> While you also make good points, I think it's wise to lean in favor of pragmatic technology choices rather than dogmatism in favor of "one side of the other".
Lets just approach this from a purely business perspective.
What you are telling me is that I should invest in technology that doesn't even comply with its own standards, that forces me to redo a perfectly well workign website just coz MS decided to re-re-re-reinvent a standard from half a decade or more ago (CSS1) ?
Or should i use a standard that changes incrementally and where each change is relatively cheap for me because it takes past versions into account and/or explains clearly where things will break?
The later is so amazingly much cheaper that only one conclusion can be left. The peopel who decide to keep IE in their companies are throwing away money, and those who keep developing exclusively for IE are wasting their and everyone elses money.
> Yahoo themselves admitted they would use more Java if they weren't locked into thousands of FreeBSD servers.
I have migrated machines between FreeBSD and Linux and the other way around.. given a few right choices, it is not that much work since virtually all the same apps are available, and you have enough control over both to make them as you want.
THere is a price to migration, but between 2 open sourced variations of Unix such costs will be relatively low when you have the knowledge to maintain a large cluster of such machines and change the source and rebuild things as desired already anyway.
So there is a matter of JAVA being important enough to overcome the costs of that and any possible disadvantages it might bring. It seems that its worth the wait for a practical distributable JAVA on FreeBSD for as far as yahoo goes..
I am using FreeBSD with JAVA right now, and with the latest patches it is quite stable and usable. There are other problems but most are not of a technical nature, but rather of the way Sun deals with JAVA and not confined to FreeBSD really. It seems tho that at some point FreeBSD can distribute JAVA binaries once their current patchwork on JAVA is deemed good enough... so there is some hope for Yahoo to get both without having to move in the end.
When you need reliable things on a large scale, you may indeed also have a preference for things that had more time to prove themselves. That however will only work favorable for such inherently older thigns if they do actually prove themselves.
Legacy is King only in certain areas of IT, and guess what, sometimes it even matters. Conservatism with regards to OS choice however has little to do with legacy but with wanting things that have proven themselves, which can at times make you end up with what you'd call legacy stuff. Its a bonus in the case of FreeBSD that it in specific cases also performs better, tho one that many people worked quite hard for.
Hmm, that may be true for Google since they generally don't have to keep state. When you have to keep track of sessions and need a substantial amount of sharing between your server instances as a result, a cluster of dual cpu machines or even quad cpu machines may turn out to perform a lot better.
At any rate, if your goal is to handle lots and lots of relatively independent queries, then yeah, large clusters of relatively small machines do a very good job.
> Yahoo is a large website operation. For most corporates the web server is a _very small_ part of the overall IT infrastructure. > Yahoo is not representative of what you'd find in a typical corporate datacentre / computer room.
They are indeed not representative, but you kindof forget one thing here..
Yahoo depends on its web servers. If they stop running, Yahoo has no business.
That means that their choice and motivation counts for a lot more then what a company says who doesn't depend on the stuff to begin with.
I used to run FreeBSD 5.2.1 with Apache 2, MySQL 4.0, PHP 4.3.7 (with Turckmmcache) and Geeklog on a dual CPU pII 333 machine with 512mb ram. (the machine is still running MySQL, but geeklog moved to a bigger machine by now)
I also tested this with Debian (current) and gentoo.
With both Linux distributions, this configuration tops at approx 10k requests/hour with 20 parallel connections.
The exact same setup on FreeBSD handles upto 12k requests/hour with 20 parallel connections.
In both cases the machine its rather busy. In neither case it runs out of memory.
In case of FreeBSD I do get a response from the shell within a few sec still, that is not the case on Linux.
As I said, purely anecdotal, but enough reason for me to believe the 'handles high load very well' claim. approx 20% more connections is not a little bit either in my world.
Cellphone with PDA? yeah, but do me a seperate PDA
on
Are PDAs Simply Finished?
·
· Score: 4, Insightful
I rather want a phone with very basic PDA functions (addressbook, agenda, possibly a notepad) and a seperate dedicated PDA.
Here is why:
I can still use my PDA while talking on the phone for things like looking up info or taking a note
PDA formfactor has a much better usable screensize, esp. for someone who is visually impaired.
