This among other reasons is why I wish the Mozilla folks would integrate CPU throttling for background tabs. Chrome and Opera both have it. It is extremely effective and drives down CPU usage greatly for those of us who normally have a handful (or more) of tabs open. Hopefully, it's being worked on and I just can't see a single shred of evidence to that end because it's all being done quietly? It'd be a MUCH better feature than this letterboxing shit sounds like.
I do know that feeling. While I'm fine with avoiding words that offend people I care for I'm less and less sympathetic to the word police who want to tell complete strangers what they can and cannot say. That's the whole reason the term "hate speech" is so common. It's an attempt to re-brand the censorship and make it seem alright. That is to say "we aren't censoring anyone - we are just stopping hate speech." It seems whole swaths of the body politic in the USA forgot that the cure for speech you don't like is.... more speech.
So what? I'm younger than you are. Don't assume everyone is a greybeard or that the greybeards are clueless. Nonetheless, What you say is true, except that just about all mainframes are behind hella firewalls and support some type of encryption with GSSAPI or Secure Shell. However, I agree there are a lot of fossilized admins who do rely too much on obscurity.
I found a lost Ubuntu user. Can someone escort him across the street back home? Gosh, maybe we need one of those signs that says "Retards playing nearby" to warn folks to slow down.
There are tons of filesystems and many are purpose built for specific instances. They perform well at those jobs. So what? I never said UFS was the best filesystem for all purposes. I pointed out FAT because it's of an age with UFS and still used widely despite being "old". BTW, there is no "cheating" with filesystems. You perform well with the workload or you go away whining about "cheating". Cache safety is completely tunable and comes with very reasonable defaults on any BSD boxes I've used. UFS is still one of the best choice for BSD systems with small memory footprints. That's why it's still in wide use on embedded systems. Then again, that's nuance that you naysayers seem to lack. Sure, ZFS is better in lots of ways. It also came along MUCH later and it's a huge memory hog and shouldn't be used on systems with less than 4GB of RAM. UFS didn't "have it's day". It's still in use for all kind of purposes by folks smart enough to know when to use the right tool for the job. Again, just because a few Linux-children didn't see it on their new Ubuntu laptop doesn't really mean anything besides "they haven't seen much of the world." It also *still* doesn't mean Samsung should have re-used the acronym, which was the original point you've fallen so far away from in your efforts to bash UFS.
Actually I support thousands of machines using UFS. I can debug the filesystem code and have done so many times. So, yeah, I'm quite familiar with it. It's absolutely still in use in many many places (mostly legacy Solaris machines or on new NetBSD and/or FreeBSD boxes with low memory). Just because *you* don't see it, doesn't mean it's not there. Sure there are better file systems with more modern ideas incorporated into them like btrees, journaling or logging, and many other items like pseudo-volume-management like ZFS does. However, I'd point out that FAT is still in pretty wide use today and it's inferior to UFS in almost every way. So, the ever-lame "OMG that's so old" is just an ignorant and tiresome reaction like some scene from Idiocracy "Water... like from the toilet?". Water is "old" too. UFS is quite fast compared with FAT file systems (and beats ext2 quite handily most of the time, too depending on the workload). It's also easily tuned for prefetch caching on most systems that support it and it's ability to handle power failures is going to depend greatly on how much write-back caching is being used (and if it's battery backed or not). UFS in FreeBSD and NetBSD support logging quite nicely and handle power failures just fine, thanks (MUCH better than, say, ext4 which in my lab tests a few years ago performed the worst in power failure testing).
The point (besides you have no idea WTF you are talking about concerning UFS) is that manufacturers will start re-using acronyms that have been heavily used by others in the past and it pollutes search results for the unrelated items. Your judgment about UFS's lifecycle (besides being mostly wrong) just sounds like the view of an igmo who really doesn't know better and thinks he knows everything because he once built a gaming PC.
