the writer meant that Mac/Windows network integration services is a small corner of the computing world. That's the only way I can read it that doesn't make the writer sound extremely isolated.
Got an app that causes a fatal library error in PHP because of a syntax error in an include? It'll die. Silently. At runtime, i.e. when your customer is using it.
This is a design issue, not a PHP issue. With a proper controller system (not part of PHP, but can be designed in PHP with things like LogiCreate) this is simply not true. Use a user-defined function to do the includes and have it trap the error, or dig deeper into error-handling in PHP.
This is the result of lazy or unclued-in thinking on the part of a PHP programmer, not PHP itself. While I wish PHP provided better ways of handling this directly, it's certainly possible to put together good design in PHP to handle things like this.
Re:PHP books *are* needed, just not all of them
on
Core PHP Programming
·
· Score: 2, Interesting
You're correct on the breaking stuff. What our problem has been is that generally the stuff that breaks doesn't even make it into the online manual. I'm still peeved at the way most changes are handled in PHP - they seem to have no concept of the fact that *millions* of people are affected by their actions, and carry on as if they were just hacking in their basement.
Re:PHP books *are* needed, just not all of them
on
Core PHP Programming
·
· Score: 4, Informative
The developer cookbook, because of the 'recipe' approach (identify problem, discuss it, show solution).
There's an older NewRiders book by, argh, lost the name - blue/purplish spine, something like "Web Application Development with PHP". One of the first books I saw on PHP where they went into advanced concepts, and didn't treat the reader as if they knew no programming. You were walked through various business problems and shown how they were addressed, and IIRC chapters built on earlier chapters for continuity.
Some WROX are OK, but it's hit and miss.
Haven't read the new Schlossnagle book yet, but it looks good too (likely more internal, gutsy type stuff not just 'here's a variable' kind of thing).
HTH
PHP books *are* needed, just not all of them
on
Core PHP Programming
·
· Score: 4, Insightful
There's a lot of people who say 'you don't need a book, the online docs are great!'. I disagree.
*Some* books are good (although I think there are too many which repeat the same information, not enough focus on particular topics in the PHP world) and necessary because they can go into greater detail than you get from the online docs.
"What about online tutorials?" Some are good, but having it all in one book, written by only one or two authors (as opposed to wrox-style 15 authors) can help keep a consistent presentation of concepts from beginning to end.
I'm not saying online sucks and all books are great - many PHP books aren't all much more useful than the online docs really. But for those that try to actually teach, rather than reprocess, I think they can be more valuable over time than *just* the online docs.
Personally, I think this 3rd edition is good, although there is, imo, too much reprocessing of the manual. You could cut 200-300 pages out of this book and not miss much of anything. What would be left is worthwhile, though. What's missing in all the reference material is details on what, if any, differences there are between PHP4 and PHP5. If it's there it's in text form, not a standard icon set to alert you of potential differences.
BTW, I have roughly the same arguments for PHP training courses, which we teach (subtle plug). "It's all online!" isn't the best answer for everyone. Many people struggle for hours or days with some concepts with only tutorials and reference pages. Put them in a classroom where they can get immediate feedback on new concepts, and they get it much quicker. Each person learns and adapts to new information in different ways, and classroom training is appropriate for some people, whether it's "only" PHP or something else.
I really have to ask - why NS 4.7 as a 'standard'. Netscape themselves don't support (on to version 7 already). It's essentially 1997 technology. I bet you don't support or push IE 3, do you? I truly don't understand people still supporting or demanding people using NS 4.7. Could you explain please?
It took 10 years to get through the majority of status quo defenders shouting
"Why would you ever do that?"
"No one needs to do that"
"X is network capable - that's all you need"
"You CAN resize the desktop with CTRL-ALT-+"
That last one was, without a doubt, the most annoying response ever. It was as if everyone repeating that line (and there were thousands) had never even *seen* Windows, which has allowed dynamic desktop resizing for years.
So, figure on a 5-10 year wait on many other things considered basic on other systems. Not to say Linux has no good points (it has many) but the majority of developers at different levels have *much* different ideas about 'normal' usage than do, well, 'normal' people.
What's missing in the 'lower TCO' factor, assuming it's true, is what effect this has on a local economy.
Yes, many businesses will feel good about 33% lower labor costs. That's over a 5 year cycle tho. So, you've generally got higher first year costs with most of that money leaving the local economy (unless you live in Redmond). Then, year after year, you can pay people in your area (employees) less money. Paying somewhat more to employees over 5 years ensures they've got money to spend - primarily locally (usually within the state at least) as well as pay more taxes (not just income taxes, but taxes on the local services).
