Here in Moscow, Russia I 100Mbps home connections are quite widespread. Well, they are actually provided by fiber + cat5 going to individual apartments. It's not overly reliable, so failures do occur sometimes, and sometimes network is somewhat overloaded, but usually it's quite tolerable. Actually you can get download speed of 3-5 Mbytes/s from many Russian sites (and up to about 1 Mbyte/s from American sites) at night. I'm paying about $30 (converted from RUR) for 6 Gb/month prepayed limit, which is quite enough for me.
It really pleases me as a (semi-)closet Lisp programmer to see such extensions, as they closely resemble something so near to my heart, such as this
or this;-)
... but of course the best use of Greasemonkey is "defacement" of some your not-so-favorite sites. When having to work full time with something like.NET makes you really angry, it pleases to see something like this
just to imagine what you would want to do with their site;-)
Here in Russia we do have similiar problems. Some time a go, Linux was chosen for schools in Volgograd city. Now Microsoft boasts of making them swith back (it's in Russian, I was unable to find any non-automatic translation of that page). The obvious reason for this decision is that some local official was dreaming of Great Happy New Year and Microsoft helped him with this dream. Financially.
Joel is way too self-assured
on
Joel On Software
·
· Score: 3, Informative
I've came across an
interesting discussion at Joel's site. Answering a question regarding Lisp, Joel writes:
Paul Graham is brilliant and I'm sure lisp was great for him and his team, but I think most of the productivity of lisp came from the fact that it's garbage collected. I don't see any reason why a lisp programmer today would have a productivity advantage over a C# or Java programmer (and yes, I know lisp.)
If I knew Lisp so little (seems like e.g. he doesn't even know about Lisp macros and has no clear idea what closures are), I'd have never written "yes, i know lisp". Later in same discussion some guys who know Lisp bring out clearly that Joel is, well, somewhat underinformed and that he'd better shrink his self-assurance a bit. Poor Joel seemingly was unable to say anything against it.
I'dont care much about AOL's version Firefox (most likely it'll be something like Firefox + some useless stuff), but does all this mean that they will bring devedge.netscape.com back online?
I don't think that there's something too platform specific in indexing/search mechanisms that doubtlessly comprise the most complex part of Google Desktop Search. Platform specific issues are (probably) fast FS traversal mechanism + stuff to make the whole app not as bad as updatedb, file format plugins and some small parts of built-in web server. But I thought that good search/indexing engine is 10000% more complex than all this stuff...
I thought Google doesn't suffer so much from bad design. Tying such app to platform is definitely very bad design choice, especially if there are plans to port it to different platforms. They could save a lot of development time by using platform abstraction instead of direct usage of Win32 API throughout the code. I wonder why Google engineers have chosen such a strange approach. Maybe they were too short of time?
I've once managed to kill my Windows XP by similiar experiment. I've wrote a simple ASPX that generated randomly damaged WMFs and displayed them in browser doing reload afterwards. I don't know how well WMF problems were fixed after several WMF-related vulnerabilities are published, but the problem is that metafile data is directed right to GDI32 via PlayMetafile() and some of not-so-checked data even manages to get through to kernel-level win32k.sys. EMF has similiar problems. During these cruel experiments I've caused IE to crash many times, and after 30 mins of testing i've got BSOD. I've rebooted Windows, and it didn't find it's registry...:( NTFS was severely damaged, and unfortunatelly the corresponding "test case" (broken wmf) was lost...
I've once talked to one guy who, being fed up with IT, leaved his programmer job and became a blacksmith. Not just blacksmith, but one who does horseshoing. Though soon after that he was hired again to work on a program that examines pictures of horse hoofs to detect various diseases...
BTW, I think this is much better than Embedded SQL which makes programs look like mess (for those who don't think that Lisp programs are mess anyway:), and can be used together with Lisp macros.
Lisps allow usage of SQL-like syntax inside programs without stuff like "SELECT X FROM Y WHERE Z = ? ORDER BY A" or "SELECT X FROM Y WHERE Z = '" + something + "' ORDER BY A". Look here and here.
So, it takes Microsoft years to convert docs from one format to another... Perhaps instead of using e.g. some simple script they've hired a guy who spends whole days clicking, dragging & dropping. Definitely The Microsoft Way.