Buwahahaha. Speaking as a guy who maintains old Unix and OpenVMS boxes for a living I can tell you that most people would have zero idea what to do if they did "get in" to a mainframe or even a mini. Those skills are very rare and that's why people pay good money for folks who have skills beyond the basics that lots of schmucks learn from using Bash on their Ubuntu machine. I watch newbies with MVS, MPE, and VMS and they can't even change directories, much less do any damage. The mother fuckers couldn't even set their terminals up for proper emulation. I've gone as far as to give people access to honeypot environments for various mainframe and minicomputer operating systems from the 1980's (long story on how I set that up) but it was hilarious watching children from China, Romania, Israel, Bulgaria, Russia, Ukraine, and even a few Americans flop around like fish once they dropped to an environment that doesn't have command line completion and on screen help to walk their ignorant asses through everything. These little tykes are too dumb to do much damage, even though all they'd have to do is READ for about 10 minutes to get familiar with things (so in other words - forget it). Hell there are even Youtube videos on how to get around on a various mainframes. Nonetheless, only the true Scotsmen seem to have the attention span to pull it off and they are too busy making money and living well to worry about breaking into old city mainframes.
Personally, *nix represents smart people doing cool things whereas smartphones represent retards running into walls and having nervous breakdowns. The fact vendors horked a few bits of code does not make them Unix operating systems, not even fucking close.
Chrome has it, Opera has it. Firefox doesn't have it and looking around online they appear to not even give a fuck, have any plugins that can do it, or plan to do it. Someone please slap a link on me to prove otherwise. I hate that the only browsers that have this nearly essential feature are likely to be spying on my browsing. I will *never* trust Chrome, I don't care how fast it is. It's like a BMW. You can put 2000HP in one and it still looks like an ugly ass mom-car.
It stands for Unix File System, it doesn't stand for "Cheap shitty phone-related stupid-shit". Unless it comes per-initialized with UFS, then Samsung, here's a big middle finger for putting more noise and chaos into google searches. Yes, I saw the lower-case "e" in there two. Learn how to create an acronym rather than how to copy Steve Jobs (and even he would know better than to use "UFS").
I used that term with a friend of mine who is mixed-race. He told me it was offensive and that it was an Italian slur that referred to a coffee drink where milk and coffee were mixed together. He said that it's not offensive because of it's content, it's because of the way people used it in the past. However, then I looked it up and it's actually a Spanish or Portuguese word with "uncertain" origins. I basically look at it like this, you shouldn't call people things they don't like. I have an unusual name and when people butcher it or alter it, I get annoyed in a hurry (mainly because it's easy to pronounce and they do it on purpose). I can imagine that a racially charged term that someone is used to being flung at them by hostile racist assholes is that much worse. However, I was glad that my friend simply explained it to me and I was able to say "Oh, well I won't be using that again, then." That's a much more pleasant method for race relations than the kind of crap the mainstream media tries to propagate. I wonder if they'd like a race war, just to improve ratings!
Interesting. Sorry you had that experience. I'm not sure what you mean by a "multi-line text widget". I can tell you that early versions of OpenMOTIF were very very buggy in my experience. You probably know this, but after OpenMOTIF was completed and revved a few times the original MOTIF code was released as open-source. Many of the bugs I'd been seeing (and some just strange visual artifacts) disappeared. I know a lot of people love QT and it's produced real apps and real results - I won't poo-poo it. However, personally, I tend to burst into flames when coding in C++ so I just have a personal problem.:-)
That's right, you don't have to be a programmer to know better, the performance speaks for itself. Plus, a simple cogent explanation of why SIMD instructions aren't easy to optimize helps, but too many folks just want to dismiss anything that's not in their experience as "OLD" and useless. They have no idea about the giant's shoulders they stand on.
Bullshit. "Security fixes" aren't the only solution to security problems that crop up. Ring fencing, firewalling, byte patching, and many other mitigation strategies exist for security issues. One of which is simply "don't use the thing that's vulnerable." Another would be "replace the software with an open source alternative." Coward, you are very narrow minded if you think waiting like sheeple for "The Almighty Vendor" to come to your rescue for money, you might be sorely disappointed. They don't have the best history of giving a fuck. But yeah, for hyper-conformists who want to run the absolute most mainstream spyware garbage (read: Android && iOS) they can find then you are pretty much depending on the paternal vendor to come save you. How's that shit been working out so far bruh? Also you are dumber than most cowards if you think sealing the fucking battery or otherwise making the hardware difficult to service doesn't destroy the lifespan of the device for most people. Sadly, you aren't alone, though. A lot of other 12 year olds feel the same way.
