Old Updates (5/98 & earlier)

5/29/98 - Lucent V.90 newer version The Digitan website now has a Lucent (Apollo & Mars) version 5.09 V.90 firmware; the 5.06 firmware has been replaced with this version. See the 5/24 update.

5/25/98 - Who we gonna sue? "All 100 million of us Web users are on the verge of suing somebody over the slow, unreliable, and expensive Internet access we get using dial-up telephone modems. Who we gonna sue?"- Written by Bob Metcalfe, inventor of Ethernet, in his 4/13/98 InfoWorld column.

5/25/98 - Lucent LT does V.90 better than 3Com/Courier! Having installed V.90 on my "free" cheap LT, I've found it makes a more reliable connection to my IBM/Prodigy POP (using 3Com server equipment) than my "top-of-the-line" Courier! Amazing! Within the last week, my Courier V.90 connections have been getting progressively worse - frequent re-trains, high error rate, and disconnections. While the LT connects at a rate 2-4k lower than the Courier, it makes a solid connection. This morning, I had a 4-1/2 hour connection with the LT that started at 38.6k, ended at 38.6k, and had 2 rx frame errors for the entire connection. I then called with the Courier, got a 46.6k connection, and was disconnected after 4-1/2 minutes on a failed re-train (i6 showing disconnect reason as "Loop loss disconnect"). The Courier had 231 blers in the 4-1/2 minutes! Most of my LT connections (that I keep) are in the 44k range - reliable and with very low error rate. Is there something wrong with my Courier, or is it the firmware??? Should I try and force the Courier down to a <44k connect rate? (I tried setting the maximum Courier rate at 44k, called and again was disconnected after 4-1/2 minutes with "Retransmit Limit" disconnect reason; 364 blers; I then re-connected with the LT at 44k for 2 hours, the call ended at 44k with 232 rx frame errors out of 42487 frames received.)

5/24/98 - Lucent LT Winmodem V.90 firmware available on the Digitan website. I received a message from a reader who has tried it successfully on his HP computer. I've tried it in my IBM-Aptiva successfully. Caution: there is more than 1 Lucent 56k chipset, and there is no sure way to tell which is which! (Although if AT I3 shows a driver version 4.xx, it is most likely compatible with this release. See this message from Ed Schulz at Lucent.) The Digitan page has 2 Lucent V.90 files - one in zip format, another is an .exe that "will extract and install the driver automatically. It will also overwrite the old Modem drivers in Windows." I would suggest downloading the zip file, and using the process in my 4/10 update to install the .vxd or .sys files. Note that there are new connect speeds with this release, so it is necessary to also use Control Panel/Modems to remove the LT; then re-install it with the don't detect/have disk... option to specify the location of the updated .inf file included in the .zip file. (For Windows95, remove the LT, reboot; plug 'n pray will take over and reinstall the modem; it should ask you for a driver location - that's where you need to specify the updated .inf file.) The Digitan/Lucent driver is located on this page - and is labeled: "DS560-446 to -452 (choice-1): ISA 56K Modem V.90 zipped file for Win-95/98/NT " After update, AT I3 will respond with: "LT V.90 Data+Fax Modem Version 5.09" I'm able to connect with all the local 3Com V.90 servers at 30-46k rates (similar to what I'm getting with my 3Com Courier modem). The LT appears to be more stable than the Courier to my challenging IBM-POP call routing. Interesting though: the LT will not make a V.90 connection to a local Flex provider running the Lucent PM3 V.90 beta! It will either make a Flex connect (46-48k), or a double-handshake v.34 (28k) connection!

.5/23/98 - IBM announces V.90 support - IBM Global Network has officially announced V.90 availability at over 500 US POPs, and has updated their access number list to reflect V.90 support.

5/20/98 - Where's the code? While 3Com/USR has made V.90 upgrades available for most of its x2 models, several of the original x2 Sportsters - the 1786-00 and 1787-00 are still not available. The 3Com web site has had availability estimates for this modem starting with March '97 dates; the latest revision came today when they changed a 5/20 estimate to 5/22; on 5/18, the estimate was changed to 5/20.

Zoom has released a firmware upgrade for their dual-mode Flex/V.90 modems, but early reports indicate it makes matters worse - a user reporting the new firmware results in no Flex or V.90 connections. The original firmware is not available on the Zoom website, so a user needing to flash-back has to contact Zoom technical support. If you have this modem and want to try, the "upgrade" is at http://www.zoomtel.com.

5/5/98 - V.90 Interoperability - Rockwell V.90 doesn't work at all to Lucent; 3Com to Lucent "sort-of" in this interoperability update.

