This seems mighty kneejerk-y. A duplicaton issue causes your contacts to disappear, therefore you should move your contacts and calendar away from a service that probably has better engineers, better backup, better redundancy, and better uptime than any one of us could possibly muster, and onto a somewhat immature product with worse functionality and integration which is by no means immune to the same kind of problem?
Yeah.. see, I don't think the author has thought their cunning plan all the way through.
I understand the anti-google sentiment is at an all time high after the Reader shutdown fiasco, but let's stay rational, here!
What about Google shutting down the open standard for syncing contact and calendar data across platforms?
I don't ask this as a form of "anti-Google sentiment," but as a legitimate question. The announcement that they would be phasing out ActiveSync support at the end of last year coupled with the announcement just made that they were doing the same for CalDAV/CardDav (when they announced the end of Reader) indicate they're "silo-ing" the service somewhat. Unless I've missed something, it's still not clear that iOS/iCal calendars are going to sync with Google's unless they "whitelist" Apple.
It's not a "cunning plan," it's self-preservation and keeping your stuff working or at the least not being at the mercy of a commercial concern who demonstrates a willingness to break it even for paying customers.
Calendaring is a big issue in enterprises, especially when you have Apple, Google and Microsoft devices with all separate calendar apps.
Many organizations, very large ones in fact, are interested in moving away from exchange as their mail platform and going with Gmail. However, the gripe is that calendaring with Gmail is simply utterly broken. iOS devices dont get updates, creating resources is a complete joke (if anyone at google reads this; whomever designed the method for implementing resources in gmail/ical should be fired, and stripped of their ability to code ANYTHING).
At the same time - companies are looking for greater integration with their calendars and resources and devices etc...
What would be ideal - is a standard calendar blob which any view can access via an agreed protocol (maybe caldav is supposed to be this - but from the way resources are created/accessible/scheduled, its broken).
The next layer that everyone I have been working with over the last two years is looking for is "my location as it relates to the resources I am attempting to use"
Things like wayfinding, occupancy, meeting-start-based-on-presence, issue reporting etc... are just some of the features people are looking to add to the calendar systems.
None of this will be possible if Google is willy-nilly changing what they do, support, implement etc..
The sad thing is that for core business services Google seems to be running away terrified - and I suspect this is due to their inability to understand customer support.
>It's not a "cunning plan," it's self-preservation
Fair enough, and I can see the value in keeping a local backup just for redundancy and convenience purposes (being locked out of my Google account for whatever reason would suck), but using it as your daily driver? Ehhhhhhh....
As far as shutdown fears, though? Calendar, Contacts, and Gmail are the core-est of the core products. They're tied directly into the Android ecosystem as well. Reader was more or less a niche product.
Google isn't going to shut down any of the holy three applications which make up the entire backbone of their mobile OS's PIM functionality. Any such fears are completely irrational.
I didn't say anything about them shutting down the services, I said they were "silo-ing" them and clearly moving away from having them interoperate with 3rd party software.
It is not at all irrational to worry for how long your enterprise users' contacts and calendars will continue to sync.
They've thrown us a lifeline for the time being and even appear to be working on a fix for Outlook 2013 so it will sync (it doesn't as of right now), but the whole process of sunsetting support for ActiveSync then CalDAV completely tips their hand that they want to control who and what interoperates with it - and for how much longer it will interoperate with platforms beside their own
Android core PIM suite =| the entire mobile and desktop universe. There are people paying for Google's enterprise apps offering that don't use Android. What is in question is for how long core services of those apps will work outside of a browser/Android device.
I barely use Calendar or Contacts. I only use Gmail for the mail -- something everything else can do, too, and it's a far cry to call gmail something impressive. By web dev standards, it's a fairly basic application.
