Same applies. 1 subscription per machine. Having a Satellite server does not remove the need to have a basic or standard subscription for each one as well....
Red Hat offer support subscriptions per *server*, not per user. And i'd get back to class pretty quickly - satellite server does not take away an obligation to buy support for your servers if you want Red Hat to respond to telephone calls, or to supply updates, on them. And Red Hat don't sell licenses at all - only support and update subscriptions.
Satellite server just acts as a cache to stop every machine pulling updates from Red Hat individually, plus the ability to do PIXie boots, to clone systems and, if needed, to be able to distribute updates without having the machines connected to the Internet - if that's what you want to do.
The differences between MS's business model and that for open source aren't that far apart - both provide functionality "as is" with no warranty nor obligation for "fitness for purpose" (if in doubt, read the EULA that came with Windows or Office). Red Hat charge for support and updates, and you can use support and install any supported version (even new ones as they arrive) while you have a valid subscription active; at the end of the subscription term, you're still free to use the software whether you renew your subcription or not (just don't expect support or updates any more if you don't pay). Microsoft sell a license to use a single major version, but expect you to pay again when a new version comes out, or to subscribe to Software Assurance to allow you to use new versions as they emerge. Very few customers tend to buy Software Assurance atm.
I always wonder why Microsoft don't sell subscriptions in the same way Red Hat do...
The term "The Network is the System" dates back to the NaC group at DEC (VP: Julius Marcus) in 1977. "The Network is the computer" little more than a late developing bastard child of Digital's foresight.
The #1 thing that Linux gives to users (that they value highly) is choice of hardware and software on commodity (read: Intel or AMD) platforms. From a commercial subscription market share perspective, it's just about game over; Red Hat is up there in 90% plus land with SUSE collecting almost all of the few remaining crumbs.
To date, Novell is stronger on PR (Google search volumes on "SUSE" are almost at Red Hat levels) but are struggling really badly to monetise this.
The best thing that Oracle could do would be to support both equally. As stupid as it sounds, everyones best interest is served by having two (or more) Linux distributions duking it out.
Tell me if my facts are wrong, but I reckon Oracle have an installed base of 9,000 companies using Oracle on Linux. MySQL deloy at the rate of 40,000 per *day*.
To date, all the big DB vendors try clutching straws; Oracle themselves define market share as "license revenue", make out that they're slugging it out for dominance with DB2, and say that only 10% of MySQL deployments are in mission-critical apps. Follow the numbers, and you still see MySQL lapping Oracle's installed base in the "mission critical" space every 3 days.
Hence MySQL are already everywhere and starting to do a traditional market discontinuity. Virtually all the nice open source tools these days go MySQL first (PHP, Ruby on Rails among them). Oracle's defence is either to move to services... or to bully their way into the low end. I wonder whether they're really trying to buy MySQL.
... before they consumated their relationship with IBM (5 initiatives on offer, including one called Pink - the domain of then Apple employee and ex-VMS engineer Roger Heinen). KO was a big fan of the Apple Desktop Bus and it's simple connectivity, but vetoed Digital Semiconductor providing the chips. Or maybe that was Jack Shields in the Executive Commitee, who drove everything vertically integrated to go eat IBM's lunch by 2007...
Fokelore in DEC at the time anyway. How much of it was true may be a different story.
I thought it was very lightweight as an article and had too many rose colo(u)red glasses around. Red Hat seems to be a Novell obsession (the Novell folks here gave me an eight page Red Hat vs Novell competitive paper, full of speeds and feeds, stuff about YAST, some very subjective positioning on partnering and heritage). So, trying to pick a fight with Red Hat when they could really be more productive focussing on the untapped 97% of the market.
Most people don't care; Novell have a competent Linux distribution that's a worthy substitute if Red Hat ever pi**es people off. Their Google adwords search volumes are comparable to Red Hats, but they have a real problem monetising the demand; Red Hat outship them in subscription volumes by a factor of 9:1. With that goes share of mind in the ISV community, and if you have the developers, you have the market.
The unfortunate side effect of the market performance is that Novell slide SUSE underneath Netware and call the result "OES", then start to muddle their "open systems" and proprietary product numbers when they go full flight on their PR and stock market statements. Their recent very congratulatory PR got reproduced everywhere (a key Novell strength), but I still wince at three particular facts behind the numbers:
1) If you took out the effect of their one off payment from Microsoft, their year on year, like for like revenue as a company appears to have dropped 35%.