I can use my PDA to read/write at places where I have to turn off my phone such as in an airplane
I keep my PDA and phone synced to my pc and to eachother using bluetooth, so I take my notes and appointments once, and have them everywhere even if I only carry my phone around.. I do end up taking my PDA with me as well tho most of the time.. its small and light enough to just fit in the pocket of a shirt.
Since my phone doesn't have to have a as big as possible screen or even color, it can me sxtremely small and have a very low energy usage, resulting in being able to carry it around for a logn time without charging.
So well, by not wanting phone functionality from my pda, and only wanting very limited pda functionality from my phone, I end up with 2 small and light devices. I can always carry my appointments and such with me in a small and light phone, and it is little bother to take a pda with me as well since it is also light and small. Since they have a wireless link I can still browse the internet and do mail on my pda, and do so with a nice well readable screen and something more comfortable then a phone keyboard.
It's kindof funny, neither my pda or phone is new, both are over 2 years old actually, but the combination ends up being very usable, and as it is, I often end up reading slashdot during my regular 5 hours long trips to Berlin by train, only depending on the availability of the cellular network, but with a repeater in the train itself that is not a problem.. and knowign that by the tiem I get there the batteries of my PDA will be somewhat drained, but my phoen will have enough power to last another week:)
Who the fuck want to carry around the huge bricks called pocketpc?
You are completely failign to get the point of a PDA.
That said, I doubt PDAs are a very big market anyway, and many who are in the market got carried away by the idea of making a pocket size computer.
A PDA must be a very convinient, EXTREMELY LIGHT, easy to use agenda, address book, and optionally a way to take notes, that also keeps working even if you forget to charge it for a week or 2.
The battery live issue excludes most phones (tho.. my phoen on a larger battery will easily last that long) and most phones lack a decent sized screen.
Weight, battery live, and over complicatedness makes pocketPC an even worse alternative then the PDA features of the average phone.
Don't get me wrogn btw, if you NEED or want a pocket sized computer, then a pocketpc may be for you, but it is simply not the same as a pda, and it doesn't fucntion very well as pda.
> However, I don't see blocking mail coming directly from home users' systems as taking away control,
I see it as taking away my control over how my mail gets delivered. Control over my own computer has nothing to do with it.
> unless their email server is hosed, which it sounds like it is. You said it was slow, which is not very specific,
I was not the original poster you replied to, I was just better at reading the parent post and can compare it to my own experiences and that of many other users of a whole variety of ISPs everywhere in the world. It is a relatively common complaint about smtp servers run by ISPs. Currently the smtp server of my ISP is doing quite well, but that hasn't always been the case.
The 'pick another ISP' argument is not a good one here either, it only works if that option is actually available in your area.
> In the end, the ISP has control over every packet leaving your network anyway.
Yes, and I pay them to get a network connection. Within the terms of service, I have control over how I use that connection.
> Running servers violates your user agreement with Comcast anyways, so what difference does it make to them if you can't run your mail server on your cablemodem? If you want to run a server, pay for the *right* service. Don't just try to be a cheap bastard.
It seems that Comcast cares enough to not just close off port 25 alltogether. In their own words, they are aware of people running legitimate mail servers on such accounts. So.. what is your point exactly when Comcast itself seems to disagree with you in word and behavior?
> Even if the mailservers are slow as hell, this should not affect anyone who is not trying to use them for business purposes, for which you should have a business account in order to comply with the AUP. Most AUPs also prohibit running a mailing list on a home account, which pretty much rules out the other purpose of sending a lot of mail from home.
Why should quick delivery not matter to a home user? WHat kind of elitst arogance is that really?
It has nothign to do with how much mail you send but with mail arriving within a reasonable time. I see no reason why I should have to wait 10 hours for a mail to get delivered because of the virus scanning of my ISP getting overloaded for example, and I don't see how you can argue that that should be acceptable either.
Again, has nothing whatsoever to do with mailinglists. (besides the fact that the ISP I use allows me to run a private mailinglist, webserver, mailserver and such and gives a static ip)
> I feel that avoiding your ISP's mailservers because you want to eliminate hops (why? does it matter?) is somewhat contrary, and unnecessarily so. Unless their server is holding mail for inordinately long periods of time, or dropping it, I don't see the issue...
The problems people were mentioning were:
Slow (as in, takes LOOOOOOOOONG to deliver mail)
Unreliable, as in: doesn't deliver mail reliably, which kindof implies it drops it every now and then eh?