When I got my n7, I considered Reader to be the core of the core of the core of that device. As soon as I powered it up, I was on Reader instantly reading new posts from tens of blogs I like to follow. I wasn't interested in the magazines they wanted to throw at me because immediate digital publishing has pretty much killed the magazine/newspaper industry. So for me, reading news and tech articles on my Android device WAS Reader. That's a HUGE part of why the device was purchased.
Do I use it to gmail? No, maybe to read mail, but not to respond to it because typing more than a paragraph on a touch interface is maddening. Do I use it to sync contacts? No, it's not a social device for me. Do I use it to do calendary stuff? No, I do that on my phone, so syncing both of those are turned off. The device is a web browser, game console, and news consumption engine. What drove those were Chrome, the app store, and Reader. One would think those things would be held up by Google as core parts of the platform, but maybe they want to push their own news web portal and Currents. Who knows.
As part of google apps, it's a mediocre solution. It's not good for collaborating with people outside your organization (clients, patients, whatever else ends in -ients) but it's just good enough to prevent me from going with a hosted Exchange solution or to go back to administering our own infrastructure. I suspect it's prioritized to remain at that level.
Google moving away from open standards is a current popular opinion in bologsphere. Google has always made it super easy to take away your data. So, if you are not happy with a company than just don't take their services.
It's not an opinion. In the case of WebDAV, they are, in fact, moving away from open standards. It's also not about data export - it's about making commitments to businesses that pay for these services under contract and in conjunction with software that may or may not work with the apps in future.
You also have to realize that nothing can be committed forever. I don't really subscribe to doomsday warnings. Open Standards is an idea and not Google's corporate policy. It will prevail as long as 'the people' care for open standards.
Just move it when/if they shutdown the service? It's not like they are going to suddenly shutdown the service or remove your ability to export contacts.
I only skimmed the article. But I agree with your worry that this looks to be instructions that could put anyone who follows them at even higher risk of pain. The aesthetic of the web site doesn't give me a warm and fuzzy feeling about the quality of the tools being promoted.
That said, I suspect a key point the author would make in response to our concern is that we can vigorously backup the server hosting our contacts and calendar when we host these ourselves. In fact, this point is precisely why I host so much of my own stuff: I can manage how data is handled. Synchronization notoriously sucks, and if some synchronization failure causes many records to suddenly vanish, it would be easier to restore those records from a self-maintained server than to plead with Google's support. I wouldn't know the first thing about recovering synchronization-scrubbed records from the devices themselves (e.g., an Android phone).
I suppose it would attack your point about "better backup" by asking what value that backup has if you can't easily leverage it to restore on demand.
Still, I feel your points about quality, redundancy, and uptime could not be easily countered.
Touché. Those are awful. I'd say Oracle should be ashamed, but I'm sure Larry Ellison is feeling really bad about himself in between counting stacks of money.
Oracle installers aren't meant to be pretty, but rather practical, they are run mostly by IT professions who know what they are doing. Remember an Enterprise solution has to conform to ADA in 57 different countries and be fully translated, must work on version current-15, must work on lynx, IE5, csh, bash, ksh, windows 31, osx and vms :)
your convenience is pretty low on their priority list...
No, the enterprise solution means it can run only on versions ancient-x, ancient-y, and ancient-z of these OSes on approved hardware and must use this JVM and accept applets and plugins. :-)
This is what happens when the buyer is not the user.
I've run my own mail server since I got my first permanent connection in 1995. It's not hard to keep it up and reliable, especially on a small scale. It's not much harder to keep a large one up (I designed and operated a huge qmail hub in the late 90's).
Everything is on hard drives just as it is at google and all the other providers. I don't use rack and data center distribution for redundancy or even RAID to be honest.
However I do have a DLT and a fire safe (which acts as an anti-theft device as it weighs 200Kg) :)
What I'm saying is there are varying grades of redundancy, backup and reliability. It might be good enough, or might give more confidence than the disclaimers Google offer.
As for functionality, Maildir with your MDA of choice gives you all kinds of rube-goldberg opportunities with text files. No silly "max connections" junk when your phone and your laptop hit IMAP at the same time.