Effectively, MS is arguing to simply extract money from local economies and pay people less. Short term, that may be fine. Long term, it only hurts. Schools/governments/etc should be *vary* cautious about this, if not downright hostile.
If even some of their time was spent preventing others from contributing working code which addressed problems or added new features (code developed by the time of others) for personal/political/spurious reasons, they deserve a certain amount of chastising.
Not saying I know the ins and outs of the project, but the above still stands in any project. When leaders become an hindrance to problem solving rather than a facilitator, that's a bad situation, especially for projects like this. It's probably good they disbanded , if this was starting to be the case.
Re:Let the conspiracy theories begin...
on
Make More Mistakes
·
· Score: 1
In general I agree with your point, but do realize he was working on a word processor, something which was easily done and redone 100 times before, with numerous market giants past and present holding 100% of the market. MS Word, WordPerfect on the commercial side and dare I say OpenOffice in the open source side (StarOffice at the time, under a commercial banner).
Was there any compelling market need for another word processor, open source or not? I don't think there was. Even if AbiWord had been a commercial product (wait, wasn't ApplixWare around that time?) it would have failed because word processors are too much of a commodity these days (and in those days too). Heck, AbiWord didn't even have table support for years - it wasn't cutting edge in any respect. It *only* had the 'open source' aspect going for it (if that).
Again, I more or less agree that open source as a business model is *tough*, but I wouldn't say it's completely impossible. It's just not something people are going to sink hundreds of millions into and get back billions.
As a result you had labs full of Macs that were like 5 years old but still perfectly usable, versus the lab down the hall with the latest PCs that had replaced the old PCs which were unusable after 2 years (couldn't run Windows 95, for example).
There's your flawed logic right there. To most people a 5 year old Mac *isn't* 'perfectly usable' because of the speed increases they've experienced on other machines in 5 years. You still *could* run Windows 95 on a P100 with 32 megs of RAM and it'd be fine for Office 95, Netscape 3 and other technology from its time. However, time marches on and people expect more from computers because, well, current computers *do* more. So it's not a case of 'ignorant masses' drawing false conclusions, it's that they want to get work done.
I worked at a place where P200s and up were common (1998) and into 1999 people were getting 400 and 500mhz machines. The Mac people generally didn't *want* to upgrade (read: dirsupt) their systems, so many of them were still using - I forget the models - 75mhz macs. SLOW as molasses to actually bring files up or do anything, but they'd sit there in front of their Photoshop sessions chugging away for hours telling themselves how much faster their machines were than the Windows guy running a 400mhz machine because, well, damnit, clockspeed doesn't really mean anything. The difference between a file being opened and ready for use in 10 seconds and 50 seconds, multiplied by dozens of people, over months of work, REALLY adds up.
Perhaps there were more than one or two conversations they learned about through research but didn't report about.
This is slanted *against* Apple as much as most Apple zealots are slanted *for* Apple, and it will all balance out in the end. Too many Applefans are prepared to push their favorite company to everyone, facts/figures be damned, and when something like this comes out, somehow the world is 'against' Apple. It's ridiculous.
I actually don't remember now if debian offered a choice or not - it was a joint install with someone else and I don't recall exactly how that went, except that I didn't keep it on longer than a week.:)
I've been reading this for months, people saying that 'distro X has KDE as the default desktop' or 'distro Y uses Gnome by default'.
EVERY distro I've installed over the last 3 years *asks* me which desktop managers I want to install. Although this decision is generally put on par with choosing whether you want to install 'games' or 'server software' or 'scientific' software, it's still a decision you're expected to make. I don't think any distro I've ever installed just puts a desktop on by default with no choice (save for Knoppix).
What have I missed in these wars where certain distros make the choice for you? I've installed mandrake, redhat, suse, plain debian, knoppix and and caldera over the years.
Someone else will respond probably too, and this point has been brought up numerous times. They should NOT teach Microsoft Office specifically (unless it's an elective class called 'MS Office' or something). What should be taught is the concepts. Teaching people not to be afraid of the computer, and to learn basic ideas like spell checking, formatting (bold, italics, etc). The basics are the same on every major platform, have been for years, and will continue to be. If there's a need to teach "Excel Macros", that's fine, but label the class as such.
What will you do when someone learned MS Office 2000 2 years ago, left school, and gets a job 2 years from now using Office 2005XP or something like that? If they've been trained to 'select the 4th option in the 3rd menu' they're screwed anyway. *I* wouldn't hire anyone like that. I would hire someone who comes across as competent *and* confident with a computer, regardless of which version of an office suite I throw at them.