I hope you aren't joking but I sense that you probably are. There is a growing retro scene out there and a number of folks who do board-level repair on C64, Amigas, STs, you name it... I personally don't think it's funny at all and I've used their services to do things like VRU or capacitor replacements many times.. I think it's awesome. I'm much excited about someone being able to repair fun & reusable "old" retro tech rather than the privacy invading garbage "intellectual property" of software megacorps and sue-happy hardware griefers who've brought us throw-away "app stores" and hardware that won't last longer than 5 years.
Not in the 040 but in the 060, yes. Check out how the branch cache works and the timings for branches. The branch prediction is like the static prediction in the 68040 where previous branch targets are assumed occupied because they are probably loops and new targets are assumed unoccupied. When the 060 sees a branch for the first time in order to aid in dynamic prediction it creates an entry in the branch cache. However, I don't know enough about the engineering they did to know if it's definitely vulnerable. Speculative execution as a concept is a bit risky anyway if you ask me.
You are so right. I'd also throw in libjpegturbo too, as it or something like it is used in most browsers. Even most IT people have little idea how much optimized, hand-rolled, asm there is in play in their daily computing experience. Then they get on threads like this and talk noise about it like compilers have it all figured out. All that tells me is that they are **not** asm coders and know exactly bupkis about the topic. It's too funny seeing them all commenting on it like they are some kind of authority and I am confident none of them have written a single ASM program in their lives. They just don't realize how ignorant they sound.
Some people are always trying to poo-poo old technology. I suspect this is a play to act like every CPU made before yesterday is no good. "Sorry folks, you're going to have to turn on all those old computers we can't control^H^H^H Uh, I mean, those vulnerable systems for your own safety." That or so some smug security weenie can sit and smirk pointing to some ridiculous "researcher" saying "it's impossible to prevent". I just think there may be more going on here than just "old stuff sucks"
But who cares how friendly RISC cores are to assembly programmers?
Me. I am one. I find it useful for my job and I use it hobby projects, too. I still do assembler on about.... 4 platforms on a regular basis. People pay good money for it.I also care because I like to code on my old SGI's and I find it frustrating at times to deal with MIPS versus 68k.
Assembler is less important than ever before, for anyone who isn't writing a compiler, because compilers keep doing more optimization.
My experience disagrees. Have you ever looked at SIMD instructions ? Compilers barely scratch the surface of their power and are notoriously terrible at optimizing for them. They *try* but they fall very short of the mark. When one understands how they work, you'll see why compilers are very very bad at trying to optimize them and if have to hint the crap out of your code, you might as well have just dropped into ASM anyway (better in fact). Write a color space transformation in C and try your absolute best to setup RGB triplets in your C code to be well-hinted and compiler friendly and then target them to some platform with appropriate SIMD instructions to help you out. Go ahead, buy ICC and get the best edge you can. I've done it. What one finds is that compilers optimize common software problems well and against specific software problems much more poorly. Read what others in the thread (about libavcodec and ffmpeg) have said: the results can be absolutely dramatic. Old machines able to do incredible tasks in realtime when properly hand-rolled in ASM. You doubt it and think compilers are godlike AI? Turn off your ASM optimized JPEG decoder in your browser (usually libjpegturbo or the like) and watch your experience CRAWL like you just won't believe. I can quote line and verse were they were hand optimized assembly code and the compiler was told to STFU for extremely good reason.
Unless you're writing code for the baseband processor in a cellphone, you're probably not going to bother with assembler. The other big examples, stuff like graphics for example, tend to just be handed off to coprocessors of some type or another.
I feel ASM is much more flexible than you imply and can be more general purpose when optimization is still important. Guys like the MenuetOS crew have shown that kind of flexibility in lots of general purpose apps for their working all-ASM operating system. Also, most modern graphics routines are often handled now by compiler optimized shaders implemented in a DSL or in C++ and less and less by hand rolled ASM due to the opaque nature of GPU hardware (unless you mean at the driver level and those are generally in C). Your dismissive attitude toward assembler makes me sad on one hand, but I'm glad only a few of us work on the metal these days comparatively. I know the true Scotsman when I meet him.:-)
Slashdot Mirror
Step 1: "When was the last time this place flooded?"
Step 2: "Oh, it has flooded before? Thanks. I'm not interested in the property."