So, those are the exact cases you mentioned I would say..
I get the feeling you are not interested in the argument at all and just want to see control taken away from the end-user in the name of fighting spam.
> Exactly. Promoting well-written websites is in their interests as they don't have to work so hard on their broken HTML parser, and they get more semantic information to work with.
Which just means that when their bot gets better information to work with, it will do a better job and the pages will be easier to find. The result you are looking for will happen without the petition and without favoring xhtml in their pageranking.
> What? Nobody would use such a thing, it's infeasible to implement part of it, and it doesn't give the incentive to create better pages (which is the whole point of the petition).
Nobody would use such a thing because they are interested in information relevant to their search, and not in whatever standard it complies with as logn as they can read it.
That is also exactly why it is against the interest of a search engine to favor one standard over another.
While no doubt you are right in your technical argument about it beign easier for google to parse correct xhtml, it is in their interest to present the most relevant information, and that simply has nothing to do with what markup language the results are in as long as they can assume the majority of their users can display it.
Or to put it differently, when a user searches for somethign on google, they are interested in the information, and that is what should determine the pagerank. WHat form the information is in is a subject for filtering (just like image searches)
If you don't understand that simple thing, then the only point of the discussion is the technical differences between html and xhtml, and I am sure you can explain those better then I can and is interesting enough. But if you send or propose to send a petition to google, take a look at where their business is. Seeing the consious efford they put into indexing things like pdf besides html makes me think that they care abotu relevance of information more then anything else.
Besides that, when the web is soemwhat difficult to index and they have developed the technology to still do it relatively well, why would it be in their interest to eliminate that advantage?
Hmm, did you try to do any tuning whatsoever on that machine? Something simple liek loading accf_http and takign a peek at memory usage to make user its not swapping? Also, did you partition your drive such that you have a small seperate root filesystem so you can also use softupdaes on the partition where squid's and Apache's data recides?
on another note: I have run many such configurations, and under very light load, Linux usually outperforms them, but as soon as I actually start using them, they end up being faster.
Hmm, and how about the fact that it gives the producer of said sofware a bad name? I am of the impression that that is the one and only thing that is at least causing a little bit of movement at Microsoft with regards to creating more secure products.
Another effect of this is that the end-user is at least a little bit aware of the problem, and doesn't fully trust his/her computer...
I think that in the end we are more secure as a result, its just nto a very direct result.
I am aware of the difference between XML and SGML, but you are as well as me aware of the fact that HTML is a severely stripped down subset of SGML, to the point of it being simplistic.
If you are talking about an XML vs an SGML parser, then yeah, XML is a lot simpler. HTML is still simpler then that and to render the html subset that is usable on a small screen you only need a very simple parser, not a full fledged sgml parser.
Btw, I quite see reasons to want xhtml, but those are not the ones mentioned in the google petition, at this moment they simply fail to make up in real world situations.
It is simply not true that a technical specification is going to make thigns better for disabled people for example, a THOUGHTFULL implementation of a specification that keeps them in mind and uses what the spec has to offer for helping them is what can make the web more accessable.
XHTML offers even better ways to support disabled people then previous versions, but seeing the utter lack of usage of already existing features is rather telling me that more and better features (tho nice to have) will not solve the problem. Getting peopel to actually use such features will help.
If people insist on making their website look cool and don't care about this at the same time, it will keep going wrong, no matter what HTML or XHTML or other markup language you want to use.
The modular setup and extendebility are two very good reasons for wanting xhtml.
If and when people write XHTML and use the specification to its possibilities, then sure, you end up with smaller pages that will work on almost anything. MIn fact, the same was true for html 4.01 + CSS 2.0 when used properly. Many sites manage to use it in a way that validates and yet won't render properly or even in a usable way with anything other then internet explorer
Looking at what is around on the web and validates as XHTML vs what validates as html 4.01 transitional, the simple result is still that esp. on a small device like a palm m505, the later seems to work a lot better, and usually seem to result in smaller transfers. That may be a result of bad design and bad usage of xhtml maybe, but that just brings us back to the point I was making regarding the google petition and disabled people, it is not a matter of what standard you comply to so much as how well you design your stuff.
