Re: gathering use cases
Hi, A server provides services to client computers. These services enable and shape the interaction of users with the client. For example, web pages are delivered to the user by a browser. However, the web pages are delivered to the browser by a service on the server side, e.g. Apache. Suppose that a deployment wants email capability (a use case). The standard server side service is pop3 and smtp. Thunderbird is a possible client-side application for these services. Most discussions on the lists are based on direct login to gmail which does not requires only httpd (Apache) on the server side. However, this means that email access only when the XO is connected to the internet. A client-side email client could enable email to be prepared offline and to be read offline. How would email capability be provided by sneaker-net with a usb drive, e.g. for someone to periodically take the usb key to an internet cafe and send/receive email? Tony On 11/09/2012 12:09 AM, Sameer Verma wrote: To clarify, by use case, we mean a way to describe the interaction of people with a system. In this case, it may be how a child interacts with the XO, and the XS (via XO) or how a teacher may interact with the XS (via XO or otherwise) by using Moodle. As you may have noticed, pretty much every response on this thread focuses on the system, with the assumption that the user end of it is understood well. I am not so sure. Take a look at http://www.gatherspace.com/static/use_case_example.html to see if it helps you understand the idea behind use cases and user stories. Here's another example of how use cases can vary by being very high level (which is what we are aiming for) and can be user centric or system centric. http://www.agilemodeling.com/artifacts/systemUseCase.htm Our focus is user centric, in a way where we would like to describe the actors (children, parents, teachers, admin, etc) and their actions (access class information, read books, send email) without the XS as the focus. Networking topology, storage, UI, LMS, DNS, etc. should flow from the storytelling exercise. We are a bunch of technologists and it is easy to get carried away by designing from the tech and not the user end. Sometimes that misses the mark. We may build it and they may not come. cheers, Sameer On Thu, Nov 8, 2012 at 4:07 PM, Sridhar Dhanapalan srid...@laptop.org.au mailto:srid...@laptop.org.au wrote: On 9 November 2012 10:19, Tony Anderson t...@olenepal.org mailto:t...@olenepal.org wrote: Hi, Sridhar Thanks for the clarification. I guess I was mislead by statements such as: The platform for the One Network server is an ARMv7-based XO, running the One Education OS (based on OLPC OS). This makes development, and deployment and support far simpler than a standalone distribution. The OS can be extended with server capabilities using a bootable USB Customisation Stick (offline) or yum. Please accept my apology if any statement I have made seemed uncivil, that was certainly not my intention. Communicating by email in certainly much more hazardous in this regard than face-to-face. Thank you, Tony. I was quite careful to take your needs into account when I wrote the design doc, so I had trouble understanding your opposition to the idea. Maybe we can make the doc clearer somehow? I structured it as: 1. context 2. Community XS design 3. One Network server The Community XS design itself is flexible enough to handle a variety of different deployments' needs. One Network server is merely one configuration of the Community XS, mentioned as an example of what can be done. Just as the deployments you are supporting have specific and urgent needs, so do the ones I am working with. I don't believe either of us is pursuing personal desires. We certainly can easily differ on which is the appropriate technical approach to solving the problems of a deployment. I think we generally want the same thing in the end. I'm happy to continue the conversation to improve the design and implementation. I sincerely believe that this design can accommodate your needs. I really appreciate this specification: * this is a flexible design, built on Fedora * it will run anywhere where x86 or ARM Fedora will work * it can be installed on top of an existing Fedora installation using 'yum groupinstall xsce' * being designed in this way provides extreme flexibility for deployments * all current features of the XS (Moodle, etc.) will be ported, but will be optional * installation and configuration will be easy, but sysadmins will be able to treat it like any Fedora installation Awesome. I hope this also satisfies your desire to have it rebased to Fedora
Re: gathering use cases
Hi, Sridhar One of the important potential benefits from gathering use cases is to clarify the roles of the Community XS project and XS-0.7. From my perspective, if you can afford to deploy a server with 2GB or more main memory and 500GB or more hard drive capacity, XS-0.7 serves well and easily supports adding or customizing services. The configuration shown in SF costs less than $350 and draws 18w (based on measurements taken by George Hunt). With a large hard drive, it is perfectly to keep Moodle even if it is not used. In my experience, the school server is installed in the school ready with XS-07 installed along with the available content. Barring hardware failure, the only interaction at the school is to turn the server on at the start of day and turn it off at the end. With modern computers, both functions require only a push on the power button. Community XS serves a valuable role where only XOs are available or where power constraints are critical. It may also have a role in supporting special network requirements. Tony On 11/08/2012 10:55 AM, devel-requ...@lists.laptop.org wrote: Message: 2 Date: Thu, 8 Nov 2012 13:57:12 +1100 From: Sridhar Dhanapalansrid...@laptop.org.au To: Holth...@laptop.org Cc: Devel's in the Detailsdevel@lists.laptop.org, Maryam Rehan El Bazmel...@mail.sfsu.edu,iaepi...@lists.sugarlabs.org, XS Devel server-de...@lists.laptop.org, Support Gangsters support-g...@laptop.org Subject: Re: gathering use cases Message-ID: CABPDnXn8Xz9BwRD_zy_PUr43CG=_rKXv8u6=zLjD0f+Juk=x...@mail.gmail.com Content-Type: text/plain; charset=ISO-8859-1 On 3 November 2012 07:46, Holth...@laptop.org wrote: Thanks much Sameer. Am includingsupport-g...