This among other reasons is why I wish the Mozilla folks would integrate CPU throttling for background tabs. Chrome and Opera both have it. It is extremely effective and drives down CPU usage greatly for those of us who normally have a handful (or more) of tabs open. Hopefully, it's being worked on and I just can't see a single shred of evidence to that end because it's all being done quietly? It'd be a MUCH better feature than this letterboxing shit sounds like.
I do know that feeling. While I'm fine with avoiding words that offend people I care for I'm less and less sympathetic to the word police who want to tell complete strangers what they can and cannot say. That's the whole reason the term "hate speech" is so common. It's an attempt to re-brand the censorship and make it seem alright. That is to say "we aren't censoring anyone - we are just stopping hate speech." It seems whole swaths of the body politic in the USA forgot that the cure for speech you don't like is.... more speech.
So what? I'm younger than you are. Don't assume everyone is a greybeard or that the greybeards are clueless. Nonetheless, What you say is true, except that just about all mainframes are behind hella firewalls and support some type of encryption with GSSAPI or Secure Shell. However, I agree there are a lot of fossilized admins who do rely too much on obscurity.
I found a lost Ubuntu user. Can someone escort him across the street back home? Gosh, maybe we need one of those signs that says "Retards playing nearby" to warn folks to slow down.
There are tons of filesystems and many are purpose built for specific instances. They perform well at those jobs. So what? I never said UFS was the best filesystem for all purposes. I pointed out FAT because it's of an age with UFS and still used widely despite being "old". BTW, there is no "cheating" with filesystems. You perform well with the workload or you go away whining about "cheating". Cache safety is completely tunable and comes with very reasonable defaults on any BSD boxes I've used. UFS is still one of the best choice for BSD systems with small memory footprints. That's why it's still in wide use on embedded systems. Then again, that's nuance that you naysayers seem to lack. Sure, ZFS is better in lots of ways. It also came along MUCH later and it's a huge memory hog and shouldn't be used on systems with less than 4GB of RAM. UFS didn't "have it's day". It's still in use for all kind of purposes by folks smart enough to know when to use the right tool for the job. Again, just because a few Linux-children didn't see it on their new Ubuntu laptop doesn't really mean anything besides "they haven't seen much of the world." It also *still* doesn't mean Samsung should have re-used the acronym, which was the original point you've fallen so far away from in your efforts to bash UFS.
Actually I support thousands of machines using UFS. I can debug the filesystem code and have done so many times. So, yeah, I'm quite familiar with it. It's absolutely still in use in many many places (mostly legacy Solaris machines or on new NetBSD and/or FreeBSD boxes with low memory). Just because *you* don't see it, doesn't mean it's not there. Sure there are better file systems with more modern ideas incorporated into them like btrees, journaling or logging, and many other items like pseudo-volume-management like ZFS does. However, I'd point out that FAT is still in pretty wide use today and it's inferior to UFS in almost every way. So, the ever-lame "OMG that's so old" is just an ignorant and tiresome reaction like some scene from Idiocracy "Water... like from the toilet?". Water is "old" too. UFS is quite fast compared with FAT file systems (and beats ext2 quite handily most of the time, too depending on the workload). It's also easily tuned for prefetch caching on most systems that support it and it's ability to handle power failures is going to depend greatly on how much write-back caching is being used (and if it's battery backed or not). UFS in FreeBSD and NetBSD support logging quite nicely and handle power failures just fine, thanks (MUCH better than, say, ext4 which in my lab tests a few years ago performed the worst in power failure testing).
The point (besides you have no idea WTF you are talking about concerning UFS) is that manufacturers will start re-using acronyms that have been heavily used by others in the past and it pollutes search results for the unrelated items. Your judgment about UFS's lifecycle (besides being mostly wrong) just sounds like the view of an igmo who really doesn't know better and thinks he knows everything because he once built a gaming PC.
I'm not able to express the intelligence of the sentence you wrote, that's for sure. Are you sure it's English?