Back to the discussion, google is a search engine. Go read what they have to say themselves and it seems that their philosophy is to be better then anyone else by doing a better job at returning relevant results. That is their livelyhood for now, and they go to the point of converting things like pdf to html to be able to index those as well.. Do you really believe they are going to consider this untill the majority of both websites and users support it (not just int heory but in reality)
A much better thing to suggest to them and to call a petition for is to add a filter to only get results that are either compatible with your browser as reported by its user-agent, or comply with a specific standard, and to display the (x)html version a page complies with as part of the search result.
What this current petition suggests is simply opposed to the one thing that made them bigger then anyone else, so no chance that they are gonna listen.
The argument regarding visually disabled people sounds better then it is btw. Google will do a better job at indexing pages that are easily accessable with a text only browser as a simpel result of their technology or so they claim (read about optimizing your site for indexing) and that makes a lot of sense also when you consider they are mostly indexing text.
They should, instead of meddling with standards by means of their search results, start generating valid html themselves (and I don't care what version)
Btw, I am visually impaired myself (have a visus of approx 20%). While I can use a display given a large enough one and decent sized fonts, and can use
Past experience tells they don't, and untill they have provided some very compellign proof of the opposite, I am not going to assume they are able to change themselves into a producer of secure software.
And honestly, Longhorn is in no way convincing. The first big problem is with their approach to what the end-user should get. It doesn't matter how unsafe it is as logn as it is convinient. Now, isolating all those unsafe components by means of 'managed code' is adding a lot of overhead, partially hiding the problems and make them even less easy to udnerstand for the casual user, while nto solving the problems.
Security is not a product, it is not a technology, and not even a process. Security starts with the right mindset, it is a matter of mentality first of all. Anythign else follows from that.
If you encourage the wrong mindset from the start, then you actively encourage an insecure environment.
Microsoft has been doing that for decades, and I have no reason to believe they changed their mind, rather the opposite actually, they go for a technical approach to what is first of all a non technical problem.
Slashdot Mirror
> While you also make good points, I think it's wise to lean in favor of pragmatic technology choices rather than dogmatism in favor of "one side of the other".
Lets just approach this from a purely business perspective.
What you are telling me is that I should invest in technology that doesn't even comply with its own standards, that forces me to redo a perfectly well workign website just coz MS decided to re-re-re-reinvent a standard from half a decade or more ago (CSS1) ?
Or should i use a standard that changes incrementally and where each change is relatively cheap for me because it takes past versions into account and/or explains clearly where things will break?
The later is so amazingly much cheaper that only one conclusion can be left. The peopel who decide to keep IE in their companies are throwing away money, and those who keep developing exclusively for IE are wasting their and everyone elses money.
> Yahoo themselves admitted they would use more Java if they weren't locked into thousands of FreeBSD servers.
I have migrated machines between FreeBSD and Linux and the other way around.. given a few right choices, it is not that much work since virtually all the same apps are available, and you have enough control over both to make them as you want.
THere is a price to migration, but between 2 open sourced variations of Unix such costs will be relatively low when you have the knowledge to maintain a large cluster of such machines and change the source and rebuild things as desired already anyway.
So there is a matter of JAVA being important enough to overcome the costs of that and any possible disadvantages it might bring. It seems that its worth the wait for a practical distributable JAVA on FreeBSD for as far as yahoo goes..
I am using FreeBSD with JAVA right now, and with the latest patches it is quite stable and usable. There are other problems but most are not of a technical nature, but rather of the way Sun deals with JAVA and not confined to FreeBSD really. It seems tho that at some point FreeBSD can distribute JAVA binaries once their current patchwork on JAVA is deemed good enough... so there is some hope for Yahoo to get both without having to move in the end.
When you need reliable things on a large scale, you may indeed also have a preference for things that had more time to prove themselves. That however will only work favorable for such inherently older thigns if they do actually prove themselves.
Legacy is King only in certain areas of IT, and guess what, sometimes it even matters. Conservatism with regards to OS choice however has little to do with legacy but with wanting things that have proven themselves, which can at times make you end up with what you'd call legacy stuff. Its a bonus in the case of FreeBSD that it in specific cases also performs better, tho one that many people worked quite hard for.
This was a few weeks ago with 2.6.6 sources from kernel.org and FreeBSD 5.2.1-current from cvsup.
You guys still type that on an office/vm commandline? whatis is way cool tho..
