|
This should now be fixed.
cheers,
Chris Maunder
CodeProject.com : C++ MVP
|
|
|
|
|
If you google for CMPXHG8 (no, I don't do that every day) one of the first links points to a CP forum message, but not to the right one[^].
Anything CP can do to fix this?
|
|
|
|
|
It's difficult because Google indexes PermaLinks, and each permalink contains text to messages not actually referred to by the permalink.
We would make it so that when Google requests a Permalink all they see is that specific question, but I'm not 100% sure this would help
cheers,
Chris Maunder
CodeProject.com : C++ MVP
|
|
|
|
|
Hi Chris,
I'm not sure I understand all this, I am still struggling with the permalink concept.
But your speed today is amazing.
|
|
|
|
|
Luc Pattyn wrote: If you google for CMPXHG8 (no, I don't do that every day)
Me neither. But I do search for CMPXCHG8 occasionally, just for fun
And in fact Google return what you asked for, because you failed to initialize ECX:EBX pair in your search query, so Google returned random garbage that was there
|
|
|
|
|
Mladen Jankovic wrote: you failed to initialize ECX:EBX pair in your search query
O, my mistake. That is what you get when using strict language compilers most of the time.
Thanks for pointing this out.
|
|
|
|
|
Luc Pattyn wrote: CMPXHG8 (no, I don't do that every day)
Me neither. I've seen the instruction years ago, never have actually found a real use for it.
"The clue train passed his station without stopping." - John Simmons / outlaw programmer
"Real programmers just throw a bunch of 1s and 0s at the computer to see what sticks" - Pete O'Hanlon
|
|
|
|
|
//Me neither. I've seen the instruction years ago, never have actually found a real use for it.//
Not sure if I should continue this thread in the wrong forum, but anyway...
The double-length compare-exchange would be very handy if vb.net would allow it to be used on a structure containing an object reference and a counter. Among other things, it would make it much more practical to use pooled objects while operating on lock-free data structures.
As a simple example (and this is one that .net actually will facilitate) suppose that you have a 64-bit number in memory and you wish to perform bitwise 'and' and 'xor' operations upon it. Even though multiple threads may attempt to perform operations simultaneously, you want to guarantee that the number will not be corrupted. If in one thread you wrote var = var or 1 and in another thread var = var or 2 there would be no guarantee as to what var would end up containing.
CompareExchange allows a solution.
Sub BitMaskLong(ByRef Dest as Long, ByVal AndMask as Long, Byval XorMask as Long)
Dim OldVal, NewVal as Long
Do
OldVal = Dest
NewVal = (OldVal and AndMask) xor XorMask
Loop While Threading.Interlocked.CompareExchange(Dest, NewVal, OldVal) <> OldVal
End Sub The compare-exchange performed in the last statement will write the newly-computed value to the destination variable unless something else has change it since it was first read. If it has been changed, it will be reread and the new value recomputed. Provided that the computation between the read and the compare-exchange is shorter than a timeslice, the compare-exchange is unlikely to fail more than once for each processor in the system (indeed, most of the time it will succeed on the first try).
|
|
|
|
|
supercat9 wrote: double-length compare-exchange would be very handy if vb.net would allow it to be used on a structure containing an object reference and a counter.
Possibly so.
"The clue train passed his station without stopping." - John Simmons / outlaw programmer
"Real programmers just throw a bunch of 1s and 0s at the computer to see what sticks" - Pete O'Hanlon
|
|
|
|
|
At the top of every article when a user is logged in you see the following information:
Print Report Article Watch Bookmark Discuss
Would it be possible to add a button for 'Publicly Recommend' next to the bookmark.
I know their is a setting to make all future bookmarks recommended, but I want to be selective of which go through and which don't. This suggestion is just for a matter of convenience, I know their are ways to do this currently.
Regards,
Thomas Stockwell
Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning.
Visit my homepage Oracle Studios
Discounted or Free Software for Students:
DreamSpark - downloads.channel8.msdn.com
MSDN Academic Alliance - www.msdnaa.com
|
|
|
|
|
What about if we removed the 'Watch' icon and replaced it with a 'Bookmark Publicly' link?
cheers,
Chris Maunder
CodeProject.com : C++ MVP
|
|
|
|
|
Wouldn't those two buttons serve two different purposes? I thought the 'Watch' feature notified the watcher of updates, replacing it would remove the watching functionality wouldn't it?
Regards,
Thomas Stockwell
Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning.
Visit my homepage Oracle Studios
Discounted or Free Software for Students:
DreamSpark - downloads.channel8.msdn.com
MSDN Academic Alliance - www.msdnaa.com
|
|
|
|
|
I was thinking that bookmarks could be 'watched' by default.
cheers,
Chris Maunder
CodeProject.com : C++ MVP
|
|
|
|
|
Sounds great
Regards,
Thomas Stockwell
Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning.
