In the cloud source control?

I'm looking to move all my code into the cloud using some sort of source control service. It must first be able to integrate with Visual Studio 2008 and 2010, and possibly other environments. Meaning something that's fairly popular among different development environments. The only source control I'm familiar with is VSS, but I think it's time to move away from that. I also prefer a free service. I know that there are only code repository services that are free, but I just don't know how that integrates with Visual Studio and source control.

I really like the idea of something in the cloud as I can access it from anywhere.

Thanks for any advice you can offer!

Oh, one other important fact. This is not open source code I'm talking about. This is my own personal archive. I don't want anyone to be able to access, except for myself and/or other people I specifically grant access to. Thanks.

So I guess there is no option for this?

Why do you think you need a 'cloud' based application for source control?

Why not simply buy the service from a remote storage provider designed around source code control requirements and let them worry about how many gnomes they need to ensure your data is available when you demand it?

Personally, I've just started using Kiln from Fog Creek - based on Mercurial); it doesn't integrate directly into VS (but it can through the external tools), but it doesn't really need to. It doesn't matter to me - or any other user - how they store the data, as long as it is secure and is available.

I use VisualSVN on my server for my repository and AnkhSVN as my integration into Visual Studio. It works really well, but sometimes AnkhSVN gets a little wonky.

AnkhSVN works with any Subversion compliant server. I don't know if there's a cloud-enabled version of Subversion yet.

Heck, I'm still trying to figure out what the heck a "cloud" really is. I always called it "the Internet". Seems just the next big tech buzzword to me.

Cloud is giving all your code for free to da Microsoft. :P

You're all wet.

Look here for a standard definition: http://csrc.nist.gov/groups/SNS/cloud-computing/

I know, this "cloud" thing is just the latest buzz word for "Internet" or "web". I'm just saying I would like to start consolidating all my projects on the web/cloud/internet, whatever you want to call it. So that from no matter where I may physically be, I can always just fire up my Visual Studio environment, and have access to my "repository" of code. I feel safer doing this as well, just in case my house burns down, I don't lose all my hard work.

I realize I could just use a "dumb" online document storage, but I'd really like to have source control, and Visual Studio integration, to make it easier, and less prone to working on the wrong copy of the source code.

For access you can use remote software VNC,remote desktop, a VPN etc.
For your house burn(i think your data will be the least of the problems), u can copy it to your hot girlfriend's pc(not on folder "latest gossips" that she has though).
Sorry this is where i put my foot down no mater what anyone says, i don't trust cloud (at least for now) for my PERSONAL data.If it's company data and it's their decision, i couldn't care less.

Isn't that what the site SourceForge is designed to do? Be an online code repository to allow anyone to work on projects? Just sign up, possibly pay some monthly/yearly fee and you have a full repository system with source control accessible by anyone of your choosing anywhere in the world?

Quote:

---

Originally Posted by Jenner

Isn't that what the site SourceForge is designed to do? Be an online code repository to allow anyone to work on projects? Just sign up, possibly pay some monthly/yearly fee and you have a full repository system with source control accessible by anyone of your choosing anywhere in the world?

---

Isn't SourceForge for open source only for anyone to work on? This is definitely proprietary stuff I'm talking about that I (or whoever I explicitly give) should have access to.

Quote:

---

Originally Posted by dbassettt74

Isn't SourceForge for open source only for anyone to work on? This is definitely proprietary stuff I'm talking about that I (or whoever I explicitly give) should have access to.

---

Yea, just looked over their site a bit and it does look open source only. How about Beanstalk, or Assembia? Both offer private SVN service and there are a lot of good integration options for SVN. I already mentioned AnkhSVN which integrates directly with Visual Studio (has an "update/revert" tab, many right-click options, icons in the solution explorer to show synchronization status of all files, etc). There's also TortoseSVN which integrates with Windows Explorer rather than Visual Studio.

Quote:

---

I know, this "cloud" thing is just the latest buzz word for "Internet" or "web".

---

It's more than that... it's also virtualization and being able to expand capabilities on demand. For instance, if I have a system whose demand and need increases, I can spin up new instances of SQL Server to take some of the load w/o having to buy new hardware. Just create the instance, config it, and off it goes.If the demand goes down, the virtual instances can be taken off line. No need to worry about wasting money on hardware that is going to sit idle.

-tg

So, you're saying it's just a more efficient cluster model where the same services rather than being virtualized on every machine in the cluster, is now virtualized on-demand according to load? I assume there's a master instance which determines load then?

