TokenBus is more appropriate to a radio situation than TokenRing, though it also isn't really designed for an "A can hear B and C but not D" situation. You could rework to to make use of knowledge of which nodes can hear other nodes to optimize the token-passing, of course. Tricky and probably not worth it in practice, but cute.
I _like_ tokenbus; I designed a TokenBus protocol long ago for the Amiga that was never implemented (complete with auto-registration and optimizations for the common small-number-of-talkers case - let the token ping-pong between two (or n out of m) nodes, and allow other (skipped over) nodes to object to being skipped, causing the token to get passed through the formerly skipped nodes.
DV cameras compress (with MPEG-2). Since this person is doing compression research, source material with compression artifacts is not a good place to start. Also, even without the compression (not that you can turn it off I believe) that level of camera might not be the best for the source quality he's looking for.
"DSL will continue growing twice as fast as cable modems." - not true (though I like DSL).
a) You can't "translate the entire cable signal into 802.11g connections and transmit those signals throughout the house." I suspect they meant to say they'd translate a single cable channel into 802.11g. The entire cable signal is circa 700MHz for most systems (and uses 256-QAM for encoding). A single standard-definition channel (digital!) is circa 2-4Mbps, an HD channel is 12-19Mbps.
b) UDP over 802.11anything is a bad idea for streaming media at high rates. Not that it's impossible, but it's a lot more tricky than it sounds. Lost packets == video glitches. Not to mention issues with neighbor-interference in cities/apts.
c) Residential phones are RJ-11, not RJ-45. Minor nit, but it shows this isn't very well thought out.
d) Comcast (using Samsung boxes) is doing a multi-TV DVR trial in Philly this summer with a system that transmits video from the central box with 802.11g - but it sends the 802.11g over the in-house cable line to the box on top of the other TV. Generally, people have TV's where there are cable drops. Not that they wouldn't mind being untethered in the future... Using 802.11g this way avoids a lot of the interference and packet-loss problems without interfering with the cable signals (which are all
e) The prime factor in the cable industry: capital investment. The reason high-end digital boxes are only being used for HDTV subs is because they cost (gasp) $300. Comcast (and others) hope to get the cost for simple digital box down to $35-50 in the next two years, and drop all analog transmissions. (See last week's MultiChannel News/Broadband Weekly.) This applies doubly for their idea of gaming, at least if the games are assumed to be classic Xbox-type videogames. Their $325/650 and $20,000,000 system cost is WAY under-speced. The $20M is probably an order of magnitude low for the sort of services they're describing, and perhaps more like 20-40x low. $325 per main box isn't too far off (perhaps $450-500), but you need to acccount for each TV - each will need a 802.11g-enabled settop, which is a minimum of $150 currently (and that's not assuming cameras/speakerphones/etc). Also, installation costs - i.e. truck rolls.
The cable companies are VERY SLOW to roll out capital-intensive items. (As my company knows well.)
f) They don't cover the cost/utilization of the downstream/upstream bandwidth (especially for things requiring BW outside the headend, like videophones and even more so video-delivered gaming).
g) TV's don't make very good videophones. Standalone is far better.
h) Where's the cost-of-goods? These services don't all appear for free after your $20M investment. Bandwidth, maintenance, Customer Support, etc.
i) Phone service involves all sorts of capital-intensive (and regulatory) items. Current estimates for VoIP (alone) over cable is that that per-sub-capital-cost is $400-600 depending on various details - not including ongoing costs.
All that said, the basic idea (deliver more marginal services over the 2-way cable infrastructure) is generally a good one, and even some of the specifics here may be good (though 802.11g isn't good for video). My company is in fact in that business (internet access over cable boxes on the TV), and it's like pulling teeth to get the cable companies to actually roll out.
You may be referring to bug 39573, which is about pages with a large number of drop-down (or other) widgets. This isn't due to legacy Mosaic code per se. The problem with recent IE's may be due to code that's meant to give fast "Back" and "Forward" speed by keeping recent pages around in laid-out form - which means a lot of memory and (OS) widgets. Spyglass (which IE2 was based on) had such features.
The problem in Mozilla is based on the number of widgets that need to be created. A solution (which requires a fair bit of coding - care to help?) is to instantiate popups and the like lazily. Another thing that would help with responsiveness would be to have a more-interruptable reflow.
If you're really interested, get involved. http://www.mozilla.org/
The DHTML bug was caused by an 1-character-incorrect backout of a patch that I did in too much of a hurry (removed the entry from the list, but didn't adjust the count). Mea Culpa. This happened 2 weeks before 1.2 final, but most testers were working on 1.3 by that time, and the ones that weren't didn't visit the type of DHTML that causes the problem (most DHTML doesn't have the problem). There was a separate problem where the wrong files were tagged (some recent fixes weren't included).
We're fixing these and will have an updated build up soon. How long would Microsoft take to fix this sort of problem?... (Let alone tell you why the problem happened.)
Re:I learned network programming from Netrek
on
Netrek
·
· Score: 5, Interesting
I indeed did write the Amiga client, which was a pretty cool trick for the day, since the Amiga didn't run X11, and all the clients were Unix/X at the time. Win32 came much later (after Win95, of course). This was around 1990 or 1991. I just wrote wrapper routines for every X11 call made in the program (and all the input focus stuff).
I added one feature still missing (I think) from all others: it would read all the messages to you so that you didn't have to look down at the message window. Important when dodging torps! The Amiga speech device made it easy.
One of the great thing about the game was that it rewarded a number of different strategies and playing styles. Also, league play was and is VERY different than pickup.
At first at Commodore we had no net connection, so we played internally Mondays at 5:30pm. After we got a 64K link, I and 1 or 2 others would play on the net (3 was about the limit). I joined a league team (the Buddies) and played for a couple of years.
