A few years ago I lived in a town called Wellington, Ohio. The LEC in the area was Verizon (now Frontier) who offered a DSL service that was officially supposed to be around 3m/768k but usually barely beat 2m/512k and would go entirely down (signal loss at the modem or no PPPoE response) for hours at a time multiple times per month. My roommate and I tried multiple modems including the Verizon-provided Westell, a Cisco 675, a few Motorolas, and an Edgemarc 200AW with no change, nor did installing a direct 35 foot Cat5e link from the telco demarc to the modem, entirely disconnecting the in-house wiring (we were cell-only and worked at a VoIP company so we had no interest in POTS).
So go to cable, you may say. Enter GLW Broadband. Basically, these guys were a local cable TV provider who did the minimum necessary to keep people mostly happy with their TV in these semi-rural areas. At some point they added broadband internet services, but it's clear that they do not take it seriously. As far as I could tell during my time as a customer of theirs as well as while working with customers of mine who had their internet services, they had a single upstream connection to Time Warner. They were somehow worse than Verizon. The modem would lose sync regularly, when it was up the speed was rarely even the 1.5/768 that it was supposed to be, and more than once static IPs just stopped working altogether and would have to be changed. One of my customers I shared with them was literally a stone's throw down the street from the GLW main office, yet they had an outage which lasted a week which they couldn't figure out. I could have run a standard ethernet cable across the lawn between the offices and it would have been entirely within spec.
There was also a local WiFi ISP, but they wouldn't even offer a non-NATed address so they were right out.
When faced with these options, the only option if you want anything approaching reliability is to get as many as you can afford and hope they're not all sucking at the same time.
Echoing this. I started with pfSense because I wanted a multi-WAN router, but still use it to this day even on single-WAN environments because it's trivial to get going on any spare PC hardware, can be easily built in to an "appliance" with a number of available embedded x86 boards, and does pretty much everything. Of course it's not as fast as hardware built for the purpose, but if you have hardware encryption available I haven't yet seen a reason to choose a PIX/ASA over it on performance grounds. Obviously support and corporate love for Cisco are other factors which may come in to play for some applications.
Why are phones, particularly the VM box that is more or less an automatic part of today's cell phone, so damn vulnerable?
In most cases, because the users are stupid. In some cases, because the telco is stupid.
The majority of the time, the user will have a stupidly weak password like 1234, 123456, 111111, etc. I do VoIP for a living and one of the platforms I support, Broadworks, can not block a user from having a password like 123456. 111111 is banned, but easy sequences can't be yet. Due to this, I have on average 3-5 cases a month of people getting their accounts hacked and someone trying to forward calls to some other country. We block this system-wide, so it just results in the user's incoming calls breaking until it's noticed, but it happens with reasonable regularity.
In a few cases, the telco is retarded and allows the user to set that calls from their phone be allowed directly in to the voicemail system. Unfortunately they do not sanity-check this to verify that the call is actually coming from that phone and instead depend entirely on caller ID. Anyone with a VoIP or PRI system and a trusting upstream carrier can send whatever caller ID they want, making it trivial to get in to the voicemail. I think T-Mobile was in that category last time I checked, no idea on the others.
The Telcoes seem to have no trouble tracking our activities in great detail if those activities are something for which we can be billed, and they also seem eminently willing to cooperate with law enforcement. Why, then, do I have absolutely no way of knowing when, and from where, my VM box was called into, and why would the VM box of a phone that is subject to police investigation be accessible from the outside at all?
When, they should easily be able to give you and if they don't its only policy. From where, that's a lot tougher, given the ease of spoofing caller ID. Also, a lot of attacks are routed through multiple systems to disguise the source(s). Most attacks I see seem to come from other PBXes, likely hacked in similar ways.
Agreed on the box being open to remote access. It's trivial on my systems to allow incoming messages but break phone access to any given box, and investigators could just access the e-mail server that stores the messages directly via IMAP.