@laptop.org to make sure we gather microdeployment and support volunteer feedback too. While our XS Community Edition experiment ain't quite ready for showtime/download yet, its white paper code repository are here: http://wiki.laptop.org/go/User:Holt/XS_Community_Edition The white paper that Adam is referring to can be found at https://docs.google.com/a/dhanapalan.com/document/pub?id=1dnhU2F6EntepVXTgN8QpkME8fZVUuPjcCoMUfAVKbcc I am the primary author, but it was written by canvassing feedback and opinions from schools and XS community members (including Jerry Vonau, George Hunt, Adam Holt and many others). The paper discusses several use cases and design goals, and outlines the underlying principles behind how we do things in OLPC Australia. There is some miscommunication being spread around that I'd like to clear up. Firstly, this is not a 'secret' project. The white paper explains our mission, and the code repo and issues trackers are open. Secondly, the intent is to create a flexible framework that can be adapted to suit local needs. This is*not*, as some keep asserting, a project to have the XS work only on the XO. The community XS will work on any x86 or ARM based hardware that works with Fedora. The white paper explains this in depth. Sridhar ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: gathering use cases
Hi Tony and all, On Thu, 2012-11-08 at 13:32 -0500, Tony Anderson wrote: Hi, Sridhar One of the important potential benefits from gathering use cases is to clarify the roles of the Community XS project and XS-0.7. Why is that an issue at at all? The Community XS (think XS-0.8 here) is going to address the current shortcomings of XS-0.7. From my perspective, if you can afford to deploy a server with 2GB or more main memory and 500GB or more hard drive capacity, XS-0.7 serves well and easily supports adding or customizing services. The configuration shown in SF costs less than $350 and draws 18w (based on measurements taken by George Hunt). With a large hard drive, it is perfectly to keep Moodle even if it is not used. In my experience, the school server is installed in the school ready with XS-07 installed along with the available content. Barring hardware failure, the only interaction at the school is to turn the server on at the start of day and turn it off at the end. With modern computers, both functions require only a push on the power button. I'm not thinking about an XO here at all but on a server that you spec'd. Tell me how do you deploy the 0.7 version on a network where you don't have control over the DNS and have everything work? I have an idea on how to pull that off with avahi, I would like to hear about how you would implement that situation. Community XS serves a valuable role where only XOs are available or where power constraints are critical. It may also have a role in supporting special network requirements. Not to mention better hardware support for the latest hardware available and ARM support by using Fedora. You can get on board with the future and have a say in your pet interests but please don't get in the way of progress. Jerry The other XS-CE developer Tony On 11/08/2012 10:55 AM, devel-requ...@lists.laptop.org wrote: Message: 2 Date: Thu, 8 Nov 2012 13:57:12 +1100 From: Sridhar Dhanapalansrid...@laptop.org.au To: Holth...@laptop.org Cc: Devel's in the Detailsdevel@lists.laptop.org, Maryam Rehan El Bazmel...@mail.sfsu.edu, iaepi...@lists.sugarlabs.org, XS Devel server-de...@lists.laptop.org,Support Gangsters support-g...@laptop.org Subject: Re: gathering use cases Message-ID: CABPDnXn8Xz9BwRD_zy_PUr43CG=_rKXv8u6=zLjD0f+Juk=x...@mail.gmail.com Content-Type: text/plain; charset=ISO-8859-1 On 3 November 2012 07:46, Holth...@laptop.org wrote: Thanks much Sameer. Am includingsupport-g...@laptop.org to make sure we gather microdeployment and support volunteer feedback too. While our XS Community Edition experiment ain't quite ready for showtime/download yet, its white paper code repository are here: http://wiki.laptop.org/go/User:Holt/XS_Community_Edition The white paper that Adam is referring to can be found at https://docs.google.com/a/dhanapalan.com/document/pub?id=1dnhU2F6EntepVXTgN8QpkME8fZVUuPjcCoMUfAVKbcc I am the primary author, but it was written by canvassing feedback and opinions from schools and XS community members (including Jerry Vonau, George Hunt, Adam Holt and many others). The paper discusses several use cases and design goals, and outlines the underlying principles behind how we do things in OLPC Australia. There is some miscommunication being spread around that I'd like to clear up. Firstly, this is not a 'secret' project. The white paper explains our mission, and the code repo and issues trackers are open. Secondly, the intent is to create a flexible framework that can be adapted to suit local needs. This is*not*, as some keep asserting, a project to have the XS work only on the XO. The community XS will work on any x86 or ARM based hardware that works with Fedora. The white paper explains this in depth. Sridhar ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: gathering use cases
On 11/08/2012 05:01 PM, Jerry Vonau wrote: Hi Tony and all, On Thu, 2012-11-08 at 13:32 -0500, Tony Anderson wrote: Hi, Sridhar One of the important potential benefits from gathering use cases is to clarify the roles of the Community XS project and XS-0.7. Why is that an issue at at all? The Community XS (think XS-0.8 here) is going to address the current shortcomings of XS-0.7. This is my greatest concern. The ability of a server to deliver content is central. It is my understanding that your Community XS does not support a LAMP stack or Moodle. Please do not refer to the Community XS as XO-0.8 until there is a chance that it can deliver the essential capabilities of XS-0.7. My other concern is that a lot of very talented people are spending a lot of time solving a non-problem. Naturally, the network problem you mention is solved by connecting the XS-0.7 to that network as the WAN. Yours, Tony From my perspective, if you can afford to deploy a server with 2GB or more main memory and 500GB or more hard drive capacity, XS-0.7 serves well and easily supports adding or customizing services. The configuration shown in SF costs less than $350 and draws 18w (based on measurements taken by George Hunt). With a large hard drive, it is perfectly to keep Moodle even if it is not used. In my experience, the school server is installed in the school ready with XS-07 installed along with the available content. Barring