The one thing I really liked when going from procmail/postfix/dovecot to Google Apps was spam filtering. I hated keeping up to date with the various new methods spammers discovered to offer me good times with farm animals.
Today, I get a lot of false positives, especially with the "this is spam because other people marked messages like this as spam" emails from my bank. I miss emails from family members, travel vendors, and even online shipment confirmation messages. I've had to create rules to catch them and move them to folders/labels as well as have a client-side process occasionally search for a whitelist and "rescue" the messages.
I never get spam. I rarely give out my address and if i do, its usually an alias unless i really trust the person. Spam? Delete the alias and get the fucker blacklisted as it deterministically points to the culprit.
I use usenet via eternal-September.org eith an alias and some junk in the address. Has worked fine for years now.
But if they lose it, do they care? It's one drop in the bucket of their analytics. The incentives I have to preserve my data are greater than the incentives they have.
On their side, they have some very bright minds and some very deep pockets, as well as a proven infrastructure. But if they lose my data, they lose some analytics which may cost them pennies or perhaps dollars of ad revenue.
On my side, I have myself and whatever advice I can cobble together from friends and the Internet, a reasonable salary, and the occasional hour or two of spare time. But if I lose my data, I lose over a decade of accumulated emails and contact information.
I will gladly allow Google to back up my data (I use gmail to archive my call logs and SMS messages right now, in fact!), but I would rather be in control of the primary copy myself.
I agree about moving contacts and calendar off Google myself. To be honest, if it's not actually somewhere you control, it's not yours.
At least if you host or manage it yourself, you are only at the mercy of your own incompetence, whereas the cloud's competence is usually not something you can determine or anticipate regardless of SLA's and architecture. This gets demonstrated regularly if you watch the front page of this site :)
I still use text files, a VCS and a couple of editor wrapper scripts as I have for at least 20 years to solve the calendar and contacts problem.
Honestly, is there anything that matches Google search quality of services? Is there any email hosting, paid or not, that is as feature rich as Gmail? What about online documents? Calendars and contacts? What about all the other little things I get, like Gtalk or Google hangouts?
Honestly, if there was any alternative that covered all of these with the quality of Googles services, I'd switch in a heartbeat, even if I had to pay $30+ a month for them all.
Actually, if anyone has recommendations, I'd love to hear them!
Bing isn't as good as google yet, but its reached the threshold of good enough. Hosted Exchange email is very feature rich. I pay outlook.com 5 per month and get wonderful integration with all my devices for calander, contacts, and ToDo. Gtalk is hard to replace because your friends use it. I use apples messages and it works fine with people on gtalk.
For my core functioms, outlook.com is head and shoulders better. People have probably just gotten used to how utter shit the gmail UI has become, but outlook.com is refreshingly well designed (the web client). And activesync support is of course killer.
Data corruption is unforgivable. My bank might have have better accountants, stronger safes, great hours, terrific customer service and round-the-clock security, but you can bet that if they randomly lost my money on a regular basis (or even once?) that I'd withdraw it and stuff it in my mattress until I found a better alternative.
So the actual article title is 'Moving your contacts and calendar away from Google', not 'Move your contacts and calendar away from Google' as it was posted here. Less of an instruction/warning than might be expected, and more of a how-to guide.
In my experience, other sources tend to have significantly better caching (google's don't even have images), (archive.org) multiple versions, but way way slower to pick up new content.
I use own blogging software on a low-end Intel Atom which keeps consistently generating pages around 20ms under load. So imho Wordpress is the fault here, not the caching plugin. Of course Wordpress is great software to use as writer and has a lot of features that are very easy to use, so I understand why you'd use it. I just think it's cool when I'm running something on 10x slower hardware than everyone else without compromising on performance.
I mean it looks very snazzy right when you first install it and start to use it, but has a lot of little issues hidden under the surface.
Couple of examples: Randomly deleting files because it got confused. Infinite loops. Essentially unusable on Windows servers (don't even try). etc.
