Quite a few Europeans are dropping their land lines, even though they are cheaper, just for the convenience of having a single phone for everything. Europeans pay for inbound roaming calls (when in a different country), it's just for the common case that you are not roaming that you don't pay for inbound.
The US and Europe are *not* adopting the same 3G standards (why change the habit of the last few decades??): the US will use CDMA2000 (mainly), and Europe will use W-CDMA (aka IMT-2000). However, I think some US operators such as AT&T will use W-CDMA so they get global roaming revenues from visitors to the US.
TDMA is quite obsolete - all the TDMA operators are converting to GSM/GPRS, see www.3gamericas.org for what used to be the TDMA trade body in the US, which now promotes GSM.
Cost isn't the only metric - being able to depend on mobile phone coverage is another key one. Ease of roaming between GSM operators is a key part of getting good coverage.
Re:Already approaching from the wrong direction
on
64kbps @ 40,000 ft.
·
· Score: 2
Another problem is that you can be in many cells at once when using a mobile phone on a plane - this confuses the network.
Windows isn't that easy to use for complete beginners - it's far too easy to reconfigure it by mistake (e.g. they frequently drag the taskbar to the side of the screen). But it is what most people already know.
It's not about whining, it's about sticking to what the GPL requires. IANAL, but I understand that when you distribute the binary of a GPLed program (such as Linux), you must also distribute the source code (or make it available, e.g. via the Net). If this is not acceptable to Lindows, they should have chosen another OS with a more permissive license, e.g. one of the BSDs. They can't have the developer base and mindshare of Linux without doing what the GPL requires.
If Lindows had licensed a commercial software component and were breaking its license terms, would that be a 'silly question'? For some reason, conforming to open source licenses is considered by some people to be an optional extra...
The only real issue IMO is whether some delay is acceptable between releasing the binary and the source, particularly for betas - this seems to happen with some projects, in practice, but if the project/business goes away in the mean time, the users are left without the source.
You seem to be confusing the Palm user interface with the bundled applications - the UI is simple and Mac-like, but doesn't constrain the applications.
For example, I use DateBk3, which is a much enhanced version of the built-in Datebook with many extra month/week type views, and lots of extra features. Similarly, you can get third party apps that let you do formatted text, databases, spreadsheets, etc, including syncing to and from Microsoft Office.
It's understandable if you prefer the Psion form factor, and EPOC is way ahead of Palm OS in stability, but the UI is one of the things that Palm got right IMO.
For a simple example of XSLT, see http://www.w3.org/2000/08/w3c-synd/home.rss - this looks like a normal web page if you view it in an XSL-capable browser (e.g. IE 5.5), but it's actually an XML page for an RSS feed of W3C news (see http://twiki.org/cgi-bin/view/Codev/RichSiteSumm ar y for links about RSS, which drives slashboxes amongst other things).
Another fun story - I tried upgrading IE 5.5 to IE6 and it broke on a web-based bug tracking system (fairly horrible system, lots of Javascript). I eventually downgraded to IE5.5 (which was not easy) and then tried Mozilla 0.9.8, which worked perfectly...
Clearly Mozilla is the natural upgrade path for IE 5.5!
Just be sure to get a tri-band GSM phone - the US uses GSM at 1900 MHz, while the rest of the world uses a mix of 900 and 1800 MHz.
Even if you get a single-band phone, you should be able to take your GSM phone's SIM card and put it in phones that you rent overseas (called SIM roaming) - but it's much more convenient to just use your phone everywhere, with all the phone numbers and other setup. I've used my GSM phone all round Europe and it worked fine in India (Mumbai) when I was there.
I went to a Windows developer conference at the end of 1993, and they were talking a lot about Cairo, including the magical OFS feature. Of course, Cairo turned out to be vapourware... Windows 95 was being developed independently, as Chicago, and managed not to slip more than a year or so. Win95 had a similar UI to Cairo on the outside, but the UI code was probably quite different internally.
The short memories of the industry are perhaps because a lot of people probably weren't in the industry in 1993:) But yes, it is a bit worrying - most things that eventually succeed are re-invented several times, without anyone realising...
Inter-carrier SMS is now enabled for some US wireless operators, and they are already seeing increases in revenues (big surprise, the European operators enabled this a long time ago...). So don't be surprised if SMS takes off in the US as well.