With.NET, I've spent many sleepless nights digging MSDN docs, various forums and lib sources (thanks to Rotor and Lutz Roeder's Reflector), looking for solutions for various problems caused by.NET bugs and design issues (BTW the worst thing is Windows.Forms). I NEVER had such bad problems with C++/Qt. Of course these also have their design problems, but without such bad implications. E.g. Qt, STL, Linux kernel, libc etc. inner workings are MUCH easier to understand that Microsoft stuff.
You probably didn't use.NET too much, or came to it from something much worse (maybe you have string VB6- background?). I also was thinking that.NET is something good 1.5 years ago. Now I have some not-so-good experience with it. It's unbelievably buggy, the design stinks - e.g. you often have to access private members (via reflection) or undocumented stuff (e.g. when traversing BindingManagerBases in BindingContext) to get necessary behavior.
Also, you didn't read enough of the article before criticizing it. E.g. MS will not add anonymous inner classes, but it will add anonymous methods.
Here's some discussion
concerning "Avalon knock-offs". If I get it correctly, Xamlon doesn't even include layout engine! So, one is supposed to enter Xs and Ys manually? Or what is the whole point of such XAML "implementations"?
That said, I don't think that XAML itself is a good idea. Apart from XAML vs XUL/SVG/etc. issues,.NET itself is fundamentally broken. I'm not sure
that it will be fixed and serious design issues will not make its way to Longhorn.
I said it figurativelly. Shipping a compiler that ICEs is like figurativelly pissing in my beer.
SuSE was better some time ago...
on
SUSE 9.2 Released
·
· Score: 3, Informative
I've started using SuSE in 2000 when I was in Germany. My previous distro was RH, which deeply offended me by shipping broken gcc 2.96 (and refusing to adopt KDE for a while before that).
I was impressed by SuSE's stability/quality, a nice feature of being able to save a list of selected packages in the middle of installation (IIRC), and also the fact that my old scanning prog kscan was included in Alpha section (I was far from being a good C++ programmer when I wrote it...)
Now I see an unpleasant tendency of including prerelease software in SuSE. As far as I remember, they were shipping a prerelease gcc 3.3, which caused problems with my (in-house) project and some prerelease of X11. Overall quality of the distro degraded. Also, I just don't get why they have Qt compiled with -DQT_NO_STL. As result, C++ programs that use STL have problems with system's Qt/KDE. This doesn't save memory/improve performance/etc., gcc shipped with SuSE has no problems with STL - so why?
I don't know whether SuSE is improving or getting worse now, as I'm currently deeply buried in.NET brain damage stuff. But next time when I'll be able to work under Linux most of time, I think I'll switch to something like Gentoo.
Slashdot Mirror
Here in Moscow, Russia I 100Mbps home connections are quite widespread. Well, they are actually provided by fiber + cat5 going to individual apartments. It's not overly reliable, so failures do occur sometimes, and sometimes network is somewhat overloaded, but usually it's quite tolerable. Actually you can get download speed of 3-5 Mbytes/s from many Russian sites (and up to about 1 Mbyte/s from American sites) at night. I'm paying about $30 (converted from RUR) for 6 Gb/month prepayed limit, which is quite enough for me.
It really pleases me as a (semi-)closet Lisp programmer to see such extensions, as they closely resemble something so near to my heart, such as this or this ;-)
... but of course the best use of Greasemonkey is "defacement" of some your not-so-favorite sites. When having to work full time with something like .NET makes you really angry, it pleases to see something like this
just to imagine what you would want to do with their site ;-)
;; Why do you think that modern Common Lisp
;; implementations are pure interpreters?
CL-USER> (declaim (optimize speed
(safety 0)
(space 0)
(debug 0)))
T
CL-USER> (defun my-func (x)
(declare (fixnum x))
(the fixnum (1+ x)))
MY-FUNC
CL-USER> (compile 'my-func)
MY-FUNC
NIL
NIL
CL-USER> (disassemble 'my-func)
;; disassembly of #<Function MY-FUNC>
;; formals:
;; code start: #x2083517c:
0: 83 c0 04 addl eax,$4
3: f8 clc
4: 8b 75 fc movl esi,[ebp-4]
7: c3 ret
; No value
;; This doesn't look like something "intermediate".
;; It's executable code which is created dynamically.
So, with this great feature, I will be unable to say (compile 'my-function) in my LISP REPL, because it creates executable code in data section?
Does the world need XML when there are s-expressions?
It's hard to find such thing exciting if you have children.