Buwahahaha. Speaking as a guy who maintains old Unix and OpenVMS boxes for a living I can tell you that most people would have zero idea what to do if they did "get in" to a mainframe or even a mini. Those skills are very rare and that's why people pay good money for folks who have skills beyond the basics that lots of schmucks learn from using Bash on their Ubuntu machine. I watch newbies with MVS, MPE, and VMS and they can't even change directories, much less do any damage. The mother fuckers couldn't even set their terminals up for proper emulation. I've gone as far as to give people access to honeypot environments for various mainframe and minicomputer operating systems from the 1980's (long story on how I set that up) but it was hilarious watching children from China, Romania, Israel, Bulgaria, Russia, Ukraine, and even a few Americans flop around like fish once they dropped to an environment that doesn't have command line completion and on screen help to walk their ignorant asses through everything. These little tykes are too dumb to do much damage, even though all they'd have to do is READ for about 10 minutes to get familiar with things (so in other words - forget it). Hell there are even Youtube videos on how to get around on a various mainframes. Nonetheless, only the true Scotsmen seem to have the attention span to pull it off and they are too busy making money and living well to worry about breaking into old city mainframes.
Personally, *nix represents smart people doing cool things whereas smartphones represent retards running into walls and having nervous breakdowns. The fact vendors horked a few bits of code does not make them Unix operating systems, not even fucking close.
Chrome has it, Opera has it. Firefox doesn't have it and looking around online they appear to not even give a fuck, have any plugins that can do it, or plan to do it. Someone please slap a link on me to prove otherwise. I hate that the only browsers that have this nearly essential feature are likely to be spying on my browsing. I will *never* trust Chrome, I don't care how fast it is. It's like a BMW. You can put 2000HP in one and it still looks like an ugly ass mom-car.
It stands for Unix File System, it doesn't stand for "Cheap shitty phone-related stupid-shit". Unless it comes per-initialized with UFS, then Samsung, here's a big middle finger for putting more noise and chaos into google searches. Yes, I saw the lower-case "e" in there two. Learn how to create an acronym rather than how to copy Steve Jobs (and even he would know better than to use "UFS").
You nexus, huh? I design you eyes!
I used that term with a friend of mine who is mixed-race. He told me it was offensive and that it was an Italian slur that referred to a coffee drink where milk and coffee were mixed together. He said that it's not offensive because of it's content, it's because of the way people used it in the past. However, then I looked it up and it's actually a Spanish or Portuguese word with "uncertain" origins. I basically look at it like this, you shouldn't call people things they don't like. I have an unusual name and when people butcher it or alter it, I get annoyed in a hurry (mainly because it's easy to pronounce and they do it on purpose). I can imagine that a racially charged term that someone is used to being flung at them by hostile racist assholes is that much worse. However, I was glad that my friend simply explained it to me and I was able to say "Oh, well I won't be using that again, then." That's a much more pleasant method for race relations than the kind of crap the mainstream media tries to propagate. I wonder if they'd like a race war, just to improve ratings!
Too late, Coward. I've already discovered that (hence my comment about DSL languages and shaders), but thanks for the snide remark, assface.
Interesting. Sorry you had that experience. I'm not sure what you mean by a "multi-line text widget". I can tell you that early versions of OpenMOTIF were very very buggy in my experience. You probably know this, but after OpenMOTIF was completed and revved a few times the original MOTIF code was released as open-source. Many of the bugs I'd been seeing (and some just strange visual artifacts) disappeared. I know a lot of people love QT and it's produced real apps and real results - I won't poo-poo it. However, personally, I tend to burst into flames when coding in C++ so I just have a personal problem. :-)
That's right, you don't have to be a programmer to know better, the performance speaks for itself. Plus, a simple cogent explanation of why SIMD instructions aren't easy to optimize helps, but too many folks just want to dismiss anything that's not in their experience as "OLD" and useless. They have no idea about the giant's shoulders they stand on.
Amazing how little memory and CPU MOTIF applications take. Once you get over the callbacks, it's actually not bad!
I agree, gimme more SIMD and you can kept branch prediction.
Bullshit. "Security fixes" aren't the only solution to security problems that crop up. Ring fencing, firewalling, byte patching, and many other mitigation strategies exist for security issues. One of which is simply "don't use the thing that's vulnerable." Another would be "replace the software with an open source alternative." Coward, you are very narrow minded if you think waiting like sheeple for "The Almighty Vendor" to come to your rescue for money, you might be sorely disappointed. They don't have the best history of giving a fuck. But yeah, for hyper-conformists who want to run the absolute most mainstream spyware garbage (read: Android && iOS) they can find then you are pretty much depending on the paternal vendor to come save you. How's that shit been working out so far bruh? Also you are dumber than most cowards if you think sealing the fucking battery or otherwise making the hardware difficult to service doesn't destroy the lifespan of the device for most people. Sadly, you aren't alone, though. A lot of other 12 year olds feel the same way.
