Cloud rather than Thin?

[jamesleeds]

As applications move more to the web, are you considering the possibility of using a Cloud Client (such as the new Google Chrome OS, or Android) as an alternative? This gives the benefits of low maintenance on the desktop - but would require all applications that are to be used to be web based (e.g. VLEs, Google Docs instead of MS Office etc)

[DarrenM]

Have thought about it but how do you know who is using the client when something inappropriate is accessed. Or perhaps I am thinking in the old ways still.

[tmcd35]

Personally always thought the half-way house of application virtualisation was the best solution (Thinstall, V-App, etc). Just never been able to find a good cost effective (read cheap) solution.

[enjay]

but would require all applications that are to be used to be web based (e.g. VLEs, Google Docs instead of MS Office etc)

 

I think you've answered your own question there - I would have thought we're a long way off every little edu-app being web-based, especially Primary ones. Not sure how video editing or interactive things like Smart and Promethean would fare in the Cloud, either.

[dhicks]

As applications move more to the web, are you considering the possibility of using a Cloud Client (such as the new Google Chrome OS, or Android) as an alternative?

 

Yes. We'd have to run web-only workstations alongside traditional fat clients at the moment, as we would for a thin client setup, as some applications are still desktop-only at the moment. I can't see any real technical issues stopping anyone re-implementing pretty much any application as web-based, though. Having people authenticate before they are allowed to use certain web sites is easy enough, you just get your filtering solution (in our case SmoothWall) to ask for a username and password before letting you on to certain sites.

 

--

David Hicks

[dhicks]

Not sure how video editing or interactive things like Smart and Promethean would fare in the Cloud, either.

 

Actually, video editing and interactive whiteboard software is perfectly possible / really quite easy to do as web-based applications. For video editing, you need a decently hefty server with lots of assorted ways of plugging in cameras / memory cards / external drives / whatever to read in video footage and covert it all to handy-sized YouTube-style clips that end users can then chop up and stick back together to make a video. If anything, video encoding should be faster with such a machine as you're putting more investement in to one system, probably with dedicated accelerating hardware for video processing, instead of trying to kit out an entire classroom with video editing workstations.

 

Interactive whiteboard software is easy to write with Flash or HTML 5 Canvas (like, half-a-dozen-lines-of-code kind of easy), and I'd suggest the Becta Common Interactive Whiteboard Format as being a sensible format to choose to get something to import / export.

 

--

David Hicks

 

--

David Hicks

[torledo]

I think you've answered your own question there - I would have thought we're a long way off every little edu-app being web-based, especially Primary ones. Not sure how video editing or interactive things like Smart and Promethean would fare in the Cloud, either.

 

that's assuming people aren't changing the way they do video editing and interactive presentation.

 

i think they are.

 

i do prefer to use the term 'web-based' although i suppose the cloud is in my view just a more sophisticated way of web enabling. so you've actually got the correct infrastructure for web services delivery. And the client doesn't have to be a web browser, could be a more bespoke RIA so long as the API's/Web services are exposed. ofcourse your not going to be able to rewrite every school desktop app into a RIA or ajax-loaded web app, but then again your not going to get rid of the fat pc altogether and the somewhat bizarre [if you think about it] idea of streaming full desktops to thin clients.

but in theory you can curtail the reliance on the traditional models.

 

i'm really not entirely sure what i'm talking about....not a programmer but i can see the bigger picture.

[torledo]

oh, by the way, to answer the OP, i think it'd be daft not to implement some form of cloud client - just not sure it needs to be entirely reliant on web communication

.....for initial deployment, and to get the most out of collaborative functions, yes. but no reason why they couldn't exist on the client as an installed and be operable offline,

albeit with a much smaller footprint than the traditional desktop model.

[dhicks]

but no reason why they couldn't exist on the client as an installed and be operable offline

 

Or, rather simpler, have locally-based servers - just because something is "web based" doesn't mean the web server can't be on your local LAN.

 

--

David Hicks

[tmcd35]

Yes. We'd have to run web-only workstations alongside traditional fat clients at the moment, as we would for a thin client setup, as some applications are still desktop-only at the moment.

oh, by the way, to answer the OP, i think it'd be daft not to implement some form of cloud client

I dunno, I feel as if I'm missing something here. Isn't your web browser (IE, Firefox, et al) your 'cloud client'? Unless you want to completely do away with the fat client, why implement any special 'cloud client' at all? Beside surely a Thin Client box launching a web browser (thinstation + firefox?) is all you'd need client side?

 

somewhat bizarre [if you think about it] idea of streaming full desktops to thin clients.

I don't see that, if it's done right. I think Wyse Zero Clients have hit the nail on the head with this one. Much simpler management and all the raw power of traditional fat clients. That said I'd much rather stream apps in their own protected virtual bubble than stream entire desktops. I also think streaming apps has more scope than cloud computing while providing much the same benefits.

 

The only time cloud based apps make sense to me is when you need to edit documents (photos, music, video, spreadsheet, whatever) from multple locations (home, school, travel lodge) and can't guarrenty always having the same software no matter where you. I don't think cloud based apps, or current internet speeds (especially with Virgin Media's throttling policy, grrrr), are good enough to make me want to use them as my primary editing tool.

 

Compare Google Docs or Zoho to OpenOffice for example.

