11/18/98 - Interoperability update - Interoperability page - updated today. Thanks to Stanley Diamond for this link to CMP's item on V.90 interoperability.

11/6/98 - FCC delays decision - See the 10/18 update below on telco efforts to get calls to your ISP treated more like toll calls; the FCC was expected to make a decision today, but delayed it as reported in this ZDNN story. Also, see this ZDNN story on FCC action to take jurisdiction over ADSL service. Even though the service is entirely local until you get to an ISP, apparently the telcos feel more comfortable with FCC regulation of this emerging access technology.

11/2/98 - I sell 3Com - On 8/15/98, as rumors circulated that 3Com might be a target of other companies, I disclosed that I had on 7/30 purchased 200 shares of 3Com in my IRA. Today, I closed the position, selling 200 shares @ 36 7/8, a gain of over 40%.

10/31/98 - WWW [World-Wide-Wait] - Until we have fiber to the curb (what are you waiting for Bill Gates?), we've got to come up with band-aid solutions. Like 56k modems. ISDN. ADSL. -Or software... I received a message asking if I'd tried Webcelerator. This is software that intends to speed up your net surfing experience by (among other things) prefetching web pages while your dial-up connection is otherwise idle. If it has downloaded a link that you click on, it pops up like you had fiber to the curb. But, it can be terribly inefficient: while you're reading this, it may be downloading pages from this site that you'll never look at and storing them on your hard drive. Within the last 48 hours, someone from Texas has been accessing my site with software that constantly downloads the same pages over and over (I don't think Webcelerator is that bad...) - over 3500 pages from my site in a 10-hour period. [And Mr. Flash.Net customer, I'd like you to stop!] This kind of solution to the world-wide-wait can only get less effective: as more people use it, the internet backbone and ISP's modem->internet backbone usage will clog with wasted data - pages people don't look at; log files that reflect pages not viewed, etc. There is a price to bandwidth - and this solution, while effective for the first few millions, is a band-aid that will not easily, or painlessly be removed. I know you're curious now - so here's the site that's going to wreak havoc on the net if too many people use it.....

10/29/98 - Site update - My hosting service and my site have been experiencing trouble for a week. I think it's finally working properly; hopefully I'll be able to make some revisions and updates to v90 interoperability. The Throughput Page has been revised with comments about the upstream modem rate achieved on 56k connections: it is often lower than you'll get with a v.34 connection - so if you are doing heavy uploading, you may be better off with 56k disabled. Note that the hit in the upstream rate varies depending upon both the client and server modems. I believe you're least likely to take an upstream rate hit with 3Com client to 3Com server. The Who Manufactured My Modem? page has been updated to indicate not all IDs (especially on some "LT Win Modems w/ Lucent chipset) are found in the database. The LT Win Modem page has a new link for the somewhat elusive 5.23 firmware revision.

10/27/98 - Zoom updates - Zoom has released new firmware for a number of its 56k models - including the 2949 that I have. This October release for the Rockwell-based modem finally allows me to get some v90 connections where the previous firmware (2.081) couldn't make 56k connects on my line; however, the rates achieved are 10+% lower than I achieve with my Courier and LT Win Modem. I also get "bad" calls with the Zoom where although a connect of 37.3k is reported, throughput is around zero. This firmware is better, but still has a long way to go. Zoom has also done one of the craziest things I've seen yet: it is now using Lucent chipsets in some modem models without changing the model number! Apparently, you could now have a 2949 or 2919 model with either a Rockwell or Lucent chipset, and of course each requires different firmware! Talk about confusing. And their revamped website (www.zoomtel.com) is rather confusing as well.

10/18/98 - Telcos paying 100's of Millions to CLECs - Local telcos are paying small CLECs (competitive local exchange carriers) money to complete 'local' call to ISPs. C|Net reports on telco efforts to get the FCC to treat these calls more like "long distance."

10/13/98 - Winmodem page - New to the site today - Winmodems.

10/12/98 - Modem market heading into death spiral? - On the same day as my Hayes chapter 11 bankruptcy update below, ZDNN published this story indicating stormy weather ahead for other modem makers.