On consumer/small business internet connections, increased latency almost always brings jitter so it's just easier to look for latency. As I specifically noted, with low jitter a high latency connection can be used.
I mostly agree, they have their purpose if they're the only option but they can't compete with any of the other broadband options. 3G cellular or a long-run DSL are certainly better choices where available even as bad as they can be. If you don't have that though, satellite is an appealing last resort.
My thoughts exactly. VoIP is my day job and I can tell you that once pings exceed 200ms things get questionable. If there's low jitter, it can work and just have a delay like old intercontinental satellite PSTN links, but usually this is not the case. Any satellite connection using fixed dishes and thus geostationary satellites (a.k.a. everything marketed to home users) has an absolute minimum latency caused by the "last mile" of 472ms. This could only be achieved at the equator, anywhere else would be farther away and have greater latency but I don't feel like doing the math for Canadian latitudes.
tl;dr: VoIP on consumer satellite internet connections is stupid, end of story.
The broadband quality in my rural area sucks, big time. I can barely stream Netflix movies, let alone consider making phone calls with it.
If you can even contemplate streaming video, you can run a phone call over it just fine. The absolute worst case bandwidth needs are 64kbit/sec in each direction, and a non-shitty VoIP provider will offer multiple codecs which need far less.
Using AMR-WB (typically used for high quality cellular calls in places that aren't the US) I can easily put four simultaneous calls through a dialup connection while matching or beating the quality of a PSTN call. Any form of broadband, as long as it's actually working (no horrible packet loss or jitter problems), is more than sufficient for phone use.
Personally, I often find myself longing for the higher sound quality of a fully-wired phone line versus that of a cell phone. Isn't that ironic?
Try a VoIP phone with G.722 support some time. It'll blow you away. I thought they were gimmicks when I first got one in for testing, but it really makes the G.711 codec used by the PSTN sound like the shit it is. The bandwidth usage is exactly the same too at full rate, so the only reason it's not more common is legacy equipment which doesn't support it. A few lower bitrate options are in the 3GPP specs and are being used by certain cellular carriers in Europe and Canada to deliver better than PSTN call quality over mobile phones.
I think you're on to something with this. Blackberries seem to mostly be used as dumbphones with good PIM and e-mail functionality rather than the typical modern smartphone which gets used as a gaming/browsing/e-mail device that can probably make phone calls.
Just a note, a lot of Android devices get better battery life with WiFi turned on as long as you're in an area with an AP you can use, as the WiFi radio uses less power than the cellular radio for background internet tasks like syncing. I don't know whether Blackberries benefit from this since I know they usually require special services from the cellular carrier to get full functionality (I assume something to do with BES, which IMO is a pointless extra step that needs to go away), but it might be worth looking in to.
When we were getting very close to releasing Half-Life 1 (less than a week or so), we found there were already some projects that we needed to start working on, but we couldn't risk checking in code to the shipping version of the game. At that point we forked off the code in VSS to be both $/Goldsrc and/$Src. Over the next few years, we used these terms internally as "Goldsource" and "Source". At least initially, the Goldsrc branch of code referred to the codebase that was currently released, and Src referred to the next set of more risky technology that we were working on. When it came down to show Half-Life 2 for the first time at E3, it was part of our internal communication to refer to the "Source" engine vs. the "Goldsource" engine, and the name stuck.
The best Blackberry, the Bold 9780, is roughly equal to the iPhone 4 in all tests, trading off for the top spot in the web browsing tests. In the 3G talk time test both the iPhone and the Blackberry are firmly beaten by a number of Android devices.
I'll also point out that the Blackberry has a small 2.4" screen compared to the 3.5-4.3" screens of the majority of the competition and has a slow 624 MHz processor, compared to ~750 MHz in the iPhone and 800-1200MHz on the Androids. On paper it should have no problem beating the others simply from having less display to illuminate and a less demanding processor.
Standby time was not tested so it may win there, but honestly is it really that much of a problem to plug a phone in when you go to bed? I have a HTC Evo, one of the most power-hungry phones on the market, and that's all I have to do. Same as I've done out of habit with every phone I've owned in the past, smart or not.