hardware failure, the only interaction at the school is to turn the server on at the start of day and turn it off at the end. With modern computers, both functions require only a push on the power button. I'm not thinking about an XO here at all but on a server that you spec'd. Tell me how do you deploy the 0.7 version on a network where you don't have control over the DNS and have everything work? I have an idea on how to pull that off with avahi, I would like to hear about how you would implement that situation. Community XS serves a valuable role where only XOs are available or where power constraints are critical. It may also have a role in supporting special network requirements. Not to mention better hardware support for the latest hardware available and ARM support by using Fedora. You can get on board with the future and have a say in your pet interests but please don't get in the way of progress. Jerry The other XS-CE developer Tony On 11/08/2012 10:55 AM, devel-requ...@lists.laptop.org wrote: Message: 2 Date: Thu, 8 Nov 2012 13:57:12 +1100 From: Sridhar Dhanapalansrid...@laptop.org.au To: Holth...@laptop.org Cc: Devel's in the Detailsdevel@lists.laptop.org, Maryam Rehan El Bazmel...@mail.sfsu.edu,iaepi...@lists.sugarlabs.org, XS Devel server-de...@lists.laptop.org, Support Gangsters support-g...@laptop.org Subject: Re: gathering use cases Message-ID: CABPDnXn8Xz9BwRD_zy_PUr43CG=_rKXv8u6=zLjD0f+Juk=x...@mail.gmail.com Content-Type: text/plain; charset=ISO-8859-1 On 3 November 2012 07:46, Holth...@laptop.org wrote: Thanks much Sameer. Am includingsupport-g...@laptop.org to make sure we gather microdeployment and support volunteer feedback too. While our XS Community Edition experiment ain't quite ready for showtime/download yet, its white paper code repository are here: http://wiki.laptop.org/go/User:Holt/XS_Community_Edition The white paper that Adam is referring to can be found at https://docs.google.com/a/dhanapalan.com/document/pub?id=1dnhU2F6EntepVXTgN8QpkME8fZVUuPjcCoMUfAVKbcc I am the primary author, but it was written by canvassing feedback and opinions from schools and XS community members (including Jerry Vonau, George Hunt, Adam Holt and many others). The paper discusses several use cases and design goals, and outlines the underlying principles behind how we do things in OLPC Australia. There is some miscommunication being spread around that I'd like to clear up. Firstly, this is not a 'secret' project. The white paper explains our mission, and the code repo and issues trackers are open. Secondly, the intent is to create a flexible framework that can be adapted to suit local needs. This is*not*, as some keep asserting, a project to have the XS work only on the XO. The community XS will work on any x86 or ARM based hardware that works with Fedora. The white paper explains this in depth. Sridhar ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: gathering use cases
On Thu, 2012-11-08 at 17:10 -0500, Tony Anderson wrote: On 11/08/2012 05:01 PM, Jerry Vonau wrote: Hi Tony and all, On Thu, 2012-11-08 at 13:32 -0500, Tony Anderson wrote: Hi, Sridhar One of the important potential benefits from gathering use cases is to clarify the roles of the Community XS project and XS-0.7. Why is that an issue at at all? The Community XS (think XS-0.8 here) is going to address the current shortcomings of XS-0.7. This is my greatest concern. The ability of a server to deliver content is central. It is my understanding that your Community XS does not support a LAMP stack or Moodle. Please do not refer to the Community XS as XO-0.8 until there is a chance that it can deliver the essential capabilities of XS-0.7. My other concern is that a lot of very talented people are spending a lot of time solving a non-problem. That's very interesting LAMP is installed and George tells me Moodle is now working. We just didn't quite have the time prior to SF to debug the issue and I confident that the work done will become 0.8. Naturally, the network problem you mention is solved by connecting the XS-0.7 to that network as the WAN. How do you deal with the name resolution of 'schoolserver' without having the ability to alter the network dns server's entries or have the server called something other than 'schoolserver' to avoid a clash with an pre-existing server on the target network? Jerry Yours, Tony From my perspective, if you can afford to deploy a server with 2GB or more main memory and 500GB or more hard drive capacity, XS-0.7 serves well and easily supports adding or customizing services. The configuration shown in SF costs less than $350 and draws 18w (based on measurements taken by George Hunt). With a large hard drive, it is perfectly to keep Moodle even if it is not used. In my experience, the school server is installed in the school ready with XS-07 installed along with the available content. Barring hardware failure, the only interaction at the school is to turn the server on at the start of day and turn it off at the end. With modern computers, both functions require only a push on the power button. I'm not thinking about an XO here at all but on a server that you spec'd. Tell me how do you deploy the 0.7 version on a network where you don't have control over the DNS and have everything work? I have an idea on how to pull that off with avahi, I would like to hear about how you would implement that situation. Community XS serves a valuable role where only XOs are available or where power constraints are critical. It may also have a role in supporting special network requirements. Not to mention better hardware support for the latest hardware available and ARM support by using Fedora. You can get on board with the future and have a say in your pet interests but please don't get in the way of progress. Jerry The other XS-CE developer Tony On 11/08/2012 10:55 AM, devel-requ...@lists.laptop.org wrote: Message: 2 Date: Thu, 8 Nov 2012 13:57:12 +1100 From: Sridhar Dhanapalansrid...@laptop.org.au To: Holth...@laptop.org Cc: Devel's in the Detailsdevel@lists.laptop.org, Maryam Rehan El Bazmel...@mail.sfsu.edu, iaepi...@lists.sugarlabs.org, XS Devel server-de...@lists.laptop.org,Support Gangsters support-g...@laptop.org Subject: Re: gathering use cases Message-ID: CABPDnXn8Xz9BwRD_zy_PUr43CG=_rKXv8u6=zLjD0f+Juk=x...@mail.gmail.com Content-Type: text/plain; charset=ISO-8859-1 On 3 November 2012 07:46, Holth...@laptop.org wrote: Thanks much Sameer. Am includingsupport-g...