> Unless you are a web designer who wants to make sure that his site looks correctly when viewed with a browser that adheres to STANDARDS,
Here we go again... Beating the tired STANDARDS drum...
Repeat after me...
IE has 95% market share.
However IE does it is the de-facto standard.
If a page works in IE but not in Mozilla/Galeon/other browser of miniscule market share, then why would a regular end user ever want to switch?
Users don't care about standards, they just want to get their job done.
You can talk all you want about all the so-called "standards" that Moz/Galeon/Gecko adhere to (and arrogantly dismiss the myriad de-facto standards), but at the end of the day, if the user experience is less than for competing browser or if regular (non-slashdot) users can't get their work done, then standards don't matter.
-Steve, who thinks Galeon is the world's best browser
I too have been surveying some of the photo album packages that are available out there ever since I picked up a digital camera.
While there are some nice packages such as Gallery, IDS, My Photo Gallery, and others, I prefer Album. (Actually, I prefer Album plus my fixes and enhancements.) The one reason that I don't like albums that have server-based components to them is that they tend to do all sorts of image resizing operations at runtime, making them horribly slow.
Album, on the other hand, recognizes that photos are static content, and takes the performance hit of generating thumbnail galleries and web-friendly images right up front (i.e. offline), and as a result, the speed that end users see is limited only by how fast Apache can serve up files.
The approach that I am taking for long-term archival is to burn Album-generated photo albums into CDs, such that the only requirement to view them will be a browser. HTML and JPGs are so pervasive, I don't see them being replaced anytime soon. I can burn a CD, send it via snail mail to my Mom, and she can browse away easily (including "themes"). The same cannot be done with these [php- | cgi- | perl-] based solutions that require some server software to be setup and configured.
The only advantage that the server-based programs provide is the ability to add comments, but that seems overrated IMHO.
-Steve
Now this one would be a bit more tricky. I only allow raw device access to root, and all writing to my mounted windows partition(s) is restricted to root: It would be tough for a virus to come pre-loaded with local root exploits for a wide variety of systems without it weighing-in around a couple of megs.
When you are in Windows, your permissions on raw device access do not matter. The software running on windows has full access to all the hardware, all the partitions, and there is effectively nothing to stop any software from doing whatever it wants to a mounted FAT32 partition, or your unmounted Linux partition.
I haven't used it in a while, but in the past there was a suite of Win32 tools called the e2tools (I sure someone else can correct my memory) that would allow you, right from your DOS prompt, to look at any file on an ext2 partition. Set an environment variable, and then type e2ls/home/foo and it would happily display the contents of/home/foo the same as if you were booted into Linux and typing at a Bash prompt.
This seems like a secure solution for a small shop, where the boxes are physically located in the same working place as the administrators.
But I don't see how this would scale to more than a few boxes. As much as we like to believe in the rock-solidness of our servers, there are many things that will cause them to crash, usually at the 3am or some other highly inopportune time.
When servers go down and then come back up, you can't have your site inoperable because the console is stuck at some password prompt, waiting for human intervention. I say this from experience: just last night one of our backend Solaris machines had some kind of trouble and rebooted itself.
Honestly, I don't know the best solution to the original question, but my experience says that it's better to automate and be able to shut down quickly when there is suspicious activity rather than relying on human intervention for routine activities such as starting up services/applications.
It is easy really, just get one of the hard-core Gnu/Linuxers to start ranting about how free doesn't really mean free, but instead means freedom.... Add in some reference to speech and beer...
One slimy telemarketer slamming trick was to call people and ask, "Can I put you on hold?"
Most folks, in being polite, would say OK, expecting to be put on hold (in the usual telephone sense).
Instead, they had their telephone service switched to another provider, and put on their H.O.L.D. program. [Sorry, I forget what the acronym stood for.] So, strictly speaking, the tele-slime could say that they asked the customer if they wanted it, and the customer did in fact say yes.
Galeon is a great example of the Unix philosophy. Find one thing, web browsing in this case, and do it well. There is no mail client, no instant messenger, or extra stuff to complicate things. Though it does depend on Mozilla libraries, it has a refreshing lightweight "feel" that Mozilla and even old Netscape lack.
