Summary: I just purchased and installed my brand, new WPNT834 router and I'm generally pleased, I suppose, especially considering the trouble I've had with routers, in general (I'll get to that later, if only to satisfy your curiousity).
The WAN performance is good, as least as far as I can tell, as the speed test at broadbandreports.com tells me I'm getting just over 9 megabits per second, which is my ISP's current limit. The LAN performance is good. The wireless performance is excellent even though my speed is limited to 54 megabits per second due to the fact that my laptop, an IBM T-41, has a built-in a/b/g wireless NIC and that's its maximum speed. The feature set is good though, as others have pointed out, there's no access to the Airgo chipset's QoS features.
Here's a list of the problems I've discovered so far:
1) I find that after a few hours of operation, I can no longer access the router's html server; I've tried several browsers and all report that the server is busy. Recycling power to the WPNT834 re-enables the html server for another few hours. While I am not able to access the router's html server, the router continues to function properly.
2) After saving the configuration (I'm not sure if this is the trigger or if it's random), the router's html server appears to lose contact with the router's internal data. All settings appear via the web interface to be the default, which is not possible as the router continues to function normally.
3) After recycling power to the WPNT834 to fix either 1) or 2), the router occasionally loses access to my ISP's DNS servers. The proper server addresses are listed on the connection status screen, but no DNS services are available. I must release and renew the router's lease to my ISP's DHCP server to solve the problem. This problem is intermittent, occurring twice in, perhaps, seven router reboots.
4) I cannot access the router's html server remotely even though I enabled and set-up the Remote Management feature to be as permissive as possible for experimental purposes. If I replace the WPNT834 with my ancient D-Link DI-704, I can access the latter's html interface remotely, so I can only assume that this is yet another firmware problem with the WPNT834.
5) My TiVo, which has a wireless NIC, is assigned an IP address by the WPNT834's DHCP server, but the TiVo doesn't show-up in the router's client list and an entry in the router's persistence table, with the TiVo's NIC's MAC address, is ignored.
I have reported all of these problems to Netgear, via detailed text at their web site. Each problem was answered to the effect that the problems could better be dealt with over the phone. When I called, I found that the level-one technician did not have a good grasp of colloquial American English. He persisted in trying to help me debug wireless speed issues, though I did not report that as a problem and, in fact, am happy with the wireless performance. Eventually, after a fruitless hour on the phone, he queued my problems to level-two. I haven't yet heard from the level-two technician, so I don't know if my problems will be solved, but I sincerely doubt that they will be.
For it is my feeling that none of the home networking companies I've dealt with (D-Link, Linksys, Netgear, Belkin) are set-up to actually solve problems with their badly written firmware, and they all have badly written firmware. Rather, they are geared toward solving problems caused by the end user and/or replacing defective hardware. The result is that if one buys a product with bad firmware, and the manufacturer decides to do nothing, there is no recourse but to return the product for a refund.
__________________________________
Now, my saga, which, I believe, points out the woeful state of affairs in the entire consumer electronics field. To compete, and to satisfy their marketing departments, manufacturers have loaded on new features at an amazing rate but they rarely if ever seem to test their new products sufficiently. There are some exceptions, but the Netgear WPNT834 is clearly not one of them.
At the dawn of the PC networking era in the 1980s, I installed a number of Netgear hubs and switches, having found them to be designed and built very well, this despite their competitive (to say the least) price. When I first got broadband access at home in the fall of 1999, however, only D-Link had home routers and I got their first four port router (the first of its type on the market, in fact), a D-Link DI-704, in December of that year. It served me well, and still serves my in-laws, this despite its WAN speed limit of 3.5 megabits per second. I replaced it just about two years ago with a D-Link DI-774 wireless router. I paid a premium for that, because it supported WiFi a/b/g, but found numerous fatal flaws in its firmware, this despite the fact that I bought it, apparently, at the end of the product life cycle. After much prodding and many threats, D-Link replaced it with a DI-784.
A week ago Wednesday night, after almost two years of nearly flawless operation, my D-Link DI-784 died. All of its panel lights came on solid and no amount of power recycling or reset button pushing will coax it to function. As it's out of warranty, I ran out the next day to replace it, installing yet another DI-704 I had laying around as a non-wireless fill-in. I bought a Linksys WRT54GX2 at Staples, having read a number of glowing reviews of it. I got it home and discovered that it had two hardware problems. One of the wired RJ-45 ports was dead and the WiFi radio kept turning itself off. I exchanged it the next day for another, which worked well enough for me to discover a myriad of design flaws. 1) While its WAN speed is good, its LAN speed is miserable, much worse than even the old D-Link DI-704. 2) It has no DHCP persistence table. Apparently, Linksys products generally do not, which, if one uses DHCP, renders application routing unreliable as one's servers change IP addresses when rebooted or reconnected. I tried switching all of my systems to static IP addresses, but the Linksys WRT54GX2 assigned them IP addresses, anyhow. These phantom addresses were not reachable, but the fact that they show-up in its client table indicates little or no firmware testing. 3) Unless the WiFi channel was set to 11 (not the default of Auto), I had a lot of trouble connecting my IBM T-41 laptop.
So, a day later, I decided to do more research and settled upon a new D-Link MIMO router, their DI-634M. Comp USA had it on their web site, but when I visited the store they hadn't yet gotten any stock, so I purchased my Netgear WPNT834, despite the much higher price, the decision based solely upon my experience with Netgear products in the early days.
I do hope my trust was not displaced and that Netgear will get their act together and fix these firmware problems ASAP. At least their box is physically solid, unlike the newer products from D-Link and Linksys (though older Linksys products are physically solid as well).
Updated
Folks,
Before I go on, I should mention that I discovered yet another problem with this router, and it's a big one: If one enables remote administration, and then disables it, the feature remains enabled. This, of course, is a major security issue. NETGEAR's response? "We cannot duplicate this in the lab."
And after weeks of documenting all of the firmware problems I discovered, and being told by all but one of the NETGEAR Level 1 and 2 technicians that none of the problems could be the fault of the firmware as no one had reported them (and I'm chopped liver, I suppose, and not someone), they issued an RMA and replaced the router with a new one.
I restored all of my settings and, as I suspected, the new router has all of the same problems as the old one. This is pretty conclusive evidence that the problems are inherent in the firmware and that the firmware was inexpertly written and barely debugged. I reported this to NETGEAR, of course, and have had no substantial response except that I could e-mail my problem list to the development team but that they would not respond to me.
This simply isn't the way the manufacturers of our electronic equipment should be treating us. But, as I say, shoddy practices, both in design and problem resolution, appear to have become the norm. It wasn't always thus, and I offer an example...
In the early 1970s, my father's company purchased a used Bell & Howell Film-o-sound 16mm projector of the type used in schools for decades. This particular model was made in the 1930s and still functioned well. When my father's company no longer had a need for it, he gave it to me and I used it to show films in my home that I borrowed from the local county library. In the late '80s, a tri-axial electrolytic capacitor in the amplifier exploded. As I had no schematic, I had no idea what to replace it with, though I assumed that I'd have to use three discrete capacitors as most tri-axials were custom made. I called Bell & Howell and was connected with an elderly technician (he was probably my current age). My request was only for the values of the exploded capacitor, but he took my name and address and sent not only a photostat of the schematic but a replacement tri-axial capacitor as well, all for free, and the warranty had expired fifty five years earlier! The projector wasn't fancy, but its design is simple and elegant. It worked perfectly the day it was built and it still works perfectly today. I will leave it to the reader to draw his or her own conclusions.
Regards,
Jeff Broido