@laptop.org to make sure we gather microdeployment and support volunteer feedback too. While our XS Community Edition experiment ain't quite ready for showtime/download yet, its white paper code repository are here: http://wiki.laptop.org/go/User:Holt/XS_Community_Edition The white paper that Adam is referring to can be found at https://docs.google.com/a/dhanapalan.com/document/pub?id=1dnhU2F6EntepVXTgN8QpkME8fZVUuPjcCoMUfAVKbcc I am the primary author, but it was written by canvassing feedback and opinions from schools and XS community members (including Jerry Vonau, George Hunt, Adam Holt and many others). The paper discusses several use cases and design goals, and outlines the underlying principles behind how we do things in OLPC Australia. There is some miscommunication being spread around that I'd like to clear up. Firstly, this is not a 'secret' project. The white paper explains our mission, and the code repo and issues trackers are open. Secondly, the intent is to create a flexible framework that can be adapted to suit local needs. This is*not*, as some keep asserting, a project to have the XS work only on the XO. The community XS will work on any x86 or ARM based hardware that works with Fedora
Re: gathering use cases
Hi, If you are re-implementing XS-0.7, one beneficial project would be to rebase it on Fedora so that it will run on Arm systems. This is currently a limitation of CentOS. The design of XS separates the LAN and WAN networks. The XOs connect to schoolserver via the LAN network. The school server has a domain name relative to the WAN network appropriate to that network. This is set up by the netsetup.sh script. XS-0.7 provides normal firewall protection to the XOs. OLE Nepal has added Dan's Guardian to provide specific protection against access to objectionable sites. Tony On 11/08/2012 05:25 PM, Jerry Vonau wrote: On Thu, 2012-11-08 at 17:10 -0500, Tony Anderson wrote: On 11/08/2012 05:01 PM, Jerry Vonau wrote: Hi Tony and all, On Thu, 2012-11-08 at 13:32 -0500, Tony Anderson wrote: Hi, Sridhar One of the important potential benefits from gathering use cases is to clarify the roles of the Community XS project and XS-0.7. Why is that an issue at at all? The Community XS (think XS-0.8 here) is going to address the current shortcomings of XS-0.7. This is my greatest concern. The ability of a server to deliver content is central. It is my understanding that your Community XS does not support a LAMP stack or Moodle. Please do not refer to the Community XS as XO-0.8 until there is a chance that it can deliver the essential capabilities of XS-0.7. My other concern is that a lot of very talented people are spending a lot of time solving a non-problem. That's very interesting LAMP is installed and George tells me Moodle is now working. We just didn't quite have the time prior to SF to debug the issue and I confident that the work done will become 0.8. Naturally, the network problem you mention is solved by connecting the XS-0.7 to that network as the WAN. How do you deal with the name resolution of 'schoolserver' without having the ability to alter the network dns server's entries or have the server called something other than 'schoolserver' to avoid a clash with an pre-existing server on the target network? Jerry Yours, Tony From my perspective, if you can afford to deploy a server with 2GB or more main memory and 500GB or more hard drive capacity, XS-0.7 serves well and easily supports adding or customizing services. The configuration shown in SF costs less than $350 and draws 18w (based on measurements taken by George Hunt). With a large hard drive, it is perfectly to keep Moodle even if it is not used. In my experience, the school server is installed in the school ready with XS-07 installed along with the available content. Barring hardware failure, the only interaction at the school is to turn the server on at the start of day and turn it off at the end. With modern computers, both functions require only a push on the power button. I'm not thinking about an XO here at all but on a server that you spec'd. Tell me how do you deploy the 0.7 version on a network where you don't have control over the DNS and have everything work? I have an idea on how to pull that off with avahi, I would like to hear about how you would implement that situation. Community XS serves a valuable role where only XOs are available or where power constraints are critical. It may also have a role in supporting special network requirements. Not to mention better hardware support for the latest hardware available and ARM support by using Fedora. You can get on board with the future and have a say in your pet interests but please don't get in the way of progress. Jerry The other XS-CE developer Tony On 11/08/2012 10:55 AM, devel-requ...@lists.laptop.org wrote: Message: 2 Date: Thu, 8 Nov 2012 13:57:12 +1100 From: Sridhar Dhanapalansrid...@laptop.org.au To: Holth...@laptop.org Cc: Devel's in the Detailsdevel@lists.laptop.org, Maryam Rehan El Bazmel...@mail.sfsu.edu,iaepi...@lists.sugarlabs.org, XS Devel server-de...@lists.laptop.org, Support Gangsters support-g...@laptop.org Subject: Re: gathering use cases Message-ID: CABPDnXn8Xz9BwRD_zy_PUr43CG=_rKXv8u6=zLjD0f+Juk=x...@mail.gmail.com Content-Type: text/plain; charset=ISO-8859-1 On 3 November 2012 07:46, Holth...@laptop.org wrote: Thanks much Sameer. Am includingsupport-g...@laptop.org to make sure we gather microdeployment and support volunteer feedback too. While our XS Community Edition experiment ain't quite ready for showtime/download yet, its white paper code repository are here: http://wiki.laptop.org/go/User:Holt/XS_Community_Edition The white paper that Adam is referring to can be found at https://docs.google.com/a/dhanapalan.com/document/pub?id=1dnhU2F6EntepVXTgN8QpkME8fZVUuPjcCoMUfAVKbcc I am the primary author, but it was written by canvassing feedback and opinions from schools and XS community members (including Jerry Vonau, George Hunt, Adam Holt and many others). The paper discusses several use cases and design goals, and outlines the underlying principles behind how
Re: gathering use cases