As I said, I like OwnCloud, and I think it has a very bright future ahead of it. But it isn't "there" yet. I'd never use it in an enterprise in its current state.
That's discomforting, because I just installed it. I don't really need or want the filesharing bits, but a shared family calendar has been a holy grail for my family for a while now, and this seemed to be the quickest and closest way of doing it cross-platform.
> a shared family calendar has been a holy grail for my family for a while now
Out of curiosity, what's wrong with clicking the "Share" button on an iCal, sorry, iCloud Calendar? My family (incl those on Windows) finds this works well.
I suppose nothing other than I didn't know that existed. I'd fall back on the old "I like owning my data," but I will take ease of setup for a company I trust. So, looking in to that tonight.
I am, and I've been using it as my personal calendar for about two weeks now. So far it seems fine enough, but a little wonky around editing events that repeat. Honestly, probably lack of training - I haven't spent more than 15 minutes with it per sitting, and stopped actively tinkering once I got CalDAV working.
I dumped ownCloud after it deleted a critical file for some unknown reason. The bug is still open. Maybe the calendar and addressbook handling is better, but I won't be trusting it.
I am not OP, but what it reportedly does is when a new file appears, if the clocks aren't exactly in sync, instead of "spreading" that file to other machines, it instead immediately deletes it because other machines are "newer" and don't have the file (or something).
There is some kind of strange race condition with the syncing. It doesn't always happen. In fact it doesn't happen very often. But most people won't accept even a "rare" random file deletion from a product they're using to protect themselves from exactly that kind of data loss.
Sorry for the problems. Obviously we never want data loss to happen but sometimes it does, whether it’s error in the code or in the setup. Can you give me a link to the bug?
Thank you for the work you've done on ownCloud. It's a fantastic project and it would be my first choice for hosting as I self-host everything. But a client that is willing to delete files instead of syncing them is just not worth risking for me.
If you just care about calendars and contacts there is single-purpose CalDAV/CardDAV server software you can use. I setup Baïkal recently and got OS X and iOS syncing with it without any problems. I haven't started migrating my existing contacts and calendars to it yet, though.
Many times grave errors like file deletion are really hard to reproduce because of special setups or other errors, that’s why we need logs and more info to be able to properly debug.
In software development term, this is probably an over engineering. Spend all the effort to setup your own data cloud before it becomes a fact that you really need it, that is, Google decides to change their cloud service so that it's no longer usable to you. Of course you could argue that what if one day Google simply disable the data exporting feature without any notice, in that case you would have no time to migrate. I think that risk is low enough for me to live with.
Google's service is already problematic in that it compromises my privacy and the privacy of my contacts. I'd rather not share my schedule with Google, and I feel uncomfortable putting additional information in my contacts knowing that it will be synced to them. I've certainly never been given permission by my friends to share their home address with Google or any other company.
This type of mentality is poisonous. You're not "polluting" your server. It doesn't add anything to the discussion; all it does is troll and set up flame-wars.
There are other good reasons to want something written in your own pet language (e.g. "I want to modify it". "I don't want to learn how to admin MySQL", even), but just claiming "pollution" isn't a good one IMO.
Why the wording wasn't great, it is a valid concern. If you already have a server running Postgres, then maintaining a separate MySQL install really is a pain, especially when it comes to being up to date on security.
Which is fine. It's the idea that software can "pollute" is absurd. Give reasons and arguments, not slander.
Already having a python/postgres stack is a valid reason to want a piece of software that runs on it. Just saying php/mysql "pollute" is a terrible mindset.
It's like saying "I won't pollute my mind with German" instead of "I have no use for German in my everyday life". One is very trollish, the other is a valid point.
I think "pollute" conveys the idea much better than the German example. Yes, it can create an emotional conflict, as you've demonstrated.