2) If you remove "OES" from their numbers (it wasn't shipping last year), their traditional Linux business appears to have grown from just north of $14m to $15m in one year.
3) The often cited UK National Health deal used to be for Netware, Identity Management and Zenworks. This got renewed for three years (with the Netware portion repurposed as OES, whether or not SUSE is physically used), with a contract value of $40m over three years. If it were Red Hat, they'd spread the subscriptions evenly over the full contract term; Novell in this case elected to count $20m as income inside the first quarter.
Are any of the above wrong? I'd be delighted to be told if this were the case.
Novell's SUSE products are excellent and given the right focus, they could monetise a lot more of the demand coming to their door. They have better pricing terms for high performance computing, some good key open source focus (with JBoss and MySQL for starters), and have relationships with many customers that would be open to inplement open source infrastructures. The main focus missing is (IMHO)... developers, developers and developers. Not just in ISVs, but also in most large end user companies - people invisible to most CIO's, and to most hardware manufacturers also.
As someone who normally enjoys reading The Register, this sort of article is a stain on their hitherto excellent site. Tantamount to peeing in their own soup. The vendetta between the Register and Wikipedia seems to be getting a life of it's own, at odds with the quality of the rest of their respective works. I hope they both snap out of it.
Some of the OS numbers used to get bent out of shape by people front ending Windows machines with Linux based cache or content distribution networks (Akamai being a notable one). Which is why MS used to show up as running Linux for www.microsoft.com. The methodology has probably improved since.
Maybe it's different in the USA, but here cells are just too big to meaningfully equate cell transitions into traffic movements on physical roads.
Motorola did a lot of work with UK company Trafficmaster on their Smartnav product, where most of the traffic data comes in realtime from roadside cameras and sensors on Motorway bridges covering 9,000 miles of UK roads. When you push the button on the unit, the mobile phone rings into a call centre with the current GPS position encoded in the Caller ID of the phone call (3 bytes), you tell them the destination, and they send the turn-by-turn instructions back into the unit. If a jam shows up on your route before you get to your destination, the central servers ring the units and offer (with an estimate of the delay if you do nothing), then download, a route around the jam (if indeed one is available). Works really well.
Getting a GPS signal communicated back and forth in real time is the only way this hack will work - at least until the cell sizes get to 3G pico sizes.
There have been experiments with floating vehicle sensors, but most companies that say they use these for real time data collection do so as a PR stunt - the number of vehicles that need to be equipped and the economics of getting the data back in real time don't make business sense today, from either an accuracy or cost standpoint. Most instead download history at the end of the day, and just pick up details of roadworks and accidents from "journalistic" sources - whether there's a jam present or not.
Until we get GPS in most handsets (and if operators allow the caller IDs to be sent through with GPS data on board, but the call terminated at either end without any money changing hands), the use of mobile phones for spotting jams, or indeed navigating around them, will be very limited.
I got told a couple of months back that Oracle had an installed base of 9,000 installations atop Linux worldwide while MySQL were increasing their footprint by 40,000 installations per day.
I just wonder if Oracle were doing to get Innodb staff to write the interface code so MySQL could be the front end for an Oracle (or Oracle RAC) based backend. Or maybe it's just a smart move when MySQL have started selling their MySQL Cluster product (albeit memory based) for circa $4000 per node.
I was on the same table as Alan Cox last night. He said his one of his MBA projects was talking to small businesses in South Wales about (Desktop) Linux adoption. Said it gave some useful insights... including the need for Sage on Linux (most popular UK SME accounting suite).
Fascinating to listen to on a wide range of subjects:-)
See ActiveGrid - provides a production environment that can scale an app over a Grid of Application Servers each running the LAMP stack. If the middleware over the top is reliable, you've got a largely fault tolerant and ever scalable setup...