Hmm, that may be true for Google since they generally don't have to keep state. When you have to keep track of sessions and need a substantial amount of sharing between your server instances as a result, a cluster of dual cpu machines or even quad cpu machines may turn out to perform a lot better.
At any rate, if your goal is to handle lots and lots of relatively independent queries, then yeah, large clusters of relatively small machines do a very good job.
> Yahoo is a large website operation. For most corporates the web server is a _very small_ part of the overall IT infrastructure.
> Yahoo is not representative of what you'd find in a typical corporate datacentre / computer room.
They are indeed not representative, but you kindof forget one thing here..
Yahoo depends on its web servers. If they stop running, Yahoo has no business.
That means that their choice and motivation counts for a lot more then what a company says who doesn't depend on the stuff to begin with.
Hmm.. dono, but I think the flamebait mod was more appropriate then the insightfull one ;P
Purely anecdotal ofcourse but..
I used to run FreeBSD 5.2.1 with Apache 2, MySQL 4.0, PHP 4.3.7 (with Turckmmcache) and Geeklog on a dual CPU pII 333 machine with 512mb ram. (the machine is still running MySQL, but geeklog moved to a bigger machine by now)
I also tested this with Debian (current) and gentoo.
With both Linux distributions, this configuration tops at approx 10k requests/hour with 20 parallel connections.
The exact same setup on FreeBSD handles upto 12k requests/hour with 20 parallel connections.
In both cases the machine its rather busy. In neither case it runs out of memory.
In case of FreeBSD I do get a response from the shell within a few sec still, that is not the case on Linux.
As I said, purely anecdotal, but enough reason for me to believe the 'handles high load very well' claim. approx 20% more connections is not a little bit either in my world.
You never visited Yahoo or many sites like it obviously.
Come back when your Dutch is about halfway as good as my English, or learn to read typo.
That is a very nice weight and size pocketpc ;P
Here is why:
I keep my PDA and phone synced to my pc and to eachother using bluetooth, so I take my notes and appointments once, and have them everywhere even if I only carry my phone around.. I do end up taking my PDA with me as well tho most of the time.. its small and light enough to just fit in the pocket of a shirt.
Since my phone doesn't have to have a as big as possible screen or even color, it can me sxtremely small and have a very low energy usage, resulting in being able to carry it around for a logn time without charging.
So well, by not wanting phone functionality from my pda, and only wanting very limited pda functionality from my phone, I end up with 2 small and light devices. I can always carry my appointments and such with me in a small and light phone, and it is little bother to take a pda with me as well since it is also light and small. Since they have a wireless link I can still browse the internet and do mail on my pda, and do so with a nice well readable screen and something more comfortable then a phone keyboard.
It's kindof funny, neither my pda or phone is new, both are over 2 years old actually, but the combination ends up being very usable, and as it is, I often end up reading slashdot during my regular 5 hours long trips to Berlin by train, only depending on the availability of the cellular network, but with a repeater in the train itself that is not a problem.. and knowign that by the tiem I get there the batteries of my PDA will be somewhat drained, but my phoen will have enough power to last another week :)
Who the fuck want to carry around the huge bricks called pocketpc?
You are completely failign to get the point of a PDA.
That said, I doubt PDAs are a very big market anyway, and many who are in the market got carried away by the idea of making a pocket size computer.
A PDA must be a very convinient, EXTREMELY LIGHT, easy to use agenda, address book, and optionally a way to take notes, that also keeps working even if you forget to charge it for a week or 2.
The battery live issue excludes most phones (tho.. my phoen on a larger battery will easily last that long) and most phones lack a decent sized screen.
Weight, battery live, and over complicatedness makes pocketPC an even worse alternative then the PDA features of the average phone.
Don't get me wrogn btw, if you NEED or want a pocket sized computer, then a pocketpc may be for you, but it is simply not the same as a pda, and it doesn't fucntion very well as pda.
> I couldn't use "merd" as a programming lauguage. This name sounds awful in italian.
*lol*
I think the French will agree with you on that.
And for as far as my webserver logs tell me, most slashdotters run windows.. and most run firefox/mozilla.
> However, I don't see blocking mail coming directly from home users' systems as taking away control,
I see it as taking away my control over how my mail gets delivered. Control over my own computer has nothing to do with it.