5/2/98 - Toll-free v.90-enabled numbers you can test your v.90-enabled modem with.

4/30/98 - Motorola sells its modem business - Motorola sells its modem business - A new company is being formed to acquire Motorola's modem business. Motorola had announced some time ago that it was seeking a buyer for the division that makes, among other products, the 56KFlex SURFR series. See this press release. (Just a few days after the announcement, the deal fell through. Motorola now has closed its modem business, but says it will still support existing users.

4/29/98 - I try the Zoom 2949 V.90 Dual Mode Modem - Can the Zoom connect to my local 3Com V.90 ISPs? How does it work with the local Flex providers? My review here.

4/29/98 - Lucent LT Win Modem 4.31 Firmware now available at ftp://ftp1.hp.com/pub/personal_computing/software/lucnt431.exe. This is not V.90 code, but it is the latest Kflex code for this modem. You can avoid complicated installation by using the instructions from this 4/10 update.

4/28/98 - 3Com Product Manager responds to ISP complaints and suggestions. Click here to see a copy of the letter which was posted in the usr-tc discussion group.

4/28/98 - New pages added to this site: Setting up HyperTerminal to talk to your modem. and If you get a CONNECT 38.4, 57.6, or 115.2k, you need to re-install your modem.

4/25/98 - v.90: improved performance The IBM POP I use was updated with V.90 on 4/24, and I am impressed! I can establish a reliable V.90 connection at 46-48k with no retrains, and a reduced bler error count. With x2, my best connections would average 4 retrains per hour with over a thousand blers/hr. I would also get a fair number of initial connects at 33-37k that were totally unusable; and some 48k connects that never got higher than 33333bps. With v.90, I get some 32k connects (that appear to stay at 32k with 0 blers), but I disconnect and go for the 46.6 or 48k connects. I've learned that part of 3Com's contribution to V.90 includes a "null speed switch": if you leave your modem speaker on after carrier, you may notice occasional "blips" interrupting the warble; during a null speed switch, you'll hear a blip, followed by a short (1/10th second) semi-silence, followed by a second blip. This allows the echo cancellers to reconverge without a retrain. This is a major improvement over the 5+ seconds of 0bps throughput which would occur during a non V.90 connection retraining.You can also hear a single blip during a link event like disconnecting: tell your modem to disconnect, and you'll hear the blip right before the warble stops and the line goes on hook, as well as when a speedshift occurs. (More RealAudio of these events coming here this weekend.) I'm finding the CONNECT reported with V.90 is real or lower than actually achieved, compared to x2 where the CONNECT reported was almost always higher than reality (for my line and telco conditions). Click to hear RealAudio of V.90 null speed switch.

4/24/98 - Zoom's Marketing VP interviewed by PCWorld OnLine's Brian McWilliams in RealAudio. Headline reads: "Don't believe the hype. True 56kbps modem interoperability is still months away, says Zoom's Terry Manning."

4/22/98 - IBM to be 100% V.90 in US by 4/30/98. PCWorld On-line report in RealAudio 4/21/98. Says Canada and the rest of the world to follow; that as of 4/21, some 400 out of 527 US POPs are already updated to v.90. [8/98- Original PCWorld report has been removed or moved.]

4/20/98 - IBM V.90 rollout update - In response to my newsgroup post, Jason Dorough, IBM.NET support, said that when x2 was rolled out, it was done overnight and their help desk was overwhelmed with calls. IBM is taking a more conservative approach with v.90. When a POP is upgraded, it is monitored to make sure it is working. When all the POPs in a country are updated, V.90 support will be officially announced. So "officially we do not support v.90."

4/18/98 -

4/18/98 - Where's the code? While 3Com has released V.90 upgrades for most of its US retail products, there are several models it is not yet available for. Owners of the USR Winmodem, and certain Sportster models are complaining as 3Com keeps pushing the date back... and back... and back. Waiting until the last date promised (4/17), 3Com that day changed the date to May 4.

4/18/98 - IBM rolling out V.90 IBM appears to be in the process of updating its all-x2 US dial-in sites to v.90; a number of areas are running V.90 now (including San Diego, Honolulu, Cincinnati, and more). You can tell if you are calling a 3Com V.90 server if your modem has the V.90 update by the distinctive bongs near the end of the handshake. Click to hear RealAudio of V.90 connections.

4/18/98 - InfoWorld article - Do Windows users get more speed from x2 or Kflex?