Just checked the UK Patents Office Site, and Vista is registered for many things, including a few software related products - by US and one German companies and trademarked across the EEC in various market segments. Vis:
EPICOR SOFTWARE CORPORATION
195 Technology Drive, Irvine, California, United States of America, 92618
ISOGON CORPORATION
330 Seventh Avenue, New York, United States of America, 1000
Gambro Inc
10810 West Collins Avenue, Lakewood, Colorado, United States of America, 80215-4498
Digital had an extension to their VAX VTX Videotex product in the 1980's called Vista. I remember it well because we got sued for using the name in the UK; it was a trademark here of a software company selling software for the printing/graphic arts industry based in London.
I'm still amazed that they announced the name without doing a simple search first.
Unfortunately our sales department tried selling it to the insurance industry, which takes way too long to purchase technology (as opposed to, oh, say, a stock broker, who would want to follow every stock he's got his clients invested in.. go figure).
And yet if my stats are correct (I work for one of the largest IT resellers in Europe), the Insurance Industry are the #1 early adopters for virtualisation software. I think something like 18 of our top 20 VMware customers are all Insurance or Financial Services companies;-}
Ian W.
In 1983, I sat in a conference room in a meeting with Microsoft (at DEC), and Scott Oki (Microsoft VP Europe at the time) was tapping all the meeting notes into a Tandy m100 silver tablet.
Now that Microsoft have got hardware contract manufacturing expertise in house (courtesy of X-Box), there's little beyond moving Windows + Windows Update to a subscription model (like Red Hat do for Linux) to stop them offering quite a compelling $100 machine.
According to Netcraft stats, the market share of Apache still outpaces IIS by something like 70:30, even with the host population growing nicely.
About the only area where MS have any meaningful share is on what Netcraft consider to be an analogue of "ecommerce sites" - measured in terms of those with a valid SSL certificate. Given MS's track record - and indeed the actual text of any security article cited by their own campaign (clue: not the headline!), some pretty significant folks are going to have to learn the hard way.
Let's face it - MS are doing all of us a favour. They're sullying their own brand reputation with the poor quality execution of this campaign, and that's even before their products bite companies where they hurt.
As a Linux advocate, I hope they keep the campaign running. Meanwhile, why doesn't someone ask Mike at Netcraft to publish the OS shares - I know he samples them.
They only need do what Red Hat does with Linux - give Windows away for free but charge for WindowsUpdate. And add telephone support as an option too. I don't understand what's taking them so long to work this out!
I seem to recall that the performance of VMS on small memory systems progressed leaps and bounds when a few hundred people at Spit Brook (VMS Engineering) got shiney new VAXstation 2000's on each of their desks with minimum memory only. It went from dog to decent performance in very short order.
Intuit have formally written to all it's UK Quicken users saying it's no longer going to supply Quicken or their tax software here - so they've sent a final CD for free and put all it's eggs into Quickbooks for the SME market. So, it looks like a good incentive to move to GNUcash now.
is repeating Geoffrey Moore's "Crossing the Chasm" and "Inside the Tornado" models, with "commercial" software on one track and FOSS on a parallel one. In the final analysis, a current "commercial" market leader will try to incrementally improve it's offerings, while the newcomers (FOSS) will try to depose it through a "paradigm shock".
If you don't want to read the whole of the two books, a one page summary can be found here.
In the Server Software space, MS has been overwhelmed everywhere bar "servers with SSL certificates on board" (a Netcraft analogue for "e-commerce web sites"). Looks like OpenOffice plus Firefox isn't quite enough of a paradigm shock to storm the desktop yet - but that may be a different story, yet to unfold.
Slashdot Mirror
Same applies. 1 subscription per machine. Having a Satellite server does not remove the need to have a basic or standard subscription for each one as well....
Satellite server just acts as a cache to stop every machine pulling updates from Red Hat individually, plus the ability to do PIXie boots, to clone systems and, if needed, to be able to distribute updates without having the machines connected to the Internet - if that's what you want to do.
The differences between MS's business model and that for open source aren't that far apart - both provide functionality "as is" with no warranty nor obligation for "fitness for purpose" (if in doubt, read the EULA that came with Windows or Office). Red Hat charge for support and updates, and you can use support and install any supported version (even new ones as they arrive) while you have a valid subscription active; at the end of the subscription term, you're still free to use the software whether you renew your subcription or not (just don't expect support or updates any more if you don't pay). Microsoft sell a license to use a single major version, but expect you to pay again when a new version comes out, or to subscribe to Software Assurance to allow you to use new versions as they emerge. Very few customers tend to buy Software Assurance atm.