Money leaving the school districts for Microsoft products when the same budgets have school lunch programs cut, textbooks not being purchased, and teachers being laid off is simply immoral.
Obviously you've not been around here long enough.:)
Of course - you HAVE to contribute in order to earn the 'right' to criticize. At that point it's not criticism tho, it's just 'innovation'.
Seriously, I'm in total agreement with you. rpm -qpl doesn't make sense based on the 'rpm --help' output, because "l" says it'll list files in a package. Why do I need to further use "p" to say that I mean a "package file"? It makes NO sense. What would help it make sense is better docs.
The best you could do would be to make a help page someplace and hope it gets googled. These issues have come up for years and nothing extra productive has ever been added to the rpm man page or --help output. rpm --showfilesinrpm would be handy, but it ain't gonna happen. Who would add in something which would make it useful to people who haven't committed their lives to arcane command-line crap?:)
Depends, but I'd say if the wall is going to have any large scale visibility to the community and be a symbol of how great MIT is, they'd probably do well to involve some of their grad students in the process, if not the actual work.
You also forgot the rule to have javascript scroll text in the bottom bar of the browser window to make sure people can't point at a link to see where it will take them.
Also more on Intellivision. Out in 1980, so Atari had 2+ years as the king, then about 2 years of sharing with Intellivision, then Coleco hit. Atari VCS was out in 1977. Coleco entered the market in 1982.
BTW, I wish people would call it the Atari VCS, which is what it originally was. It wasn't referred to as a '2600' untill there was a '5200' to upgrade to.
Not on our install here, which was a stock install. Also, after the *longest* linux install in the past 12 months, we still didn't get something as basic as 'locate'. Hrm...
Where are all these mythical companies writing commercial Gnome/GTK apps? I don't see them. I see a few KDE (kompany) but no commercial GTK apps. So the license issue doesn't seem to be driving people away from QT/KDE.
I dunno where this comes from. Last I checked, the default SuSE distro took out things from the KDE control panel like the 'information' tree (devices/dma channels/etc). Why that was completely removed is just a mystery to me. SuSE feels slightly slower than Mandrake does (tested in our labs on identical hardware).
If you could post those here, it'd be really good reading. Your paraphrase was probably close, but it doesn't make sense (of course, the original emails didn't either!):)
Slashdot Mirror
the writer meant that Mac/Windows network integration services is a small corner of the computing world. That's the only way I can read it that doesn't make the writer sound extremely isolated.
Got an app that causes a fatal library error in PHP because of a syntax error in an include? It'll die. Silently. At runtime, i.e. when your customer is using it.
This is a design issue, not a PHP issue. With a proper controller system (not part of PHP, but can be designed in PHP with things like LogiCreate) this is simply not true. Use a user-defined function to do the includes and have it trap the error, or dig deeper into error-handling in PHP.
This is the result of lazy or unclued-in thinking on the part of a PHP programmer, not PHP itself. While I wish PHP provided better ways of handling this directly, it's certainly possible to put together good design in PHP to handle things like this.
You're correct on the breaking stuff. What our problem has been is that generally the stuff that breaks doesn't even make it into the online manual. I'm still peeved at the way most changes are handled in PHP - they seem to have no concept of the fact that *millions* of people are affected by their actions, and carry on as if they were just hacking in their basement.
The developer cookbook, because of the 'recipe' approach (identify problem, discuss it, show solution).
There's an older NewRiders book by, argh, lost the name - blue/purplish spine, something like "Web Application Development with PHP". One of the first books I saw on PHP where they went into advanced concepts, and didn't treat the reader as if they knew no programming. You were walked through various business problems and shown how they were addressed, and IIRC chapters built on earlier chapters for continuity.
Some WROX are OK, but it's hit and miss.
Haven't read the new Schlossnagle book yet, but it looks good too (likely more internal, gutsy type stuff not just 'here's a variable' kind of thing).
HTH
There's a lot of people who say 'you don't need a book, the online docs are great!'. I disagree.
*Some* books are good (although I think there are too many which repeat the same information, not enough focus on particular topics in the PHP world) and necessary because they can go into greater detail than you get from the online docs.
"What about online tutorials?" Some are good, but having it all in one book, written by only one or two authors (as opposed to wrox-style 15 authors) can help keep a consistent presentation of concepts from beginning to end.
I'm not saying online sucks and all books are great - many PHP books aren't all much more useful than the online docs really. But for those that try to actually teach, rather than reprocess, I think they can be more valuable over time than *just* the online docs.