As mentioned, we were one of the first to make use of signed clients and things like that (the wonders of RSA).
I can't imagine the number of hours I spent playing, often coming in for 8-16 or more hours on weekends to play.
Every once in a while I'll fire up a client. One additional problem last time I looked was that the metaserver (again, greatly predates gamespy/etc) moved or was down.
-- Randell Jesup, ex-Amiga OS group
Absolutely wonderful in HD
on
The Rise of CSI
·
· Score: 4, Interesting
CSI is gorgeous in HD - the night shots of Vegas from the air, with all the color; the dark exteriors and interiors which would wash into a blur on a regular TV; the closeups of evidence, etc are wonderful in HD. HD does such a good job on color and low-light reproduction compared to NTSC that people who see it at my house are amazed, and CSI is a great example. I think the transfers or camera work has gotten better too since it started.
And everything said in the article is true - it's a riviting drama where science is often the star, for more so than the old detective-story-ish Quincy was.
I'm shocked it ever made it to the screen, and hope it'll be there for a Long Time.
The "Deployment Required" section is an illusion. "shall attain high speed data capability" - that means that they have at least 1 384kbps connection available in each CO. It doesn't mean you personally can get one. Also, lots of people are served by satellite co's and repeaters that don't work with DSL.
More to the point, this is of little import to the ILEC's. They all planned to roll out DSL service to all their CO's within 5 years anyways. Why wouldn't they? It's just the old Tom Sawyer trick "please don't force us to to roll out these nice lucrative services".
The "helps consumers" parts are either bait-and-switch, things that they would have done anyways, or (wait for it) provisions the ILEC lawyers are certain they can get thrown out or ignored, just like they did with the 1996 act. And it gets rid of any credible competition for good.
The closest thing to a "helps consumers" part is this:
(Sec. 5) Requires each incumbent local exchange carrier to provide: (1) Internet users with the ability to subscribe to and have access to any Internet service provider that interconnects with such carrier's high speed data service; (2) any Internet service provider with the right to acquire necessary facilities and services to facilitate such interconnection; (3) any Internet service provider with the ability to collocate equipment in order to achieve such interconnection; and (4) any provider of high speed service, Internet backbone service, or Internet access service with special access for the provision of Internet access service within a period that is no longer than the period in which such local incumbent exchange carrier provides special access to itself or any affiliate for the provision of such service.
As a former Commodore/Amiga OS engineer (AmigaDos/SCSI/IDE/FS/networking/etc) and former Scala engineer, I think I'm qualified to comment.;-)
The preferred term back in the Old Days was "Desktop Video". Still not something that most PC's even with TV-outs do super-well, though it's partly a software issue.
Scala was founded in Norway, but a US office was created, and now it's primarily a US operation, with almost all the remaining engineers being ex-Commodore engineers hired right before or after Commodore went under. The CEO is now Jeff Porter, who was head of hardware engineering for a number of years and later head of "special projects" (i.e. pushed aside by cronies of Mehdi Ali, the person who ran Commodore into the ground).
Scala has been growing in popularity, though the focus is mostly on larger commercial installations (HR depts, airports, point-of-sale displays, waiting rooms, cable companies, etc, etc). The local cable company (Comcast) has a local-info channel thats run by Scala software.
Scala has been proving since the days of Pentium 60's and 486DX66's that with the right software (and video card), you can do some very professional amiga-like smooth video generation, and have a really nice UI for building it. Of course, it's a lot easier nowadays than it was in '94.
Scala uses "MMOS", a multimedia dynamic-OO OS we built that runs on top of the host OS (which in most cases is NT/2000/XP nowadays). It's a pretty clean design, given that the people who designed it were mostly from the core of the OS team at Amiga, along with some talented Scala engineers.
Digital cable is LOWER resolution then even normal cable. In fact, because the displays are large and sharp, digital cable often looks far worse on an HDTV than on a smaller regular TV.
Quality will improve, but not dramatically - the limits are size and (for RPTV's) gun size.
Prices have already come down lots - I paid $5k, but that was 2 years ago for a top-of-the-line Pioneer Elite. Now effectively the same TV (minus lacquer) is $2500.
HDTV is really hitting it's stride nowadays
on
To HDTV or Not to HDTV?
·
· Score: 5, Informative
I got an HDTV almost 2 years ago, when the amount available was pretty low other than Jay Leno and some football and a few other specials. Even so, it was stunning. Nowadays, it's not even that expensive, and some cable companies are starting to carry it.
Now, there's a wealth of HDTV material available:
- Most of HBO, including Sopranos and Band of Brothers (wow).
- Some of Showtime (increasing)
- HDNet on DirectTV (Marc Cuban of broadcast.com and Dallas Mavericks fame - sports/etc HD channel that will carry much of the winter Olympics in HD).
- PBS (several Nova/Nature/etc shows a month, plus many stations show repeats of HD material)
- Almost all of CBS including CSI, District, Alias, etc, also US Open tennis, football playoffs, etc
- Much of ABC including movies of the week, The Practice, etc
- NBC is starting to get on the bandwagon after being first with Leno; they now have Crossing Jordan and more are coming.
- Fox isn't interested in HD, but they do some in 480p widescreen like Ally, X-Files, etc.
CBS lets viewers not in range of a station or in the area of ones they own (most big cities) view the HD feeds from either NYC or LA via Dish Network.
Dish Network has 24-hr PPV in HD; DirectTV has partial-day. Dish has a 24-hr HD promo channel.
Comcast and Time-Warner are starting to roll out carrying HD HBO/Showtime and local HD channels (most of Philly has it now).