10/9/98 - Hayes files for bankruptcy - The newswires are running stories of today's filing by Hayes for bankruptcy. Hayes was the early leader in the PC-modem business, inventing the AT-command method we all take for granted today in controlling our modems. See more here. Who's next? - How about Zoom? [Any company that makes primarily [analog] modems is a contender for going bust.] Over the weekend, news:comp.dcom.modems was inundated with posts from a Zoom tech support employee - primarily with repeated 'boilerplate' information. I found particularly interesting Zoom's suggestion that the user try a very long list of initialization strings, and the comment of one user that he had better things to do than spend the hours it would take to follow this suggestion.

10/9/98 - Dell notebook modems - The Dell Inspiron™ Notebook is available with an LT Win Modem with Lucent chipset. As you can see from this user page, (or the Dell site), there is a lack of support from Dell for the modem, but Dell owners are finding my LT Modem page. Actiontec apparently is the OEM supplier of the modem to Dell. The Actiontec site is confusing, doesn't have the latest firmware, and doesn't mention Dell. Is this any way to sell and support a high-end notebook?

10/9/98 - GTE.net ending unlimited plan - GTE Internet is ending it's unlimited access plan effective November 1, and "heavy" users are starting the exodus to other ISPs. While this site is about modems, I want to comment: I think all users should move from a provider switching to a limited plan: even if you use less than 100 hours, there may come a time when your on-line time increases. While the future may see the demise of unlimited Internet access, the harder we make it for companies imposing limits to be successful, the further away that metered future becomes.

10/5/98 - New firmware for I-Courier - 3Com has released new V.90 firmware for the internal Courier-I (ISDN) modem - version 2.7.7. 10/2/98.

10/2/98 - v.91 is coming - Point-Topic reports that the ITU has determined an "all-digital-mode" (v.adm) extension to V.90 and has assigned this new, forthcoming standard v.91. As I understand it, this extension would work for customers that have ISDN or all-digital trunk connection to the telephone network, allowing a 56k+ speed both up and downstream. Today's analog modems (including the Courier) would not be (easily) upgradeable to this standard as the modem would be putting out a digital stream to an ISDN or similar interface instead of an analog signal to a POTS (plain-old-telephone-service) line. Since many phone companies charge more for digital-data ISDN calls, and many ISPs charge more for an ISDN-account, v.91 would allow customers who have ISDN service to get near ISDN speeds with a call that appears to the telephone network and the ISP to be a POTS call.

9/29/98 - Lucent releases V.90 for Portmaster - After 21 beta releases of v90 code, Lucent/Livingston last week released v90 firmware for the Portmaster access equipment used by many ISPs. A number of ISPs are asking "what's changed" in modem support from the beta code, and while the answer is unclear, it may be "nothing". In the list I monitor, ISPs are still complaining about 56k/v90 interoperability, as well a v.34 interoperability problems with this release. Perhaps Lucent felt that since V.90 was officially ratified, they had to release "non-beta" code.

9/25/98 - 56k modem forecast - Item at 56k.com led me to an interesting site - Point.Topic. The 56k news page contains some interesting items including a V.90 modem sales forecast, and an 'extension' to V.90 - "all digital mode" which apparently will become a separate ITU standard, now referred to as v.ADM. Point-Topic's forecast for 56k analog modems is very rosy - but, like many in the industry, I think they wrongly attribute poor 56k modem sales to date to the 'lack of a standard', overlooking the failure of the technology to work for a significant number of users!

9/25/98 - How cheap for 56k modems? - How about $28.98 for a "56K Lucent PCI Modem"? Robert Lowrey alerted me to this modem (I wonder how Lucent would feel about them implying it is a 'Lucent' modem?) to a shopping site with this modem. BEWARE - while the 'price is right' - shipping is outrageous! They want $14.75 shipping for regular UPS surface service to a mainland address, and "can't compute" shipping to my Hawaii address. Robert indicates the modem is actually made by Encore, and comes with [old] V.90 firmware on CD; the Encore web-site has even older firmware. Robert is using the latest 5.20 firmware and getting 52k connects every time. More on the Lucent-based modems here. (9/30/98 - The description of the modem has been changed to "56k PCI modem with Lucent chipset")

9/24/98 - 3Com reports 49% 1Q Profit Decline - CNNfn reports 3Com "...hurt by price cuts on computer modems..." beat analyst estimates for the first quarter despite a 12% revenue reduction.