On 9 November 2012 09:10, Tony Anderson t...@olenepal.org wrote: This is my greatest concern. The ability of a server to deliver content is central. It is my understanding that your Community XS does not support a LAMP stack or Moodle. Please do not refer to the Community XS as XO-0.8 until there is a chance that it can deliver the essential capabilities of XS-0.7. My other concern is that a lot of very talented people are spending a lot of time solving a non-problem. Naturally, the network problem you mention is solved by connecting the XS-0.7 to that network as the WAN. Tony, It appears that no matter what we say, you are cemented in some strange notions about the community XS: * that it is intended to run only on XOs * that it cannot (and will not) serve content * that your personal desires from an XS are shared by every other deployment in the world I think we've been quite clear about what we're intending to achieve. If you're going to criticise, at least be civil enough to do so based on the facts: * this is a flexible design, built on Fedora * it will run anywhere where x86 or ARM Fedora will work * it can be installed on top of an existing Fedora installation using 'yum groupinstall xsce' * being designed in this way provides extreme flexibility for deployments * all current features of the XS (Moodle, etc.) will be ported, but will be optional * installation and configuration will be easy, but sysadmins will be able to treat it like any Fedora installation We've tried hard to be inclusive and constructive. I ask you to do the same. Regards, Sridhar ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: gathering use cases
Hi, Sridhar Thanks for the clarification. I guess I was mislead by statements such as: The platform for the One Network server is an ARMv7-based XO, running the One Education OS (based on OLPC OS). This makes development, and deployment and support far simpler than a standalone distribution. The OS can be extended with server capabilities using a bootable USB Customisation Stick (offline) or yum. Please accept my apology if any statement I have made seemed uncivil, that was certainly not my intention. Communicating by email in certainly much more hazardous in this regard than face-to-face. Just as the deployments you are supporting have specific and urgent needs, so do the ones I am working with. I don't believe either of us is pursuing personal desires. We certainly can easily differ on which is the appropriate technical approach to solving the problems of a deployment. I really appreciate this specification: * this is a flexible design, built on Fedora * it will run anywhere where x86 or ARM Fedora will work * it can be installed on top of an existing Fedora installation using 'yum groupinstall xsce' * being designed in this way provides extreme flexibility for deployments * all current features of the XS (Moodle, etc.) will be ported, but will be optional * installation and configuration will be easy, but sysadmins will be able to treat it like any Fedora installation There is clearly a great deal to be gained by a community taking responsibility for the ongoing development and maintenance of the school server as neither Daniel Drake nor Martin Langhoff are likely to have adequate time for this in the foreseeable future. Tony Tony On 11/08/2012 05:59 PM, Sridhar Dhanapalan wrote: On 9 November 2012 09:10, Tony Anderson t...@olenepal.org wrote: This is my greatest concern. The ability of a server to deliver content is central. It is my understanding that your Community XS does not support a LAMP stack or Moodle. Please do not refer to the Community XS as XO-0.8 until there is a chance that it can deliver the essential capabilities of XS-0.7. My other concern is that a lot of very talented people are spending a lot of time solving a non-problem. Naturally, the network problem you mention is solved by connecting the XS-0.7 to that network as the WAN. Tony, It appears that no matter what we say, you are cemented in some strange notions about the community XS: * that it is intended to run only on XOs * that it cannot (and will not) serve content * that your personal desires from an XS are shared by every other deployment in the world I think we've been quite clear about what we're intending to achieve. If you're going to criticise, at least be civil enough to do so based on the facts: * this is a flexible design, built on Fedora * it will run anywhere where x86 or ARM Fedora will work * it can be installed on top of an existing Fedora installation using 'yum groupinstall xsce' * being designed in this way provides extreme flexibility for deployments * all current features of the XS (Moodle, etc.) will be ported, but will be optional * installation and configuration will be easy, but sysadmins will be able to treat it like any Fedora installation We've tried hard to be inclusive and constructive. I ask you to do the same. Regards, Sridhar ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: gathering use cases
On 9 November 2012 10:19, Tony Anderson t...@olenepal.org wrote: Hi, Sridhar Thanks for the clarification. I guess I was mislead by statements such as: The platform for the One Network server is an ARMv7-based XO, running the One Education OS (based on OLPC OS). This makes development, and deployment and support far simpler than a standalone distribution. The OS can be extended with server capabilities using a bootable USB Customisation Stick (offline) or yum. Please accept my apology if any statement I have made seemed uncivil, that was certainly not my intention. Communicating by email in certainly much more hazardous in this regard than face-to-face. Thank you, Tony. I was quite careful to take your needs into account when I wrote the design doc, so I had trouble understanding your opposition to the idea. Maybe we can make the doc clearer somehow? I structured it as: 1. context 2. Community XS design 3. One Network server The Community XS design itself is flexible enough to handle a variety of different deployments' needs. One Network server is merely one configuration of the Community XS, mentioned as an example of what can be done. Just as the deployments you are supporting have specific and urgent needs, so do the ones I am working with. I don't believe either of us is pursuing personal desires. We certainly can easily differ on which is the appropriate technical approach to solving the problems of a deployment. I think we generally want the same thing in the end. I'm happy to continue the conversation to improve the design and implementation. I sincerely believe that this design can accommodate your needs. I really appreciate this specification: * this is a flexible design, built on Fedora * it will run anywhere where x86 