If I have to manually configure new instances... then there's nothing new about this "cloud" model. I've been able to do that for the last 7 years since Server 2003's release with clusters. Virtual or not, I'd still only be running one instance per physical machine. It would make no sense for a single machine to be running 2 virtualized instances of the same database server since it's running under the same hardware; the performance will be the same - less that of overhead.

I'm still not buying into the buzz because I don't see anything new and exciting at all. Just a bunch of salespeople who finally "got" an understanding of the tech that's been around for some time now.

Let me use the example they gave us at PDC last year.... you own a pizza joint.... Super Bowl Sunday is coming up, and you want to make sure you can handle the extra load of everyone ordering pizza for the game... the old school of though, you buy more hardware and expand your SQL Server and Web Farms... you have to set up the replication and hook it all up and hope that it holds. Then after the weekend, demand slacks off.... now you have a bunch of hardware you just paid for that you are no longer utilizing.

With the web, on the other hand, when the demand starts to increase, you login, fire up an other virtualized server, click a couple buttons, and moments later you have added virtualware. When the demand dies down, you turn off some of the virtualizations as needed.

Granted it really isn't new... it's about as new as Web2.0 was when it came out. The technology has been there, but now it's being used in a new way and taken to another level.

In the case of an Aszure database with Microsoft. they have multiple data centers around the world (5 in total at the moment) ... all replicated. So, here in the US, I'd hit the center in Waashington. While in India, I would hit the center in say Mumbai... it also means in the extreme situation if one center goes down, there's 4 others with the same data and can be re-routed. It's server farming and virutalization that's being taken to a level where companies don't need to do it themselves or invest in the hardware/expertise themselves.

-tg

Oh! Wait, I think I understand now... this is a SERVICE! This isn't anything my company owns... this is a service provided by Microsoft or whoever where I can run virtualized machines on their hardware... (i.e. rent-a-hardware).

Me and my company don't run a "cloud" like we would a server cluster, we just buy the service we need from whoever owns the "cloud" system.

Bingo! There you go! Now you've got it... exactly.

-tg

Interesting idea - if I understand it right.

But how does it work performance wise? If your database is in a cloud somewhere on the internet and your IIS server is at your company, doesn't the bandwidth of your internet connection throttle your performance?

Hosting Subversion in the Cloud with Live Mesh/

Quote:

---

Originally Posted by dolot

...doesn't the bandwidth of your internet connection throttle your performance?

---

That's my concern too. Most companies internet connection are 100x to 1000x+ slower than their internal LAN speed. CAD on a cloud would be a joke with regular 40meg files, not to mention that in a CAD assembly, you might have 30 of these open at once (crawling over a 1Mbps T1 that shared with 30 other office employees? Ouch).

I could see source control too; but I work with some projects with resources upwards of 100Mb. At least good source control systems like Subversion only update to you the files that are different than your local copy. You'd only have to pull the painful 120Mb head version once.

Yea, I don't know how you'd handle large database transfers like inventory stock levels. Right now, we have small data-puller programs that update the off-site web system once a day at 5:30PM. It takes 10 minutes to pull the data, zip it, FTP it over, and update the database on the website with the new info. I could put the inventory on the cloud, but then pulling bulk data we use in-house daily (like for some work order generation and reporting) would take an eternity.

I've just started using BitBucket (www.bitbucket.org). It is free, though it doesn't integrate with VS IDE, it does seem to work fairly well. The process is just a tad "dense." It took me several hours to get it going.

Dick

Quote:

---

Originally Posted by Jenner

That's my concern too. Most companies internet connection are 100x to 1000x+ slower than their internal LAN speed. CAD on a cloud would be a joke with regular 40meg files, not to mention that in a CAD assembly, you might have 30 of these open at once (crawling over a 1Mbps T1 that shared with 30 other office employees? Ouch).

I could see source control too; but I work with some projects with resources upwards of 100Mb. At least good source control systems like Subversion only update to you the files that are different than your local copy. You'd only have to pull the painful 120Mb head version once.

Yea, I don't know how you'd handle large database transfers like inventory stock levels. Right now, we have small data-puller programs that update the off-site web system once a day at 5:30PM. It takes 10 minutes to pull the data, zip it, FTP it over, and update the database on the website with the new info. I could put the inventory on the cloud, but then pulling bulk data we use in-house daily (like for some work order generation and reporting) would take an eternity.