Though not the first to implement the feature, Galeon sets the standard for tabbed browsing.
Actually, it didn't work for Apple. They are still stuck in the education market, and that has not translated into any measurable difference in market penetration beyond their core graphics constituents.
This is as repugnant as the 'bit tax' proposed in the early nineties. If you don't remember, legislators were considering taxing every digital transmission that related to business. That's right, every fax transmission, every X-Y-ZModem upload and download, every corporate communication on CompuServe or Delphi. It never made it into law because everyone realized how insidious it was.
Oh, c'mon now. That bit tax was a hoax which had gullible users everywhere forwarding the same alarmist e-mails, over and over and over. Source please?
Perhaps its because I am looking out from the ivory tower, see my web page if the email address doesn't make it immediately obvious, but there isn't anything on the web that I actually want to use that doesn't render perfectly happily in Navigator 4.x. So why am I loosing anyuthing at that point?
Actually, you make my point. Netscape 4.x does render pages that are not in strict conformance with <insert standard here>. This is a good thing for users.
Why has the web got anything to do with what business wants?
Perhaps that was a poor choice of a name on my part. I should have said, "But at the end of the day, the browser which just "works" for Joe User is the one which wins." I was not trying to make a claim that what business wants it what all web users should get.
Your priorities aren't my priorities, your real world doesn't necessarily correspond to mine.
Sidestepping thorny philosophical arguments about your reality versus my reality, it is an undisputible fact that the web if full of broken and non-standard HTML. I only argue that it should be a requirement for a browser's rendering engine to make a "best effort" to render non-compliant markup, and not to give up and huffily refuse to continue until the markup meets some quality bar. By sticking to the standards, and refusing to recognize that most, if not all, web sites out there contain "illegal" markup, you do the project a disservice, and relegate it to niche status. Nobody has refuted this.
argument about flash
Personally, I think Flash is trash, but that is beyond the scope of the argument.
If AOL rolls out a Gecko-based client, it will take a long long time to do so. Further, they will then have to migrate 24 million accounts from IE to (likely less-functional) Netscape 6.x. Having had to do migrations before, this is a very difficult task.
Bottom line, even if AOL were able to get all of its customers to switch (anyone want to take bets on this?), it would take several years. Since MSN is adding accounts at a faster rate than AOL, I claim that getting to 40% market share (as you claim) would be extremely difficult, if not impossible. But even if they were able to do so, it would still be a minority position, and they would not be able to command the web to rewrite all the pages in a "standards-compliant" way.
So the fact that Gecko fails to render non-standards compliant HTML is a good thing.
No, it is not. The genie is long out of the bottle, and there is no going back. Non-standard HTML is just way it is.
By maintaining an ivory tower postion, where only so-called standards-compliant HTML gets rendered, and ignoring the web-wide reality of broken HTML generators, common HTML errors, and commonly used HTML extensions, we embark on a losing strategy.
This strategy leaves you a marginal player, whining on the sidelines, out of any consideration of being a real player. Yes, you get to feel smug about taking the "high road" and pontificate about how standards-compliant you are, and how stoopid web authors are to blame for users' troubles.
But at the end of the day, the browser which just "works" for Joe Business is the one which wins. Simple as that. No arguments about standards. No lectures about speach and beer. It either works, or it doesn't.
Reminds me of that saying: In theory, theory and practice are the same. In practice, however, they are different. Time to get away from theory and have a look at real-world practice.
I dunno abour your predictions, but AOL switching to a Gecko rendering base will do wonders for web standards compliance.
This sounds nice in theory, but the simple fact of the matter is that however Internet Explorer does it is the de-facto standard. With IE commanding over 90% of the market, Mozilla, Galeon [my favorite], Konqueror, and others are all down in the noise.
What am I going to tell my Mom when she asks why she can't access some non-standards-compliant site, which used to work quite nicely with IE, but now is completely unusable? "Well, Mom, you should feel good because you are using a browser that is fully compliant with all of those three-letter acronym standards, and this bad site does not comply. No, Mom, you can't use this site anymore. Yes, Mom, your old AOL account would have worked fine."