9/21/98 - RealAudio - A number of users have reported trouble playing the RealAudio clips at this site. I've made a change that I hope will eliminate the problem... and I hope it doesn't cause a new group to experience trouble....

9/18/98 - Interoperability - A number of ISPs using the latest 3Com server 'modems' - the HiPerARC/DSP - are reporting that users of Lucent LT Winmodems can no longer connect (the modems do an 'endless handhsake'), and that they didn't have the problem until they upgraded and replaced their 3Com Quad modems. Another problem of premature disconnects is also being reported. It is clear this is an interoperability problem - it is unclear as to whose problem (Lucent or 3Com) it is. 3Com is promising an ER (engineering release) of code for its server-side products "in the near future" to address as many of the Rockwell and Lucent interoperability issues that it is able to with fixes to 3Com server code.

9/17/98 - FCC to reconsider '56k restriction' - CNN reports today that the FCC yesterday proposed changing its rules to allow 56k modems to operate at "their maximum speeds." See the CNN story here.

9/17/98 - V.90 Standard approved - CMP/Computer Retail Week reports that the ITU has given the official approval to the V.90 standard in Geneva on September 15.  3Com has issued a press release applauding the ITU ratification of V.90, suggesting it will encourage adoption of V.90 with "universal compatibility". [Note: the 3Com release is marketing spin: while V.90 modems may be "universally compatible", they are not universally able to achieve greater than v.34 (33.6k) connections, nor does the standard do anything to solve these problems that occur in the diverse deployment of telephone network facilities.]

9/17/98 - Windows98 & modem problems - I've been getting quite a few messages from users of Win98 and various modems who are having difficulty using the modem at all with Win98. This site won't help you with those problems - I haven't even seen Win98 (except in a box) - and the problem isn't 56k-specific (ie, it's an OS/configuration issue). Note: chances of having this kind of a problem are greatly reduced if you purchase an external modem.

9/17/98 - Limiting connect speed - Users who are experiencing variability in 56k connect rates may want to automate the process. The 'Limiting Your Connect Speed' page has been updated with Jim Bell's contribution on how to do it with Lucent and Rockwell-based modems.

9/7/98 - 3Com/USR Sportster firmware - 3Com has released a new version (4.10.x) of firmware for some Sportster/USR x2/v90 modems. I'm not sure what is changed, or what models the firmware is available for; however, I do know that it does not address the RBS-related problem I encounter. Inasmuch as 3Com does not provide an archive so you can flash-back to the old version if you develop problems, I'd only recommend upgrading if you presently get lousy or no 56k connects. There is also no easy (free) way I'm aware of to check to see if your USR/Sportster modem has the latest firmware - you can navigate the confusing 3Com site all you want, but I think you'll have to use the 'Upgrade Wizard' and make a toll call to find out.

9/6/98 - Testing... Curiosity finally motivated me to unplug everything from my Aptiva with LT Win Modem and take it to the location where the Hayes/Lucent PCMCIA modem wouldn't work (not in my CO). I also got the Zoom piece-of-junk modem back (the vendor refused my return). The results confirm my earlier suspicion that there is a significant difference in firmware/56k performance for the Lucent LT Win Modem vs. the Lucent Venus chipset; the Zoom is a paperweight with the latest firmware - it worked better with the unavailable original firmware. The LT and 3Com modems perform almost the same; there is a difference in upstream rates depending upon the make of server equipment called. I present the results here.

9/5/98 - 56k Modem Chat & RBS - Today's 56k Modem chat contained some Q&A on robbed-bit-signaling. Jim Bell, who like me, gets a 1-in-6 good 56k connections, myself, and Ed from Lucent excerpt:

<JimBell> We were wondering about RBS
<edschulz2> What about RBS specifically?
<JimBell> Richard and I believe it is messing up our connections"remote to GTD5 to DMS
<edschulz2> Probably right.
> can it be fixed?
<EdSchulz> What's "it?"
<JimBell> RBS is it.
> ie, is it client modem firmware? server?
<edschulz2> Downstream, the client is boss....
> to fix it is to make 6-out of 6 ~48k
<edschulz2> Impossible, if my theory is right. All we can hope for is that you 34k calls get 40k or so.
<edschulz2> Your good calls will still be good, I hope.
> Doesn't sound like the best of news to me, ed....