However, I would say that adding PHP and MySQL to a public server just increased the attack surface as well as operations management overhead. In my eyes, that is an increase in risk/cost/etc. for that server. That justifies "pollute", "infect", "degrade", etc.
One of the problems with using a self-hosted alternative like this is the fact that Google open up a lot of their services through APIs which other applications then consume. A web app I wrote a couple of years ago is one example (meetingShed). This easy integration with other apps opens up a lot of possibilities, but is there an alternative with ownCloud? It would be really interesting to see if a solution could be developed to expose self-hosted services (dynamically located) through a publicly accessible API (statically located).
As micampe points out in this thread, Google is shutting down their CalDAV API [0] in favor of their own proprietary one [1] OwnCloud supports the open WebDAV/CalDAV standard, as do several other self-hosted implementations [2].
The solution is trivial - just have some json in a specific format. The problem is getting everyone to agree on a standard and to get big players like Google, MS, Yahoo to use the standard.
To be honest though, there doesn't really seem to be much effort to even try - maybe someone famous will step forward and get the ball rolling?
Can't read the article right now, but for contacts and calendars we have standard public APIs: CardDAV and CalDAV. Too bad Google just discontinued the latter, weirdly enough just a few months after finally supporting the former.
Unless something has changed recently, there isn't much information about Google's CardDAV implementation. I had to find a reference on a mailing list for a CardDAV implementation to even figure out what Google's URL structure looked like.
I couldn't figure out a way to list all contact lists. You had to know the name of the list and use it in the URL. Even at that, you can only access contacts in "My Contacts" or other groups. There's (seemingly) no way to access "All Contacts".
CardDAV is 'supported,' but not very well. It's definitely a second-class citizen.
"Click “Advanced” and select your database options. A MySQL DB will be way faster than a SQLite DB."
Do you really think you can really notice the microsecond that MySQL will be faster than SQLite? We are talking about a database that is going to host what, maybe 5 megabytes worth of data?
We are talking about calendars and contacts here, this is not stuff that you will sync every second, and will not contain large amounts of records.
Hosting this stuff in an SQLite database makes things like backup and security a whole lot easier. You don't really need a complicated database server just to store your phone numbers...
Still the performance of the database is going to be negligible if you take the network (probably wifi/cellular) and mostly background syncing into account.
While it's patent/licensing encumbered, I still prefer ActiveSync based solutions, just because they do good push for iOS/Android, are easy to manage, give you "free" lightweight MDM functionality, etc.
There are a few open source ActiveSync tools (which may be in violation of Microsoft IP), but I just use a commercial one (for work). Still thinking of screwing with the free ones for a personal server vs. strictly IMAP.
It's $2 per account, you can host multiple domains under one control panel, it has email, contacts, and a calendar under one roof, and offers iPhone/iPad provisioning as well. You can also create subadmin accounts if you want to delegate responsibility for a certain domain to someone else.
Thanks for the info. I'd happily for a service like this than rely on Google's "free" stuff that they might the plug on any time they want to push their useless G+ crap.
I've self-hosted my calendar using DAViCal (http://www.davical.org/) since I got my first iPod Touch. My wife and I use it to host shared calendars that we access from our phones. I run it on a cheap box I own but it would be just as easy to host it on an inexpensive VPS. Backup is a Postgres dump taken daily stored locally on the server and rsync'd to two other machines in separate physical locations. It took a couple hours to set up but has been trouble and attention-free for years.
Any similar setup that uses IOS phone instead of Android out there? I am tired of contacts going missing from IOS and Google or 4 copies of all my contacts showing up in both IOS and Gmail. I lost all my contacts in IOS when I upgraded to the iPhone 5. Then I had a bunch of contacts that I did not want in my contacts populate my IOS contacts when I linked my Facebook account on my IPhone 5. It is even worse on my Windows 8 notebook as I have 5-6 duplicate contacts for each single contact as it imports from IOS, Outlook, and Gmail.
Loosing contacts can happen. Just this past week my father-in-laws contact info disappeared from Google contacts. This isn't the first time this has happened to me either.