---

That's precisely the pros and cons kind of stuff one needs to consider when (or if) moving to the cloud. It's not for everyone... and it's primary capability is for scaling... high volume, high transaction kind of operations...

-tg

Quote:

---

Originally Posted by dolot

But how does it work performance wise? If your database is in a cloud somewhere on the internet and your IIS server is at your company, doesn't the bandwidth of your internet connection throttle your performance?

---

Normally you'd only have the "client" in-house (or at your customers' sites). So for a Web application you'd run IIS, Apache, etc. inside the cloud. Within the cloud bandwidth is high and ranges from very cheap to free.

To Customer Bill's Chrome browser it just looks like any other Web application.

It works the same way for smart client and thin client applications, with as much of the data digestion as possible happening in the cloud where bandwidth is fast and cheap.

"Hybrid clouds" are where you get into trouble with bandwidth. That would be the case where you have indexed and blob storage, message queuing, etc. in the cloud and your IIS server on-premise.


The nice thing is that once they get cloud pricing tuned properly it becomes attractive for doing pilot projects or small LOB applications without having to get on your knees before the centralized IT bureaucracy to present a mound of resource justification documents and the blessings of three upper managers. You'll buy cloud services like ordering office supplies.

Eventually the "IT" guys will all be selling used cars once there is no need for in-house infrastructure. The idea is to eliminate DBAs as well. ;)


How much of this will happen is anybody's guess. However NetFlix had moved much of its video streaming and billing to AWS the last I heard.

Quote:

---

Originally Posted by Jenner

I could see source control too; but I work with some projects with resources upwards of 100Mb. At least good source control systems like Subversion only update to you the files that are different than your local copy. You'd only have to pull the painful 120Mb head version once.

---

This is somewhat off thread, but one thing I've discovered recently about SVN is that it has a way of doing a binary diff when it transfers files. So only the part that is changed is actually transferred.

The other day I retrieved a 3meg .mdb file off our svn server, but svn told me it only copied 2K over - and I believe it since it only took 2 seconds over my cable modem. But when I checked the file, there was the change my dev lead put in there where he checked it in.

How that works is a mystery to me.

Quote:

---

Originally Posted by dilettante

Normally you'd only have the "client" in-house (or at your customers' sites). So for a Web application you'd run IIS, Apache, etc. inside the cloud. Within the cloud bandwidth is high and ranges from very cheap to free.

---

Now that is starting to make sense.

Quote:

---

Originally Posted by dilettante

Eventually the "IT" guys will all be selling used cars once there is no need for in-house infrastructure. The idea is to eliminate DBAs as well. ;)

---

I doubt they will be selling used cars, although in today's economy that may make more $. :) Probably, though, they will just start working for the cloud companies.

And let's not forget all those pesky desktop issues they have to deal with at the client side. :)

Not everything will go to the cloud. Can you imagine a 1.5 TB database on SQL Azure DB. Never.

Quote:

---

Originally Posted by Jenner

Most companies internet connection are 100x to 1000x+ slower than their internal LAN speed. CAD on a cloud would be a joke with regular 40meg files, not to mention that in a CAD assembly, you might have 30 of these open at once (crawling over a 1Mbps T1 that shared with 30 other office employees? Ouch).

---

September 30, 2010: AutoCAD in the Cloud Launches Today

So let's say a company is doing it's daily heavy db manipulations and suddenly the phones are off(happens a lot here).
I imagine everyone hooking their mobiles to the network to create the supermobile internet communicator :D
Or loosing 5-10000 Euros.

Quote:

---

Originally Posted by dilettante

September 30, 2010: AutoCAD in the Cloud Launches Today

---

Notice it's only talking about AutoCAD DWG files which are nothing more than line-drawings. That's old-school 2D CAD. There's no mention of Autodesk Inventor; their 3D platform. We stopped using "kiddieCAD" in 1992. :)

You're not going to see ProEngineer solid models, MATLAB simulations, or Algor Finite Element Analysis runs in the cloud any time soon.

3D assembly model of an ink dispensing machine: 635MB
FEA analysis of the pressure load on a pump chamber: 175MB

No, I agree with techgnome. There are trade offs with using Cloud style computing. This would be one of them. I know it takes approximately 1.5 hours to download 650MB at work. Putting something like our engineering data in the cloud would shut down our engineering department.

I can certainly see it for web-based systems but not even the internet itself is fast enough to start putting that kinda data on it.