HD is here to stay and has come WAY down in price. It looks even better than in the showrooms - they turn the brightness up too high; they often try to sell it using DVD's (which look great but not near as good as real HD material); they don't bother to converge the sets, etc.
Prices are way down - I've seen $13xx in Best Buy for a small 4x3; and $1800 for an open-box 38" RCA CRT HDTV, 16x9, with HD DirectTV built-in.
Don't forget to get an HD receiver; $400-600 currently, but if you're a new DIsh Network subscriber it can be cheaper.
In short: Buy one. Subscribe to HDTVmagazine.com ($35 lifetime; daily email newsletter with the day's HDTV lineup, upcoming news, reviews, etc). You will enjoy it for a long time to come, and you'll have lots of excuses to have people over for parties (starting this winter with playoffs and the Olympics).
I think you're confused. Some banks charge for "online bill pay" systems (like First Onion), but that's not the same thing as what PayPal is doing. That charge is for using their website to send out electronic transfers or physical checks (depending on the recipient). It's possible you bank charges for automated withdrawls, but I've never heard of that.
Paypal (and a lot of places that will do online bill payment) effectively just enter a transfer into the automated clearing-house system from your account to theirs.
Sunset provisions don't help, and may even hurt. Look at Great Britain: in the 70's they passed an anti-terrorism act that allowed indefinite detention without charges (kind of hard to prove you didn't do something if you aren't charged and have your day in court). It had a 1-year sunset provision, so it's only temporary, right?
Wrong. No politician (or very few) are ever going to vote in favor of "letting terrorists go free" (as their next opponent would term it). So the law has been renewed every year since.
The same would happen, in all likelyhood, with this law. The sunset provision would discourage many from objecting, thinking they'd get their rights back someday. But come 2004 or 5 or 6, when the provision runs out, you can bet your booties that the intelligence community (having gotten used to being able to spy on whomever they like) will tell lawmakers in a well-covered hearing that not renewing the law would let terrorists go free, compromise investigations, and destroy motherhood and apple pie, and the legislators would renew it with only a few objectors.
Back a Long Time Ago (1997? 98?) someone on eBay was selling non-existant Japanese swords, offering to repair people's swords, taking them and then selling them or simply disappearing, etc. A bunch of us from the sword mailing list and eBay bidders got together and worked to track him down.
I identified an auction under a new name that was obvious fraud (the image was a link to someone's random web page). He also sent me email claiming to have not know anything about blades using his new email address - but signed it with his real first name, and used technical terms no layperson would know.
We arranged for the deputy(!) he ripped off to be the winner on the bid and got a snail-mail address from him for the postal order. (We'd tracked him down, but he'd moved to another state, and we needed his new address.) When he tried to pick up his check at the PO box, he was arrested. About a dozen blades were recovered, and he was convicted.
Banding together made us FAR more able to get action; this was probably one of the earliest arrests of an eBay seller. Of course, things have changed since then I imagine. Back then I bought a $2500 sword via eBay by personal check (and he mailed it before he received payment - the catch is that he knew who I was from the sword community).
To a certain degree some people on eBay are like people driving around in a van saying "psst: want to get a great deal on some speakers" or "hey, genuine rolex, only $100". Why shuld you trust them? Photos are hardly evidence the item ever existed.
The AOL protocol described is a modification of the old (1984) PlayNet error-correction and data communication protocol I devised (with some input from Steve Bohram, but it was mostly my design based on the Tannenbaum networking book).
CRC-16 was used because modems (300 baud) didn't have any error correction, and we could use tables to process the data 16 bits at a time without using too much memory or CPU (the servers were 12MHz 68010's).
Packets all ended in hex 0D because we were using Telenet and Tymnet X25 dial-in pads in line-buffered mode, because we were charged by the packet. We also munged the other fields to avoid 0D (that may be gone now). Also, they were limited to 256 byte lines; thus the length byte instead of something longer.
Bytes 6 & 7 (which the author doesn't understand) are sequence numbers used in the sliding-window error-correction protocol.
The two-character ASCII prefixes were the actual message types for data packets, and were the input to a multi-tasking state-machine language. EM for example was (IIRC) part of email, perhaps to turn on the 'MAIL' icon. (I forget all the codes, I'm afraid).
Z on the front seems to be an AOL addition.
I was at PlayNet from Feb '84 to Feb '86 (when we declared bankruptcy). AOL licensed the PlayNet software from us for a song when we were running out of money, and rebranded it QuantumLink (and made minor mods, many of which we did for them).
PlayNet ran out of money in Feb '86, though the service continued to remain up for the 1500-3000 subs for another year or two.
PlayNet got a cut of AOL gross revenues until they finally wiggled out of it right before launching America Online (a port of the software to the PC with considerable enhancement), at which point PlayNet's bankruptcy was closed.
The servers were Stratus fault-tolerant machines, and as of 3 years ago they were still using them.
They didn't manage to change the 10-character limit on usernames until a few years ago. That limit was because of the 40-character width of the C64 screen, a ',' between each name, 16(?) characters for the room name plus a space, and we wanted N (12? 15?) users in a chat room. The result was that there were 10 characters available for the username.
The algorithm in AOL for selecting usernames that resulted in JohnQ12345 was also part of the old PlayNet (server) software. Also the default initial passwords for "marketing" accounts (i.e. the free disks) of "word-word" is another thing thought up over lunch at PlayNet that still hasn't changed.
Many things have been added & changed - but far more than I ever expected remains the same. I figured they'd dropped the ECC protocol ages ago.
In fact, having small children at home can be a serious minus for telecommuting, especially from the employer's view. If they feel you're going to spend your time reading stories, changing diapers, playing games, etc, they'll be far less willing to allow it. If you have (small) children, I'd make sure you agree to have someone else take care of them while you're working.