The US market is different, largely due to regulation as you point out, but people are increasingly mobile, particularly teenagers, so SMS and some mobile internet applications (chat, dating, multi-player games, etc.) are probably going to take off in the US as well as Europe.
As for an exclusive license - don't know where you get that from, the Japanese competitors to i-mode have similar services and I think now have the ability to interoperate with i-mode to some extent. What's harder to replicate is the end-to-end specification of phones, protocols, servers, billing models and content services so that they all work together.
I mostly agree with you about i-mode vs WAP (see my post elsewhere about WAP's brokenness). The phones in the UK are nowhere near i-mode or i-appli in ease of use, and FOMA is further ahead still. The new 3G phones coming out this year in Europe may give FOMA a run for its money, but ease of use is still the critical issue - i-mode and FOMA seem to be designed on the MacOS model, i.e. an end to end design that is highly usable. European and US 3G won't have that level of integration.
However, GPRS is packet-mode and now rolled out on 3 out of 4 UK networks (not sure about one2one), and in many other European countries. It's faster than i-mode's PDC-P (which is 9.6 Kbps), up to 30-50 Kbps depending on cell site, following wind, etc. And of course GPRS is always-on, like PDC-P.
I've used WAP since the Nokia 7110 came out, and I still find it has huge problems, even on my T68 - WAP gateways break, WAP browsers refuse to display pages, and WAP sites make it hard to do simple things like 'Back' in the browser (they actually control that sort of feature...). Most importantly, even when WAP works, the sites are frequently very hard to use due to poor design.
GPRS makes WAP easier to use, because it's always-on, but it's not any faster for my provider - in fact a side-by-side test at the same time on the same site showed GSM was a bit faster (though more expensive and not really a noticeable difference).
J2ME will help matters, particularly the model of keeping a local data cache on the phone, but it's a lot harder to write a J2ME midlet than a WAP or i-mode page, so they will not be so common IMO.
I think that i-mode has a huge future if it addresses the failings of WAP before the M-Services and other initiatives deliver. It's more like TV than a PC, i.e. 'it just works' whereas WAP is a painful experience akin to the web browser in the mid-90s.
i-mode is not a protocol on the level of GPRS, in fact it can use GPRS (and UMTS [3G] and the Japanese PDC-P) as a transport. In fact i-mode is being deployed by KPN in the Netherlands and will run over GPRS there (some other deployments should also happen e.g. in Germany).
i-mode includes its own HTML variant (cHTML, quite close to standard HTML unlike WAP's WML), and some sort of application level protocol (proprietary to NTT I think). This runs over a packet mode layer, which is analogous to IP but designed for wireless (shorter headers etc - spectrum is expensive). i-mode also includes (crucially) a content-based billing system for official providers (NTT approved), and data transfer based (i.e. Kbytes) billing for unofficial providers (NTT makes a lot of money out of these).
Most importantly, NTT designed the phones and told the manufacturers to make them - the NTT logo is on the phone, the manufacturer isn't (hard to replicate anywhere except Japan). The end to end i-mode architecture is a bit like MacOS - quite proprietary but it's very easy to use (see http://www.useit.com/ for an article on this). Also, NTT takes only 9% of the official content providers' revenues, compared to the Euro wireless providers who typically take about half for premium SMS services. Guess which one has more providers, more revenues and more profits...
3G isn't a protocol either, but CDMA2000 and UMTS are suites of protocols that conform to the ITU-T's '3G vision' - aka IMT-2000, which is even more fluffy and vaporous than 3G is!
Good post, but I'd just like to point out that you mean 'developing countries' not 'foreign countries' - the latter includes Europe and Japan, where people are quite able to afford the latest hardware...
Hmmm - according to http://slashdot.org/comments.pl?sid=27788&cid=2986 922 (post from a UK Treo GPRS user), the Treo is a GPRS multi-slot class 10 device, aka 4+2 slots, meaning 53.6 Kbps downstream. I hope the latter is right... I think the UK/Europe device is 900/1800 MHz so it could be a different chipset that's the explanation.
In the UK, at least one operator (Vodafone) lets you have two SIM cards (that's your 'SMID card', SIMs are what they are usually card) on a single account and single phone number. This is meant to address exactly this problem. Not sure what happens on incoming calls, perhaps one is permanently the 'incoming' device.