or ARM Fedora will work * it can be installed on top of an existing Fedora installation using 'yum groupinstall xsce' * being designed in this way provides extreme flexibility for deployments * all current features of the XS (Moodle, etc.) will be ported, but will be optional * installation and configuration will be easy, but sysadmins will be able to treat it like any Fedora installation Awesome. I hope this also satisfies your desire to have it rebased to Fedora to work on ARM systems. That's a key goal of this project, while maintaining compatibility with x86. There is clearly a great deal to be gained by a community taking responsibility for the ongoing development and maintenance of the school server as neither Daniel Drake nor Martin Langhoff are likely to have adequate time for this in the foreseeable future. Indeed. A motivating factor was to take some of the load off some of these prolific people and spread it out to the community in a sustainable way. Cheers, Sridhar ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: gathering use cases
To clarify, by use case, we mean a way to describe the interaction of people with a system. In this case, it may be how a child interacts with the XO, and the XS (via XO) or how a teacher may interact with the XS (via XO or otherwise) by using Moodle. As you may have noticed, pretty much every response on this thread focuses on the system, with the assumption that the user end of it is understood well. I am not so sure. Take a look at http://www.gatherspace.com/static/use_case_example.html to see if it helps you understand the idea behind use cases and user stories. Here's another example of how use cases can vary by being very high level (which is what we are aiming for) and can be user centric or system centric. http://www.agilemodeling.com/artifacts/systemUseCase.htm Our focus is user centric, in a way where we would like to describe the actors (children, parents, teachers, admin, etc) and their actions (access class information, read books, send email) without the XS as the focus. Networking topology, storage, UI, LMS, DNS, etc. should flow from the storytelling exercise. We are a bunch of technologists and it is easy to get carried away by designing from the tech and not the user end. Sometimes that misses the mark. We may build it and they may not come. cheers, Sameer On Thu, Nov 8, 2012 at 4:07 PM, Sridhar Dhanapalan srid...@laptop.org.auwrote: On 9 November 2012 10:19, Tony Anderson t...@olenepal.org wrote: Hi, Sridhar Thanks for the clarification. I guess I was mislead by statements such as: The platform for the One Network server is an ARMv7-based XO, running the One Education OS (based on OLPC OS). This makes development, and deployment and support far simpler than a standalone distribution. The OS can be extended with server capabilities using a bootable USB Customisation Stick (offline) or yum. Please accept my apology if any statement I have made seemed uncivil, that was certainly not my intention. Communicating by email in certainly much more hazardous in this regard than face-to-face. Thank you, Tony. I was quite careful to take your needs into account when I wrote the design doc, so I had trouble understanding your opposition to the idea. Maybe we can make the doc clearer somehow? I structured it as: 1. context 2. Community XS design 3. One Network server The Community XS design itself is flexible enough to handle a variety of different deployments' needs. One Network server is merely one configuration of the Community XS, mentioned as an example of what can be done. Just as the deployments you are supporting have specific and urgent needs, so do the ones I am working with. I don't believe either of us is pursuing personal desires. We certainly can easily differ on which is the appropriate technical approach to solving the problems of a deployment. I think we generally want the same thing in the end. I'm happy to continue the conversation to improve the design and implementation. I sincerely believe that this design can accommodate your needs. I really appreciate this specification: * this is a flexible design, built on Fedora * it will run anywhere where x86 or ARM Fedora will work * it can be installed on top of an existing Fedora installation using 'yum groupinstall xsce' * being designed in this way provides extreme flexibility for deployments * all current features of the XS (Moodle, etc.) will be ported, but will be optional * installation and configuration will be easy, but sysadmins will be able to treat it like any Fedora installation Awesome. I hope this also satisfies your desire to have it rebased to Fedora to work on ARM systems. That's a key goal of this project, while maintaining compatibility with x86. There is clearly a great deal to be gained by a community taking responsibility for the ongoing development and maintenance of the school server as neither Daniel Drake nor Martin Langhoff are likely to have adequate time for this in the foreseeable future. Indeed. A motivating factor was to take some of the load off some of these prolific people and spread it out to the community in a sustainable way. Cheers, Sridhar ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Server-devel] gathering use cases
Sameer, Thanks for doing this work. Here are some thoughts In the deployment I have been managing we have used the server in two ways: 1. As a jabber server only. We have about 140 XO-1s, and without the jabber server, the cross-talk slowed down the devices dramatically when more than about 30 were running simultaneously. 2. As a wireless network access point. This was especially important in the dark times before our school reconfigured our wireless network to work properly with the XOs. When we began, Moodle was less important to us, but now that I have been Moodle with my 7th and 8th grade students (who are not using the XOs), I can see the benefit of using it with an XO deployment. I hope this is helpful. Thanks again. Gerald On Sat, Nov 3, 2012 at 12:07 AM, Chris Leonard cjlhomeaddr...@gmail.comwrote: Sameer, You should talk to SomosAzucar and Aleksey Lim about the use case for Sugar Network being developed for a pilot in LimaNorte. cjl On Fri, Nov 2, 2012 at 4:15 PM, Sameer Verma sve...