) as system could be
same old distro and have same old packaging system in
place as it always was just delivered to the user in a
slightly bigger chunks.
So when user is updating he ether have all of the new packages or not. Same goes for apps, a bigger chunk of all dependency packages, ether fully applied to the system or not. System could not be in a broken state ever. Nor in the situation where user have to figure out a way out of the dependency hell.
“Revisited Linux System” does not have to be btrfs based. ZFS has same functionality required and Docker at the moment achieves the similar with UFS (writing this from the top of my head as I am familiar and worked with those, but I bet there are other ones who can or could have be up to the task).
“Revisited Linux System” could be compatible with but actually share zero similarity with RHEL. Different distros could coexists on a same box and be totally oblivious of each other.
Imagine… User pops up on the IRC and have some issue… I ask him for snapshot label details… User say x.y and a.z. I type few commands and have the EXACT same situation as user. I hope that one day I could have such a tool. :)
Hi nadrimajstor – back in a previous show, we took a (somewhat superficial) look at the Nix package management system (and NixOS, the demonstration distro built on top of it). Nix appears to offer much of the same functionality that Poettering is talking about, and even takes a declarative approach towards dependency specification – something also adopted by systemd. Which begs the question: why is Lennart so hell bent on reinventing the wheel?
A charitable reading of the situation would be that he (and Red Hat/GnomeOS) are simply again suffering from NIH syndrome. However, I believe a more plausible answer goes back to the point that I attempted to make on the last show; that this is as much – or more – about the politics of power, rather than a technical solution to a technical requirement.
While making my first debian package for Ubuntu, and battling/learning all the little intricacies of making a .deb I did take almost thorough look at Nix to see how can one make a package and stay sane. :)
Point that I could buy in “Revisited Linux System” is that I could use Nix to package e.g. MATE and deliver a snapshot of that whole stack to all other distros (including those that are totally oblivious to the existence of Nix).
Speaking of power, “Revisited Linux System” will be pushed forward and only question is will one participate or stay aside.
Paddy’s referrence to the ‘GnomeOS’ term in the show got me thinking about the social/political aspects of the campaign by the key systemd proponents’ to promote acceptance of their ‘vision’ across the many different Linux communities. (my take on their ‘vision’ = move Linux Distros away from POSIX/Classic Unix interfaces and consolidate them around them around systemd/Gnome interfaces).
I ran across this 2013 audio interview (30 min) with Lennart where he was advocating for the ‘GnomeOS’ label: http://faif.us/cast/2014/sep/11/0x4D/
I also ran across this rather lengthy but IMHO informative video (1hr 45m) of a presentation Lennart did in 2012 where he using the ‘Core OS’ label: http://www.youtube.com/watch?v=9UnEV9SPuw8
Interesting to note that neither of those labels has gotten much traction although that doesn’t seem to have hindered the systemd project’s propogation into Distros. Ironically, the ‘Core OS’ label has even been usurped by an entirely different project.
Despite my reservations about some of the technical aspects of systemd, observing the social/political aspects of the project’s campaign for acceptance has been highly enlightening into the dynamics of the Distro’s organizational eco-systems for me. As fragmented and diverse as the various communities are, I must admit surprise that anyone outside of Linus himself could push such impactful changes as widely throughout the Linux eco-systems.
Hi SoN – thanks for Polish video link; that was a new one for me. Shame about the audio quality and tape changing, but it was an interesting watch. As you say, it was informative to see Lennart explain how certain features (or ‘gimmicks’, as he inadvertently slipped out at one stage) were added to win over specific stakeholders. I thought it equally telling that he was prepared to proselytize in front of such a small gathering.
The thing about Lennart’s talk is that so much of it was reasonable – there really is a genuine need for a decent resource management framework in the server space, for example. In fact, if I had no dog in the hunt, I’d be agreeing with virtually everything he expounded. I have no issue with somebody taking the Linux kernel and building something rather unlike our traditional systems on top of it – as Google has done for a couple of different form factors so far. Where my hackles start to raise is that rather than making this a clearly stated goal (although, to be fair, they haven’t exactly hidden their intent, it’s just that few have been paying adequate attention) of a /new/ project, the folks behind all of the ongoing shenanigans have clearly decided that the optimal way to leverage wider resources is to slowly tip-toe towards the destination, dragging everyone along with them, and burning bridges as they go.
But it’s certainly interesting to watch play out; both from a technical perspective and, as you suggest, seeing how the various communities are nudged in the desired direction.
I agree completely Paddy, Lennart makes some good arguments for certain capabilities of systemd, capabilities that may well demand breaking away from POSIX and classic ‘Unix Way’ traditions. Exploiting the potential of the kernel’s cgroups and container capabilities are among the most compeling to me.
However, Lennart’s argument that systemd’s modular structure is proof that it won’t lead to monolithic interdepencies I find unconvincing. His claims that use of most of the systemd modules is optional seems dubvious to me as the modules appear to be becoming more coupled by interdependcies over time. Alternative implementations of systemd modules seems extrodinarily difficult due to those same interdepencies, and the evolving systemd APIs. Throw GNOME dependencies on top, and I can foresee the potential for the coming of the generic and monolithic Linux Distro(s) that you have often alluded to in your podcast.
I’ve seen more then one Lennart presentation where he is explicit about his belief that the fragmentation (variety) of Linux Distros is a major impediment to more wide spread Linux adoption. I get the impression that he doesn’t see a dominant single flavor Linux as a problem. How much that could stifle the innovation and evolution that the ‘Unix Way’ has enabled is a trade-off question we appear headed to see answered.