You can get an Ericsson T68 GPRS phone in the UK for less than 100 UKP on contract, or 400 UKP off contract... This has a colour screen and Bluetooth, so it's sort-of equivalent in positioning. So you are right about that sort of phone - but I bet the price is more than $549 when it hits Europe! There are definitely cheaper GPRS phones available than the T68, too.
Treos in the UK (on test) have GPRS already, which gives you always-on at the network level (no phone calls). Handspring have said they are developing software that will do Blackberry-style always-on email soon, I think by Q2 this year.
Even without this extra software, checking email over GPRS is easily done - my Ericsson T68 phone can check every 5 minutes without making phone calls.
See www.gsmworld.com for info on GSM around the world - about 70% of mobile subscribers worldwide use GSM, so it is pretty much everywhere except Japan. Some providers won't have GPRS, though, and those that do won't have enabled GPRS roaming, but you can at least make phone calls wherever you are - I've used my GSM phone in India as well as all over Europe. It even worked a week ago while I was skiing in the Alps (mountaintop base stations...)
If you are staying some time in a country, it's worth getting a local GSM SIM card (subscriber card) that's pre-paid - means you pay for local calls at the normal rate, though your number changes.
GSM/GPRS should be a fairly safe bet for some time, though of course 3G is just around the corner so some obsolescence is inevitable.
It would have been good to include Bluetooth instead of a Springboard slot - at least then you could network through to a different device, e.g. your laptop to get onto a wireless LAN. There may even be a market for wireless routers that talk Bluetooth on one side and CDMA/GSM/UMTS wireless on the other - that way you can decouple your user interface device (web pad, PDA, toaster) from the wireless standards.
Thanks for the info, here are some calculations...
According to http://www.compaq.com/products/quickspecs/10903_na/10903_na.HTML and some other pages I found, multi-slot class 10 means 4+2 slots, i.e. 4 downlink and 2 uplink. The slot capacity is dependent on the coding scheme, which in turn varies depending on radio conditions - this looks like the CS-2 scheme, which is 13.4 Kbits per slot, since 4 x 13.4 is 53.6 Kbps downstream. (See http://www.nuntius.com/solutions22.html ) Upstream should be 26.6 Kbps of course since there are half as many slots. This probably doesn't include IP overhead, and may also not include the various layer 2 and below stuff above the actual radio link (there's a lot of tunnelling between the GPRS device and the network, to enable roaming).
Anyway, YMMV depending on number and activity of other GPRS and GSM-voice users in the cell - if they're busy you won't get the full 4 slots. Also, as the radio conditions worsen you'll drop down to CS-1, which has more error correction but maxes out at 4 x 9.05 = 36.2 Kbps downstream for a 4+2 device.
Fortunately, if you are using an efficient protocol, such as WAP or perhaps a custom protocol on top of IP (like some of the compressing/transcoding proxy setups), you can get pretty good response times out of a GPRS device, but it's not really meant for big downloads. On some tariffs, it's cheaper to use HSCSD (two GSM phone calls at once) for downloads, but ultimately GPRS will become a commodity and should end up being cheaper, with HSCSD the high-end service for those who can afford it.
At present, you may actually find WAP over a GSM call is a bit faster (I did a side by side test with Ericsson T68 vs Nokia 7110 on the same Orange UK WAP site) - but GPRS is much more convenient particularly for quickly checking a website on your Treo, or for sending an email without firing up your PC or making a data call on your phone. In the longer term, there'll be higher-spec phones, up to 8 slots downstream for 115 Kbps, but there'll be a price in battery life and perhaps overheating.
I don't think the country matters - GPRS works the same way regardless of frequency, so the US/Canada's GPRS at 1900 MHz will have the same data rates IMO as the UK's 1800 and 900 MHz (subject to radio conditions of course).
For a technical intro to GPRS, have a look at http://ww1.comsoc.org/pubs/surveys/3q99issue/pdf/S alkintzis.pdf
Page ranking mainly works by finding documents that are linked *to*, and therefore are popular, authoritative, or whatever. As long as the intranet sites have links to.DOC files etc, Google should continue to work OK - admittedly, there won't be links between the DOC files, but that's just another reason to convert them to HTML, where they can perhaps be auto-linked based on keywords etc.
Quite a few Europeans are dropping their land lines, even though they are cheaper, just for the convenience of having a single phone for everything. Europeans pay for inbound roaming calls (when in a different country), it's just for the common case that you are not roaming that you don't pay for inbound.
