Lotus Notes might be nice for administration, but as a mail system it sucks massively. The UI is atrocious - quite possibly the worst I've ever seen. The speed is a joke, and yes, I know it's a database underneath because it exposes this fact (which the users shouldn't need to know about) every chance it gets. You're right, mail doesn't go missing, in fact it arrives multiple times, days apart.
How anyone has the bare-faced cheek to sell this as a mail system is beyond me, it's so awkward to use that since being forced to use it we're seeing mail volumes drop considerably as people find it easier and quicker to hand write messages and stick them in the internal post or use services like HotMail for sending internal mail !
This reminds me of the articles of why "NT is good for the enterprise developer" 'cos it saves a few cycles here and there, but misses the point that as a developer it's so hard to find out what's going on beneath the veneer that those precious few cycles fade into insignificance for those who don't work for MS, IBM, Sybase or similar (hence a thriving industry in "internals", "secrets", "undocumented areas" etc.).
I don't want a "powerful database merging programmable blah blah blah", I just want to send mail to people (please). Anyone got an Emacs RMAIL to Lotus Notes binding lying around ?
The original poster goes on to state that "he's just repeating what he's heard" - he makes claims about code quality when he states twice he can't code, he makes claims for compatibility or lack thereof and then admits he doesn't know what he's talking about, and he launches personal attacks on anyone criticising his rambling logic.
Someone moderate it down to -1, please. I don't doubt it's well intentioned, but MS mis-information is bad enough, I'd hope we can not lower ourselves to that level round here.
Re:OT: Memory leaks? GC no help for the real probl
on
Mozilla M5 Released
·
· Score: 1
>> You still haven't adressed the issue
I don't want to sound rude, but I think I have.
I address the real, root-cause bugs: inconsistent state.
Adding an external hack to hide the more benign instances of this bug gains me nothing, I'm curing a trivial symptom and leaving the cause way open. Hiding bugs doesn't count in my world... fixing them does.
I use a heap walker to show me a list of memory leaks, but I don't then just add a loop to free them all - I want to know what logic faults left them there.
If you want to continue the chat (it is rather OT) then feel free to mail me
OT: Memory leaks? GC no help for the real problem
on
Mozilla M5 Released
·
· Score: 3
>> I really can't understand why anyone would like to punish themselves by not using a GC.
Because memory leaks are only one instance (and are the trivial instance) of a generalised problem, and GC does nothing to solve the more general problem.
The real problem is (I'm going to say objects, but this applies equally to non-OO code, just terminology) when two objects disagree over state. A memory leak is just a single case (one thinks it's free, the other doesn't), but the real serious problems occur when objects aren't initialised, they're initialised twice, they're not saved, they have poorly designed "flux states", they don't notify changes properly, they blindly propogate changes throughout the object-space, etc.
GC does absolutely nothing for this, and, more importantly, no GC "Band-Aid" techniques help with these problems, so I try to write code (and debug) to fix the general problem: the memory leaks are the trivial case - easiest to detect and simplest to fix.
Find me any production program where memory leaks are "the biggest code problem", and I'll show you a project which doesn't even realise where they've really got problems.
And I thought the only British public acknowledgement of AT was a small blue plaque on a building I pass every day (near Maida Vale in London) recording the fact he was born there. That in itself was a real shock - as a good Manchester lad myself I'm astonished to find that Alan Turing was in fact a "Southern Jessie" !
I ran my HttpSniffer script over a couple of slashdot sessions to see if you were chunk-encoding stuff (it's ok, you're not AFAIK) in your replies, but also you're not using content-coding (where you can return gzip'ed or deflate'd content that will be unpacked by the client).
I've never used content-coding myself, can anyone comment on how successful it is ?
Most of the bandwidth you're using is relatively plain text, I would have thought gzip on a low setting could reduce the number of bytes to be sent by a fair bit IF you've got some CPU power to spare AND the chunk of text to be returned is a decent size (say > 10K). The browser request indicates which encodings are allowed (Accept-Encoding header field).
I haven't looked at the slashdot perl source recently, but I would think that it would be a pretty small mod to use Compress::Zlib to do an in-memory deflate of large text bodes for users with browsers that will accept it.
>>... and I suggest you read John Searle's "Chinese Room Arguement" that debunked just about
Spare me, not Searle and his "foolish schoolboy error".
As for Penrose, he might be a good mathematician (and friends at Oxford tell me he is) but his "E-N-M" book is the biggest load of twaddle I've seen in a long time. This is the only book I've _ever_ taken back to a book shop and asked for my money back, 500 pages of "my version of a popular intro to Quantum Mechanics" followed by some garbage about "a brand new physics" based on the basic Searle error on page 27 or so.... file it along with Edward De Bono's simplistic and patronising works.
This is sitting on my shelf, but I haven't got round to it so far .
I've read and enjoy Hofstadter, anyone care to compare and contrast this to Hofstatdter and Dennet's "The Minds I" or Dennet's "Conciousness Explained" ??
Wait until they announce the Plus pack (to be released same day) which contains some "groovy extra features" (aka better User Interface to key functionality, such as scripting for dial-up networking).
I always thought these were the real rip-off, but no-one ever remembers to factor in these costs.
MS started charging for IE in the same way about a year ago, you got "IE Basic" for free, but they released an "IE Plus Pack" (I forget the exact name, Gold Edition or something) for about $80 that added not very much, but gave you a nice box.
>>I know that the artifacts are small and _most_ of the time they aren't noticeable
Decent encoding/mastering gets rid of this, expect things to get better. My wife runs a DVD encoding operation, and they spend a lot of time tuning the encodings for specific problems - assisted by the encoding software getting much better too.
All products are compromises, non-lossy compression at a similar quality would have to spin the disc so fast (and such a big disc too) to make the product useless.