4/10/98 - Lucent LT 4.28 firmware available - ftp://ftp.cybmax.com/modem/s56c the file is CPILU428.EXE. These are self-extracting files. Before you update the modem, I recommend that you copy the existing modem drivers (lt*.vxd or lt*.sys) to a backup folder. All that is necessary to upgrade the modem is:

(For WIN95): Copy the *.VXD files extracted by the download to the \WINDOWS\SYSTEM directory, then re-boot the computer.

(For WIN-NT) Copy the LTMODEM.SYS file to your \NT\SYSTEM32 directory, then re-boot the computer.

You can downgrade to any prior version by following the same procedure with the older .vxd/.sys files. (AT I3 will verify your version.) I've achieved a 50k connect to local Flex provider with this version; before, I maxed out at 34k and once saw a 36k.

4/8/98 - Big x2 Problems with Prodigy Internet - Hundreds of newsgroup posts from Prodigy users having trouble connecting to various Prodigy local access numbers with a wide range of problems. Most of these posts are from users calling into a "Split Rock POP". Split Rock is a national access provider that Prodigy contracts with to provide its local access service in various cities across the country. The big difference: Split Rock is the only major x2-provider using Bay Networks server equipment as opposed to 3Com/USR equipment. There appear to be some serious compatibility problems in the Bay firmware causing problems not only for x2/V.90 users, but for v.34 users as well! There is a known problem with Bay firmware not being able to negotiate a connection with a 3Com V.90 modem. In many of these situations, it is impossible for a customer to get anything over 28.8 even with V.90 disabled. On pre V.90 3Com modems, some x2 connections can be obtained, but it appears that the Bay equipment is unable to deal with some network routing that 3Com server equipment has no problem with. So even with pre V.90 3Com firmware, some users will not be able to get a reliable x2 connection to the Bay equipment. The Prodigy 56k rollout is way behind schedule. Many users are screaming for 56k access numbers, and many of those who have them are screaming because they can't connect at 56k speeds if they can connect at all! It should be noted that Prodigy also uses IBM POPs in a number of areas, all of which are x2-capable and all of which use 3Com server equipment. These problems do not occur on the IBM POPs. (Even though all IBM POPs are x2-enabled, Prodigy only updated its access list to reflect the x2 capability on my local POP very recently, and had many other IBM-provided POPs not listed as supporting x2. It appears that they've updated their list to reflect x2 on all the IBM POPs.)

4/8/98 - V.90 status - A number of ISPs across the country have upgraded their server equipment to 3Com's V.90 release. It appears to provide more reliable connections for most users, and greatly improved performance for some of those who had been experiencing poor x2 performance. One ISP in my local area has upgraded to V.90 - but it happens to be the only one that I could get consistently reliable x2 connections to (at about 48k); my V.90 performance with them is about the same, but, the initial V.90 connect speed is usually lower than the actual rate I achieve; with x2, the reported connect rate is usually higher than the actual rate achieved. Even though the x2 and V.90 connect rates are almost the same, V.90 gives slightly better throughput and feels much snappier browsing web pages. One local user of that ISP that I am familiar with is getting much poorer performance with v.90: his Sportster connects at 36k with v.90, but if he disables v.90, he gets a 48-50k x2 connection! There are still unresolved issues with 3Com V.90 firmware that "are being worked on."

4/8/98 - KFlex problems - Many users are experiencing throughput and disconnection problems with 56KFlex modems. These problems are due to firmware problems in the client and/or server modems, and the solution somewhat more complicated. In the x2 world, all the client firmware originates from 3Com. In the Flex world, there are 2 flavors - Rockwell and Lucent. (The disconnect problem appears to be with Rockwell client firmware; Lucent firmware doesn't have the disconnect problem, but rarely achieves decent speeds.) With x2, 3Com has a near-monopoly on server equipment as well (so, except for Prodigy users, few x2 users experience the Bay Networks problem). With Flex, there is a greater mix of server equipment and firmware as well as the client. On a positive note, it appears that Lucent is nearing release of updated client firmware (version 4.28), and I've seen some encouraging information that would indicate a great improvement in speed for many users of Lucent-based Flex modems.

3/28/98 - 3Com V.90 for ISP servers released - The 3Com web site now has firmware available to ISP customers using its Total Control server equipment to upgrade to v.90. This means that any x2-enabled gateway using 3Com's Total Control equipment can be upgraded right now to support v.90! A message in the Usr-tc mailing list indicates that it might not be without problems: the owner of Roanoke Virginia Net indicates he is getting "major disconnects" since upgrading his equipment to V.90 early this morning.

3/22/98 - V.90 modem handshake To hear (RealAudio) V.90 modem handshake - click here.