[CyberNerd]

Citrix running over a java web app on an amazon cloud ?

[torledo]

when i think of a cloud client, i think of the devices that are based on android or iphone os or potentially windows phone 7. rightly or wrongly. here, not all apps need to exist within the browser. that's why i think stuff like the methods for developing native apps and ria technologies on such devices can provide richer experience not purely bound by the browser. ofcourse you can still use the browser, but there other ways to develop apps. the other ways just leverage web services to make them connected.

 

I don't see that, if it's done right. I think Wyse Zero Clients have hit the nail on the head with this one. Much simpler management and all the raw power of traditional fat clients. That said I'd much rather stream apps in their own protected virtual bubble than stream entire desktops. I also think streaming apps has more scope than cloud computing while providing much the same benefits.

 

i don't know the technology all that well. but it's still about working in a more conventional manner. conventional form factor, conventional apps....the lower maintenance aspect of the client device is a bonus. but at what cost as far as the backend infrastructure and the cost of client devices ? low maintenance and over the air control will be available for the mobile devices which rely on more web-centric services.

 

The only time cloud based apps make sense to me is when you need to edit documents (photos, music, video, spreadsheet, whatever) from multple locations (home, school, travel lodge) and can't guarrenty always having the same software no matter where you. I don't think cloud based apps, or current internet speeds (especially with Virgin Media's throttling policy, grrrr), are good enough to make me want to use them as my primary editing tool.

 

Compare Google Docs or Zoho to OpenOffice for example.

 

i tend to agree with dhicks, that hosting an app on your own servers is a fine solution. i'm not advocating a 100% move to this amorphous 'cloud'.....merely that people are changing the way they work productively. once upon a time it was local email clients, now most people are happy with most webmail offerings which removed the need for locally installed clients. i'm sure the online web processing apps work fine for most people, but even with a small footprint device you can have locally installed versions of productivity apps. for what most people do, do you really need a several hundre megabyte or multi-gigabit local install. sometimes yes. but then that's why you haven't thrown your desktop or laptop away [or your thin client infrastructure if you want it to appear local] ;0)

 

the point is as working behaviours change, how much more will you be able to accomplish with the new breed of devices and ostensibly web-based solutions. a suprisingly large amount i think. no one thinks anything of updating their blog, editing a wiki entry, making a voip call, organsing their emails on some form of webmail service - all from a variety of devices, some with comparatively little processing power.....now it's editing documents, collaborating on projects, quick and easy upload and edit of video. and they adjust to the reduced horsepower available for such tasks. but increasingly there's more horsepower available within the 'cloud' and the services improve even through a not as quick as it should be broadband service. but i agree it's not a magic bullet to every problem that needs solving. and for a large site, on-premise app hosting can make a lot of sense.

[AntiThesis]

Given our fairly slow internet and that we're around 6000 miles away from the closest clouds it's certainly not something we're looking at now. Once we start making proper strides in internet access for schools here it should be quite interesting.

 

I've been playing with the idea of using two spare servers to make my own private cloud to mess around with. Eucalyptus is starting to look quite attractive really.

[tmcd35]

when i think of a cloud client, i think of the devices that are based on android or iphone os or potentially windows phone 7. rightly or wrongly. here, not all apps need to exist within the browser. that's why i think stuff like the methods for developing native apps and ria technologies on such devices can provide richer experience not purely bound by the browser. ofcourse you can still use the browser, but there other ways to develop apps. the other ways just leverage web services to make them connected.

 

That's a very interesting take on cloud computing and one I never really considers. My brain as always done - cloud = inside a web browser. I can see cloud computing for devices like the iPhone and iPad, HP Slate, etc. But by their nature they are not you primary computing device.

 

i tend to agree with dhicks, that hosting an app on your own servers is a fine solution. i'm not advocating a 100% move to this amorphous 'cloud'.....merely that people are changing the way they work productively. once upon a time it was local email clients, now most people are happy with most webmail offerings which removed the need for locally installed clients. i'm sure the online web processing apps work fine for most people, but even with a small footprint device you can have locally installed versions of productivity apps. for what most people do, do you really need a several hundre megabyte or multi-gigabit local install.

 

I think cloud computing needs a few things to happen before it can win the hearts and minds of the masses. As someone who tried his best, and failed, to get users to migrate from MS Word to Open Office Writer - I know the current state of web based productivity apps really would not cut the mustard.

 

- Internet speeds need to be around the 100mbps so file transfers appear as seemless as possible. In fact the end user should see no difference uploading a photo from a camera to web server storage than storing on their local hard drive.

- An extention of that is WebOS's need to be a bit clever'er with storage. The best method is to use WebDAV, or similar tech, and map the remote storage as a local hard drive so the end user interact with the storage with the same file management tools (windows explorer or apple finder or equivalent) they use to interact with their local hard drives.

- At the very least the main three productivity apps - word processor, spreadsheet, presentation - needs to be at a feature parity level with OpenOffice. Microsoft's Webbased Office 10 may be the one to watch here. Given the to proviso's above it may just be able to pull it off.

 

I still think we are probably around 10 years away from Cloud computing being able to effectively replace the traditional desktop model and until then I really see little sense in rushing in to it. It's a good solution for somethings but your not going to convince your users to switch to it perminatly for a while yet.