Clearly, Lennart was not dependent on my approval to be successful at pushing systemd into the major Distros :-) Looks like I’ll just be along for the ride. Just hoping that my reservations are over-blown.
Hello and thanks for your extensive discussion about ownCloud and syncing :-)
Re: FLOSS alternative to BitTorrent Sync
Have you heard of / tried Syncthing? Me and a friend had used BTSync before for a podcasting project (use case was ergo syncing both many small files like Audacity raw data and very large ones like finished tracks) across Mac, Linux and Windows. It worked OK for several months, but recently we ran into problems: one of the machines started pushing old file versions and deleted file to the others, so we switched to Syncthing.
1st impressions: The setup is a bit more complicated (needs sharing of device _and_ folder IDs; BTSync needs only the folder secret), it has fewer folder config options and the finding of peers seems to be a bit slower (min’s rather than sec’s, due to slower indexing maybe?!?). However, the speed of actual file transfer makes up for all of that. For us, it was at the physical or protocol limit of the network hardware en route :-) Therefore: Not sure if production-ready, but a definite recommendation to keep an eye on and to test it.
PS: Regarding P2P sync vs. always-on devices. How about this angle: the former can help decelerate our lives! I mean, which syncs do we really need to happen quickly?
I have been using Syncthing for a while and it works well for me. One improvement I would like is to stop a master being able to delete files from a “slave”. Very useful if you accidentally delete a whole raft of things you didn’t mean to. It sort of does it at the moment, as you can enable versioning, so if you delete something by accident, you can restore the previous version.
Maybe setting this repo to “Master” on both devices achieves this? I think your’s is an idea worth posting to their issue tracker ;-)
Thanks for the show Guys. Just a small comment on Wine. Another interest of mine is Amateur Radio. Nearly all radio apps are written for Windows and my main machine does not have Windows, only Linux. I’ve no objection to dual-booting or Windows, but Wine gives me the option of accessing apps for my other major interest. Dosbox is another good app for those obscure, old, non-updated apps. I have NO interest in games, but there are other areas for which we need something like Wine.
About a month ago, a new Knoppix came out.
And absolutely nobody’s covering it.
Good show Joe, Jesse, Paddy.
I don’t think I’ve heard of any show which did an adequate test on OwnCloud other than your own.
Also the news and discussion sections, OS testing sections, are just the best in the podcast business.
Keep up the good work.
Re: OwnCloud. You hit the nail on the head. It’s a LOT of PHP. Just like WordPress, but a whole lot more code a whole lot more quickly.
If you think about it, PHP is still compelling for these projects because every shared-hosting service and VPN can run PHP with MySQL. You don’t need to figure out Rails, Node, Django or any other “modern” technology to get a PHP app running on a server.
Sure it’s “easier” to run Node or Rails on a PaaS system like OpenShift or Amazon Elastic Beanstalk, but that’s DevOps territory, and the average person would rather just shove a bunch of PHP files onto a server via FTP and watch it go.
You are all right about OwnCloud — they need to get the basics right.
I ran it a few releases ago, and aside from WebDAV being prone to losing my file changes, the sync client wasn’t terribly reliable. And just like you, I’m not so confident in my own sysadmin-security skills that I wouldn’t rather outsource the job to Dropbox and/or Google.
You didn’t mention alternative collaboration tools like EtherPad. That, at least, is free and open. And it works pretty darn well.
Hi Steven – we didn’t talk about Etherpad (or Ethersheet, SparkleShare, Seafile etc.), and only just touched on Syncthing, as we intend to revisit this entire territory in a future show. There’s clearly a demand for these sort of services/products out there, and we intend to give them a proper going over before reporting back.
I am really in two minds about ownCloud.
As for having a sensible sync solution, I have recently installed Baikal: http://baikal-server.com/ – it offers CalDAV and CardDAV services, is quite lightweight and works on my cheap hosting PHP server.
I pull it into my Android setup.
Morten (add other names here) – Thanks for mentioning Baïkal. I’m now wondering if I could get it running on OpenShift. Spoken as a true Red Hat employee, eh? :D
I am also in two minds about ownCloud. It seems to me that too much has been crammed in within a short space of time. For documentation collaboration I would prefer to have seen Etherpad Lite used instead of a GUI collaborative editor. As for file synchronisation, I am currently using the proprietary Copy, but would like to move to Seafile instead. Last time I checked, a Seafile client was not available for Fedora, which is my main Linux distribution.
I like segmenting my tools into my various
subdomains, so I am good with
one-service-one-function. I have a Wallabag page,
a Shaarli page, a Kanboard page, a Zenphoto page,
a Kimai time tracking page. A bit of
What makes ownCloud interesting for me is being able to put it on whatever hosting I happen to have. Etherpad requires system access. Did you come across any tools like it for a shared hosting system?
I really enjoyed your discussion of OwnCloud, and thought I might suggest a possible alternative: git-annex (https://git-annex.branchable.com/). I had looked into using it for work quite some time ago, but couldn’t deploy it because of its lack of Windows support. I unfortunately have not had time to tinker with it on my home network, but would be interested to hear how it compares to OwnCloud (which I ended up deploying at work instead >.>).
I’m not a big eBay user, but I saw this link and thought about you:
Comments are now closed.
The content of this website, and that of the podcasts produced by the website owners, is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License.