Slashdot Mirror
Lotus Notes might be nice for administration, but as a mail system it sucks massively. The UI is atrocious - quite possibly the worst I've ever seen. The speed is a joke, and yes, I know it's a database underneath because it exposes this fact (which the users shouldn't need to know about) every chance it gets. You're right, mail doesn't go missing, in fact it arrives multiple times, days apart.
How anyone has the bare-faced cheek to sell this as a mail system is beyond me, it's so awkward to use that since being forced to use it we're seeing mail volumes drop considerably as people find it easier and quicker to hand write messages and stick them in the internal post or use services like HotMail for sending internal mail !
This reminds me of the articles of why "NT is good for the enterprise developer" 'cos it saves a few cycles here and there, but misses the point that as a developer it's so hard to find out what's going on beneath the veneer that those precious few cycles fade into insignificance for those who don't work for MS, IBM, Sybase or similar (hence a thriving industry in "internals", "secrets", "undocumented areas" etc.).
I don't want a "powerful database merging programmable blah blah blah", I just want to send mail to people (please). Anyone got an Emacs RMAIL to Lotus Notes binding lying around ?
Trying not to troll but...
The original poster goes on to state that "he's just repeating what he's heard" - he makes claims about code quality when he states twice he can't code, he makes claims for compatibility or lack thereof and then admits he doesn't know what he's talking about, and he launches personal attacks on anyone criticising his rambling logic.
Someone moderate it down to -1, please. I don't doubt it's well intentioned, but MS mis-information is bad enough, I'd hope we can not lower ourselves to that level round here.
>> You still haven't adressed the issue
I don't want to sound rude, but I think I have.
I address the real, root-cause bugs: inconsistent state.
Adding an external hack to hide the more benign instances of this bug gains me nothing, I'm curing a trivial symptom and leaving the cause way open. Hiding bugs doesn't count in my world... fixing them does.
I use a heap walker to show me a list of memory leaks, but I don't then just add a loop to free them all - I want to know what logic faults left them there.
If you want to continue the chat (it is rather OT) then feel free to mail me
>> I really can't understand why anyone would like to punish themselves by not using a GC.
Because memory leaks are only one instance (and are the trivial instance) of a generalised problem, and GC does nothing to solve the more general problem.
The real problem is (I'm going to say objects, but this applies equally to non-OO code, just terminology) when two objects disagree over state. A memory leak is just a single case (one thinks it's free, the other doesn't), but the real serious problems occur when objects aren't initialised, they're initialised twice, they're not saved, they have poorly designed "flux states", they don't notify changes properly, they blindly propogate changes throughout the object-space, etc.
GC does absolutely nothing for this, and, more importantly, no GC "Band-Aid" techniques help with these problems, so I try to write code (and debug) to fix the general problem: the memory leaks are the trivial case - easiest to detect and simplest to fix.
Find me any production program where memory leaks are "the biggest code problem", and I'll show you a project which doesn't even realise where they've really got problems.
Too late, too stupid...
And I thought the only British public acknowledgement of AT was a small blue plaque on a building I pass every day (near Maida Vale in London) recording the fact he was born there.
That in itself was a real shock - as a good Manchester lad myself I'm astonished to find that Alan Turing was in fact a "Southern Jessie" !
I ran my HttpSniffer script over a couple of slashdot sessions to see if you were chunk-encoding stuff (it's ok, you're not AFAIK) in your replies, but also you're not using content-coding (where you can return gzip'ed or deflate'd content that will be unpacked by the client).
I've never used content-coding myself, can anyone comment on how successful it is ?
Most of the bandwidth you're using is relatively plain text, I would have thought gzip on a low setting could reduce the number of bytes to be sent by a fair bit IF you've got some CPU power to spare AND the chunk of text to be returned is a decent size (say > 10K). The browser request indicates which encodings are allowed (Accept-Encoding header field).
I haven't looked at the slashdot perl source recently, but I would think that it would be a pretty small mod to use Compress::Zlib to do an in-memory deflate of large text bodes for users with browsers that will accept it.
>>... and I suggest you read John Searle's "Chinese Room Arguement" that debunked just about
.... file it along with Edward De Bono's simplistic and patronising works.
Spare me, not Searle and his "foolish schoolboy error".
As for Penrose, he might be a good mathematician (and friends at Oxford tell me he is) but his "E-N-M" book is the biggest load of twaddle I've seen in a long time.
This is the only book I've _ever_ taken back to a book shop and asked for my money back, 500 pages of "my version of a popular intro to Quantum Mechanics" followed by some garbage about "a brand new physics" based on the basic Searle error on page 27 or so
Tim
This is sitting on my shelf, but I haven't got round to it so far .
I've read and enjoy Hofstadter, anyone care to compare and contrast this to Hofstatdter and Dennet's "The Minds I" or Dennet's "Conciousness Explained" ??
Does it say much new compared to those two ??
Wait until they announce the Plus pack (to be released same day) which contains some "groovy extra features" (aka better User Interface to key functionality, such as scripting for dial-up networking).
I always thought these were the real rip-off, but no-one ever remembers to factor in these costs.
MS started charging for IE in the same way about a year ago, you got "IE Basic" for free, but they released an "IE Plus Pack" (I forget the exact name, Gold Edition or something) for about $80 that added not very much, but gave you a nice box.
>>I know that the artifacts are small and _most_ of the time they aren't noticeable
Decent encoding/mastering gets rid of this, expect things to get better. My wife runs a DVD encoding operation, and they spend a lot of time tuning the encodings for specific problems - assisted by the encoding software getting much better too.
All products are compromises, non-lossy compression at a similar quality would have to spin the disc so fast (and such a big disc too) to make the product useless.