The US and Europe are *not* adopting the same 3G standards (why change the habit of the last few decades??): the US will use CDMA2000 (mainly), and Europe will use W-CDMA (aka IMT-2000). However, I think some US operators such as AT&T will use W-CDMA so they get global roaming revenues from visitors to the US.
TDMA is quite obsolete - all the TDMA operators are converting to GSM/GPRS, see www.3gamericas.org for what used to be the TDMA trade body in the US, which now promotes GSM.
Cost isn't the only metric - being able to depend on mobile phone coverage is another key one. Ease of roaming between GSM operators is a key part of getting good coverage.
Another problem is that you can be in many cells at once when using a mobile phone on a plane - this confuses the network.
Windows isn't that easy to use for complete beginners - it's far too easy to reconfigure it by mistake (e.g. they frequently drag the taskbar to the side of the screen). But it is what most people already know.
It's not about whining, it's about sticking to what the GPL requires. IANAL, but I understand that when you distribute the binary of a GPLed program (such as Linux), you must also distribute the source code (or make it available, e.g. via the Net). If this is not acceptable to Lindows, they should have chosen another OS with a more permissive license, e.g. one of the BSDs. They can't have the developer base and mindshare of Linux without doing what the GPL requires.
If Lindows had licensed a commercial software component and were breaking its license terms, would that be a 'silly question'? For some reason, conforming to open source licenses is considered by some people to be an optional extra...
The only real issue IMO is whether some delay is acceptable between releasing the binary and the source, particularly for betas - this seems to happen with some projects, in practice, but if the project/business goes away in the mean time, the users are left without the source.
You seem to be confusing the Palm user interface with the bundled applications - the UI is simple and Mac-like, but doesn't constrain the applications.
For example, I use DateBk3, which is a much enhanced version of the built-in Datebook with many extra month/week type views, and lots of extra features. Similarly, you can get third party apps that let you do formatted text, databases, spreadsheets, etc, including syncing to and from Microsoft Office.
It's understandable if you prefer the Psion form factor, and EPOC is way ahead of Palm OS in stability, but the UI is one of the things that Palm got right IMO.
For a simple example of XSLT, see http://www.w3.org/2000/08/w3c-synd/home.rss - this looks like a normal web page if you view it in an XSL-capable browser (e.g. IE 5.5), but it's actually an XML page for an RSS feed of W3C news (seem ar y for links about RSS, which drives slashboxes amongst other things).
http://twiki.org/cgi-bin/view/Codev/RichSiteSum
Another fun story - I tried upgrading IE 5.5 to IE6 and it broke on a web-based bug tracking system (fairly horrible system, lots of Javascript). I eventually downgraded to IE5.5 (which was not easy) and then tried Mozilla 0.9.8, which worked perfectly...
Clearly Mozilla is the natural upgrade path for IE 5.5!
Just be sure to get a tri-band GSM phone - the US uses GSM at 1900 MHz, while the rest of the world uses a mix of 900 and 1800 MHz.
Even if you get a single-band phone, you should be able to take your GSM phone's SIM card and put it in phones that you rent overseas (called SIM roaming) - but it's much more convenient to just use your phone everywhere, with all the phone numbers and other setup. I've used my GSM phone all round Europe and it worked fine in India (Mumbai) when I was there.
Somebody mod this up!
:) But yes, it is a bit worrying - most things that eventually succeed are re-invented several times, without anyone realising...
I went to a Windows developer conference at the end of 1993, and they were talking a lot about Cairo, including the magical OFS feature. Of course, Cairo turned out to be vapourware... Windows 95 was being developed independently, as Chicago, and managed not to slip more than a year or so. Win95 had a similar UI to Cairo on the outside, but the UI code was probably quite different internally.
The short memories of the industry are perhaps because a lot of people probably weren't in the industry in 1993
Inter-carrier SMS is now enabled for some US wireless operators, and they are already seeing increases in revenues (big surprise, the European operators enabled this a long time ago...). So don't be surprised if SMS takes off in the US as well.
The US market is different, largely due to regulation as you point out, but people are increasingly mobile, particularly teenagers, so SMS and some mobile internet applications (chat, dating, multi-player games, etc.) are probably going to take off in the US as well as Europe.
As for an exclusive license - don't know where you get that from, the Japanese competitors to i-mode have similar services and I think now have the ability to interoperate with i-mode to some extent. What's harder to replicate is the end-to-end specification of phones, protocols, servers, billing models and content services so that they all work together.