My cousin worked for DEC/Compaq, and she worked from home 3 days a week - but on those days she had a sitter/nanny/whatever to take care of her two kids while she was working. It worked out great.
"facts"? MAngelo's "facts" are mostly misunderstandings (intentional or not) of how a development team and bug database actually work. He makes an assumption that somehow bugzilla bugs of a given severity are all objectively rated, that the bugzilla database is totally up-to-date with actual development at a given point, that all bugs in the project have been found and are in the database, how obscure the bug is, etc.
I've reported dozens of bugs in bugzilla - just because I reported them doesn't make it buggier than before I reported them, but by his measures it does, even though many of them have been fixed.
mangelo simply wants to find ways to present Mozilla in the worst light possible, and will root around until he finds "proof" of his assertion. Lies, damn lies, and statistics, after all, can be used to "prove" anything.
Does he lie? Maybeso, but probably not. He certainly does see everything through colored glasses, and either misinterprets things, or purposely misleads (take your pick). He's decided he doesn't like something about it, and he's going to do his best to make sure everyone else doesn't like it either. Obviously you decided long ago also ("late, buggy, and ugly", etc), and so anything posted by someone who says "it's getting pretty good" will be discounted as "flaming by the mozilla faithful".
Unlike either you or mangelo, I actually try to make things better. And it is AMAZINGLY less buggy than the NS6.0 release, and is pretty darn solid. Perfect - hardly; and nor is IE. But it's gotten pretty darn good and stable, and keeps getting better.
Lack of templates or some equivalent is a real problem. I'd sent Walter some email, but his address seems to be well hidden on that website.
I'd suggest he take a look at Ada Generics, which is part of where C++ templates came from. That might show a way to include the important functionality without too many funny interactions or complexities.
As a long-term sword collector and person interested in metallurgy, that article is a mixture of silly mysticism and legends and pure posturing. Read the SciAm article, or other serious discussions of the topic. Much of what's in the article is long-debunked myth and legend. There've also been good articles in the Japanese Sword Society journal.
The claims for the properties of Damascus steel are overblown. It's the search for the "lost secret" again; a classic story (or more to the point legend). They were good. Very good compared to Western European blades of the day. But not magical. In metallurgic terms they had a very interesting microstructure, but metallurgically good japanese swords were superior. Both, however, where the best of their part of the world, and both designed for their specific uses in warfare.
The reporter obviously didn't research things much (for a circa $14K+ total expenditure). He also apparently didn't have a good salesperson. For example, apparently no one told him that Dish Network has 4 HDTV channels (now 5): HBO, Showtime, a demo channel, and a full-time PPV channel (it soon will have 2 CBS HDTV channels as well - if you're not near a CBS affiliate, though NYC and LA people should be able to get it as well).
There are many sets shipping with decoders, and have been since the start. Even more ship without them, which is a good thing, because it means if you want (say) Dish Network or DirectTV HD you don't have a second, unused decoder in the box. It also means that you can wait to buy a decoder (and use it as an amazing DVD-viewing set in the meantime) or replace an early-generation decoder with a newer one. Eventually, the decoders will be cheap enough it won't matter.
The standards are set (though the Sinclair group keeps bitching that they want to service to moving vehicles). 8VSB is available in almost any location that gets OK (or even poor) NTSC reception. Newer settops generally are much better at pulling in signals in tough locations or indoors.
I've had HDTV for 1.5 years now and love it. PBS, CBS (CSI is great, hell they even have a soap in HD now), HBO (including the Sopranos), etc.
Umm... sorry, but it DOES have to be spelled out, very explicitly, that the box remains theirs.
It was explicitly spelled out in my agreement. they have every right to ask for it back in that situation. If you were "switched over" from another service, and they sent you a telocity DSL modem (people who are "switched over" might not be), then I imagine they notified you somewhere that it was still theirs.
This is Old News. Check the rec.*.tvro newsgroup archives. This has been going on for decades, not just years. Distributing via the net is the only new twist, and it's not that new (though perhaps worrisome for the networks).
I used to have B5 watching parties a week before the episodes aired locally. I'd record them in the morning on SVHS, then play them back over dinner with friends at 8pm.
As for open mikes, that's also been something going on for decades. Some announcers know about this, and some don't. Some even will talk to the sat viewers during breaks (such as the old ESPN Formula 1 hosts, and some other car-racing hosts).
I've met Hal a number of times, and he taught science to my brother. (His real name is Harry Stubbs, and he was a teacher at a prep school for many years.)
Hal is definitely one of the most pure of hard-SF writers. Partly that's a result of the time when he began, partly (or mostly) that's a result of that fact that he simply and truely loves science and thought experiments.
If you're looking for extensive character development and truely alien psychologies, read C. J. Cherryh. If you're looking for really interesting thought experiments, especially about bizarre planetary environments (and their effects on biology and ecosystems), read Hal Clement.
Mission of Gravity is old, but good. He's written quite a few since then (with one small series that were not hard-SF - Eye of the Needle), and has been a regular at SF conventions for many years. He's also one of the nicer people I've met, and always has a smile on it seems.
The ILECs (baby bells mostly) do make it HARD for people to use CLECs like Covad. (They can also do a job on themselves, but that's because they're just incompetant as customer support - witness Verizon....)
Covad had some of the same problems other CLECs did, plus some perhaps. A friend and I both tried to get Covad. I _fought_ to get them to believe my house was where it was, and not 1 mile further from town (they had a mapping DB (mapquest I think) that had the numbers going the wrong way for about 1.5 miles). Even after multiple physical loop length tests, there just wasn't a way to get their system to stop cancelling the order. Finally I went with Telocity, which is a DSL ISP (with their own DSL modems) that in this area uses Verizon. Telocity was a totally smooth install (true ADSL, no new pair needed), and real customer support unlike Verizon Hell. (Verizon has some of the worst support known to mankind.)