Keyboard you're probably right, though personally I have never found a smartphone keyboard I liked. Likewise having not used Blackberry e-mail in years I'll refrain from judgement, though I can't say I have any complaints about either iOS or Android in that regard.
Many competitions have rules regarding dishonorable behavior. If he did use substantial amounts of open-source code without crediting the original authors as seems to be the case, that would be plainly dishonorable and thus grounds for disqualification.
As for the open source aspect, in this case a quick skim of the Fruity site seems to indicate (it's not very clearly worded) that it was initially open source during development and then went to a closed/commercial model after it had some wins under its belt. This makes sense to me if the developer started working on it as more of a personal challenge rather than for competition and then just ended up being good in competition.
On the point of telling your competitors your secrets, arguably Chrome and Firefox are competitors yet they're both fully open. If you're open source, your competitors may learn a trick or two from you but they might implement it in a completely different way. This may be better or worse. If they made different fundamental design decisions, they might not be able to make use of anything you did directly. If your main goal is to push the field forward and winning competitions is just a bonus, who cares if your competitors can see your code?
Re:Always show your work
on
Happy Tau Day
·
· Score: 1
Just maybe the teachers were preparing you for high school and higher level maths. Education isn't just about showing how fucking clever you are to teachers, it's accepting that there are strategies and techniques that, once ingrained, allow you to learn better in the long run.
Ok, so I did it once, I know how to do it, why do I have to keep doing it on stuff I don't need it for? When I got to the harder math I couldn't do entirely in my head, guess what? I wrote it out!
But, obviously, you're such a fucking genius you still do everything in your head, right?
Nope, I use a calculator and write down intermediate results as necessary, which makes sense. You apparently missed the entire point, which was not to say that it's always pointless but that forcing a student to do it in situations where they don't need to only results in a pissed off student. "the phantom" above got it, what's so hard for you?
Re:Always show your work
on
Happy Tau Day
·
· Score: 1
Then you teacher failed. Showing your work is about knowing the procedure to do something.
Then once you've showed it they should fuck off and let you just do mental math.
I absolutely despised math teachers who enforced showing of work. Pretty much everything done in a math class below the high school level I could do in my head nearly instantly. Writing out the whole process was a waste of 30-60 seconds per question. On a long test that could be half the class period where I could be reading rather than writing out how I managed to figure out that 13 * 24 = 312.
I understand that it is useful both as a teaching aid (to see where the student got it wrong) and as a deterrent to cheating (harder to copy and of course calculators don't tell you how they got there), but that still didn't mean it should be used without exception. If the student can prove they know the material, let them stop having to prove it every damn time.
HTML 5, CSS 3, massively improved Javascript engines...
Many of the standards in use today are either not fully supported or supported with quirks. Where the development resources are focused depends on how those technologies are used. This is demonstrated by Javascript in that to my knowledge the actual language has not changed in any significant way, but every major browser has completely reworked how they handle it to improve performance.
Go find a copy of IE 7, Firefox 3.0.x, or some old version of Chrome and see how it compares to their modern equivalents. It won't be as bad as IE 6, but it still will be a major step backwards. It's not like browser developers are changing things willy-nilly, they're doing it because it's an evolving market where the priorities can change.
Again I agree with the article. If you need a stable, unchanging platform the web is not for you.
I run all of these extensions and have been using FF5 since it became available on the beta channel. No problems.
Rarely does a plugin or extension actually break with a new release. The problem is that Mozilla's addon system allows the developers to be idiots and set a max version even if it's not needed. Max version, to me, is something to be set once you know it breaks rather than preemptively.
The whole point was that if they have such a long cycle for a web browser, they're doing it wrong. I agree with that. The web is a fast-moving platform, old browsers are bad.
You missed the second part about there not really being anything which only works in these versions. I'm sure someone can come up with something obscure, but it's not like the IE6 stragglers who actually have an excuse.