@sfsu.edu wrote: There are several use cases that may or may not get addressed when designing a particular software stack to address a requirement. The XS 0.7 is designed to be a single image install and comes with Moodle. Given that I work with Moodle everyday, I see the pros and cons of it being central on the XS. I am in fact fairly happy with its current design, but also realize that it was built for a specific use case or three that OLPC needed at the time. There is also an effort (currently dubbed XS Community Edition) that is attempting to address certain other use cases where Moodle and other services could possibly become optional. We saw this at the OLPC SF Community Summit. I hope it will grow up to be the next XS (but that's another thread). My concern is that perhaps, if we don't do our homework right, we will once again build something that will fail to address a use case or two. Can one design address all use cases? Maybe not. But it's good to know what those use cases are. To this end, I would like to collect data on different possible use cases from all kinds of deployments. I have a student (cc'd) who is working on this project currently. She will gather data from various deployments (suitcase, boutique, MoE etc) as much as possible (with the cooperation of projects, of course) and write a report on what we see out there. We'll gladly make the report available once it is done. Is this useful? What should the scope be? Initially I had thought of the server side, but that may be limiting. What should we gather? Location, school, size, personnel, skills, electricity, Internet access, language, sugar version, ... Feedback? cheers, Sameer -- Sameer Verma, Ph.D. Professor, Information Systems San Francisco State University http://verma.sfsu.edu/ http://commons.sfsu.edu/ http://olpcsf.org/ http://olpcjamaica.org.jm/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Server-devel mailing list server-de...@lists.laptop.org http://lists.laptop.org/listinfo/server-devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
gathering use cases
There are several use cases that may or may not get addressed when designing a particular software stack to address a requirement. The XS 0.7 is designed to be a single image install and comes with Moodle. Given that I work with Moodle everyday, I see the pros and cons of it being central on the XS. I am in fact fairly happy with its current design, but also realize that it was built for a specific use case or three that OLPC needed at the time. There is also an effort (currently dubbed XS Community Edition) that is attempting to address certain other use cases where Moodle and other services could possibly become optional. We saw this at the OLPC SF Community Summit. I hope it will grow up to be the next XS (but that's another thread). My concern is that perhaps, if we don't do our homework right, we will once again build something that will fail to address a use case or two. Can one design address all use cases? Maybe not. But it's good to know what those use cases are. To this end, I would like to collect data on different possible use cases from all kinds of deployments. I have a student (cc'd) who is working on this project currently. She will gather data from various deployments (suitcase, boutique, MoE etc) as much as possible (with the cooperation of projects, of course) and write a report on what we see out there. We'll gladly make the report available once it is done. Is this useful? What should the scope be? Initially I had thought of the server side, but that may be limiting. What should we gather? Location, school, size, personnel, skills, electricity, Internet access, language, sugar version, ... Feedback? cheers, Sameer -- Sameer Verma, Ph.D. Professor, Information Systems San Francisco State University http://verma.sfsu.edu/ http://commons.sfsu.edu/ http://olpcsf.org/ http://olpcjamaica.org.jm/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: gathering use cases
Thanks much Sameer. Am including support-g...@laptop.org to make sure we gather microdeployment and support volunteer feedback too. While our XS Community Edition experiment ain't quite ready for showtime/download yet, its white paper code repository are here: http://wiki.laptop.org/go/User:Holt/XS_Community_Edition Feedback will be especially interesting this week as many of us will reconvene outside Toronto Nov 10-18 to advance this work, with nightly Skype calls then for those with a strong interest in contributing. Daniel Drake helped George a lot in SF, but note SF Summit presenter (one of XSCE's lead developers, George Hunt) is part of the 3 million households / 6-10 million people lacking electricity around NY/NJ due to the Hurricane Sandy, so he won't be able to respond immediately -- please don't let that stop you from carefully reviewing his/our work -- responses will certainly be forthcoming later in the week! On 11/2/2012 4:15 PM, Sameer Verma wrote: There are several use cases that may or may not get addressed when designing a particular software stack to address a requirement. The XS 0.7 is designed to be a single image install and comes with Moodle. Given that I work with Moodle everyday, I see the pros and cons of it being central on the XS. I am in fact fairly happy with its current design, but also realize that it was built for a specific use case or three that OLPC needed at the time. There is also an effort (currently dubbed XS Community Edition) that is attempting to address certain other use cases where Moodle and other services could possibly become optional. We saw this at the OLPC SF Community Summit. I hope it will grow up to be the next XS (but that's another thread). My concern is that perhaps, if we don't do our homework right, we will once again build something that will fail to address a use case or two. Can one design address all use cases? Maybe not. But it's good to know what those use cases are. To this end, I would like to collect data on different possible use cases from all kinds of deployments. I have a student (cc'd) who is working on this project currently. She will gather data from various deployments (suitcase, boutique, MoE etc) as much as possible (with the cooperation of projects, of course) and write a report on what we see out there. We'll gladly make the report available once it is done. Is this useful? What should the scope be? Initially I had thought of the server side, but that may be limiting. What should we gather? Location, school, size, personnel, skills, electricity, Internet access, language, sugar version, ... Feedback? cheers, Sameer -- Sameer Verma, Ph.D. Professor, Information Systems San Francisco State University http://verma.sfsu.edu/ http://commons.sfsu.edu/ http://olpcsf.org/ http://olpcjamaica.org.jm/ -- Help kids everywhere map their world, at http://olpcMAP.net ! ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: gathering use cases