My friend ended up out of range for any DSL (even though he's quite close to the CO).
The Bells have wire DBs that are full of errors as well, and they've been swamped with demands for installs. In the northeast, right when DSL took off they had a strike, and got even further behind.
I'm happy with Telocity (and DSL) now that I finally have it installed; I've had it for most of a year.
Ignoring tricks/bugs that cause the machine to violate the partitioning, there are always other ways to get around VMware in theory. Data could be passed in display adapter memory, in "garbage" registers or memory, or you can even pass data between VMware/etc processes on a totally secure machine by having them affect task switch time (admittedly, this is a very low-bandwidth channel, but it does work).
Slashdot Mirror
TokenBus is more appropriate to a radio situation than TokenRing, though it also isn't really designed for an "A can hear B and C but not D" situation. You could rework to to make use of knowledge of which nodes can hear other nodes to optimize the token-passing, of course. Tricky and probably not worth it in practice, but cute.
I _like_ tokenbus; I designed a TokenBus protocol long ago for the Amiga that was never implemented (complete with auto-registration and optimizations for the common small-number-of-talkers case - let the token ping-pong between two (or n out of m) nodes, and allow other (skipped over) nodes to object to being skipped, causing the token to get passed through the formerly skipped nodes.
DV cameras compress (with MPEG-2). Since this person is doing compression research, source material with compression artifacts is not a good place to start. Also, even without the compression (not that you can turn it off I believe) that level of camera might not be the best for the source quality he's looking for.
a) You can't "translate the entire cable signal into 802.11g connections and transmit those signals throughout the house." I suspect they meant to say they'd translate a single cable channel into 802.11g. The entire cable signal is circa 700MHz for most systems (and uses 256-QAM for encoding). A single standard-definition channel (digital!) is circa 2-4Mbps, an HD channel is 12-19Mbps.
b) UDP over 802.11anything is a bad idea for streaming media at high rates. Not that it's impossible, but it's a lot more tricky than it sounds. Lost packets == video glitches. Not to mention issues with neighbor-interference in cities/apts.
c) Residential phones are RJ-11, not RJ-45. Minor nit, but it shows this isn't very well thought out.
d) Comcast (using Samsung boxes) is doing a multi-TV DVR trial in Philly this summer with a system that transmits video from the central box with 802.11g - but it sends the 802.11g over the in-house cable line to the box on top of the other TV. Generally, people have TV's where there are cable drops. Not that they wouldn't mind being untethered in the future... Using 802.11g this way avoids a lot of the interference and packet-loss problems without interfering with the cable signals (which are all e) The prime factor in the cable industry: capital investment. The reason high-end digital boxes are only being used for HDTV subs is because they cost (gasp) $300. Comcast (and others) hope to get the cost for simple digital box down to $35-50 in the next two years, and drop all analog transmissions. (See last week's MultiChannel News/Broadband Weekly.) This applies doubly for their idea of gaming, at least if the games are assumed to be classic Xbox-type videogames. Their $325/650 and $20,000,000 system cost is WAY under-speced. The $20M is probably an order of magnitude low for the sort of services they're describing, and perhaps more like 20-40x low. $325 per main box isn't too far off (perhaps $450-500), but you need to acccount for each TV - each will need a 802.11g-enabled settop, which is a minimum of $150 currently (and that's not assuming cameras/speakerphones/etc). Also, installation costs - i.e. truck rolls.
The cable companies are VERY SLOW to roll out capital-intensive items. (As my company knows well.)
f) They don't cover the cost/utilization of the downstream/upstream bandwidth (especially for things requiring BW outside the headend, like videophones and even more so video-delivered gaming).
g) TV's don't make very good videophones. Standalone is far better.
h) Where's the cost-of-goods? These services don't all appear for free after your $20M investment. Bandwidth, maintenance, Customer Support, etc.
i) Phone service involves all sorts of capital-intensive (and regulatory) items. Current estimates for VoIP (alone) over cable is that that per-sub-capital-cost is $400-600 depending on various details - not including ongoing costs.
All that said, the basic idea (deliver more marginal services over the 2-way cable infrastructure) is generally a good one, and even some of the specifics here may be good (though 802.11g isn't good for video). My company is in fact in that business (internet access over cable boxes on the TV), and it's like pulling teeth to get the cable companies to actually roll out.
You may be referring to bug 39573, which is about pages with a large number of drop-down (or other) widgets. This isn't due to legacy Mosaic code per se. The problem with recent IE's may be due to code that's meant to give fast "Back" and "Forward" speed by keeping recent pages around in laid-out form - which means a lot of memory and (OS) widgets. Spyglass (which IE2 was based on) had such features.
The problem in Mozilla is based on the number of widgets that need to be created. A solution (which requires a fair bit of coding - care to help?) is to instantiate popups and the like lazily. Another thing that would help with responsiveness would be to have a more-interruptable reflow.
If you're really interested, get involved. http://www.mozilla.org/
The DHTML bug was caused by an 1-character-incorrect backout of a patch that I did in too much of a hurry (removed the entry from the list, but didn't adjust the count). Mea Culpa. This happened 2 weeks before 1.2 final, but most testers were working on 1.3 by that time, and the ones that weren't didn't visit the type of DHTML that causes the problem (most DHTML doesn't have the problem). There was a separate problem where the wrong files were tagged (some recent fixes weren't included).
We're fixing these and will have an updated build up soon. How long would Microsoft take to fix this sort of problem?... (Let alone tell you why the problem happened.)