3/22/98 - V.90 "Total kill"? Having upgraded my Courier to V.90 firmware, I've only been able to try long-distance V.90 servers. (The USR Line Test at 1-888-877-9248 is V.90 enabled.) My results have been unimpressive - the "total kill" of digital line impairments apparently isn't "total". I've been able to get connections ranging from v.34 (28.8) to V.90 from 32k to 50k to the USR test line; the V.90 connections still have a high error (bler) rate. I've also tried calling the Raleigh, NC IBM & Mindspring V.90 POPs (919-878-4801 and 919-425-7500), and have not been able to get a successful V.90 connection at all with 10 attempts.The improvements to the x2 portion of the code give me a big improvement with my local x2 connections. Since the most important factor for me (and most users) is local call performance, most of us have to wait to see how V.90 client to V.90 server will work on the local telephone network facilities. My guess is the "total kill" still needs improvement. (see 3Com message re: Total Kill.)

3/20/98 -

3/20/98 - Here's the (Courier) code: The 3Com/USR V.90 firmware for Courier (analog) modems (U.S.) has been released and is available in a toll call wizard version.and a web self-extracting version. [If you want instructions on using the free version, click here.] Although there are virtually no V.90 servers to call, there are improvements to the x2 code in this release. Some people (like myself) that have been unable to achieve good x2 connections may find that this release is a big improvement. So far, I've gone from over 90% unusable x2 connections to over 50% usable with some connections giving me a 4.5k throughput. I find that some calls connect unreliably (high error rate), but that disconnecting and re-dialing will often get a much better connection. If you have a Courier, check to see if you have the 7/97 release of the firmware as well - you can always flash it back to the earlier release if it doesn't give you better results. If you don't have the 7/97 release, at the present time it is still available for download here.(Note: there are V.90 servers set up in Raleigh-Durham, NC: IBM Global Network at 919-878-4801 and Mindspring at 919-425-7500)

3/16/98 - Bogus x2 connect speed? Check out this item at 56k.com. An unidentified source tells how x2 modems report their initial connect speed: it is only an estimate - the modem estimates what it thinks it will be able to achieve, then actually connects at 32000bps and trains upwards - if it can! (I think this account is accurate based upon personal experiences.) The link above is part of coverage resulting from Boardwatch's article about x2 and Flex, as covered in my 3/6 update.

3/15/98 - What's your throughput? I've added a new page showing how to use Net.Medic to monitor your modem throughput. Throughput (not initial connect speed) is really what matters!

3/14/98 -

3/14/98 - Where's the code? I would guess that bugs and problems have changed the release schedule for V.90 firmware upgrades for 3Com/USR modems. Some owners of Sportsters for which V.90 code has been released are reporting problems ranging from not being able to get an x2 connect, to not being able to connect, to getting a 'brain-dead' modem after running the wizard. 'Brain-dead' modems aren't recognized by the computer any more, and cannot be 'repaired' by the user. Unless you are already having trouble with a Sportster modem, I would advise holding off on upgrading your modem for now even if the V.90 code is available for your model. Most models with an original 3/9 availability date have now been changed to a 3/31 estimated date on 3Com's 56k site.

3/12/98

3/12/98 - How 56k Modems Work (and don't work) - My latest simple explanation of what's gone wrong with x2 and Flex - but maybe there's reason to be hopeful V.90 is the cure. Based in part on 3Com's admitting the x2-model of PAD compensation will not work (in some instances), and how its V.90 implementation is different.

3/11/98 - V.90 trouble reports appear in newsgroups. As noted in earlier updates, 3Com V.90 modems lack backward compatibility with some x2-enabled servers. An update to previously release V.90 code for the new 3Com modems has made the problem even worse for some. Certain Sportster models were scheduled to have V.90 upgrade code available on 3/9/98; the 3Com web site indicated the update was available; a number of customers made the toll-call to Skokie only to be informed they already had the latest code. 3Com blamed the problem on the weather, and reportedly told some customers it would be available by 3/12. Now, the web site indicates 3/15 (Sunday?).

3/8/98 - 3Com Responds to a message I posted. A 3Com engineer posts a lengthy response acknowledging x2 issues and digital pads, and thinks the problem will be licked with their release of v.90. I urge you to read the full text of this message. 3/8/98 - 3Com Responds to a message I posted. A 3Com engineer posts a lengthy response acknowledging x2 issues and digital pads, and thinks the problem will be licked with their release of v.90. I urge you to read the full text of this message.