Personally, I think this 3rd edition is good, although there is, imo, too much reprocessing of the manual. You could cut 200-300 pages out of this book and not miss much of anything. What would be left is worthwhile, though. What's missing in all the reference material is details on what, if any, differences there are between PHP4 and PHP5. If it's there it's in text form, not a standard icon set to alert you of potential differences.
BTW, I have roughly the same arguments for PHP training courses, which we teach (subtle plug). "It's all online!" isn't the best answer for everyone. Many people struggle for hours or days with some concepts with only tutorials and reference pages. Put them in a classroom where they can get immediate feedback on new concepts, and they get it much quicker. Each person learns and adapts to new information in different ways, and classroom training is appropriate for some people, whether it's "only" PHP or something else.
I really have to ask - why NS 4.7 as a 'standard'. Netscape themselves don't support (on to version 7 already). It's essentially 1997 technology. I bet you don't support or push IE 3, do you? I truly don't understand people still supporting or demanding people using NS 4.7. Could you explain please?
It took 10 years to get through the majority of status quo defenders shouting
"Why would you ever do that?"
"No one needs to do that"
"X is network capable - that's all you need"
"You CAN resize the desktop with CTRL-ALT-+"
That last one was, without a doubt, the most annoying response ever. It was as if everyone repeating that line (and there were thousands) had never even *seen* Windows, which has allowed dynamic desktop resizing for years.
So, figure on a 5-10 year wait on many other things considered basic on other systems. Not to say Linux has no good points (it has many) but the majority of developers at different levels have *much* different ideas about 'normal' usage than do, well, 'normal' people.
What's missing in the 'lower TCO' factor, assuming it's true, is what effect this has on a local economy.
Yes, many businesses will feel good about 33% lower labor costs. That's over a 5 year cycle tho. So, you've generally got higher first year costs with most of that money leaving the local economy (unless you live in Redmond). Then, year after year, you can pay people in your area (employees) less money. Paying somewhat more to employees over 5 years ensures they've got money to spend - primarily locally (usually within the state at least) as well as pay more taxes (not just income taxes, but taxes on the local services).
Effectively, MS is arguing to simply extract money from local economies and pay people less. Short term, that may be fine. Long term, it only hurts. Schools/governments/etc should be *vary* cautious about this, if not downright hostile.
If even some of their time was spent preventing others from contributing working code which addressed problems or added new features (code developed by the time of others) for personal/political/spurious reasons, they deserve a certain amount of chastising.
Not saying I know the ins and outs of the project, but the above still stands in any project. When leaders become an hindrance to problem solving rather than a facilitator, that's a bad situation, especially for projects like this. It's probably good they disbanded , if this was starting to be the case.
In general I agree with your point, but do realize he was working on a word processor, something which was easily done and redone 100 times before, with numerous market giants past and present holding 100% of the market. MS Word, WordPerfect on the commercial side and dare I say OpenOffice in the open source side (StarOffice at the time, under a commercial banner).
Was there any compelling market need for another word processor, open source or not? I don't think there was. Even if AbiWord had been a commercial product (wait, wasn't ApplixWare around that time?) it would have failed because word processors are too much of a commodity these days (and in those days too). Heck, AbiWord didn't even have table support for years - it wasn't cutting edge in any respect. It *only* had the 'open source' aspect going for it (if that).
Again, I more or less agree that open source as a business model is *tough*, but I wouldn't say it's completely impossible. It's just not something people are going to sink hundreds of millions into and get back billions.
As a result you had labs full of Macs that were like 5 years old but still perfectly usable, versus the lab down the hall with the latest PCs that had replaced the old PCs which were unusable after 2 years (couldn't run Windows 95, for example).
There's your flawed logic right there. To most people a 5 year old Mac *isn't* 'perfectly usable' because of the speed increases they've experienced on other machines in 5 years. You still *could* run Windows 95 on a P100 with 32 megs of RAM and it'd be fine for Office 95, Netscape 3 and other technology from its time. However, time marches on and people expect more from computers because, well, current computers *do* more. So it's not a case of 'ignorant masses' drawing false conclusions, it's that they want to get work done.
I worked at a place where P200s and up were common (1998) and into 1999 people were getting 400 and 500mhz machines. The Mac people generally didn't *want* to upgrade (read: dirsupt) their systems, so many of them were still using - I forget the models - 75mhz macs. SLOW as molasses to actually bring files up or do anything, but they'd sit there in front of their Photoshop sessions chugging away for hours telling themselves how much faster their machines were than the Windows guy running a 400mhz machine because, well, damnit, clockspeed doesn't really mean anything. The difference between a file being opened and ready for use in 10 seconds and 50 seconds, multiplied by dozens of people, over months of work, REALLY adds up.