With 2% or less of market share, we ignore the installed base of IE at our own peril.
Offtopic rant: if Gecko is so darn fast, why is Mozilla so s l o o o o w?
Re:Check this code to see why RMI is so dangerous
on
Java RMI
·
· Score: 1
While it is indeed cool that invocation of a remote call is syntactically equivalent to that of a local call, it masks the true "cost" that you pay.
It is a fact that remote calls are more expensive in terms of network bandwidth, latency, and response time compared with a local in-process procedure call.
The problem is that RMI makes expensive remote calls look like cheap local calls. This can lead to performance problems which are hard to find, since the offending code looks so innocuous.
Secondly, VA/OSDN are for-profit. If subscriptions are successful, and they get more than they need, will the subscriptions be extended? Or will Taco, Hemos, ESR & Larry Augustin pocket the money?
So what?
Why do you think that these people, who have worked hard to build a business, should not reap the the fruits of their labors?
If subscriptions are successful, and they get more than they need, I hope they get stinking rich, and go buy fast sports cars, Aeron chairs, and lots of toys!
Re:A new FUD campaign, I swear
on
HTTP's Days Numbered
·
· Score: 2, Informative
Hell, I'd love to see telnet-over-HTTP done while we're at this.
Outside the Seattle area, the primary electric utility is Puget Sound Energy. About a year ago, they received approval from the Washington State Utilities Commission to implement time-sensitive rates. In essence, they charge a higher rate when consumption is higher, and a lower rate when consumption is lower.
Since it is supposed to be most cost effective to produce electricity at a constant rate, rather than larger fluctuations, this was an attempt to do "traffic shaping".
Despite some initial misgivings about the unknowns, it has worked pretty well so far. Lots of folks (my family included) now make it a habit to delay starting the laundry or dishwasher until after 9pm to get the better rate.
I am not sure exactly which technology Puget Sound Energy is using to transmit the readings back to the home office. On each monthly bill, they have a bar graph indicating how much electricity was consumed in each of the rate periods.
One thing that is both a blessing and a curse is that it does not (at least, our internal versions do not) come with a repository system a la Visual Age
Well thank God for that.
That sucking sound you hear when using Visual Age is your code being pulled into the repository, never to be heard from again. I will stick with CVS, thank you, and avoid lock-in where our source code is held hostage.
While the repository had some cool features, it sure seemed like a trap door to me.
And while we are at it, let's get rid of the annoying, arrogant info pages.
How many times have you done
man somecommand
only to be scolded that this man page is begrudgingly provided to you as a "courtesy," and that to get the full documentation, you have to learn someone's nonstandard, failed attempt at hypertext documentation.
Like it or not, 'man' is the standard for Unix documentation.
Like it or not, HTML is the standard for linked documents.
Slashdot Mirror
AOL browser is hacked version of IE.
A 1.3% market share browser doesn't call the shots. A 1.3% market share browser doesn't make standards.
As for upsdell.com:
Here we go again... Beating the tired STANDARDS drum...
Repeat after me...
- IE has 95% market share.
- However IE does it is the de-facto standard.
- If a page works in IE but not in Mozilla/Galeon/other browser of miniscule market share, then why would a regular end user ever want to switch?
- Users don't care about standards, they just want to get their job done.
You can talk all you want about all the so-called "standards" that Moz/Galeon/Gecko adhere to (and arrogantly dismiss the myriad de-facto standards), but at the end of the day, if the user experience is less than for competing browser or if regular (non-slashdot) users can't get their work done, then standards don't matter.-Steve, who thinks Galeon is the world's best browser
Um, how about libsafe?
-Steve
While there are some nice packages such as Gallery, IDS, My Photo Gallery, and others, I prefer Album. (Actually, I prefer Album plus my fixes and enhancements.) The one reason that I don't like albums that have server-based components to them is that they tend to do all sorts of image resizing operations at runtime, making them horribly slow.
Album, on the other hand, recognizes that photos are static content, and takes the performance hit of generating thumbnail galleries and web-friendly images right up front (i.e. offline), and as a result, the speed that end users see is limited only by how fast Apache can serve up files.