> unless their email server is hosed, which it sounds like it is. You said it was slow, which is not very specific,
I was not the original poster you replied to, I was just better at reading the parent post and can compare it to my own experiences and that of many other users of a whole variety of ISPs everywhere in the world. It is a relatively common complaint about smtp servers run by ISPs. Currently the smtp server of my ISP is doing quite well, but that hasn't always been the case.
The 'pick another ISP' argument is not a good one here either, it only works if that option is actually available in your area.
> In the end, the ISP has control over every packet leaving your network anyway.
Yes, and I pay them to get a network connection. Within the terms of service, I have control over how I use that connection.
> Running servers violates your user agreement with Comcast anyways, so what difference does it make to them if you can't run your mail server on your cablemodem? If you want to run a server, pay for the *right* service. Don't just try to be a cheap bastard.
It seems that Comcast cares enough to not just close off port 25 alltogether. In their own words, they are aware of people running legitimate mail servers on such accounts. So.. what is your point exactly when Comcast itself seems to disagree with you in word and behavior?
> Even if the mailservers are slow as hell, this should not affect anyone who is not trying to use them for business purposes, for which you should have a business account in order to comply with the AUP. Most AUPs also prohibit running a mailing list on a home account, which pretty much rules out the other purpose of sending a lot of mail from home.
Why should quick delivery not matter to a home user? WHat kind of elitst arogance is that really?
It has nothign to do with how much mail you send but with mail arriving within a reasonable time. I see no reason why I should have to wait 10 hours for a mail to get delivered because of the virus scanning of my ISP getting overloaded for example, and I don't see how you can argue that that should be acceptable either.
Again, has nothing whatsoever to do with mailinglists. (besides the fact that the ISP I use allows me to run a private mailinglist, webserver, mailserver and such and gives a static ip)
> I feel that avoiding your ISP's mailservers because you want to eliminate hops (why? does it matter?) is somewhat contrary, and unnecessarily so. Unless their server is holding mail for inordinately long periods of time, or dropping it, I don't see the issue...
The problems people were mentioning were:
Slow (as in, takes LOOOOOOOOONG to deliver mail)
Unreliable, as in: doesn't deliver mail reliably, which kindof implies it drops it every now and then eh?
So, those are the exact cases you mentioned I would say..
I get the feeling you are not interested in the argument at all and just want to see control taken away from the end-user in the name of fighting spam.
> Exactly. Promoting well-written websites is in their interests as they don't have to work so hard on their broken HTML parser, and they get more semantic information to work with.
Which just means that when their bot gets better information to work with, it will do a better job and the pages will be easier to find. The result you are looking for will happen without the petition and without favoring xhtml in their pageranking.
> What? Nobody would use such a thing, it's infeasible to implement part of it, and it doesn't give the incentive to create better pages (which is the whole point of the petition).
Nobody would use such a thing because they are interested in information relevant to their search, and not in whatever standard it complies with as logn as they can read it.
That is also exactly why it is against the interest of a search engine to favor one standard over another.
While no doubt you are right in your technical argument about it beign easier for google to parse correct xhtml, it is in their interest to present the most relevant information, and that simply has nothing to do with what markup language the results are in as long as they can assume the majority of their users can display it.
Or to put it differently, when a user searches for somethign on google, they are interested in the information, and that is what should determine the pagerank. WHat form the information is in is a subject for filtering (just like image searches)
If you don't understand that simple thing, then the only point of the discussion is the technical differences between html and xhtml, and I am sure you can explain those better then I can and is interesting enough. But if you send or propose to send a petition to google, take a look at where their business is. Seeing the consious efford they put into indexing things like pdf besides html makes me think that they care abotu relevance of information more then anything else.
Besides that, when the web is soemwhat difficult to index and they have developed the technology to still do it relatively well, why would it be in their interest to eliminate that advantage?
Hmm, did you try to do any tuning whatsoever on that machine? Something simple liek loading accf_http and takign a peek at memory usage to make user its not swapping? Also, did you partition your drive such that you have a small seperate root filesystem so you can also use softupdaes on the partition where squid's and Apache's data recides?
on another note:
I have run many such configurations, and under very light load, Linux usually outperforms them, but as soon as I actually start using them, they end up being faster.
> It's not remotely an indication of FreeBSD's quality.
Nah, a counter that wraps where it shouldn't isn't lack fo quality, I agree.