Visit my homepage Oracle Studios
Discounted or Free Software for Students:
DreamSpark - downloads.channel8.msdn.com
MSDN Academic Alliance - www.msdnaa.com
|
|
|
|
|
Chris Maunder wrote: bookmarks could be 'watched' by default
That would be cool. Been using the bookmark feature more and more, and it's great
"The clue train passed his station without stopping." - John Simmons / outlaw programmer
"Real programmers just throw a bunch of 1s and 0s at the computer to see what sticks" - Pete O'Hanlon
|
|
|
|
|
This user[^] has been cross-post spamming multiple forums today, and appears to have been subtly spamming article forums for several days.
Citizen 20.1.01 'The question is,' said Humpty Dumpty, 'which is to be master - that's all.'
|
|
|
|
|
You beat me to it.
Shog9 wrote: and appears to have been subtly spamming article forums for several days.
Yeah, I guess the myriad of "Nice!" and "Thanks!" comments to arbitrary articles with a link to his blog wasn't enough, so he's upped the antey and is blantently spamming now.
|
|
|
|
|
Leslie Sanford wrote: Yeah, I guess the myriad of "Nice!" and "Thanks!" comments to arbitrary articles with a link to his blog wasn't enough, so he's upped the antey and is blantently spamming now.
I guess a blog full of links to other people's work just wasn't bringing in the ad revenue...
Chris, ever thought of adding rel="nofollow" server-side?
Citizen 20.1.01 'The question is,' said Humpty Dumpty, 'which is to be master - that's all.'
|
|
|
|
|
Shog9 wrote: adding rel="nofollow" server-side
Possibly a good thing.
"The clue train passed his station without stopping." - John Simmons / outlaw programmer
"Real programmers just throw a bunch of 1s and 0s at the computer to see what sticks" - Pete O'Hanlon
|
|
|
|
|
Only effective if the spammer knows it happens, understands what it means, and, most importantly, gives a toss.
As a rule I haven't found spammers to be the types who really care about the subtle nuances of how a site operates.
cheers,
Chris Maunder
CodeProject.com : C++ MVP
|
|
|
|
|
Chris Maunder wrote: haven't found spammers to be the types who really care about the subtle nuances of how a site operates
No, they are pretty much the bottom of the barrel, intellectually speaking.
"The clue train passed his station without stopping." - John Simmons / outlaw programmer
"Real programmers just throw a bunch of 1s and 0s at the computer to see what sticks" - Pete O'Hanlon
|
|
|
|
|
Chris Maunder wrote: Only effective if the spammer knows it happens, understands what it means, and, most importantly, gives a toss.
Well, yeah, as far as reducing spam goes. AFAIK, it'll still prevent them from using your site to bump up their PageRank, regardless of whether they ever notice or not.
Citizen 20.1.01 'The question is,' said Humpty Dumpty, 'which is to be master - that's all.'
|
|
|
|
|
It appears that, occasionally, someone decides to spend the day posting new threads on CP. Regardless of how interesting he might find the seventy-two links to "My Life as a CP Hamster," or how many times he emphasizes that his homework is "RGNT", or how quickly he leaves the ranks of bronze members, there comes a point where the forum begins to look like his/her personal blog.
If a relatively simple way could be imposed that would keep the user from starting more than say a half-dozen threads in a single day (or X threads in a Y length of time) it might go a long way towards restoring civility in all the forums, including SB.
It's a thought and probably worth only what you paid for it.
Jon
Smith & Wesson: The original point and click interface
|
|
|
|
|
While we are at it, why not limit the number of messages one can post in Lounge and Soapbox
to the number of messages the same person has already posted in programming forums and article sections.
Should this site really spend a significant part of its facilities to people who don't contribute
to the programming forums and article sections?
|
|
|
|
|
Luc Pattyn wrote: While we are at it, why not limit the number of messages one can post in Lounge and Soapbox to the number of messages the same person has already posted in programming forums and article sections.
Why not limit the number of messages one can post in programming forums and article sections to the number of messages the same person has already posted in Lounge and Soapbox.
The answer, imho, in both cases would be that the programming forums and the message-board forums serve different purposes. N'est-ce pas?
By the way, how much credit would you give, if any, to someone who carefully reviewed a new article, including downloading the attached code, running and possibly stepping through it, and performing repeated googles to insure a lack of plagiarism? It could easily take an hour or two, before reporting or approving an article. Dashing off a few answers, especially one to the "RGNT PLS HLP" folks would take considerably less time and effort, don't you think?
Jon
Smith & Wesson: The original point and click interface
|
|
|
|