The approach that I am taking for long-term archival is to burn Album-generated photo albums into CDs, such that the only requirement to view them will be a browser. HTML and JPGs are so pervasive, I don't see them being replaced anytime soon. I can burn a CD, send it via snail mail to my Mom, and she can browse away easily (including "themes"). The same cannot be done with these [php- | cgi- | perl-] based solutions that require some server software to be setup and configured.
The only advantage that the server-based programs provide is the ability to add comments, but that seems overrated IMHO. -Steve
I'm with you, Mozilla 1.0 looks better (on my Viewsonic tube). Perhaps it is only LCD screens where there is a noticable difference?
-Steve
When you are in Windows, your permissions on raw device access do not matter. The software running on windows has full access to all the hardware, all the partitions, and there is effectively nothing to stop any software from doing whatever it wants to a mounted FAT32 partition, or your unmounted Linux partition.
I haven't used it in a while, but in the past there was a suite of Win32 tools called the e2tools (I sure someone else can correct my memory) that would allow you, right from your DOS prompt, to look at any file on an ext2 partition. Set an environment variable, and then type e2ls /home/foo and it would happily display the contents of /home/foo the same as if you were booted into Linux and typing at a Bash prompt.
Could be done. Could be done pretty easily.
-Steve
But I don't see how this would scale to more than a few boxes. As much as we like to believe in the rock-solidness of our servers, there are many things that will cause them to crash, usually at the 3am or some other highly inopportune time.
When servers go down and then come back up, you can't have your site inoperable because the console is stuck at some password prompt, waiting for human intervention. I say this from experience: just last night one of our backend Solaris machines had some kind of trouble and rebooted itself.
Honestly, I don't know the best solution to the original question, but my experience says that it's better to automate and be able to shut down quickly when there is suspicious activity rather than relying on human intervention for routine activities such as starting up services/applications.
-Steve
It is easy really, just get one of the hard-core Gnu/Linuxers to start ranting about how free doesn't really mean free, but instead means freedom.... Add in some reference to speech and beer...
<ducking>
One slimy telemarketer slamming trick was to call people and ask, "Can I put you on hold?"
Most folks, in being polite, would say OK, expecting to be put on hold (in the usual telephone sense).
Instead, they had their telephone service switched to another provider, and put on their H.O.L.D. program. [Sorry, I forget what the acronym stood for.] So, strictly speaking, the tele-slime could say that they asked the customer if they wanted it, and the customer did in fact say yes.
The customer did, in fact, get slammed.
Galeon is a great example of the Unix philosophy. Find one thing, web browsing in this case, and do it well. There is no mail client, no instant messenger, or extra stuff to complicate things. Though it does depend on Mozilla libraries, it has a refreshing lightweight "feel" that Mozilla and even old Netscape lack.
Though not the first to implement the feature, Galeon sets the standard for tabbed browsing.
Actually, it didn't work for Apple. They are still stuck in the education market, and that has not translated into any measurable difference in market penetration beyond their core graphics constituents.
Oh, c'mon now. That bit tax was a hoax which had gullible users everywhere forwarding the same alarmist e-mails, over and over and over. Source please?
Actually, you make my point. Netscape 4.x does render pages that are not in strict conformance with <insert standard here>. This is a good thing for users.
Why has the web got anything to do with what business wants?
Perhaps that was a poor choice of a name on my part. I should have said, "But at the end of the day, the browser which just "works" for Joe User is the one which wins." I was not trying to make a claim that what business wants it what all web users should get.
Your priorities aren't my priorities, your real world doesn't necessarily correspond to mine.
Sidestepping thorny philosophical arguments about your reality versus my reality, it is an undisputible fact that the web if full of broken and non-standard HTML. I only argue that it should be a requirement for a browser's rendering engine to make a "best effort" to render non-compliant markup, and not to give up and huffily refuse to continue until the markup meets some quality bar. By sticking to the standards, and refusing to recognize that most, if not all, web sites out there contain "illegal" markup, you do the project a disservice, and relegate it to niche status. Nobody has refuted this.
argument about flash
Personally, I think Flash is trash, but that is beyond the scope of the argument.