More on my "RBS Theories" - here.

9/3/98 - Prodigy starting V.90 rollout - the only major provider using modems/concentrators from Bay Networks has begun deploying V.90 at some of its POPs. I've seen posts from users in the Atlanta and Tucson areas indicating they are getting v90 connects to Prodigy's network through its provider SplitRock. 9/6/98 - Confirmation that v90 is at the following Super POPs: Atlanta, Chicago, Salt Lake City, Oklahoma City, Phoenix, Tucson, and Dallas.

9/3/98 - 56k patent dispute - C|Net reports today on patent disputes brewing on PCM (56k) modem technology.

9/3/98 - Lack of communication - I've been fortunate to be able to communicate privately with people who know parts of the 56k puzzle - from 3Com, Lucent, GTE, Siemens, ANS, and others, as well as plenty of "ordinary" users of 56k modems. I was surprised recently when tech support from one company contacted me about a problem mentioned on my site. [I despise dealing with tech support for modem companies, because it's unlikely that I'll get to communicate with a person that knows what they're talking about!]. Notably absent from my contacts is Rockwell, and vendors of Rockwell-based modems. I've never received a reply to my support request from Zoom. What's wrong with you guys? I say the product sucks, and you don't disagree? My feedback form is there for you, too, and what you say is confidential if you check the "Do Not Publish" box. Here's a comment without the publish restriction:

Got a new Gateway 333MHZ, but unfortunately it came with a Telepath X2, whereas my ISP supported 56KFlex. I went in search of a 56Kflex modem. The first one was an Internal PCI ZOOM Flex. It would sporadically connect from 28K to occasionally 42K. But, it would not hang up and I had to pull the plug to reset it. I took it back, Next was a Diamond Supra, External 56K Flex. It would stop receiving every 2 minutes or so for up to 45 seconds. Had no better luck with an external ZOOM and an Internal Supra. Then I read your evaluation of the Lucent Technologies chipset and bought an internal PCI ActionTech, 56K Flex with a free upgrade, whenever my ISP switches over to V 9.0. (Including tax - $74.35) Plugged it in and have connected ever since at 44K !!!! Unreal. Downloads of JPG files run from 4.7K/Sec to 5.3K/Sec. I am totally in awe. Thanks for giving us technically challenged people an unbiased picture of what is out there. Without someome to steer you in the right direction, the options which are available are overwhelming to the novice.

Thanks again. Steered several of my friends onto your site.

9/1/98 - IBM Global Network for sale - The Wall Street Journal reported today that IBM has hired Merrill Lynch to auction the IBM Global Network in a transaction that could fetch over $2billion. In addition to offering consumer ISP accounts, the network handles corporate data for about 30,000 customers with POPs in 900 cities across 50 countries. See the story as reported by ZDNN.

8/31/98 - Analog->Digital conversions - 56k modems won't work with more than 1. If you have a 3Com/USR modem, you can determine if there is more than 1. Why don't other vendors show similar data? See the updated A/D Conversion Check page.

8/31/98 - Troubleshooting Page Updated - I've updated the Troubleshooting Page as well as various other pages in an attempt to update them with my current understanding of the state of "v.Unreliable".

8/29/98- Windows98 note - Various pages at this site refer to modemlog.txt - the file created by Win95/NT DUNS (dial-up-networking) connections. It should be noted that Windows98 creates a modem log file with a <your modem name>.log file name.

8/29/98 - Interoperability - Lucent/Livingston has released yet another V.90 'open' beta release for ISPs using the Portmaster equipment. Some ISP comments include: "The best yet, but still not ready for prime time." The latest release is 3.8b20. (Yes, that is 20 beta releases - not all of them 'open'.) What some ISPs miss in evaluating server code (whether it be from Lucent, 3Com, Ascend, etc.) is that variability in telco routing as well as 'flaws' in client modem firmware is a major factor for connectivity and interoperability. There are so many variations and possibilities and moving targets - one gigantic 56k mess. One observant PM3 user noticed that there is no change to the modem code with the new release - only in the core OS.