Why should Google continue to support outdated and/or broken browsers when all but a tiny handful of the remaining users of those browsers don't actually have a reason to and are just lazy about updating?
People also seem to think that Google will be putting up a wall like some stupidly designed sites do, attempting to block older browsers from even accessing the site. This is not the case, rather they won't be testing in those browsers nor will they work around bugs in those browsers. It'll probably keep working for some time.
Why should we have to support old versions of free software where in all cases the user would have had to go out of their way to disable automatic updates?
Any Firefox 3.5 user can update to 3.6. No system requirements changed and it's hard to believe anyone would have gone IE6-style and coded anything specifically to it. Same with IE7->IE8 and Safari 3->4.
If the users can upgrade at the press of a button and a problem is caused entirely by a bug in their outdated software, why should I fix it rather than then just updating to a version without the bug?
It's a point release, not a full new version. There's not really significant testing needed. It's also a web browser which is not integrated with the system in any way, making it absolutely trivial that if a user manages to stumble upon something that's actually broken in any important way, rolling back is trivial.
I agree that some software can not be rapidly upgraded in a sane business environment, but this is not one of those cases. It's a very minor upgrade that's been out for 17 months and even superseded by a major revision. A conservative testing plan would be to sit back for a few weeks initially and let the first patches show up for those issues that slip through testing but show up quickly upon general release. During this time a few bored people in IT should upgrade and test the basics of anything critical. Once it looks stable as a whole, deploy to all of IT and possibly some savvier users in each department to get a sampling of the webapps used by the company. If no stoppers show up in a month roll it out fully.
Something like an operating system deserves months of testing. A browser, not so much. The three browsers that Google is no longer supporting were carefully chosen, as there are no technical reasons that anyone using one of those three could not move up to at least the next one in line. IE9 is Vista/7 only and the Mac versions of both Firefox 4 and Safari 5 require OS X 10.5, so reaching the current version of any of these browsers could require an OS upgrade, but there are still upgrades available for all users of the dropped browsers. None of them have the IE6 problem of piles of shitty webapps written to their quirks either.
If it's that much of a hassle to run a 64 bit browser, why bother? I don't know about you but I get concerned if my browser's memory usage hits 2GB much less 4, I can't imagine it needs to do math on really big numbers very often, and the plugin situation is just easier to deal with. There's no real gain I see to a 64 bit browser, so I haven't ever figured out why people complain so much about Mozilla not releasing 64 bit binaries. Maybe it's different on other architectures, but on x86 and PPC there is no performance penalty from running 32 bit apps. Hell, a lot of Windows Vista and 7 installs are 64 bit including my dual boot to 7, but other than drivers and the kernel itself I can't think of any 64 bit apps I have installed.
These are all versions which you have no excuse to still be on, so Google is simply saying that they will no longer test for those browsers or fix bugs found only in those browsers.
Anyone on FF3.5 can move to 3.6, IE7 to 8, and Safari 3 to 4, so there are no valid reasons to still use them. Those browsers don't have any IE6-like hold on a market due to shitty apps coded explicitly for them, nor are there major changes which could throw off a user.
AT&T has deployed HSPA+ to virtually 100% of our nation's fastest mobile broadband network, which enables 4G speeds when combined with enhanced backhaul.
Translation:
AT&T has enabled your phone or wireless modem to connect to the tower at HSPA speeds. Good luck using that speed when most of the towers still use a small number of T1s to get to the internet.
It's like having a 512/128kbit entry-level DSL internet connection and upgrading your WiFi access point from an old 802.11b unit to a new MIMO 802.11n device. Sure, the air link is faster, but the internet connection is still the bottleneck. Those lucky enough to have towers with upgraded backhaul connections may be able to get interesting speeds, but most won't.
Slashdot Mirror
Shitty service.
A few years ago I lived in a town called Wellington, Ohio. The LEC in the area was Verizon (now Frontier) who offered a DSL service that was officially supposed to be around 3m/768k but usually barely beat 2m/512k and would go entirely down (signal loss at the modem or no PPPoE response) for hours at a time multiple times per month. My roommate and I tried multiple modems including the Verizon-provided Westell, a Cisco 675, a few Motorolas, and an Edgemarc 200AW with no change, nor did installing a direct 35 foot Cat5e link from the telco demarc to the modem, entirely disconnecting the in-house wiring (we were cell-only and worked at a VoIP company so we had no interest in POTS).
So go to cable, you may say. Enter GLW Broadband. Basically, these guys were a local cable TV provider who did the minimum necessary to keep people mostly happy with their TV in these semi-rural areas. At some point they added broadband internet services, but it's clear that they do not take it seriously. As far as I could tell during my time as a customer of theirs as well as while working with customers of mine who had their internet services, they had a single upstream connection to Time Warner. They were somehow worse than Verizon. The modem would lose sync regularly, when it was up the speed was rarely even the 1.5/768 that it was supposed to be, and more than once static IPs just stopped working altogether and would have to be changed. One of my customers I shared with them was literally a stone's throw down the street from the GLW main office, yet they had an outage which lasted a week which they couldn't figure out. I could have run a standard ethernet cable across the lawn between the offices and it would have been entirely within spec.
There was also a local WiFi ISP, but they wouldn't even offer a non-NATed address so they were right out.
When faced with these options, the only option if you want anything approaching reliability is to get as many as you can afford and hope they're not all sucking at the same time.
Echoing this. I started with pfSense because I wanted a multi-WAN router, but still use it to this day even on single-WAN environments because it's trivial to get going on any spare PC hardware, can be easily built in to an "appliance" with a number of available embedded x86 boards, and does pretty much everything. Of course it's not as fast as hardware built for the purpose, but if you have hardware encryption available I haven't yet seen a reason to choose a PIX/ASA over it on performance grounds. Obviously support and corporate love for Cisco are other factors which may come in to play for some applications.
Why are phones, particularly the VM box that is more or less an automatic part of today's cell phone, so damn vulnerable?
In most cases, because the users are stupid. In some cases, because the telco is stupid.
The majority of the time, the user will have a stupidly weak password like 1234, 123456, 111111, etc. I do VoIP for a living and one of the platforms I support, Broadworks, can not block a user from having a password like 123456. 111111 is banned, but easy sequences can't be yet. Due to this, I have on average 3-5 cases a month of people getting their accounts hacked and someone trying to forward calls to some other country. We block this system-wide, so it just results in the user's incoming calls breaking until it's noticed, but it happens with reasonable regularity.
In a few cases, the telco is retarded and allows the user to set that calls from their phone be allowed directly in to the voicemail system. Unfortunately they do not sanity-check this to verify that the call is actually coming from that phone and instead depend entirely on caller ID. Anyone with a VoIP or PRI system and a trusting upstream carrier can send whatever caller ID they want, making it trivial to get in to the voicemail. I think T-Mobile was in that category last time I checked, no idea on the others.
The Telcoes seem to have no trouble tracking our activities in great detail if those activities are something for which we can be billed, and they also seem eminently willing to cooperate with law enforcement. Why, then, do I have absolutely no way of knowing when, and from where, my VM box was called into, and why would the VM box of a phone that is subject to police investigation be accessible from the outside at all?
When, they should easily be able to give you and if they don't its only policy. From where, that's a lot tougher, given the ease of spoofing caller ID. Also, a lot of attacks are routed through multiple systems to disguise the source(s). Most attacks I see seem to come from other PBXes, likely hacked in similar ways.
Agreed on the box being open to remote access. It's trivial on my systems to allow incoming messages but break phone access to any given box, and investigators could just access the e-mail server that stores the messages directly via IMAP.
On consumer/small business internet connections, increased latency almost always brings jitter so it's just easier to look for latency. As I specifically noted, with low jitter a high latency connection can be used.
That was a commercial system with predictable timing and QoS measurement. Not exactly comparable to a consumer satellite internet system.
I mostly agree, they have their purpose if they're the only option but they can't compete with any of the other broadband options. 3G cellular or a long-run DSL are certainly better choices where available even as bad as they can be. If you don't have that though, satellite is an appealing last resort.
My thoughts exactly. VoIP is my day job and I can tell you that once pings exceed 200ms things get questionable. If there's low jitter, it can work and just have a delay like old intercontinental satellite PSTN links, but usually this is not the case. Any satellite connection using fixed dishes and thus geostationary satellites (a.k.a. everything marketed to home users) has an absolute minimum latency caused by the "last mile" of 472ms. This could only be achieved at the equator, anywhere else would be farther away and have greater latency but I don't feel like doing the math for Canadian latitudes.
tl;dr: VoIP on consumer satellite internet connections is stupid, end of story.
The broadband quality in my rural area sucks, big time. I can barely stream Netflix movies, let alone consider making phone calls with it.
If you can even contemplate streaming video, you can run a phone call over it just fine. The absolute worst case bandwidth needs are 64kbit/sec in each direction, and a non-shitty VoIP provider will offer multiple codecs which need far less.
Using AMR-WB (typically used for high quality cellular calls in places that aren't the US) I can easily put four simultaneous calls through a dialup connection while matching or beating the quality of a PSTN call. Any form of broadband, as long as it's actually working (no horrible packet loss or jitter problems), is more than sufficient for phone use.
Personally, I often find myself longing for the higher sound quality of a fully-wired phone line versus that of a cell phone. Isn't that ironic?
Try a VoIP phone with G.722 support some time. It'll blow you away. I thought they were gimmicks when I first got one in for testing, but it really makes the G.711 codec used by the PSTN sound like the shit it is. The bandwidth usage is exactly the same too at full rate, so the only reason it's not more common is legacy equipment which doesn't support it. A few lower bitrate options are in the 3GPP specs and are being used by certain cellular carriers in Europe and Canada to deliver better than PSTN call quality over mobile phones.
I think you're on to something with this. Blackberries seem to mostly be used as dumbphones with good PIM and e-mail functionality rather than the typical modern smartphone which gets used as a gaming/browsing/e-mail device that can probably make phone calls.
Just a note, a lot of Android devices get better battery life with WiFi turned on as long as you're in an area with an AP you can use, as the WiFi radio uses less power than the cellular radio for background internet tasks like syncing. I don't know whether Blackberries benefit from this since I know they usually require special services from the cellular carrier to get full functionality (I assume something to do with BES, which IMO is a pointless extra step that needs to go away), but it might be worth looking in to.
You are correct. Quoting Erik Johnson at Valve:
When we were getting very close to releasing Half-Life 1 (less than a week or so), we found there were already some projects that we needed to start working on, but we couldn't risk checking in code to the shipping version of the game. At that point we forked off the code in VSS to be both $/Goldsrc and /$Src. Over the next few years, we used these terms internally as "Goldsource" and "Source". At least initially, the Goldsrc branch of code referred to the codebase that was currently released, and Src referred to the next set of more risky technology that we were working on. When it came down to show Half-Life 2 for the first time at E3, it was part of our internal communication to refer to the "Source" engine vs. the "Goldsource" engine, and the name stuck.
On the battery side of things, you're wrong.
Check out the charts here: http://www.anandtech.com/show/4471/htc-sensation-4g-review-a-sensational-smartphone/10
The best Blackberry, the Bold 9780, is roughly equal to the iPhone 4 in all tests, trading off for the top spot in the web browsing tests. In the 3G talk time test both the iPhone and the Blackberry are firmly beaten by a number of Android devices.
I'll also point out that the Blackberry has a small 2.4" screen compared to the 3.5-4.3" screens of the majority of the competition and has a slow 624 MHz processor, compared to ~750 MHz in the iPhone and 800-1200MHz on the Androids. On paper it should have no problem beating the others simply from having less display to illuminate and a less demanding processor.
Standby time was not tested so it may win there, but honestly is it really that much of a problem to plug a phone in when you go to bed? I have a HTC Evo, one of the most power-hungry phones on the market, and that's all I have to do. Same as I've done out of habit with every phone I've owned in the past, smart or not.
Keyboard you're probably right, though personally I have never found a smartphone keyboard I liked. Likewise having not used Blackberry e-mail in years I'll refrain from judgement, though I can't say I have any complaints about either iOS or Android in that regard.
Oops, my brain mashed up the two likely ripped off engines in to one. Crafty and Fruit, not Fruity.
Many competitions have rules regarding dishonorable behavior. If he did use substantial amounts of open-source code without crediting the original authors as seems to be the case, that would be plainly dishonorable and thus grounds for disqualification.
As for the open source aspect, in this case a quick skim of the Fruity site seems to indicate (it's not very clearly worded) that it was initially open source during development and then went to a closed/commercial model after it had some wins under its belt. This makes sense to me if the developer started working on it as more of a personal challenge rather than for competition and then just ended up being good in competition.
On the point of telling your competitors your secrets, arguably Chrome and Firefox are competitors yet they're both fully open. If you're open source, your competitors may learn a trick or two from you but they might implement it in a completely different way. This may be better or worse. If they made different fundamental design decisions, they might not be able to make use of anything you did directly. If your main goal is to push the field forward and winning competitions is just a bonus, who cares if your competitors can see your code?
Just maybe the teachers were preparing you for high school and higher level maths. Education isn't just about showing how fucking clever you are to teachers, it's accepting that there are strategies and techniques that, once ingrained, allow you to learn better in the long run.
Ok, so I did it once, I know how to do it, why do I have to keep doing it on stuff I don't need it for? When I got to the harder math I couldn't do entirely in my head, guess what? I wrote it out!
But, obviously, you're such a fucking genius you still do everything in your head, right?
Nope, I use a calculator and write down intermediate results as necessary, which makes sense. You apparently missed the entire point, which was not to say that it's always pointless but that forcing a student to do it in situations where they don't need to only results in a pissed off student. "the phantom" above got it, what's so hard for you?
Then you teacher failed. Showing your work is about knowing the procedure to do something.
Then once you've showed it they should fuck off and let you just do mental math.
I absolutely despised math teachers who enforced showing of work. Pretty much everything done in a math class below the high school level I could do in my head nearly instantly. Writing out the whole process was a waste of 30-60 seconds per question. On a long test that could be half the class period where I could be reading rather than writing out how I managed to figure out that 13 * 24 = 312.
I understand that it is useful both as a teaching aid (to see where the student got it wrong) and as a deterrent to cheating (harder to copy and of course calculators don't tell you how they got there), but that still didn't mean it should be used without exception. If the student can prove they know the material, let them stop having to prove it every damn time.
HTML 5, CSS 3, massively improved Javascript engines...
Many of the standards in use today are either not fully supported or supported with quirks. Where the development resources are focused depends on how those technologies are used. This is demonstrated by Javascript in that to my knowledge the actual language has not changed in any significant way, but every major browser has completely reworked how they handle it to improve performance.
Go find a copy of IE 7, Firefox 3.0.x, or some old version of Chrome and see how it compares to their modern equivalents. It won't be as bad as IE 6, but it still will be a major step backwards. It's not like browser developers are changing things willy-nilly, they're doing it because it's an evolving market where the priorities can change.
Again I agree with the article. If you need a stable, unchanging platform the web is not for you.
I run all of these extensions and have been using FF5 since it became available on the beta channel. No problems.
Rarely does a plugin or extension actually break with a new release. The problem is that Mozilla's addon system allows the developers to be idiots and set a max version even if it's not needed. Max version, to me, is something to be set once you know it breaks rather than preemptively.
The whole point was that if they have such a long cycle for a web browser, they're doing it wrong. I agree with that. The web is a fast-moving platform, old browsers are bad.