Sameer, You should talk to SomosAzucar and Aleksey Lim about the use case for Sugar Network being developed for a pilot in LimaNorte. cjl On Fri, Nov 2, 2012 at 4:15 PM, Sameer Verma sve...@sfsu.edu wrote: There are several use cases that may or may not get addressed when designing a particular software stack to address a requirement. The XS 0.7 is designed to be a single image install and comes with Moodle. Given that I work with Moodle everyday, I see the pros and cons of it being central on the XS. I am in fact fairly happy with its current design, but also realize that it was built for a specific use case or three that OLPC needed at the time. There is also an effort (currently dubbed XS Community Edition) that is attempting to address certain other use cases where Moodle and other services could possibly become optional. We saw this at the OLPC SF Community Summit. I hope it will grow up to be the next XS (but that's another thread). My concern is that perhaps, if we don't do our homework right, we will once again build something that will fail to address a use case or two. Can one design address all use cases? Maybe not. But it's good to know what those use cases are. To this end, I would like to collect data on different possible use cases from all kinds of deployments. I have a student (cc'd) who is working on this project currently. She will gather data from various deployments (suitcase, boutique, MoE etc) as much as possible (with the cooperation of projects, of course) and write a report on what we see out there. We'll gladly make the report available once it is done. Is this useful? What should the scope be? Initially I had thought of the server side, but that may be limiting. What should we gather? Location, school, size, personnel, skills, electricity, Internet access, language, sugar version, ... Feedback? cheers, Sameer -- Sameer Verma, Ph.D. Professor, Information Systems San Francisco State University http://verma.sfsu.edu/ http://commons.sfsu.edu/ http://olpcsf.org/ http://olpcjamaica.org.jm/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
[Server-devel] gathering use cases
There are several use cases that may or may not get addressed when designing a particular software stack to address a requirement. The XS 0.7 is designed to be a single image install and comes with Moodle. Given that I work with Moodle everyday, I see the pros and cons of it being central on the XS. I am in fact fairly happy with its current design, but also realize that it was built for a specific use case or three that OLPC needed at the time. There is also an effort (currently dubbed XS Community Edition) that is attempting to address certain other use cases where Moodle and other services could possibly become optional. We saw this at the OLPC SF Community Summit. I hope it will grow up to be the next XS (but that's another thread). My concern is that perhaps, if we don't do our homework right, we will once again build something that will fail to address a use case or two. Can one design address all use cases? Maybe not. But it's good to know what those use cases are. To this end, I would like to collect data on different possible use cases from all kinds of deployments. I have a student (cc'd) who is working on this project currently. She will gather data from various deployments (suitcase, boutique, MoE etc) as much as possible (with the cooperation of projects, of course) and write a report on what we see out there. We'll gladly make the report available once it is done. Is this useful? What should the scope be? Initially I had thought of the server side, but that may be limiting. What should we gather? Location, school, size, personnel, skills, electricity, Internet access, language, sugar version, ... Feedback? cheers, Sameer -- Sameer Verma, Ph.D. Professor, Information Systems San Francisco State University http://verma.sfsu.edu/ http://commons.sfsu.edu/ http://olpcsf.org/ http://olpcjamaica.org.jm/ ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] gathering use cases
Sameer, You should talk to SomosAzucar and Aleksey Lim about the use case for Sugar Network being developed for a pilot in LimaNorte. cjl On Fri, Nov 2, 2012 at 4:15 PM, Sameer Verma sve...@sfsu.edu wrote: There are several use cases that may or may not get addressed when designing a particular software stack to address a requirement. The XS 0.7 is designed to be a single image install and comes with Moodle. Given that I work with Moodle everyday, I see the pros and cons of it being central on the XS. I am in fact fairly happy with its current design, but also realize that it was built for a specific use case or three that OLPC needed at the time. There is also an effort (currently dubbed XS Community Edition) that is attempting to address certain other use cases where Moodle and other services could possibly become optional. We saw this at the OLPC SF Community Summit. I hope it will grow up to be the next XS (but that's another thread). My concern is that perhaps, if we don't do our homework right, we will once again build something that will fail to address a use case or two. Can one design address all use cases? Maybe not. But it's good to know what those use cases are. To this end, I would like to collect data on different possible use cases from all kinds of deployments. I have a student (cc'd) who is working on this project currently. She will gather data from various deployments (suitcase, boutique, MoE etc) as much as possible (with the cooperation of projects, of course) and write a report on what we see out there. We'll gladly make the report available once it is done. Is this useful? What should the scope be? Initially I had thought of the server side, but that may be limiting. What should we gather? Location, school, size, personnel, skills, electricity, Internet access, language, sugar version, ... Feedback? cheers, Sameer -- Sameer Verma, Ph.D. Professor, Information Systems San Francisco State University http://verma.sfsu.edu/ http://commons.sfsu.edu/ http://olpcsf.org/ http://olpcjamaica.org.jm/ ___ Devel mailing list de...@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel