I see comments EXACTLY like this:
i=i+1;/* Add one to i */
ALL THE TIME.
I see comments SIMILAR to that one frequently.
It doesn't have to be that obvious to be
annoying; sometimes there's a bunch of text which,
when you mentally decode it, boils down to exactly what the
class name/function name/code snippet/whatever says.
The only sane thing to do is to remove them
and move on.
The programmer's attention is a finite resource that is often better spent on something other than memory management, especially given that garbage collection performs quite adequately for many programs.
Yes, but note that you don't have to bother much with memory management in C++
either, these days. I maintain really sucky C++ code for a living, and memory
management is the least of my problems. (People using C++ as C and/or Java are my biggest problems.)
A Perl, Java, or Lisp programmer isn't an idiot for not doing his own memory management any more a person who doesn't make his own shoes is an idiot.
True, although the languages which I like in that category (Perl, Python)
tend to use reference counting rather than Java-ish garbage collection...
Is that why Postfix is miles easier to configure than Sendmail- and is faster, more secure, and just as flexible if not moreso?
It's not that simple.
Sendmail has improved, and configuration-wise,
serving mail will always be tricky to configure.
That said, Postfix serves my mail, and I've not used
sendmail since the 1990s.
accessibility doesn't have ANYTHING to do in a STORAGE FORMAT. this is purely a software issue, with lots of money involved.
Yes it does. Exhibit A: a PDF, or Postscript, or fscked-up HTML document,
which cannot really be represented as anything but pixels on a screen.
Exhibit B: a plain text file, which can be fed directly into your standard speech synthesis software.
Seems to me that if some ideology has the upper hand here,
it's Unix. Not open source or closed source.
To write fast code today you have to be able to write code that can break down into numerous discrete chunks that can all work in parallel.
Or simply write code that uses the right data structures and
the right algorithms. Most of us don't write ray tracers or
encryption algorithms, but things that should not be CPU-bound
on any machine made in this millennium...
Will the use of overloading operators...
*) reduce developement time?
*) reduce the number of bugs?
*) improve maintainability?
In most cases, the answer is at best murky.
Then it follows that avoiding
operator overloading is subject to the same test.
That's not an argument against operator overloading.
I've seen lots of sucky C++ code, and operator overloading
has never been one of the problems -- maybe because
sucky C++ code tends to be written by either C or Java programmers...
I'm pretty certain his point was not optimizing. It was more like "do we actually still need the optimizations (as in far better performance at the time C was invented) of pointer arithmetic at the cost of its complexity and lower maintainability compared to array-indexing"?
I often (not always!) find pointer arithmetic more readable than array indexing.
But the point is: yes, this choice should be made based on readability and correctness, not execution speed.
There's still a place for obscure optimizations, but
this is not (or almost never) it.
I thought we were moving away from fat client technology.
Seriously, what is fatter than a bloody web browser?
I've never understood why I'm expected to prefer web interfaces
to other things. Especially when the web interfaces suck as badly
as they do.
Well, I haven't used Bugzilla, but I've used many others
which my employer doubtlessly paid megabucks for, and which were
clearly made and marketed as practical jokes. By unusability experts.
We went from decentralized, to centralized back to decentralized...... now back?
Centralization is orthogonal to the browser/local application issue.
As a user, if there is a single, central database and a single, non-programmable interface, I'm screwed.
Are you sure that wasn't just early-beta cruft? Microsoft has been making a push for C and C++ standards compliance with VS2k3 and 2k5. And if functions like strlen() are failing, I doubt Microsoft would be able to use the compiler as dogfood.
Of course, if we're talking C++, strlen() is obsoleted
in favor for std::strlen(const char *).
And by the way, marking std::strlen as unsafe by default is moronic.
It is safe; you just have to know a bit about its argument.
Like, that it actually is a string...
At the place I currently work they need to produce an awfull lot of documentation and other paperwork, so they got a couple of big xerox machines. These machines are of such calibre that they need a complete PC to control them. Guess the OS that runs them? No not linux, Solaris an unix that is way way harder to use.\
Unless it is reconfigured so that it doesn't suck, that is. All the good stuff is there in Solaris, but hidden away and not configured properly.
Of course, noone bothers to actually do that, and every day I see people struggle in CDE with the crappy Solaris terminal emulator with hopeless fonts, because there's no xterm icon on the desktop...
What is most interesting with Dicrocoelium dendriticum is that it hijacks the ant and make it climb on grass during full daylight and hold tight with it's mandible so it has a greatest chance of being eaten by a passing sheep - quite un-ant like..
There's also this other parasite, a fungus, which does the same thing to ants. Only in this case, it is to distribute its spores over a larger area to infect more ants.
References?
I saw it on the television.
Attenborough? (sp?)
I still have some documents stored in Final Writer format, Final Writer being an Amiga word processor from the early--mid 1990s. Final Writer is a fitting name, because nothing or noone else will ever write to those documents again. Or read them.
Then there was the records of the regional Ornithologist's Club, basically bird sightings from the early 1990s. My brother requested them for research purposes, but it turned out they were stored in RapidFile (an ancient personal database) and the RapidFile installation itself was long gone.
With a hex editor and some Python scripting, I managed to retrieve most (but probably not all) of the information. If the format hadn't been so simple internally, it would have been gone forever. Even if there is a theoretical chance that someone has the original hardware and software (or a DOS emulator), noone will bother unless the data is really vital.
It doesn't seem to be universally known that DST rules vary not only across the country but around the world. Starting and stopping times vary by country, and as we all ought to know there are places in the US that don't have it.
What does that mean?
This is a non-issue.
Yeah. Sweden changed the DST start/end rules
back in the 1990s (to conform with the EU, I believe). The only thing I noticed was that my ancient Windows95 installation was an hour off some of the time. I'm sure there was an upgrade available from somewhere, but I never bothered -- my copy of RedHat 4 had the updated rules.
There was no discussion and no Y2K-style panic before the switch.
There were no visible effects after it.
XML was designed specifically from the start to be parser-friendly. There are very simple APIs which handle any XML file such as SAX./etc is a very good example of what is wrong with plain text files - no single program or API could ever handle the range of formats of files there!
And why on earth should there be one?
You have to have domain knowledge to parse a file format
and do something meaningful with it. Sure, if the file format is a specialization of XML, it helps you find some broken syntax on the XML/DTD level, but the penalty is this horrible tag soup
which is never parseable with standard tools like grep, cat, awk etc.
I cannot see why XML people find it worth all that trouble.
One language for one purpose is fine with me. This applies to make too -- the language fits the purpose of the tool well, and the TAB issue has never gotten in my way during the maybe twelve years I've used it.
Only the DOS (file system) portion of the OS was written in BCPL. All the core libraries (graphics, windowing, drivers) in the Kickstart ROM were written in assembler.
You are right about dos.library, although it
included device I/O in general (with the file systems as separate modules) and was rewritten in C at some point.
But the rest being written in assembler? I would be very surprised if most of it (except the bootup code, the scheduler and hardware driver stuff) wasn't written in C. At least around AmigaDOS 2.0, it wouldn't make sense to write it all in assembly, even though all of it had to fit in an expensive 512K ROM capsule.
Of course, the entry points to the libraries were not normal C entry points (branch to the address given at a specific negative offset off the library's base address, with parameters in certain address and data registers), and Commodore provided assembly macros for interfacing to the OS.
It was a feasible to do real work in assembly on the Amiga,
but most people used C.
I think he means they both gang up and hit a site at the same time;)
People make jokes like this all the time, yet the geeks I know are about as likely to be browsing with Opera as with Firefox, often alternating between the two.
Personally, I use Opera (registered) since back in the times
of Opera 5.
Maybe Opera is less widely spread in the US? I'm in.se.
Linux has the luxury of time, broad acceptance over a large geek audience, and the benefit of being one of the first successful open source, collaborative endeavors.
I don't think that Linux counts as 'new' in this sense.
It was just another Unix, and the Unix user base was used to
the small differences, and to tweak and recompile their favorite programs.
Anyone trying to jump start the same thing now is in for astronomical challenges.
A new Unix wouldn't have a problem with the user base, at least. But for really new, innovative, incompatible OSes, I agree with you.
How will you know what language a user wants (first time, and on subsequent pages)? The user should be able to select/change their preference (though you could use their browser-reported locale as a guess) [...]
Pet peeve: the locales reported by a browser isn't just a guess;
it's the standard way for a user to tell what languages she prefers to read. I realize that web developers mostly ignore this,
but IMHO not using it (in the absence of other information, of course) is a bug.
Now if I could only figure out how to keep seedlings alive and fungus-gnat-free:)
No organic material in the soil is a solution that works
for me.
But the only plants I grow from seed are cacti; I suspect
most other plants resent the brick-and-gravel treament;-)
My point was science fiction has become a kind of faith that brings about a sense of well being that generates a positive hope for the future - not unlike various religions I guess I shouldn't name.
Maybe "has become" -- I don't read modern SF -- but it wasn't always like that.
It seems to me that there were always parallell tracks in the genre. The techno-optimistic stuff that was intended to educate and to promote science and engineering, and on the other hand the more gloomy stuff. Nuclear holocaust in particular has been a big theme in SF ever since 1946 -- see Theodore Sturgeon's Thunder and Roses, for an early example.