3/6/98 - The 56k Modem Battle - 3/6/98 - The 56k Modem Battle - March cover story by Jack Rickard in Boardroom. Extensive and interesting article which also produces test results calling 445 POPs around the country. Only problem is this test is useless: as the author points out, the way we use our modems has changed: we make local calls to primarily one ISP number. The tests done by Boardroom used 444 toll numbers. x2 tends to work much more consistently on toll calls, Flex may do a better job on local calls. The article also talks about USR's x2 development and testing - it is consistent with developing a technology that works differently than the way we use the product: it works on toll calls (which is how USR tested and developed it), but doesn't work very well on many local calls - which is how we use it.

3/5/98 - Excellent update on v.90/56k issues in Brian McWilliams' Mar. 5 PCWorld NewsRadio Report. 3Com confirms some bugs and V.90 issues, and also indicates it will make updated firmware available without charge to ISP customers that do not have a service contract. Presently, contracts costing thousands of dollars are required for ISP customers to get the software needed to fix bugs in the x2/V.90 3Com products. This is certainly a step in the right direction. I urge you to listen to this report for yourself. (A text version of this item is also available.) 3/5/98 - Excellent update on v.90/56k issues in Brian McWilliams' Mar. 5 PCWorld NewsRadio Report. 3Com confirms some bugs and V.90 issues, and also indicates it will make updated firmware available without charge to ISP customers that do not have a service contract. Presently, contracts costing thousands of dollars are required for ISP customers to get the software needed to fix bugs in the x2/V.90 3Com products. This is certainly a step in the right direction. I urge you to listen to this report for yourself. (A text version of this item is also available.)

3/5/98 - I install 3Com V.90 update in a Sportster 56k Internal Faxmodem for a friend. I was on a social call, and she went on the Internet. Connected at 33.6k to IBM's x2 line. There also appeared to be a big slowdown due to retrains or errors, but it's so hard to tell with an internal modem! I called other local numbers, and found she was getting wild variations in initial connects - 33.6 to 51k, but poor throughput on all. I downloaded the 3Com wizard from the web. The wizard called Skokie on a call that lasted about 6 minutes. When it was finished, the modem (which had 3/97 code) was updated to v.90. Calls then became 48k - 51k and the throughput difference was dramatic. I can hardly wait to try the new Courier code - which now supposedly will be released before the end of this month. While I have blasted 3Com, 56k, x2 and v.90, it does appear they are making progress. I still, however, believe this protocol will mark the end of analog modems and will never be rid of some serious problems before all have gone full digital. 3/5/98 - I install 3Com V.90 update in a Sportster 56k Internal Faxmodem for a friend. I was on a social call, and she went on the Internet. Connected at 33.6k to IBM's x2 line. There also appeared to be a big slowdown due to retrains or errors, but it's so hard to tell with an internal modem! I called other local numbers, and found she was getting wild variations in initial connects - 33.6 to 51k, but poor throughput on all. I downloaded the 3Com wizard from the web. The wizard called Skokie on a call that lasted about 6 minutes. When it was finished, the modem (which had 3/97 code) was updated to v.90. Calls then became 48k - 51k and the throughput difference was dramatic. I can hardly wait to try the new Courier code - which now supposedly will be released before the end of this month. While I have blasted 3Com, 56k, x2 and v.90, it does appear they are making progress. I still, however, believe this protocol will mark the end of analog modems and will never be rid of some serious problems before all have gone full digital.

3/3/98 - 3Com V.90 bugs? Message from Eric at Delaware On-Line 3/2/98: "We have had three customers in the past two days call because they upgraded their x2 modems with the wizard from the 3com site and it apparently screwed up their x2 connections. When they call into our x2 modems they get screens full of gibberish when i had then open a terminal ascreen after dialing in win 95. They are able to connect to our flex modems but any attempt to make x2 connections results in failure. This is on multiple chassis in different locations." 3/3/98 - 3Com V.90 bugs? Message from Eric at Delaware On-Line 3/2/98: "We have had three customers in the past two days call because they upgraded their x2 modems with the wizard from the 3com site and it apparently screwed up their x2 connections. When they call into our x2 modems they get screens full of gibberish when i had then open a terminal ascreen after dialing in win 95. They are able to connect to our flex modems but any attempt to make x2 connections results in failure. This is on multiple chassis in different locations."

3/3/98 - Should you upgrade your Sportster? Because 3Com wants to restrict access to Sportster firmware, owners of such modems take a risk when upgrading. If the new firmware causes problems, you can't flash the modem back to the earlier releases without contacting 3Com and having them send you the old code. 3Com is encouraging x2 owners to upgrade their firmware to V.90 even though there is no server firmware because it includes updates to the x2 firmware as well. Brian McWilliams' PCWorld Radio News 3/2/98 has a story with soundbite from 3Com on this subject. 3/3/98 - Should you upgrade your Sportster? Because 3Com wants to restrict access to Sportster firmware, owners of such modems take a risk when upgrading. If the new firmware causes problems, you can't flash the modem back to the earlier releases without contacting 3Com and having them send you the old code. 3Com is encouraging x2 owners to upgrade their firmware to V.90 even though there is no server firmware because it includes updates to the x2 firmware as well. Brian McWilliams' PCWorld Radio News 3/2/98 has a story with soundbite from 3Com on this subject.