I mostly agree with you about i-mode vs WAP (see my post elsewhere about WAP's brokenness). The phones in the UK are nowhere near i-mode or i-appli in ease of use, and FOMA is further ahead still. The new 3G phones coming out this year in Europe may give FOMA a run for its money, but ease of use is still the critical issue - i-mode and FOMA seem to be designed on the MacOS model, i.e. an end to end design that is highly usable. European and US 3G won't have that level of integration.
However, GPRS is packet-mode and now rolled out on 3 out of 4 UK networks (not sure about one2one), and in many other European countries. It's faster than i-mode's PDC-P (which is 9.6 Kbps), up to 30-50 Kbps depending on cell site, following wind, etc. And of course GPRS is always-on, like PDC-P.
I've used WAP since the Nokia 7110 came out, and I still find it has huge problems, even on my T68 - WAP gateways break, WAP browsers refuse to display pages, and WAP sites make it hard to do simple things like 'Back' in the browser (they actually control that sort of feature...). Most importantly, even when WAP works, the sites are frequently very hard to use due to poor design.
GPRS makes WAP easier to use, because it's always-on, but it's not any faster for my provider - in fact a side-by-side test at the same time on the same site showed GSM was a bit faster (though more expensive and not really a noticeable difference).
J2ME will help matters, particularly the model of keeping a local data cache on the phone, but it's a lot harder to write a J2ME midlet than a WAP or i-mode page, so they will not be so common IMO.
I think that i-mode has a huge future if it addresses the failings of WAP before the M-Services and other initiatives deliver. It's more like TV than a PC, i.e. 'it just works' whereas WAP is a painful experience akin to the web browser in the mid-90s.
i-mode is not a protocol on the level of GPRS, in fact it can use GPRS (and UMTS [3G] and the Japanese PDC-P) as a transport. In fact i-mode is being deployed by KPN in the Netherlands and will run over GPRS there (some other deployments should also happen e.g. in Germany).
i-mode includes its own HTML variant (cHTML, quite close to standard HTML unlike WAP's WML), and some sort of application level protocol (proprietary to NTT I think). This runs over a packet mode layer, which is analogous to IP but designed for wireless (shorter headers etc - spectrum is expensive). i-mode also includes (crucially) a content-based billing system for official providers (NTT approved), and data transfer based (i.e. Kbytes) billing for unofficial providers (NTT makes a lot of money out of these).
Most importantly, NTT designed the phones and told the manufacturers to make them - the NTT logo is on the phone, the manufacturer isn't (hard to replicate anywhere except Japan). The end to end i-mode architecture is a bit like MacOS - quite proprietary but it's very easy to use (see http://www.useit.com/ for an article on this). Also, NTT takes only 9% of the official content providers' revenues, compared to the Euro wireless providers who typically take about half for premium SMS services. Guess which one has more providers, more revenues and more profits...
3G isn't a protocol either, but CDMA2000 and UMTS are suites of protocols that conform to the ITU-T's '3G vision' - aka IMT-2000, which is even more fluffy and vaporous than 3G is!
There's at least one RPG compiler for Linux - see http://www.californiasoftware.com/calsw/rvb_unibol 400.htm
You need to improve your googling skills - the third item or so on googling for Unibol was http://www.californiasoftware.com/calsw/rvb_unibol 400.htm .
This is an RPG/400 compiler for Unices including Linux, based on much older Unibol versions (read: mature and reliable!).
Also, have a look at http://www.well.com/~jax/rcfb/as400.html - links all about Open Source and the AS/400.
Good post, but I'd just like to point out that you mean 'developing countries' not 'foreign countries' - the latter includes Europe and Japan, where people are quite able to afford the latest hardware...
Hmmm - according to http://slashdot.org/comments.pl?sid=27788&cid=2986 922 (post from a UK Treo GPRS user), the Treo is a GPRS multi-slot class 10 device, aka 4+2 slots, meaning 53.6 Kbps downstream. I hope the latter is right... I think the UK/Europe device is 900/1800 MHz so it could be a different chipset that's the explanation.
In the UK, at least one operator (Vodafone) lets you have two SIM cards (that's your 'SMID card', SIMs are what they are usually card) on a single account and single phone number. This is meant to address exactly this problem. Not sure what happens on incoming calls, perhaps one is permanently the 'incoming' device.