Growing up in the 1960s I thought there was a one-to-one relationship between the love of science fiction and a passion for space exploration. I was obviously wrong.
No, SF readers were supposed to think like that back then.
I think science fiction did influence the early rocket pioneers, and later space scientists, but all of that is in the past. I'm wondering if the Mundane SF people are not asking writers to write stories that will inspire new generations of engineers to work on practical forms of space exploration.
If they do (and I think you're right) then they are trying to revive the "hard SF" idea, where the genre is supposed to be in touch with real science and engineering. I have no problem with that.
But that's not the same as saying that writing that doesn't fit into the manifesto is just "harmless fun". Of all the best SF I've read, most have not met the manifesto's criteria. They have simply used classical SF premises as a tool to present a good story.
And yet, most of them are definitely not just harmless fun.
Try RefBase, and get everyone in your department to use it. Then you'll have a dept-wide database of references, just a few clicks away, and easily inserted into your LaTeX documents via BiBTeX. Its the way.
I think I'd prefer to have a dept-wide BibTeX database instead (under revision control using CVS or whatever).
That's the way people have worked since the seventies (first with refer(1), then increasingly with BibTeX) and I would hate if some web-based thingy broke that toolchain. Hopefully this RefBase thing doesn't.
Slashdot Mirror
i=i+1;
ALL THE TIME.
I see comments SIMILAR to that one frequently. It doesn't have to be that obvious to be annoying; sometimes there's a bunch of text which, when you mentally decode it, boils down to exactly what the class name/function name/code snippet/whatever says.
The only sane thing to do is to remove them and move on.
Yes, but note that you don't have to bother much with memory management in C++ either, these days. I maintain really sucky C++ code for a living, and memory management is the least of my problems. (People using C++ as C and/or Java are my biggest problems.)
A Perl, Java, or Lisp programmer isn't an idiot for not doing his own memory management any more a person who doesn't make his own shoes is an idiot.
True, although the languages which I like in that category (Perl, Python) tend to use reference counting rather than Java-ish garbage collection ...
No; the PC thing on Slashdot these days is to make fun of Stephenson's writing. Don't worry; you're conformant.
It's not that simple. Sendmail has improved, and configuration-wise, serving mail will always be tricky to configure. That said, Postfix serves my mail, and I've not used sendmail since the 1990s.
Also note that there is a -q-1 Vorbis (i.e. you can say oggenc -q-1 ....
That's smaller still, and IMHO good enough even for music.
Yes it does. Exhibit A: a PDF, or Postscript, or fscked-up HTML document, which cannot really be represented as anything but pixels on a screen. Exhibit B: a plain text file, which can be fed directly into your standard speech synthesis software.
Seems to me that if some ideology has the upper hand here, it's Unix. Not open source or closed source.
Or simply write code that uses the right data structures and the right algorithms. Most of us don't write ray tracers or encryption algorithms, but things that should not be CPU-bound on any machine made in this millennium ...
*) reduce developement time?
*) reduce the number of bugs?
*) improve maintainability?
In most cases, the answer is at best murky.
Then it follows that avoiding operator overloading is subject to the same test. That's not an argument against operator overloading.
I've seen lots of sucky C++ code, and operator overloading has never been one of the problems -- maybe because sucky C++ code tends to be written by either C or Java programmers ...
I often (not always!) find pointer arithmetic more readable than array indexing.
But the point is: yes, this choice should be made based on readability and correctness, not execution speed. There's still a place for obscure optimizations, but this is not (or almost never) it.
You are thinking of the A500 or A1000. There were many Amigas, and from -92 or so, 800x600 was a perfectly normal resolution.
Seriously, what is fatter than a bloody web browser? I've never understood why I'm expected to prefer web interfaces to other things. Especially when the web interfaces suck as badly as they do.
Well, I haven't used Bugzilla, but I've used many others which my employer doubtlessly paid megabucks for, and which were clearly made and marketed as practical jokes. By unusability experts.
We went from decentralized, to centralized back to decentralized...... now back?
Centralization is orthogonal to the browser/local application issue. As a user, if there is a single, central database and a single, non-programmable interface, I'm screwed.
Of course, if we're talking C++, strlen() is obsoleted in favor for std::strlen(const char *).
And by the way, marking std::strlen as unsafe by default is moronic. It is safe; you just have to know a bit about its argument. Like, that it actually is a string ...
Unless it is reconfigured so that it doesn't suck, that is. All the good stuff is there in Solaris, but hidden away and not configured properly.
Of course, noone bothers to actually do that, and every day I see people struggle in CDE with the crappy Solaris terminal emulator with hopeless fonts, because there's no xterm icon on the desktop ...