I always wonder why Microsoft don't sell subscriptions in the same way Red Hat do...
Ian W.
The term "The Network is the System" dates back to the NaC group at DEC (VP: Julius Marcus) in 1977. "The Network is the computer" little more than a late developing bastard child of Digital's foresight.
To date, Novell is stronger on PR (Google search volumes on "SUSE" are almost at Red Hat levels) but are struggling really badly to monetise this.
The best thing that Oracle could do would be to support both equally. As stupid as it sounds, everyones best interest is served by having two (or more) Linux distributions duking it out.
To date, all the big DB vendors try clutching straws; Oracle themselves define market share as "license revenue", make out that they're slugging it out for dominance with DB2, and say that only 10% of MySQL deployments are in mission-critical apps. Follow the numbers, and you still see MySQL lapping Oracle's installed base in the "mission critical" space every 3 days.
Hence MySQL are already everywhere and starting to do a traditional market discontinuity. Virtually all the nice open source tools these days go MySQL first (PHP, Ruby on Rails among them). Oracle's defence is either to move to services... or to bully their way into the low end. I wonder whether they're really trying to buy MySQL.
Ian W.
... before they consumated their relationship with IBM (5 initiatives on offer, including one called Pink - the domain of then Apple employee and ex-VMS engineer Roger Heinen). KO was a big fan of the Apple Desktop Bus and it's simple connectivity, but vetoed Digital Semiconductor providing the chips. Or maybe that was Jack Shields in the Executive Commitee, who drove everything vertically integrated to go eat IBM's lunch by 2007...
Fokelore in DEC at the time anyway. How much of it was true may be a different story.
Ian W.
Most people don't care; Novell have a competent Linux distribution that's a worthy substitute if Red Hat ever pi**es people off. Their Google adwords search volumes are comparable to Red Hats, but they have a real problem monetising the demand; Red Hat outship them in subscription volumes by a factor of 9:1. With that goes share of mind in the ISV community, and if you have the developers, you have the market.
The unfortunate side effect of the market performance is that Novell slide SUSE underneath Netware and call the result "OES", then start to muddle their "open systems" and proprietary product numbers when they go full flight on their PR and stock market statements. Their recent very congratulatory PR got reproduced everywhere (a key Novell strength), but I still wince at three particular facts behind the numbers:
1) If you took out the effect of their one off payment from Microsoft, their year on year, like for like revenue as a company appears to have dropped 35%.
2) If you remove "OES" from their numbers (it wasn't shipping last year), their traditional Linux business appears to have grown from just north of $14m to $15m in one year.
3) The often cited UK National Health deal used to be for Netware, Identity Management and Zenworks. This got renewed for three years (with the Netware portion repurposed as OES, whether or not SUSE is physically used), with a contract value of $40m over three years. If it were Red Hat, they'd spread the subscriptions evenly over the full contract term; Novell in this case elected to count $20m as income inside the first quarter.
Are any of the above wrong? I'd be delighted to be told if this were the case.
Novell's SUSE products are excellent and given the right focus, they could monetise a lot more of the demand coming to their door. They have better pricing terms for high performance computing, some good key open source focus (with JBoss and MySQL for starters), and have relationships with many customers that would be open to inplement open source infrastructures. The main focus missing is (IMHO)... developers, developers and developers. Not just in ISVs, but also in most large end user companies - people invisible to most CIO's, and to most hardware manufacturers also.
As someone who normally enjoys reading The Register, this sort of article is a stain on their hitherto excellent site. Tantamount to peeing in their own soup. The vendetta between the Register and Wikipedia seems to be getting a life of it's own, at odds with the quality of the rest of their respective works. I hope they both snap out of it.
Ian W.
Motorola did a lot of work with UK company Trafficmaster on their Smartnav product, where most of the traffic data comes in realtime from roadside cameras and sensors on Motorway bridges covering 9,000 miles of UK roads. When you push the button on the unit, the mobile phone rings into a call centre with the current GPS position encoded in the Caller ID of the phone call (3 bytes), you tell them the destination, and they send the turn-by-turn instructions back into the unit. If a jam shows up on your route before you get to your destination, the central servers ring the units and offer (with an estimate of the delay if you do nothing), then download, a route around the jam (if indeed one is available). Works really well.