You can get an Ericsson T68 GPRS phone in the UK for less than 100 UKP on contract, or 400 UKP off contract... This has a colour screen and Bluetooth, so it's sort-of equivalent in positioning. So you are right about that sort of phone - but I bet the price is more than $549 when it hits Europe! There are definitely cheaper GPRS phones available than the T68, too.
Treos in the UK (on test) have GPRS already, which gives you always-on at the network level (no phone calls). Handspring have said they are developing software that will do Blackberry-style always-on email soon, I think by Q2 this year.
Even without this extra software, checking email over GPRS is easily done - my Ericsson T68 phone can check every 5 minutes without making phone calls.
Still a pain though - the newer Ericsson phones are all tri-band so they work on all frequencies in Europe and North America.
See www.gsmworld.com for info on GSM around the world - about 70% of mobile subscribers worldwide use GSM, so it is pretty much everywhere except Japan. Some providers won't have GPRS, though, and those that do won't have enabled GPRS roaming, but you can at least make phone calls wherever you are - I've used my GSM phone in India as well as all over Europe. It even worked a week ago while I was skiing in the Alps (mountaintop base stations...)
If you are staying some time in a country, it's worth getting a local GSM SIM card (subscriber card) that's pre-paid - means you pay for local calls at the normal rate, though your number changes.
GSM/GPRS should be a fairly safe bet for some time, though of course 3G is just around the corner so some obsolescence is inevitable.
It would have been good to include Bluetooth instead of a Springboard slot - at least then you could network through to a different device, e.g. your laptop to get onto a wireless LAN. There may even be a market for wireless routers that talk Bluetooth on one side and CDMA/GSM/UMTS wireless on the other - that way you can decouple your user interface device (web pad, PDA, toaster) from the wireless standards.
Thanks for the info, here are some calculations...
a /10903_na.HTML and some other pages I found, multi-slot class 10 means 4+2 slots, i.e. 4 downlink and 2 uplink. The slot capacity is dependent on the coding scheme, which in turn varies depending on radio conditions - this looks like the CS-2 scheme, which is 13.4 Kbits per slot, since 4 x 13.4 is 53.6 Kbps downstream. (See http://www.nuntius.com/solutions22.html ) Upstream should be 26.6 Kbps of course since there are half as many slots. This probably doesn't include IP overhead, and may also not include the various layer 2 and below stuff above the actual radio link (there's a lot of tunnelling between the GPRS device and the network, to enable roaming).
S alkintzis.pdf
According to http://www.compaq.com/products/quickspecs/10903_n
Anyway, YMMV depending on number and activity of other GPRS and GSM-voice users in the cell - if they're busy you won't get the full 4 slots. Also, as the radio conditions worsen you'll drop down to CS-1, which has more error correction but maxes out at 4 x 9.05 = 36.2 Kbps downstream for a 4+2 device.
Fortunately, if you are using an efficient protocol, such as WAP or perhaps a custom protocol on top of IP (like some of the compressing/transcoding proxy setups), you can get pretty good response times out of a GPRS device, but it's not really meant for big downloads. On some tariffs, it's cheaper to use HSCSD (two GSM phone calls at once) for downloads, but ultimately GPRS will become a commodity and should end up being cheaper, with HSCSD the high-end service for those who can afford it.
At present, you may actually find WAP over a GSM call is a bit faster (I did a side by side test with Ericsson T68 vs Nokia 7110 on the same Orange UK WAP site) - but GPRS is much more convenient particularly for quickly checking a website on your Treo, or for sending an email without firing up your PC or making a data call on your phone. In the longer term, there'll be higher-spec phones, up to 8 slots downstream for 115 Kbps, but there'll be a price in battery life and perhaps overheating.
I don't think the country matters - GPRS works the same way regardless of frequency, so the US/Canada's GPRS at 1900 MHz will have the same data rates IMO as the UK's 1800 and 900 MHz (subject to radio conditions of course).
For a technical intro to GPRS, have a look at http://ww1.comsoc.org/pubs/surveys/3q99issue/pdf/
Page ranking mainly works by finding documents that are linked *to*, and therefore are popular, authoritative, or whatever. As long as the intranet sites have links to .DOC files etc, Google should continue to work OK - admittedly, there won't be links between the DOC files, but that's just another reason to convert them to HTML, where they can perhaps be auto-linked based on keywords etc.