I hope you aren't joking but I sense that you probably are. There is a growing retro scene out there and a number of folks who do board-level repair on C64, Amigas, STs, you name it... I personally don't think it's funny at all and I've used their services to do things like VRU or capacitor replacements many times.. I think it's awesome. I'm much excited about someone being able to repair fun & reusable "old" retro tech rather than the privacy invading garbage "intellectual property" of software megacorps and sue-happy hardware griefers who've brought us throw-away "app stores" and hardware that won't last longer than 5 years.
Not in the 040 but in the 060, yes. Check out how the branch cache works and the timings for branches. The branch prediction is like the static prediction in the 68040 where previous branch targets are assumed occupied because they are probably loops and new targets are assumed unoccupied. When the 060 sees a branch for the first time in order to aid in dynamic prediction it creates an entry in the branch cache. However, I don't know enough about the engineering they did to know if it's definitely vulnerable. Speculative execution as a concept is a bit risky anyway if you ask me.
You are so right. I'd also throw in libjpegturbo too, as it or something like it is used in most browsers. Even most IT people have little idea how much optimized, hand-rolled, asm there is in play in their daily computing experience. Then they get on threads like this and talk noise about it like compilers have it all figured out. All that tells me is that they are **not** asm coders and know exactly bupkis about the topic. It's too funny seeing them all commenting on it like they are some kind of authority and I am confident none of them have written a single ASM program in their lives. They just don't realize how ignorant they sound.
Some people are always trying to poo-poo old technology. I suspect this is a play to act like every CPU made before yesterday is no good. "Sorry folks, you're going to have to turn on all those old computers we can't control^H^H^H Uh, I mean, those vulnerable systems for your own safety." That or so some smug security weenie can sit and smirk pointing to some ridiculous "researcher" saying "it's impossible to prevent". I just think there may be more going on here than just "old stuff sucks"
But who cares how friendly RISC cores are to assembly programmers?
Me. I am one. I find it useful for my job and I use it hobby projects, too. I still do assembler on about.... 4 platforms on a regular basis. People pay good money for it.I also care because I like to code on my old SGI's and I find it frustrating at times to deal with MIPS versus 68k.
Assembler is less important than ever before, for anyone who isn't writing a compiler, because compilers keep doing more optimization.
My experience disagrees. Have you ever looked at SIMD instructions ? Compilers barely scratch the surface of their power and are notoriously terrible at optimizing for them. They *try* but they fall very short of the mark. When one understands how they work, you'll see why compilers are very very bad at trying to optimize them and if have to hint the crap out of your code, you might as well have just dropped into ASM anyway (better in fact). Write a color space transformation in C and try your absolute best to setup RGB triplets in your C code to be well-hinted and compiler friendly and then target them to some platform with appropriate SIMD instructions to help you out. Go ahead, buy ICC and get the best edge you can. I've done it. What one finds is that compilers optimize common software problems well and against specific software problems much more poorly. Read what others in the thread (about libavcodec and ffmpeg) have said: the results can be absolutely dramatic. Old machines able to do incredible tasks in realtime when properly hand-rolled in ASM. You doubt it and think compilers are godlike AI? Turn off your ASM optimized JPEG decoder in your browser (usually libjpegturbo or the like) and watch your experience CRAWL like you just won't believe. I can quote line and verse were they were hand optimized assembly code and the compiler was told to STFU for extremely good reason.
Unless you're writing code for the baseband processor in a cellphone, you're probably not going to bother with assembler. The other big examples, stuff like graphics for example, tend to just be handed off to coprocessors of some type or another.
I feel ASM is much more flexible than you imply and can be more general purpose when optimization is still important. Guys like the MenuetOS crew have shown that kind of flexibility in lots of general purpose apps for their working all-ASM operating system. Also, most modern graphics routines are often handled now by compiler optimized shaders implemented in a DSL or in C++ and less and less by hand rolled ASM due to the opaque nature of GPU hardware (unless you mean at the driver level and those are generally in C). Your dismissive attitude toward assembler makes me sad on one hand, but I'm glad only a few of us work on the metal these days comparatively. I know the true Scotsman when I meet him. :-)