Here in Russia we do have similiar problems. Some time a go, Linux was chosen for schools in Volgograd city. Now Microsoft boasts of making them swith back (it's in Russian, I was unable to find any non-automatic translation of that page). The obvious reason for this decision is that some local official was dreaming of Great Happy New Year and Microsoft helped him with this dream. Financially.
I'dont care much about AOL's version Firefox (most likely it'll be something like Firefox + some useless stuff), but does all this mean that they will bring devedge.netscape.com back online?
I don't think that there's something too platform specific in indexing/search mechanisms that doubtlessly comprise the most complex part of Google Desktop Search. Platform specific issues are (probably) fast FS traversal mechanism + stuff to make the whole app not as bad as updatedb, file format plugins and some small parts of built-in web server. But I thought that good search/indexing engine is 10000% more complex than all this stuff...
I thought Google doesn't suffer so much from bad design. Tying such app to platform is definitely very bad design choice, especially if there are plans to port it to different platforms. They could save a lot of development time by using platform abstraction instead of direct usage of Win32 API throughout the code. I wonder why Google engineers have chosen such a strange approach. Maybe they were too short of time?
The have SX-56 which is probably 7 times more powerful but unfortunatelly runs Windows CE... ;-(
I've once managed to kill my Windows XP by similiar experiment. I've wrote a simple ASPX that generated randomly damaged WMFs and displayed them in browser doing reload afterwards. I don't know how well WMF problems were fixed after several WMF-related vulnerabilities are published, but the problem is that metafile data is directed right to GDI32 via PlayMetafile() and some of not-so-checked data even manages to get through to kernel-level win32k.sys. EMF has similiar problems. During these cruel experiments I've caused IE to crash many times, and after 30 mins of testing i've got BSOD. I've rebooted Windows, and it didn't find it's registry ... :( NTFS was severely damaged, and unfortunatelly the corresponding "test case" (broken wmf) was lost...
s/leaved/left/ -- excuse my poor English
I've once talked to one guy who, being fed up with IT, leaved his programmer job and became a blacksmith. Not just blacksmith, but one who does horseshoing. Though soon after that he was hired again to work on a program that examines pictures of horse hoofs to detect various diseases...
BTW, I think this is much better than Embedded SQL which makes programs look like mess (for those who don't think that Lisp programs are mess anyway :), and can be used together with Lisp macros.
Lisps allow usage of SQL-like syntax inside programs without stuff like "SELECT X FROM Y WHERE Z = ? ORDER BY A" or "SELECT X FROM Y WHERE Z = '" + something + "' ORDER BY A". Look here and here.
So, it takes Microsoft years to convert docs from one format to another... Perhaps instead of using e.g. some simple script they've hired a guy who spends whole days clicking, dragging & dropping. Definitely The Microsoft Way.
With .NET, I've spent many sleepless nights digging MSDN docs, various forums and lib sources (thanks to Rotor and Lutz Roeder's Reflector), .NET bugs and design issues (BTW the worst thing is Windows.Forms). I NEVER had such bad problems with C++/Qt. Of course these also have their design problems, but without such bad implications. E.g. Qt, STL, Linux kernel, libc etc. inner workings are MUCH easier to understand that Microsoft stuff.
looking for solutions for various problems caused by
s/string VB6- background/strong VB6- background/
Also, you didn't read enough of the article before criticizing it. E.g. MS will not add anonymous inner classes, but it will add anonymous methods.
That said, I don't think that XAML itself is a good idea. Apart from XAML vs XUL/SVG/etc. issues, .NET itself is fundamentally broken. I'm not sure
that it will be fixed and serious design issues will not make its way to Longhorn.
I said it figurativelly. Shipping a compiler that ICEs is like figurativelly pissing in my beer.
Now I see an unpleasant tendency of including prerelease software in SuSE. As far as I remember, they were shipping a prerelease gcc 3.3, which caused problems with my (in-house) project and some prerelease of X11. Overall quality of the distro degraded. Also, I just don't get why they have Qt compiled with -DQT_NO_STL. As result, C++ programs that use STL have problems with system's Qt/KDE. This doesn't save memory/improve performance/etc., gcc shipped with SuSE has no problems with STL - so why?
I don't know whether SuSE is improving or getting worse now, as I'm currently deeply buried in .NET brain damage stuff. But next time when I'll be able to work under Linux most of time, I think I'll switch to something like Gentoo.