Perhaps there were more than one or two conversations they learned about through research but didn't report about.
This is slanted *against* Apple as much as most Apple zealots are slanted *for* Apple, and it will all balance out in the end. Too many Applefans are prepared to push their favorite company to everyone, facts/figures be damned, and when something like this comes out, somehow the world is 'against' Apple. It's ridiculous.
And yes, I own a Mac.
I actually don't remember now if debian offered a choice or not - it was a joint install with someone else and I don't recall exactly how that went, except that I didn't keep it on longer than a week. :)
I've been reading this for months, people saying that 'distro X has KDE as the default desktop' or 'distro Y uses Gnome by default'.
EVERY distro I've installed over the last 3 years *asks* me which desktop managers I want to install. Although this decision is generally put on par with choosing whether you want to install 'games' or 'server software' or 'scientific' software, it's still a decision you're expected to make. I don't think any distro I've ever installed just puts a desktop on by default with no choice (save for Knoppix).
What have I missed in these wars where certain distros make the choice for you? I've installed mandrake, redhat, suse, plain debian, knoppix and and caldera over the years.
Someone else will respond probably too, and this point has been brought up numerous times. They should NOT teach Microsoft Office specifically (unless it's an elective class called 'MS Office' or something). What should be taught is the concepts. Teaching people not to be afraid of the computer, and to learn basic ideas like spell checking, formatting (bold, italics, etc). The basics are the same on every major platform, have been for years, and will continue to be. If there's a need to teach "Excel Macros", that's fine, but label the class as such.
What will you do when someone learned MS Office 2000 2 years ago, left school, and gets a job 2 years from now using Office 2005XP or something like that? If they've been trained to 'select the 4th option in the 3rd menu' they're screwed anyway. *I* wouldn't hire anyone like that. I would hire someone who comes across as competent *and* confident with a computer, regardless of which version of an office suite I throw at them.
Money leaving the school districts for Microsoft products when the same budgets have school lunch programs cut, textbooks not being purchased, and teachers being laid off is simply immoral.
Obviously you've not been around here long enough. :)
:)
Of course - you HAVE to contribute in order to earn the 'right' to criticize. At that point it's not criticism tho, it's just 'innovation'.
Seriously, I'm in total agreement with you. rpm -qpl doesn't make sense based on the 'rpm --help' output, because "l" says it'll list files in a package. Why do I need to further use "p" to say that I mean a "package file"? It makes NO sense. What would help it make sense is better docs.
The best you could do would be to make a help page someplace and hope it gets googled. These issues have come up for years and nothing extra productive has ever been added to the rpm man page or --help output. rpm --showfilesinrpm would be handy, but it ain't gonna happen. Who would add in something which would make it useful to people who haven't committed their lives to arcane command-line crap?
Depends, but I'd say if the wall is going to have any large scale visibility to the community and be a symbol of how great MIT is, they'd probably do well to involve some of their grad students in the process, if not the actual work.
You also forgot the rule to have javascript scroll text in the bottom bar of the browser window to make sure people can't point at a link to see where it will take them.
Atari and Intellivision had been going for a few years before Coleco got in the game. I believe Adventure for Atari VCS was out in 1978.
http://www.atariage.com/2600/history.html
has a bit more.
Also more
on Intellivision. Out in 1980, so Atari had 2+ years as the king, then about 2 years of sharing with Intellivision, then Coleco hit.
Atari VCS was out in 1977. Coleco entered the market in 1982.
BTW, I wish people would call it the Atari VCS, which is what it originally was. It wasn't referred to as a '2600' untill there was a '5200' to upgrade to.
Not on our install here, which was a stock install. Also, after the *longest* linux install in the past 12 months, we still didn't get something as basic as 'locate'. Hrm...
Where are all these mythical companies writing commercial Gnome/GTK apps? I don't see them. I see a few KDE (kompany) but no commercial GTK apps. So the license issue doesn't seem to be driving people away from QT/KDE.
I dunno where this comes from. Last I checked, the default SuSE distro took out things from the KDE control panel like the 'information' tree (devices/dma channels/etc). Why that was completely removed is just a mystery to me. SuSE feels slightly slower than Mandrake does (tested in our labs on identical hardware).
If you could post those here, it'd be really good reading. Your paraphrase was probably close, but it doesn't make sense (of course, the original emails didn't either!) :)
Hey!
:)
I taught a 3 day PHP course up in Moncton earlier this year - were either of you there by any chance? It'd be quite a small world if so.
Excellent insight!