I've never seen that one attributed to anyone but Fred Brooks. (But it's still funny.)
There's also this other parasite, a fungus, which does the same thing to ants. Only in this case, it is to distribute its spores over a larger area to infect more ants.
References? I saw it on the television. Attenborough? (sp?)
Then there was the records of the regional Ornithologist's Club, basically bird sightings from the early 1990s. My brother requested them for research purposes, but it turned out they were stored in RapidFile (an ancient personal database) and the RapidFile installation itself was long gone.
With a hex editor and some Python scripting, I managed to retrieve most (but probably not all) of the information. If the format hadn't been so simple internally, it would have been gone forever. Even if there is a theoretical chance that someone has the original hardware and software (or a DOS emulator), noone will bother unless the data is really vital.
Yeah. Sweden changed the DST start/end rules back in the 1990s (to conform with the EU, I believe). The only thing I noticed was that my ancient Windows95 installation was an hour off some of the time. I'm sure there was an upgrade available from somewhere, but I never bothered -- my copy of RedHat 4 had the updated rules.
There was no discussion and no Y2K-style panic before the switch. There were no visible effects after it.
And why on earth should there be one? You have to have domain knowledge to parse a file format and do something meaningful with it. Sure, if the file format is a specialization of XML, it helps you find some broken syntax on the XML/DTD level, but the penalty is this horrible tag soup which is never parseable with standard tools like grep, cat, awk etc.
I cannot see why XML people find it worth all that trouble. One language for one purpose is fine with me. This applies to make too -- the language fits the purpose of the tool well, and the TAB issue has never gotten in my way during the maybe twelve years I've used it.
You are right about dos.library, although it included device I/O in general (with the file systems as separate modules) and was rewritten in C at some point. But the rest being written in assembler? I would be very surprised if most of it (except the bootup code, the scheduler and hardware driver stuff) wasn't written in C. At least around AmigaDOS 2.0, it wouldn't make sense to write it all in assembly, even though all of it had to fit in an expensive 512K ROM capsule.
Of course, the entry points to the libraries were not normal C entry points (branch to the address given at a specific negative offset off the library's base address, with parameters in certain address and data registers), and Commodore provided assembly macros for interfacing to the OS. It was a feasible to do real work in assembly on the Amiga, but most people used C.
People make jokes like this all the time, yet the geeks I know are about as likely to be browsing with Opera as with Firefox, often alternating between the two. Personally, I use Opera (registered) since back in the times of Opera 5.
Maybe Opera is less widely spread in the US? I'm in .se.
I don't think that Linux counts as 'new' in this sense. It was just another Unix, and the Unix user base was used to the small differences, and to tweak and recompile their favorite programs.
Anyone trying to jump start the same thing now is in for astronomical challenges.
A new Unix wouldn't have a problem with the user base, at least. But for really new, innovative, incompatible OSes, I agree with you.
Pet peeve: the locales reported by a browser isn't just a guess; it's the standard way for a user to tell what languages she prefers to read. I realize that web developers mostly ignore this, but IMHO not using it (in the absence of other information, of course) is a bug.
No organic material in the soil is a solution that works for me. But the only plants I grow from seed are cacti; I suspect most other plants resent the brick-and-gravel treament ;-)
Maybe "has become" -- I don't read modern SF -- but it wasn't always like that. It seems to me that there were always parallell tracks in the genre. The techno-optimistic stuff that was intended to educate and to promote science and engineering, and on the other hand the more gloomy stuff. Nuclear holocaust in particular has been a big theme in SF ever since 1946 -- see Theodore Sturgeon's Thunder and Roses, for an early example.
Growing up in the 1960s I thought there was a one-to-one relationship between the love of science fiction and a passion for space exploration. I was obviously wrong.
No, SF readers were supposed to think like that back then.
I think science fiction did influence the early rocket pioneers, and later space scientists, but all of that is in the past. I'm wondering if the Mundane SF people are not asking writers to write stories that will inspire new generations of engineers to work on practical forms of space exploration.
If they do (and I think you're right) then they are trying to revive the "hard SF" idea, where the genre is supposed to be in touch with real science and engineering. I have no problem with that.
But that's not the same as saying that writing that doesn't fit into the manifesto is just "harmless fun". Of all the best SF I've read, most have not met the manifesto's criteria. They have simply used classical SF premises as a tool to present a good story. And yet, most of them are definitely not just harmless fun.
I think I'd prefer to have a dept-wide BibTeX database instead (under revision control using CVS or whatever). That's the way people have worked since the seventies (first with refer(1), then increasingly with BibTeX) and I would hate if some web-based thingy broke that toolchain. Hopefully this RefBase thing doesn't.