3/3/98 - Where's the Code? Some 3/3/98 - Where's the Code? Some

3/2/98 - 3Com V.90 modem - NO MORE MONEY-BACK GUARANTEE? - The 3Com V.90 modems are on the shelf at the Computer City in Honolulu. I was able to personally examine the package. Conspicuously absent from the new V.90 modem is the 30-day money-back guarantee. The product has a "lifetime limited warranty". Old Sportsters still on the shelf have a 3Com-backed money-back guarantee. Maybe the folks at 3Com have figured out they aren't going to get anywhere near a 95% success rate for 56k connections so they just can't afford to give a money-back guarantee. The picture of the V.90 external modem on the box looks identical to the Sportster x2 except for cosmetic (name & graphics) differences and are priced $10 higher than the already pricey Sportsters. A healthy stock of external 28.8 Couriers (at $269) on the shelf as well - but no 56k Couriers at all. [See prior Courier item in 2/21/98 update]3/2/98 - 3Com V.90 modem - NO MORE MONEY-BACK GUARANTEE? - The 3Com V.90 modems are on the shelf at the Computer City in Honolulu. I was able to personally examine the package. Conspicuously absent from the new V.90 modem is the 30-day money-back guarantee. The product has a "lifetime limited warranty". Old Sportsters still on the shelf have a 3Com-backed money-back guarantee. Maybe the folks at 3Com have figured out they aren't going to get anywhere near a 95% success rate for 56k connections so they just can't afford to give a money-back guarantee. The picture of the V.90 external modem on the box looks identical to the Sportster x2 except for cosmetic (name & graphics) differences and are priced $10 higher than the already pricey Sportsters. A healthy stock of external 28.8 Couriers (at $269) on the shelf as well - but no 56k Couriers at all. [See prior Courier item in 2/21/98 update]

3/2/98 - Where's the Code? 3Com site says 2/27/98: The 3Com site for V.90 upgrades has been changed, and changed and changed - and my prediction about no code being available for Courier & Sportster x2 owners has held true. A check of the 3Com site still produces "check back soon" for Courier models; certain Sportster models (including 1787-01 and 1789-01) give an availability date of 2/27/98, and the text asks you to check back after that date! Other Sportster models have different dates or messages (like 1785 = est. date of 3/15/98). The 3Com site in my opinion is terrible - you have to go through many hoops and pages to get to this message - you can't bookmark the page for your information. If you want to see for yourself what it looks like at this very minute - here's the3Com link. (Also - see prior updates on v.90/where's the code below) 3/2/98 - Where's the Code? 3Com site says 2/27/98: The 3Com site for V.90 upgrades has been changed, and changed and changed - and my prediction about no code being available for Courier & Sportster x2 owners has held true. A check of the 3Com site still produces "check back soon" for Courier models; certain Sportster models (including 1787-01 and 1789-01) give an availability date of 2/27/98, and the text asks you to check back after that date! Other Sportster models have different dates or messages (like 1785 = est. date of 3/15/98). The 3Com site in my opinion is terrible - you have to go through many hoops and pages to get to this message - you can't bookmark the page for your information. If you want to see for yourself what it looks like at this very minute - here's the3Com link. (Also - see prior updates on v.90/where's the code below)

2/26/98 - 2/27/98 - Interesting article on 56k modems at PCWorld - indicates about 20% of survey respondents getting a 50% or better speed improvement over v.34; 20% getting no gain, and 14.5% feeling cheated - calling it a "scam". Also, Brian McWilliams at PCWorld News Radio has interesting 56k items in his 2/24, 2/25, and 2/26 broadcasts. 2/26/98 - 2/27/98 - Interesting article on 56k modems at PCWorld - indicates about 20% of survey respondents getting a 50% or better speed improvement over v.34; 20% getting no gain, and 14.5% feeling cheated - calling it a "scam". Also, Brian McWilliams at PCWorld News Radio has interesting 56k items in his 2/24, 2/25, and 2/26 broadcasts.

2/26/98 - V.90 problem solved - the problem with Dean's new USR V.90 modem solved by disabling v.90. (S32=66). Until 3Com, Bay and Split Rock fix the bug, this will affect users of the new USR V.90 modems dialing into some x2 providers.