Getting a GPS signal communicated back and forth in real time is the only way this hack will work - at least until the cell sizes get to 3G pico sizes.
There have been experiments with floating vehicle sensors, but most companies that say they use these for real time data collection do so as a PR stunt - the number of vehicles that need to be equipped and the economics of getting the data back in real time don't make business sense today, from either an accuracy or cost standpoint. Most instead download history at the end of the day, and just pick up details of roadworks and accidents from "journalistic" sources - whether there's a jam present or not.
Until we get GPS in most handsets (and if operators allow the caller IDs to be sent through with GPS data on board, but the call terminated at either end without any money changing hands), the use of mobile phones for spotting jams, or indeed navigating around them, will be very limited.
Ian W.
I just wonder if Oracle were doing to get Innodb staff to write the interface code so MySQL could be the front end for an Oracle (or Oracle RAC) based backend. Or maybe it's just a smart move when MySQL have started selling their MySQL Cluster product (albeit memory based) for circa $4000 per node.
Be interesting to see how this plays out.
Ian W.
Fascinating to listen to on a wide range of subjects :-)
Ian W.
Oh the joy, Microsoft mutilating themselves. Maybe they've not renewed their Harvard Business School Magazine subscriptions - they should read http://hbswk.hbs.edu/item.jhtml?id=4980&t=marketin g
Nah. How long into the future do you reckon you'll have one CPU to one person? Anonymity is a deskside cluster...
Ian W.
EPICOR SOFTWARE CORPORATION 195 Technology Drive, Irvine, California, United States of America, 92618
ISOGON CORPORATION 330 Seventh Avenue, New York, United States of America, 1000
Gambro Inc 10810 West Collins Avenue, Lakewood, Colorado, United States of America, 80215-4498
Transtechnik Lichtsysteme GmbH & Co. KG Ohmstr. 1-3, Holzkirchen, Germany, 83607
See here
I'm still amazed that they announced the name without doing a simple search first.
Ian W.
And yet if my stats are correct (I work for one of the largest IT resellers in Europe), the Insurance Industry are the #1 early adopters for virtualisation software. I think something like 18 of our top 20 VMware customers are all Insurance or Financial Services companies ;-}
Ian W.
HP are to bundle Ubuntu on some of their laptops as a default install, with all the ACPI, Wifi support etc working out of the box.
Now that Microsoft have got hardware contract manufacturing expertise in house (courtesy of X-Box), there's little beyond moving Windows + Windows Update to a subscription model (like Red Hat do for Linux) to stop them offering quite a compelling $100 machine.
Ian W.
About the only area where MS have any meaningful share is on what Netcraft consider to be an analogue of "ecommerce sites" - measured in terms of those with a valid SSL certificate. Given MS's track record - and indeed the actual text of any security article cited by their own campaign (clue: not the headline!), some pretty significant folks are going to have to learn the hard way.
Let's face it - MS are doing all of us a favour. They're sullying their own brand reputation with the poor quality execution of this campaign, and that's even before their products bite companies where they hurt.
As a Linux advocate, I hope they keep the campaign running. Meanwhile, why doesn't someone ask Mike at Netcraft to publish the OS shares - I know he samples them.
Ian W.
They only need do what Red Hat does with Linux - give Windows away for free but charge for WindowsUpdate. And add telephone support as an option too. I don't understand what's taking them so long to work this out!
I seem to recall that the performance of VMS on small memory systems progressed leaps and bounds when a few hundred people at Spit Brook (VMS Engineering) got shiney new VAXstation 2000's on each of their desks with minimum memory only. It went from dog to decent performance in very short order.
Intuit have formally written to all it's UK Quicken users saying it's no longer going to supply Quicken or their tax software here - so they've sent a final CD for free and put all it's eggs into Quickbooks for the SME market. So, it looks like a good incentive to move to GNUcash now.
If you don't want to read the whole of the two books, a one page summary can be found here.
In the Server Software space, MS has been overwhelmed everywhere bar "servers with SSL certificates on board" (a Netcraft analogue for "e-commerce web sites"). Looks like OpenOffice plus Firefox isn't quite enough of a paradigm shock to storm the desktop yet - but that may be a different story, yet to unfold.
Ian W.