8/27/98 - Lucent client modems - I'm going to make a revision to Should You Buy a 56k Modem to qualify my "praise" for Lucent client modems. It appears based upon both personal experience (with a Hayes Acura PCMCIA modem) and from some messages I've received, that the firmware for the Venus chipset (used in Lucent-based external modems) is not nearly as good as the firmware for the LT Win Modem. Based on this, the only external v90 modem I would want to have at this point is one made by 3Com. As with all 56k modems, the firmware for Lucent external modems will work on at least some line conditions.... Here's a portion of 1 message I received:

Carrier speed commonly reported by my Hayes 4703 modem(s):
24000/60000/V.90/PROTOCOL: NONE/COMPRESSION: NONE
Yes, that's 60000! And it occurs with multiple ISP's, including ibm. (and as you'd expect, with no error correction protocol, the connection is so error-full, that PPP startup negotiation is guaranteed to fail).

Hayes finally replaced the modem after pulling a mouthful of teeth, and it still did this and other odd things. When I sent them mail saying they hadn't sent me the UPS return label that they said they would, they responded by sending me another modem (again without a UPS return label)! This one also does the same and similar unreliable things. This is with the modem set to factory defaults, except that two modems needed S109=2 in order not to choose K56FLEX over V.90 when talking to an ISP that can do both. This is despite all three now having the same (latest) firmware revision level. So... All three are now stacked one atop another alongside my fully functional Zoom 2948 (same as 2949, but comes with 2 extra diskettes which I can't use since I don't own a mac) which "zooms" along at anywhere from 45333 to 49333.

The user who sent me the above message also had some advice, and refusal of return for refund from Hayes:

"If the modem will connect with Flex then leave it that way until the V.90 standard has time to mature. We can not refund your money until we work on the problem further. "

8/21/98 - RBS & 56k - How RBS is messing up some 56k connections revised and updated to reflect my evolving understanding of 56k impairments.

8/19/98 - Speed Unknown - Looking at a priced higher specially for Hawaii Radio Shack flyer, I read the fine print in the ad for the Compaq Creative Learning Series 5032 ($1599/cpu only):

"ITU V.90 modems are designed only to allow faster downloads from K56 flex or V.90 compliant digital sources. Maximum achievable download transmission rates are currently unknown, may not reach 56Kbps, and will vary with line conditions"

Today's wording seems a little less rosy than it used to be....

8/17/98 - Cisco #1? - An 8/4 Reuters News Item headlined "Cisco takes top spot in Internet-dial up from 3Com" suggests that Cisco has become the #1 supplier of dial-up Internet access equipment. Cisco CEO says the company's dial-up access business has been growing far faster than competitors 3Com and Ascend Communications Inc.

Bi-weekly 56k modem chat -Starting 8/22, I'll be hosting a 56k modem chat every other Saturday starting at 11AM Pacific (2PM Eastern, 7PM GMT). You can find the IRC chat room here.

8/15/98 - 3Com for sale? Rumors circulated this week that another company (Intel, Lucent, IBM) may be interested in purchasing 3Com. I wish to disclose that on 7/30/98 (unaware of any rumors), I purchased 200 shares of 3Com stock in my self-directed IRA at $25-1/4. I made that purchase because even though I dislike many of the things 3Com/USR has done, I thought the company was undervalued. If 3Com is undervalued, it would seem logical that it may be a target for acquisition given Northern Telcom's purchase of Bay Networks, etc.

8/15/98 - Hayes Hurting due to 56k modem fiasco - Will close analog modem manufacturing (a "3rd party" will make Hayes 56k modems) to concentrate on the 'digital future'. With nearly all modem companies losing money and blaming 'poor sales of 56k modems', when will they 'fess up that poor sales are due in large part to their making products that just aren't ready?!!!

8/11/98 - Interoperability update - New Lucent/Livingston V.90 server beta & new problems and updated info on the Interoperability Page.