You missed the second part about there not really being anything which only works in these versions. I'm sure someone can come up with something obscure, but it's not like the IE6 stragglers who actually have an excuse.
Why should Google continue to support outdated and/or broken browsers when all but a tiny handful of the remaining users of those browsers don't actually have a reason to and are just lazy about updating?
People also seem to think that Google will be putting up a wall like some stupidly designed sites do, attempting to block older browsers from even accessing the site. This is not the case, rather they won't be testing in those browsers nor will they work around bugs in those browsers. It'll probably keep working for some time.
Why should we have to support old versions of free software where in all cases the user would have had to go out of their way to disable automatic updates?
Any Firefox 3.5 user can update to 3.6. No system requirements changed and it's hard to believe anyone would have gone IE6-style and coded anything specifically to it. Same with IE7->IE8 and Safari 3->4.
If the users can upgrade at the press of a button and a problem is caused entirely by a bug in their outdated software, why should I fix it rather than then just updating to a version without the bug?
It's a point release, not a full new version. There's not really significant testing needed. It's also a web browser which is not integrated with the system in any way, making it absolutely trivial that if a user manages to stumble upon something that's actually broken in any important way, rolling back is trivial.
I agree that some software can not be rapidly upgraded in a sane business environment, but this is not one of those cases. It's a very minor upgrade that's been out for 17 months and even superseded by a major revision. A conservative testing plan would be to sit back for a few weeks initially and let the first patches show up for those issues that slip through testing but show up quickly upon general release. During this time a few bored people in IT should upgrade and test the basics of anything critical. Once it looks stable as a whole, deploy to all of IT and possibly some savvier users in each department to get a sampling of the webapps used by the company. If no stoppers show up in a month roll it out fully.
Something like an operating system deserves months of testing. A browser, not so much. The three browsers that Google is no longer supporting were carefully chosen, as there are no technical reasons that anyone using one of those three could not move up to at least the next one in line. IE9 is Vista/7 only and the Mac versions of both Firefox 4 and Safari 5 require OS X 10.5, so reaching the current version of any of these browsers could require an OS upgrade, but there are still upgrades available for all users of the dropped browsers. None of them have the IE6 problem of piles of shitty webapps written to their quirks either.
If it's that much of a hassle to run a 64 bit browser, why bother? I don't know about you but I get concerned if my browser's memory usage hits 2GB much less 4, I can't imagine it needs to do math on really big numbers very often, and the plugin situation is just easier to deal with. There's no real gain I see to a 64 bit browser, so I haven't ever figured out why people complain so much about Mozilla not releasing 64 bit binaries. Maybe it's different on other architectures, but on x86 and PPC there is no performance penalty from running 32 bit apps. Hell, a lot of Windows Vista and 7 installs are 64 bit including my dual boot to 7, but other than drivers and the kernel itself I can't think of any 64 bit apps I have installed.
These are all versions which you have no excuse to still be on, so Google is simply saying that they will no longer test for those browsers or fix bugs found only in those browsers.
Anyone on FF3.5 can move to 3.6, IE7 to 8, and Safari 3 to 4, so there are no valid reasons to still use them. Those browsers don't have any IE6-like hold on a market due to shitty apps coded explicitly for them, nor are there major changes which could throw off a user.
It's brilliant deception through marketing.
AT&T has deployed HSPA+ to virtually 100% of our nation's fastest mobile broadband network, which enables 4G speeds when combined with enhanced backhaul.
Translation:
AT&T has enabled your phone or wireless modem to connect to the tower at HSPA speeds. Good luck using that speed when most of the towers still use a small number of T1s to get to the internet.
It's like having a 512/128kbit entry-level DSL internet connection and upgrading your WiFi access point from an old 802.11b unit to a new MIMO 802.11n device. Sure, the air link is faster, but the internet connection is still the bottleneck. Those lucky enough to have towers with upgraded backhaul connections may be able to get interesting speeds, but most won't.