I indeed did write the Amiga client, which was a pretty cool trick for the day, since the Amiga didn't run X11, and all the clients were Unix/X at the time. Win32 came much later (after Win95, of course). This was around 1990 or 1991. I just wrote wrapper routines for every X11 call made in the program (and all the input focus stuff).
I added one feature still missing (I think) from all others: it would read all the messages to you so that you didn't have to look down at the message window. Important when dodging torps! The Amiga speech device made it easy.
One of the great thing about the game was that it rewarded a number of different strategies and playing styles. Also, league play was and is VERY different than pickup.
At first at Commodore we had no net connection, so we played internally Mondays at 5:30pm. After we got a 64K link, I and 1 or 2 others would play on the net (3 was about the limit). I joined a league team (the Buddies) and played for a couple of years.
As mentioned, we were one of the first to make use of signed clients and things like that (the wonders of RSA).
I can't imagine the number of hours I spent playing, often coming in for 8-16 or more hours on weekends to play.
Every once in a while I'll fire up a client. One additional problem last time I looked was that the metaserver (again, greatly predates gamespy/etc) moved or was down.
-- Randell Jesup, ex-Amiga OS group
CSI is gorgeous in HD - the night shots of Vegas from the air, with all the color; the dark exteriors and interiors which would wash into a blur on a regular TV; the closeups of evidence, etc are wonderful in HD. HD does such a good job on color and low-light reproduction compared to NTSC that people who see it at my house are amazed, and CSI is a great example. I think the transfers or camera work has gotten better too since it started.
And everything said in the article is true - it's a riviting drama where science is often the star, for more so than the old detective-story-ish Quincy was.
I'm shocked it ever made it to the screen, and hope it'll be there for a Long Time.
More to the point, this is of little import to the ILEC's. They all planned to roll out DSL service to all their CO's within 5 years anyways. Why wouldn't they? It's just the old Tom Sawyer trick "please don't force us to to roll out these nice lucrative services".
The "helps consumers" parts are either bait-and-switch, things that they would have done anyways, or (wait for it) provisions the ILEC lawyers are certain they can get thrown out or ignored, just like they did with the 1996 act. And it gets rid of any credible competition for good.
The closest thing to a "helps consumers" part is this:
As a former Commodore/Amiga OS engineer (AmigaDos/SCSI/IDE/FS/networking/etc) and former Scala engineer, I think I'm qualified to comment. ;-)
The preferred term back in the Old Days was "Desktop Video". Still not something that most PC's even with TV-outs do super-well, though it's partly a software issue.
Scala was founded in Norway, but a US office was created, and now it's primarily a US operation, with almost all the remaining engineers being ex-Commodore engineers hired right before or after Commodore went under. The CEO is now Jeff Porter, who was head of hardware engineering for a number of years and later head of "special projects" (i.e. pushed aside by cronies of Mehdi Ali, the person who ran Commodore into the ground).
Scala has been growing in popularity, though the focus is mostly on larger commercial installations (HR depts, airports, point-of-sale displays, waiting rooms, cable companies, etc, etc). The local cable company (Comcast) has a local-info channel thats run by Scala software.
Scala has been proving since the days of Pentium 60's and 486DX66's that with the right software (and video card), you can do some very professional amiga-like smooth video generation, and have a really nice UI for building it. Of course, it's a lot easier nowadays than it was in '94.
Scala uses "MMOS", a multimedia dynamic-OO OS we built that runs on top of the host OS (which in most cases is NT/2000/XP nowadays). It's a pretty clean design, given that the people who designed it were mostly from the core of the OS team at Amiga, along with some talented Scala engineers.
Digital cable is LOWER resolution then even normal cable. In fact, because the displays are large and sharp, digital cable often looks far worse on an HDTV than on a smaller regular TV.
Quality will improve, but not dramatically - the limits are size and (for RPTV's) gun size.
Prices have already come down lots - I paid $5k, but that was 2 years ago for a top-of-the-line Pioneer Elite. Now effectively the same TV (minus lacquer) is $2500.
Now, there's a wealth of HDTV material available:
- Most of HBO, including Sopranos and Band of Brothers (wow).
- Some of Showtime (increasing)
- HDNet on DirectTV (Marc Cuban of broadcast.com and Dallas Mavericks fame - sports/etc HD channel that will carry much of the winter Olympics in HD).
- PBS (several Nova/Nature/etc shows a month, plus many stations show repeats of HD material)
- Almost all of CBS including CSI, District, Alias, etc, also US Open tennis, football playoffs, etc
- Much of ABC including movies of the week, The Practice, etc
- NBC is starting to get on the bandwagon after being first with Leno; they now have Crossing Jordan and more are coming. - Fox isn't interested in HD, but they do some in 480p widescreen like Ally, X-Files, etc.
CBS lets viewers not in range of a station or in the area of ones they own (most big cities) view the HD feeds from either NYC or LA via Dish Network.
Dish Network has 24-hr PPV in HD; DirectTV has partial-day. Dish has a 24-hr HD promo channel.
Comcast and Time-Warner are starting to roll out carrying HD HBO/Showtime and local HD channels (most of Philly has it now).
HD is here to stay and has come WAY down in price. It looks even better than in the showrooms - they turn the brightness up too high; they often try to sell it using DVD's (which look great but not near as good as real HD material); they don't bother to converge the sets, etc.
Prices are way down - I've seen $13xx in Best Buy for a small 4x3; and $1800 for an open-box 38" RCA CRT HDTV, 16x9, with HD DirectTV built-in.
Don't forget to get an HD receiver; $400-600 currently, but if you're a new DIsh Network subscriber it can be cheaper.