If AOL rolls out a Gecko-based client, it will take a long long time to do so. Further, they will then have to migrate 24 million accounts from IE to (likely less-functional) Netscape 6.x. Having had to do migrations before, this is a very difficult task.
Bottom line, even if AOL were able to get all of its customers to switch (anyone want to take bets on this?), it would take several years. Since MSN is adding accounts at a faster rate than AOL, I claim that getting to 40% market share (as you claim) would be extremely difficult, if not impossible. But even if they were able to do so, it would still be a minority position, and they would not be able to command the web to rewrite all the pages in a "standards-compliant" way.
No, it is not. The genie is long out of the bottle, and there is no going back. Non-standard HTML is just way it is.
By maintaining an ivory tower postion, where only so-called standards-compliant HTML gets rendered, and ignoring the web-wide reality of broken HTML generators, common HTML errors, and commonly used HTML extensions, we embark on a losing strategy.
This strategy leaves you a marginal player, whining on the sidelines, out of any consideration of being a real player. Yes, you get to feel smug about taking the "high road" and pontificate about how standards-compliant you are, and how stoopid web authors are to blame for users' troubles.
But at the end of the day, the browser which just "works" for Joe Business is the one which wins. Simple as that. No arguments about standards. No lectures about speach and beer. It either works, or it doesn't.
Reminds me of that saying: In theory, theory and practice are the same. In practice, however, they are different. Time to get away from theory and have a look at real-world practice.
This sounds nice in theory, but the simple fact of the matter is that however Internet Explorer does it is the de-facto standard. With IE commanding over 90% of the market, Mozilla, Galeon [my favorite], Konqueror, and others are all down in the noise.
What am I going to tell my Mom when she asks why she can't access some non-standards-compliant site, which used to work quite nicely with IE, but now is completely unusable? "Well, Mom, you should feel good because you are using a browser that is fully compliant with all of those three-letter acronym standards, and this bad site does not comply. No, Mom, you can't use this site anymore. Yes, Mom, your old AOL account would have worked fine."
With 2% or less of market share, we ignore the installed base of IE at our own peril.
Offtopic rant: if Gecko is so darn fast, why is Mozilla so s l o o o o w?
It is a fact that remote calls are more expensive in terms of network bandwidth, latency, and response time compared with a local in-process procedure call.
The problem is that RMI makes expensive remote calls look like cheap local calls. This can lead to performance problems which are hard to find, since the offending code looks so innocuous.
Intel's demo showed Skamania (cute name) hopping between 802.11, 802.3 and fixed Ethernet, ....
Skamania is the name of a county along Washington State's southern border with Oregon, by the Columbia River.
Intel has a facility nearby.
So what?
Why do you think that these people, who have worked hard to build a business, should not reap the the fruits of their labors?
If subscriptions are successful, and they get more than they need, I hope they get stinking rich, and go buy fast sports cars, Aeron chairs, and lots of toys!
Well, OK.
Since it is supposed to be most cost effective to produce electricity at a constant rate, rather than larger fluctuations, this was an attempt to do "traffic shaping".
Despite some initial misgivings about the unknowns, it has worked pretty well so far. Lots of folks (my family included) now make it a habit to delay starting the laundry or dishwasher until after 9pm to get the better rate.
I am not sure exactly which technology Puget Sound Energy is using to transmit the readings back to the home office. On each monthly bill, they have a bar graph indicating how much electricity was consumed in each of the rate periods.
-Steve
Well thank God for that.
That sucking sound you hear when using Visual Age is your code being pulled into the repository, never to be heard from again. I will stick with CVS, thank you, and avoid lock-in where our source code is held hostage.
While the repository had some cool features, it sure seemed like a trap door to me.
How many times have you done
man somecommand
only to be scolded that this man page is begrudgingly provided to you as a "courtesy," and that to get the full documentation, you have to learn someone's nonstandard, failed attempt at hypertext documentation.
Like it or not, 'man' is the standard for Unix documentation.
Like it or not, HTML is the standard for linked documents.
TeamOn Systems web-based e-mail service has been aggregating all kinds of proprietary e-mail for some time now, including AOL and Hotmail.