2/26/98 2/27/98- Where's the code? Many sources on the web including the on-line press still have reports that 3Com/USR V.90 upgrade code (at least for Couriers) will be available by the end of February. 3Com's site has been changed over and over, and the best you can find now is " The V.90 software has not been released, please check back soon!" In a Reuters feature 2/22/98 by Neil Winston, some awful web sites were mentioned - one of those singled out was 3Com: the author complained "An enquiry to the 3Com page about Palm Pilot problems triggered a lengthy automated response seeking answers to questions already provided in the original communication." 2/26/98 2/27/98- Where's the code? Many sources on the web including the on-line press still have reports that 3Com/USR V.90 upgrade code (at least for Couriers) will be available by the end of February. 3Com's site has been changed over and over, and the best you can find now is " The V.90 software has not been released, please check back soon!" In a Reuters feature 2/22/98 by Neil Winston, some awful web sites were mentioned - one of those singled out was 3Com: the author complained "An enquiry to the 3Com page about Palm Pilot problems triggered a lengthy automated response seeking answers to questions already provided in the original communication."2/26/98 2/27/98- Where's the code? Many sources on the web including the on-line press still have reports that 3Com/USR V.90 upgrade code (at least for Couriers) will be available by the end of February. 3Com's site has been changed over and over, and the best you can find now is " The V.90 software has not been released, please check back soon!" In a Reuters feature 2/22/98 by Neil Winston, some awful web sites were mentioned - one of those singled out was 3Com: the author complained "An enquiry to the 3Com page about Palm Pilot problems triggered a lengthy automated response seeking answers to questions already provided in the original communication."

2/26/98 - 3Com is a criminal! At least, that's what I believe. The release of x2 to the market has now past 1 year. I got a glimpse of 3Com's upcoming marketing strategy by following a link to an Adweek article Les Jones' http://www.56k.com/ The article indicates 3Com is less than satisfied with 56k modem sales to date, and gives an idea of the marketing hype we're going to be getting from them soon. It just seems to me that they're still making a big mistake - if they came right out with all they know about when 56k works and when it doesn't, and marketed it for what it is - "maybe better than a 33.6 modem - and subject to change for better or worse" - and fixed it so it at least gives reliable v.34 when 56k fails, they'd be much better off. Instead of spending all the bucks for hype, put it into a plan that lets users try the modem with a risk/cost-free trial, and provide some real support. As it stands right now, I never want to buy or recommend any 3Com product. In my opinion, their marketing of 56k modems to date has been fraudulent and it appears to me a class action is the only chance of getting any relief. (In addition to the stockholder class action, two class-action suits were filed in Marin County, Calif., superior court -- one against USR and the other against the companies supporting K56Flex -- claiming false and deceptive advertising. This suit was filed about a year ago, and I have heard nothing about its current status. Link to 4/97 PCWorld report on the lawsuit.) 2/26/98 - 3Com is a criminal! At least, that's what I believe. The release of x2 to the market has now past 1 year. I got a glimpse of 3Com's upcoming marketing strategy by following a link to an Adweek article Les Jones' http://www.56k.com/ The article indicates 3Com is less than satisfied with 56k modem sales to date, and gives an idea of the marketing hype we're going to be getting from them soon. It just seems to me that they're still making a big mistake - if they came right out with all they know about when 56k works and when it doesn't, and marketed it for what it is - "maybe better than a 33.6 modem - and subject to change for better or worse" - and fixed it so it at least gives reliable v.34 when 56k fails, they'd be much better off. Instead of spending all the bucks for hype, put it into a plan that lets users try the modem with a risk/cost-free trial, and provide some real support. As it stands right now, I never want to buy or recommend any 3Com product. In my opinion, their marketing of 56k modems to date has been fraudulent and it appears to me a class action is the only chance of getting any relief. (In addition to the stockholder class action, two class-action suits were filed in Marin County, Calif., superior court -- one against USR and the other against the companies supporting K56Flex -- claiming false and deceptive advertising. This suit was filed about a year ago, and I have heard nothing about its current status. Link to 4/97 PCWorld report on the lawsuit.)

2/25/98 - V.90 3Com modem in some stores - no server side V.90 support yet, so the modem is supposed to connect with x2. I've exchanged e-mail with a purchaser of this new modem (USR #5687) who no longer can get a 56k connection to his local ISP. (He does get a connect - at 54666 - on long-distance to USR's BBS and IBM's x2 toll-free access lines.) Click here to read the details. 2/25/98 - V.90 3Com modem in some stores - no server side V.90 support yet, so the modem is supposed to connect with x2. I've exchanged e-mail with a purchaser of this new modem (USR #5687) who no longer can get a 56k connection to his local ISP. (He does get a connect - at 54666 - on long-distance to USR's BBS and IBM's x2 toll-free access lines.) Click here to read the details.