In short: Buy one. Subscribe to HDTVmagazine.com ($35 lifetime; daily email newsletter with the day's HDTV lineup, upcoming news, reviews, etc). You will enjoy it for a long time to come, and you'll have lots of excuses to have people over for parties (starting this winter with playoffs and the Olympics).
I think you're confused. Some banks charge for "online bill pay" systems (like First Onion), but that's not the same thing as what PayPal is doing. That charge is for using their website to send out electronic transfers or physical checks (depending on the recipient). It's possible you bank charges for automated withdrawls, but I've never heard of that.
Paypal (and a lot of places that will do online bill payment) effectively just enter a transfer into the automated clearing-house system from your account to theirs.
Sunset provisions don't help, and may even hurt. Look at Great Britain: in the 70's they passed an anti-terrorism act that allowed indefinite detention without charges (kind of hard to prove you didn't do something if you aren't charged and have your day in court). It had a 1-year sunset provision, so it's only temporary, right?
Wrong. No politician (or very few) are ever going to vote in favor of "letting terrorists go free" (as their next opponent would term it). So the law has been renewed every year since.
The same would happen, in all likelyhood, with this law. The sunset provision would discourage many from objecting, thinking they'd get their rights back someday. But come 2004 or 5 or 6, when the provision runs out, you can bet your booties that the intelligence community (having gotten used to being able to spy on whomever they like) will tell lawmakers in a well-covered hearing that not renewing the law would let terrorists go free, compromise investigations, and destroy motherhood and apple pie, and the legislators would renew it with only a few objectors.
Back a Long Time Ago (1997? 98?) someone on eBay was selling non-existant Japanese swords, offering to repair people's swords, taking them and then selling them or simply disappearing, etc. A bunch of us from the sword mailing list and eBay bidders got together and worked to track him down.
I identified an auction under a new name that was obvious fraud (the image was a link to someone's random web page). He also sent me email claiming to have not know anything about blades using his new email address - but signed it with his real first name, and used technical terms no layperson would know.
We arranged for the deputy(!) he ripped off to be the winner on the bid and got a snail-mail address from him for the postal order. (We'd tracked him down, but he'd moved to another state, and we needed his new address.) When he tried to pick up his check at the PO box, he was arrested. About a dozen blades were recovered, and he was convicted.
Banding together made us FAR more able to get action; this was probably one of the earliest arrests of an eBay seller. Of course, things have changed since then I imagine. Back then I bought a $2500 sword via eBay by personal check (and he mailed it before he received payment - the catch is that he knew who I was from the sword community).
To a certain degree some people on eBay are like people driving around in a van saying "psst: want to get a great deal on some speakers" or "hey, genuine rolex, only $100". Why shuld you trust them? Photos are hardly evidence the item ever existed.
The AOL protocol described is a modification of the old (1984) PlayNet error-correction and data communication protocol I devised (with some input from Steve Bohram, but it was mostly my design based on the Tannenbaum networking book).
CRC-16 was used because modems (300 baud) didn't have any error correction, and we could use tables to process the data 16 bits at a time without using too much memory or CPU (the servers were 12MHz 68010's).
Packets all ended in hex 0D because we were using Telenet and Tymnet X25 dial-in pads in line-buffered mode, because we were charged by the packet. We also munged the other fields to avoid 0D (that may be gone now). Also, they were limited to 256 byte lines; thus the length byte instead of something longer.
Bytes 6 & 7 (which the author doesn't understand) are sequence numbers used in the sliding-window error-correction protocol.
The two-character ASCII prefixes were the actual message types for data packets, and were the input to a multi-tasking state-machine language. EM for example was (IIRC) part of email, perhaps to turn on the 'MAIL' icon. (I forget all the codes, I'm afraid).
Z on the front seems to be an AOL addition.
I was at PlayNet from Feb '84 to Feb '86 (when we declared bankruptcy). AOL licensed the PlayNet software from us for a song when we were running out of money, and rebranded it QuantumLink (and made minor mods, many of which we did for them).
PlayNet ran out of money in Feb '86, though the service continued to remain up for the 1500-3000 subs for another year or two.
PlayNet got a cut of AOL gross revenues until they finally wiggled out of it right before launching America Online (a port of the software to the PC with considerable enhancement), at which point PlayNet's bankruptcy was closed.
The servers were Stratus fault-tolerant machines, and as of 3 years ago they were still using them.
They didn't manage to change the 10-character limit on usernames until a few years ago. That limit was because of the 40-character width of the C64 screen, a ',' between each name, 16(?) characters for the room name plus a space, and we wanted N (12? 15?) users in a chat room. The result was that there were 10 characters available for the username.
The algorithm in AOL for selecting usernames that resulted in JohnQ12345 was also part of the old PlayNet (server) software. Also the default initial passwords for "marketing" accounts (i.e. the free disks) of "word-word" is another thing thought up over lunch at PlayNet that still hasn't changed.
Many things have been added & changed - but far more than I ever expected remains the same. I figured they'd dropped the ECC protocol ages ago.
-- Randell Jesup
In fact, having small children at home can be a serious minus for telecommuting, especially from the employer's view. If they feel you're going to spend your time reading stories, changing diapers, playing games, etc, they'll be far less willing to allow it. If you have (small) children, I'd make sure you agree to have someone else take care of them while you're working.
My cousin worked for DEC/Compaq, and she worked from home 3 days a week - but on those days she had a sitter/nanny/whatever to take care of her two kids while she was working. It worked out great.
I've reported dozens of bugs in bugzilla - just because I reported them doesn't make it buggier than before I reported them, but by his measures it does, even though many of them have been fixed.
mangelo simply wants to find ways to present Mozilla in the worst light possible, and will root around until he finds "proof" of his assertion. Lies, damn lies, and statistics, after all, can be used to "prove" anything.