Wtf, I've never had this happen. I would instantly stop using any service where this happened (especially where it's not possible to decently contact support).
Thanks for the article! I didn't hear about ownCloud (or similar projects) before, I'm already reading manual (http://doc.owncloud.org/server/5.0/admin_manual/) and will be installing ownCloud on my VPS today. I had doubts about giving Google (or any other company) so much of my data (not only emails, but schedule, contacts etc.) - now I found a great solution. I love HN.
What you're looking for is syncml: https://en.wikipedia.org/wiki/SyncML the wikipedia page lists various clients and servers. On Linux (well, ubuntu at least) sync-ui will sync evolution with a syncml server. After a lot of searching for solutions to use with my n900 I chose memotoo as a provider. There is an preconfigured syncml android client for memotoo.
Nice idea. I have just gone through the minor hassle of switching my blog off of Blogger and relegating Gmail to my backup/secondary email service - all in the spirit of controlling my own stuff. I only calendar share with my wife, so setting up something like this is probably something I will do also.
Any JVM based open source projects? I would rather not deal with PHP.
Not sure that's a smart move, if you care about your data. Do you think you can achieve better uptime and reliability than Google does with their engineers and servers and distributed data centres? I mean, it sounds cool and everything to do it yourself, but if you just dump all your data onto your local server (or EC2 instance), it's just a centralised storage that can fail any time, not a "cloud" any more.
The only thing really stopping me from migrating to my own platform is email. I've had many terrible experiences and heard many horror stories about running mail servers, I just don't want to deal with it.
Mmm. I'm sure people have had bad experiences. However I run my own mail server (and no others who also do) for myself, my company and a few other domains I host with almost no effort. It's true I have fewer than a dozen users, but I'm sure most people would be the same way.
I just upgraded it this weekend, otherwise I haven't actually had to touch it in the 5 years since I originally set it up. Support ends for the Ubuntu LTS release it was based on next month, and I chose to migrate to new server software (dovecot) without much difficulty at all. I think anyone moderately skilled with sysadmin tendencies could manage it.
But I confess I personally have a gmail account that I use most of all. I receive lots of mail daily, and none of the open-source clients (webmail or otherwise) have managed to compare in terms of usability. Priority Inbox is the most recent example of something that I would struggle to live without.
What about other people's spam filters? That's the biggest problem I've heard about is being blacklisted because of being an unknown sender or on some VPS service or something like that.
What can we do for translated versions? I'm not a dev but my father uses google calendar a lot because it's easy and intuitive but he knows nothing about english...
You would have to learn quite a bit. Pick a language that you know almost nothing about, preferably one whose alphabet you can't recognize without looking it up. Now use a website built in that language. Is this potentially an exaggeration? Yes. Is it still relevant? Yes.
this post is the worst.. A) never really had contacts "disappear", although i definitely encounter conflicts with weird results (kind of expected).
B) use a shared host (justhost.com. lol) and php over google's infrastructure? fat chance this guy is competent enough to know what the fuck he's in for down the road..
ownCloud does sync with iPhone, yes. I am still basically on Google myself, but did a trial install of ownCloud and it worked pretty much instantly. Just add as CalDAV and CardDAV accounts on the iPhone, the correct addresses to use is shown in ownCloud somewhere.
Android gained popularity because it was tied to Calendar and Gmail. These two services were pretty pervasive already, and so became the default in Android. When that happened, those two services became crucial to Android.
The fear that Contacts and Calendar will go away like Reader did (or, rather, will) is irrational. Reader barely had a mobile client version of it. With Google reigning in the branding on Android and requiring that it be tied even tighter to Google services, Gmail and Calendar became a dependency of Google's mobile OS.
Yeah.. see, I don't think the author has thought their cunning plan all the way through.
I understand the anti-google sentiment is at an all time high after the Reader shutdown fiasco, but let's stay rational, here!