7/23/98 - New page for LT Win Modem - Information on the modem and firmware upgrades added today.

7/17/98 - Lucent LT Win Modem V.90 version 5.14 - Available from the Multiwave Innovation site. With 5.15 pulled from the Lucent site and still among the missing, 5.14 appears to be the latest currently available firmware version.

7/15/98 (7/10/98)- Lucent 5.15 firmware (sort of) - The Lucent web-site with firmware for the LT Win Modem has been changed. The 4.28Flex-only firmware is gone; currently, the link to the download page from their main LT page has been disabled. However, the link in the 6/14 update (which was posted in a newsgroup message on 6/14 by Ed Schulz of Lucent) provides a page to download V.90 version 5.15 drivers. It appears that Lucent isn't ready to 'release' these as the self-extracting zip file has errors, and different files have been downloaded by a reader who alerted me to this and myself. I was able to use pkunzip and get the WindowsNT driver (the Win95 .vxd file appeared corrupt). I have been able to use the 5.15 firmware: an immediate difference is a unique new DIL (digital impairment learning) tone. RealAudio handshake with DIL available here. This new firmware does not improve my performance - I still get an average of only 1 out of 6 calls with good connect speeds and throughput. 7/15 Note: As of 7/14 the page that contained the 5.15 firmware has been removed from the Lucent site; the latest available-for-download LT firmware now goes back to 5.10 - see the 6/28 update.

7/1/98 - More on RBS and 56k - How RBS is messing up some 56k connections and details on how 56k works (and doesn't work) - the 6/28 text is now updated and revised.

6/29/98 - Rockwell to spin-off chip division - Rockwell, Int'l. announced today it will cut 3800 jobs, and spin-off its semiconductor business to stockholders. The division will post a loss, attributed in part to a "weak PC modem market." 

6/28/98 - How RBS is messing up some 56k connections - If this impairment isn't identified properly (see 6/25 update), what happen?. My attempt to figure it out.

6/28/98 - Lucent LT Win Modem V.90 version 5.10 drivers at the Multiwave Innovation site. Note: This is a zip file, and indicates it is for the PCI version of the Lucent modem. It will work on the ISA version as well, using the driver copy method referred to in the 5/24 and earlier updates below. The zip file also contains a .pdf format document containing documentation for Flex (non v90)-specific commands and settings. I have found no difference in performance with these drivers and the 5.09 version. This site also has a WinNT driver for their 3Com-licensed x2 Winmodem (CommWave 56ES-x2). [See also 6/19 update re: WinNT/Winmodem driver.]

6/27/98 - Easy Throughput Check - Hawaiian Style throughput check. New today.

6/27/98 - PCWeek flawed V.90 review - 6/24 David Bardes PC Labs test of V.90 interoperability is seriously flawed. A central office emulator and a local loop emulator were used to conduct the tests. Problem is, these emulators do a lousy job of duplicating the real-world telephone circuits most of us use to access ISPs.

6/25/98 - GTE Customers - I am pretty certain that there is a large universe of GTE Telephone customers who will get good 56k connections on 1 out of 6 connection attempts. With x2, a false high rate will be reported on nearly all calls, but only 1 out of 6 (on average) will be good. With V.90 and KFlex, 5 out of 6 calls will connect in the low-mid 30's; 1 out of 6 in the upper 40's to low 50's. This situation will occur when: the customer is served by a GTD5 switch and integrated SLC, or by a remote GTD5 and is making a local call to an ISP in a DMS100 office. In some circumstances, it will also occur when the ISP is in a GTD5 office. Flex and V.90 users (that are on flat-rate calling plans) should limit your connection speed. The reason that 1 in 6 calls is good is that the odds of having 2 RBS-links [robbed-bit-signaling] align on any particular call is 1-in-6. When the links align, they rob a bit from the same frame; when they do not align, 2 bits out of every 48 instead of 1 is robbed. This extra robbed bit makes all the difference in a good vs. bad 56k connection with every release of 56k firmware I am aware of: Rockwell, Lucent and 3Com's x2, V.90, and KFlex. For a technical explanation of RBS alignment, see this archive item authored by Ed Schulz from Lucent.