2/21/98 - 3Com announces they've shipped V.90 ITU standard modems - and they're "on the trucks rolling towards you." Why then are they unable to make the firmware available to users who already have their "guaranteed upgrade" to v.90? Or even give a date? (See more below.) I've seen inventory levels of a major national distributor - before any V.90 modems were shipped, they had in stock over 46,000 Courier modems, and over 17,000 Sportsters. Unless the 46,000 Couriers are re-packaged and re-flashed, it will be a long, long time before you'll see any V.90 Couriers! (The Courier inventory is way out of balance - the distributor has around a year's supply of Couriers! About 20% of this Courier inventory is 28.8 models! Could this have any relationship to the 3Com stockholder suit alleging misrepresentation by 3Com on [inventory] problems relating to its purchase of US Robotics?)

Meanwhile, one of their press releases states in part:

Real World Performance

3Com's V.90 analog modems include a new and advanced line-probing sequence, developed by 3Com, that more accurately predicts phone line conditions. Though the initial connect speed of a V.90 call may appear to be lower than an x2 modem call, the throughput may actually be higher. The improved probing sequence is possible because of 3Com's understanding and experience with the global telephone network. This new technology will provide 3Com customers the most stable and consistent performance in real-world applications. (See http://my.excite.com/news/bw/980217/3com-2)

Be prepared to find this 56k technology, supposedly delivering nearly twice the speed of 28.8k v.34 modems, delivering about the same performance as v.34 in a large number of cases. Say goodbye to phony high connect rates; say hello to new, phony, less exaggerated connect rates.

2/19/98 - I install a new, second phone line - It gives me the same poor results as my old line - about 36k maximum 'reliable' connect rate - still with significant number of blers, retrains, and speedshifts. I take my modem to a neighbor, and get even worse results - 33.3k retraining to v.34 rates. I try an account with the one ISP in the Kona CO; get connected at 46.6 to 48k always, but the throughput in about the same as my 37.3k connects to Hilo - due in large part to the greater number of blers, speedshifts and retrains on the higher-speed connection.

2/16/98 - New x2 server modem code 3Com has released new firmware for ISP server x2 modems (version 5.8). Their "Total Service" site has documentation dated 2/8/98 for this release. (I can't read it, because it requires a service contract to access.) Notice a rise in newsgroup postings of people 'suddenly' having problems with their x2 connections. On the plus side, I am now, for the first time ever, able to get an x2 connection with the local ISP served in a GTD5 switch. And I'm getting 46-48k - higher than I'm getting to all the other ISPs in the DMS100 office! I took my modem to another site 5 miles away served by a different GTD5 central office. I was able to get 46-48k to ALL the ISPs in DMS100 and GTD5 local offices. I took my modem to a third site 1/2-mile from me and served by the same CO, and got the same results I get on my line: 33 - 37k to the DMS100 and 46-48k to the GTD5 ISPs. Now, it's also possible that the phone company has changed something because I am also now able to get a Flex (albeit poor - 32k which downtrains to v.34 shortly after connect) to the GTD5-served Flex ISP. The Flex performance to the DMS100 is also poor 32-34k.

AFTER ITU action, 3COM/USR's site last week had a new page for "Upgrades" which said to check back on February 16 for more information. As of February 16, the site still has no information, but the message has been updated to check back "later this month" for information on upgrades to the V.90 standard. By about February 27, I expect it to be changed again - and my bet would be there still won't be any information - it will just tell you to check back in March.... or April.... or next year... on next century.... Update 2/21: Well, the USR site (http://www.3com.com/56k/upgrade.html) already has been changed - now it is "check back soon". AFTER ITU action, 3COM/USR's site last week had a new page for "Upgrades" which said to check back on February 16 for more information. As of February 16, the site still has no information, but the message has been updated to check back "later this month" for information on upgrades to the V.90 standard. By about February 27, I expect it to be changed again - and my bet would be there still won't be any information - it will just tell you to check back in March.... or April.... or next year... on next century.... Update 2/21: Well, the USR site (http://www.3com.com/56k/upgrade.html) already has been changed - now it is "check back soon".

Update: 2/10/98 2/10/98 Expect more confusion and new problems as new firmware is introduced to support the forthcoming 56k standard officially labeled v.90. Only time will tell how V.90 will work, but it's my bet that we have a new standard in standards: a modem protocol that will work for only an ever changing subset of users.