(not that it is a big issue btw)
> You mean exactly what most of them get after signing with a label?
Sortof... but then without the labels telling them what to produce.
Hmm, and how about the fact that it gives the producer of said sofware a bad name? I am of the impression that that is the one and only thing that is at least causing a little bit of movement at Microsoft with regards to creating more secure products.
Another effect of this is that the end-user is at least a little bit aware of the problem, and doesn't fully trust his/her computer...
I think that in the end we are more secure as a result, its just nto a very direct result.
I am aware of the difference between XML and SGML, but you are as well as me aware of the fact that HTML is a severely stripped down subset of SGML, to the point of it being simplistic.
If you are talking about an XML vs an SGML parser, then yeah, XML is a lot simpler. HTML is still simpler then that and to render the html subset that is usable on a small screen you only need a very simple parser, not a full fledged sgml parser.
Btw, I quite see reasons to want xhtml, but those are not the ones mentioned in the google petition, at this moment they simply fail to make up in real world situations.
It is simply not true that a technical specification is going to make thigns better for disabled people for example, a THOUGHTFULL implementation of a specification that keeps them in mind and uses what the spec has to offer for helping them is what can make the web more accessable.
XHTML offers even better ways to support disabled people then previous versions, but seeing the utter lack of usage of already existing features is rather telling me that more and better features (tho nice to have) will not solve the problem.
Getting peopel to actually use such features will help.
If people insist on making their website look cool and don't care about this at the same time, it will keep going wrong, no matter what HTML or XHTML or other markup language you want to use.
The modular setup and extendebility are two very good reasons for wanting xhtml.
If and when people write XHTML and use the specification to its possibilities, then sure, you end up with smaller pages that will work on almost anything. MIn fact, the same was true for html 4.01 + CSS 2.0 when used properly. Many sites manage to use it in a way that validates and yet won't render properly or even in a usable way with anything other then internet explorer
Looking at what is around on the web and validates as XHTML vs what validates as html 4.01 transitional, the simple result is still that esp. on a small device like a palm m505, the later seems to work a lot better, and usually seem to result in smaller transfers. That may be a result of bad design and bad usage of xhtml maybe, but that just brings us back to the point I was making regarding the google petition and disabled people, it is not a matter of what standard you comply to so much as how well you design your stuff.
Back to the discussion, google is a search engine. Go read what they have to say themselves and it seems that their philosophy is to be better then anyone else by doing a better job at returning relevant results. That is their livelyhood for now, and they go to the point of converting things like pdf to html to be able to index those as well.. Do you really believe they are going to consider this untill the majority of both websites and users support it (not just int heory but in reality)
A much better thing to suggest to them and to call a petition for is to add a filter to only get results that are either compatible with your browser as reported by its user-agent, or comply with a specific standard, and to display the (x)html version a page complies with as part of the search result.
What this current petition suggests is simply opposed to the one thing that made them bigger then anyone else, so no chance that they are gonna listen.
The argument regarding visually disabled people sounds better then it is btw. Google will do a better job at indexing pages that are easily accessable with a text only browser as a simpel result of their technology or so they claim (read about optimizing your site for indexing) and that makes a lot of sense also when you consider they are mostly indexing text.
They should, instead of meddling with standards by means of their search results, start generating valid html themselves (and I don't care what version)
Btw, I am visually impaired myself (have a visus of approx 20%). While I can use a display given a large enough one and decent sized fonts, and can use
I hope they do take it serious as well.
Past experience tells they don't, and untill they have provided some very compellign proof of the opposite, I am not going to assume they are able to change themselves into a producer of secure software.
And honestly, Longhorn is in no way convincing.
The first big problem is with their approach to what the end-user should get. It doesn't matter how unsafe it is as logn as it is convinient. Now, isolating all those unsafe components by means of 'managed code' is adding a lot of overhead, partially hiding the problems and make them even less easy to udnerstand for the casual user, while nto solving the problems.
Security is not a product, it is not a technology, and not even a process.
Security starts with the right mindset, it is a matter of mentality first of all. Anythign else follows from that.
If you encourage the wrong mindset from the start, then you actively encourage an insecure environment.
Microsoft has been doing that for decades, and I have no reason to believe they changed their mind, rather the opposite actually, they go for a technical approach to what is first of all a non technical problem.