6/25/98 - 3Com growth hampered by slow V.90 adoption - See this CRN story by Shawn Willett. (Comment from one ISP who saw the story: "Excuse me! We are freakin waiting for a damn server code that actually works.")

6/25/98 - Throughput! Update to the throughput page for Win95 Port Settings. I cannot stress enough that throughput, throughput, not connect speed, is really what matters!

6/23/98 - Worldcom/UUNet announces V.90 - UUNet/Worldcom today announced the "broadest deployment of the new 56kbps standard" saying that V.90 is now available at more than 700 US POPs. OK - I tried 4 POPs in Hawaii (a part of the US), and found V.90 "worked" at only 1 - the connect rate to that 1 (using Worldcom/MCI's horrible long-distance service) was 30,666bps! Your results may vary. The UUNet press release is here. (I tried again with the more reliable AT&T long-distance to Honolulu and got a 32k and a 48k connect. Many of my MCI voice toll calls sound like a flashback to the 1960's overseas toll circuits - and this is on a 150-mile call!)

6/19/98 - Sportster Winmodem Driver for NT - Many have asked over the past year how to run a USR Sportster Winmodem in WindowsNT. The stock answer of "experts" was you can't do it (some would even say "it's a WINmodem [stupid]", as if WindowsNT is not Windows. I've believed for some time there was such a driver- and there is now proof there is: Micron WindowsNT 4.0 driver for USR/Sportster Winmodem. (Micron calls the Sportster Winmodem a "Greyhound".) At this time, it is available for download; but, if you want to try this driver, I'd suggest getting it quickly as 3Com usually makes anyone who posts their firmware yank it once they find out about it. It is available in V.90 flavor for Windows95 and WindowsNT.

6/19/98 - 3Com/Sportster firmware without toll call - On the Micron site. The files (for Micron's USR/Sportster modems) appear to allow flash update to V.90 without making the toll call as 3Com/USR requires for its retail versions. If you are interested in these files, I'd suggest getting to them quickly, as 3Com may force Micron to yank it. The readme file indicates that the software can be used to flash up to 100 times before expiring 4/6/99.

6/19/98 - Where's the code (that works?) 3Com finally released V.90 code for some models last week, and there are a number of reports of problems: many users who were getting x2-connects no longer able to get any 56k connects. Check this newsgroup post with a recommendation to users having this problem. Zoom, already having dual-mode modems that won't make a V.90 connect, has released V.90 upgrades for some of its prior models. User reports indicate that this firmware is 'less than capable' of V.90 connects as well - your results may vary.

6/14/98 - Lucent adds LT Win Modem drivers to website - While Lucent sells chipsets to modem makers as opposed to modems to end-users and has told users they have to get firmware from the supplier of the modem, they recently added (*old*) firmware to their website. Note that V.90 firmware is available from a number of Lucent partners (see 5/24 update below), while the Lucent site currently has only Flex version 4.28 firmware. (See the 7/10 update)

6/14/98 - Should you buy a 56k modem? New to this site today.

6/9/98 - IBM says v.Unreliable is "normal" - IBM's Internet Help Desk says that if 56k worked for you and suddenly stopped, or never worked and suddenly starts working, all is "normal". See the item on their help page.

6/8/98 - Bugnet/MSNBC "line noise" with V.90 - Check this 6/5 story on V.90 modems. The author appears to attribute all V.90/56k problems to "line noise" - even calling an a/d conversion "line noise". The author, Bruce Brown, is either clueless, or is trying to make the technical reality sound simple, but the bottom line is: 56k=v.Unreliable. (I'd vote for Brown being clueless - you know something's wrong when they call the modems 56.6 and have a survey asking for your connect rate with choices of 56.6, 33.3 or 28.8!) [8/98-Story has been removed or moved from MSNBC site.]

6/6/98 - Flaky modem I found that my Courier v.Everything modem was "bad" - causing my recent problems (see 5/25 update ). 3Com replaced my Courier, and now I'm back to decent 48k V.90 connections (although it still takes an average of 6 or so calls for me to get 'routing' that permits a 46-48k rate).