Does he lie? Maybeso, but probably not. He certainly does see everything through colored glasses, and either misinterprets things, or purposely misleads (take your pick). He's decided he doesn't like something about it, and he's going to do his best to make sure everyone else doesn't like it either. Obviously you decided long ago also ("late, buggy, and ugly", etc), and so anything posted by someone who says "it's getting pretty good" will be discounted as "flaming by the mozilla faithful".
Unlike either you or mangelo, I actually try to make things better. And it is AMAZINGLY less buggy than the NS6.0 release, and is pretty darn solid. Perfect - hardly; and nor is IE. But it's gotten pretty darn good and stable, and keeps getting better.
Lack of templates or some equivalent is a real problem. I'd sent Walter some email, but his address seems to be well hidden on that website.
I'd suggest he take a look at Ada Generics, which is part of where C++ templates came from. That might show a way to include the important functionality without too many funny interactions or complexities.
As a long-term sword collector and person interested in metallurgy, that article is a mixture of silly mysticism and legends and pure posturing. Read the SciAm article, or other serious discussions of the topic. Much of what's in the article is long-debunked myth and legend. There've also been good articles in the Japanese Sword Society journal.
The claims for the properties of Damascus steel are overblown. It's the search for the "lost secret" again; a classic story (or more to the point legend). They were good. Very good compared to Western European blades of the day. But not magical. In metallurgic terms they had a very interesting microstructure, but metallurgically good japanese swords were superior. Both, however, where the best of their part of the world, and both designed for their specific uses in warfare.
The reporter obviously didn't research things much (for a circa $14K+ total expenditure). He also apparently didn't have a good salesperson. For example, apparently no one told him that Dish Network has 4 HDTV channels (now 5): HBO, Showtime, a demo channel, and a full-time PPV channel (it soon will have 2 CBS HDTV channels as well - if you're not near a CBS affiliate, though NYC and LA people should be able to get it as well).
There are many sets shipping with decoders, and have been since the start. Even more ship without them, which is a good thing, because it means if you want (say) Dish Network or DirectTV HD you don't have a second, unused decoder in the box. It also means that you can wait to buy a decoder (and use it as an amazing DVD-viewing set in the meantime) or replace an early-generation decoder with a newer one. Eventually, the decoders will be cheap enough it won't matter.
The standards are set (though the Sinclair group keeps bitching that they want to service to moving vehicles). 8VSB is available in almost any location that gets OK (or even poor) NTSC reception. Newer settops generally are much better at pulling in signals in tough locations or indoors.
I've had HDTV for 1.5 years now and love it. PBS, CBS (CSI is great, hell they even have a soap in HD now), HBO (including the Sopranos), etc.
Umm... sorry, but it DOES have to be spelled out, very explicitly, that the box remains theirs.
It was explicitly spelled out in my agreement. they have every right to ask for it back in that situation. If you were "switched over" from another service, and they sent you a telocity DSL modem (people who are "switched over" might not be), then I imagine they notified you somewhere that it was still theirs.
This is Old News. Check the rec.*.tvro newsgroup archives. This has been going on for decades, not just years. Distributing via the net is the only new twist, and it's not that new (though perhaps worrisome for the networks).
I used to have B5 watching parties a week before the episodes aired locally. I'd record them in the morning on SVHS, then play them back over dinner with friends at 8pm.
As for open mikes, that's also been something going on for decades. Some announcers know about this, and some don't. Some even will talk to the sat viewers during breaks (such as the old ESPN Formula 1 hosts, and some other car-racing hosts).
- Disclaimer:
I've met Hal a number of times, and he taught science to my brother. (His real name is Harry Stubbs, and he was a teacher at a prep school for many years.)Hal is definitely one of the most pure of hard-SF writers. Partly that's a result of the time when he began, partly (or mostly) that's a result of that fact that he simply and truely loves science and thought experiments.
If you're looking for extensive character development and truely alien psychologies, read C. J. Cherryh. If you're looking for really interesting thought experiments, especially about bizarre planetary environments (and their effects on biology and ecosystems), read Hal Clement.
Mission of Gravity is old, but good. He's written quite a few since then (with one small series that were not hard-SF - Eye of the Needle), and has been a regular at SF conventions for many years. He's also one of the nicer people I've met, and always has a smile on it seems.
The ILECs (baby bells mostly) do make it HARD for people to use CLECs like Covad. (They can also do a job on themselves, but that's because they're just incompetant as customer support - witness Verizon....)
Covad had some of the same problems other CLECs did, plus some perhaps. A friend and I both tried to get Covad. I _fought_ to get them to believe my house was where it was, and not 1 mile further from town (they had a mapping DB (mapquest I think) that had the numbers going the wrong way for about 1.5 miles). Even after multiple physical loop length tests, there just wasn't a way to get their system to stop cancelling the order. Finally I went with Telocity, which is a DSL ISP (with their own DSL modems) that in this area uses Verizon. Telocity was a totally smooth install (true ADSL, no new pair needed), and real customer support unlike Verizon Hell. (Verizon has some of the worst support known to mankind.)
My friend ended up out of range for any DSL (even though he's quite close to the CO).
The Bells have wire DBs that are full of errors as well, and they've been swamped with demands for installs. In the northeast, right when DSL took off they had a strike, and got even further behind.
I'm happy with Telocity (and DSL) now that I finally have it installed; I've had it for most of a year.
Ignoring tricks/bugs that cause the machine to violate the partitioning, there are always other ways to get around VMware in theory. Data could be passed in display adapter memory, in "garbage" registers or memory, or you can even pass data between VMware/etc processes on a totally secure machine by having them affect task switch time (admittedly, this is a very low-bandwidth channel, but it does work).