What we did was just providing a additional ways for the programmer to shoot himself in the foot, and nothing else. The so-called "flexibility" is in fact only in the bullet's caliber. Again, there was no real added value in what we provided. Just a lot more complexity on both our side and the users, and less safety.
How many open source project suffer from this type of problem? They decide to be increadibly flexible, and try and please everyone, but in the end they end up being too complicated. Commercial products have external pressurers (eg, deadlines) that help them avoid this.
Not that I'm saying flexibility is bad - just flexibility should be measured against ease of implementaion, ease of use and maintainability.
Well.. it depends what you mean. Do I think that task manager is a totally accurate picture of memory usage? No, I don't. But do I think MS has somehow rigged it to make it seem like IE isn't running when it is? No, I'm not that paranoid.
Just in case you aren't convinced, here's an edited dll usage thing (Just showing dlls I was watching for):
With IE loaded:
0x4550000 0x100000 * 5.00.2919.6307 11/5/99 12:00 AM C:\WINNT\system32\MSHTML.TLB
0x1650000 0x188000 * 8/8/00 11:09 PM C:\WINNT\Profiles\Administrator\Local Settings\Temporary Internet Files\Content.IE5\index.dat
0x70C30000 0x241000 5.00.2919.6307 11/5/99 12:00 AM C:\WINNT\System32\mshtml.dll
0x70EB0000 0x3C000 5.00.2919.6307 11/5/99 12:00 AM C:\WINNT\System32\mshtmled.dll
0x711F0000 0x77000 5.01.0000.4615 11/5/99 12:00 AM C:\WINNT\System32\jscript.dll
0x400000 0x12000 5.00.2919.6304 11/5/99 12:00 AM C:\PROGRA~1\Plus!\MICROS~1\iexplore.exe
(Sorry about the formatting)
Now, with not running, guess what? Those Dll's aren't loaded!?!? Wow! Amazing, hey!
Now this tool wasn't written by MS, but maybe they rigged the API so it would hide IE dlls, right?
Okay - so what I did was close IE, and then renamed jscript.dll (which windows wouldn't let me do if it was in use). That was okay.. Lets try another... hmmm mshtml.dll! I bet it can't hide using that. Nup... that was okay too. Well... maybe they have a hidden file system with copies of those files... so I start up IE.. nup.. big crash.
My conclusion: If IE is running on my computer when it says it isn't, then I sure want to know how, because it is one of the most impressive bits of windows programming I've seen and I bet I could make a lot of money duplicating it.
Sorry if I sound a bit over the top here... I just think the MS paranoia goes a little far sometimes.
People keep claiming this, and when you have Active Desktop running, the arguement probably has merit.
However, on my computer with NT4 SP6 there is no basis for this. Yes, some DLLs used by IE are already loaded into memory, and this does account for some of the speed up, but actual application itself isn't. I know you are going to say "The actual application isn't the major part of IE" - well, that isn't really true either. IEXPLORE.EXE on my computer (which I guess is the UI to the browser, the support functionality - bookmarks etc -, the rendering engine, and maybe some networking code) is taking up over 8Meg rendering Slashdot. If I close it, that application disapperars, and I get 8Meg back.
This is comparable to the 10Meg or so it takes Netscape to do the same task.
Therefore, I conclude than IE isn't loaded into memory, at least on my computer.
I actually have good things to say about Mozilla!
on
Mozilla M17 Is Out
·
· Score: 5
I'm a big IE5 fan. I think it is stable, fast, and reasonably standards compliant - despite the general view on Slashdot. I doubt anyone can argue that it is better than Netscape 4, anyway.
I tried Galeon the other day. If you haven't heard of this, it is a GNOME/Gtk based browser, which uses the Mozilla "Gecko" rendering engine. I was very impressed. You remember all that stuff everyone was saying about how Mozilla was going to be the fastest render on the block? Well.. it's a close call, but I'd say Galeon is pretty close to achievig that now (This with is M16 - it might be quicker now).
As for startup time (always the problem with normal Mozillas) - it is now much, much quicker than Netscape 4 on either Windows of Linux, and getting close to IE5 (which is the fastest starting browser I've ever seen, except maybe Lynx or something).
It's not quite ready for everyone to use - there are a few niggling problems (no animated "Page Loading" icon so you know when it is doing something for instance).
The cool thing is that using this doesn't stop you from using XUL as well. For instance, I used it to ftp a file, and was surprised to see a new window pop up with a XUL based download monitor.
Very Impressive. If you can't wait for the Mozilla UI to get better, give this a go. At least it will start putting a few "Mozilla 6" in web server logs around the place.
Oh yeah.. I found you needed Helix-Gnome for it to work correctly on Redhat 6.2
is well over 90% of the population, and cellular use is one of the highest per head in the world (just after Scandinavian countries, but ahead of Japan, I believe). This is in a country only slightly smaller than the US.
Of course, we have somewhat unique population distribution, so the phone companies can provide coverage along the east coast and a few other centers, and that's enough to get the 90% coverage they are talking about. OTOH, it is probably only about 20-25% land coverage, which I guess would be less than in the US. Our cellular system is mostly GSM, but you can get CDMA as well.
Has this level of phone penetarion made any difference? Socially, yes, and it has made big profits for phone service providers, but it hasn't made any difference to our technical capabilites. We have one phone manufacture. We have a few research labs (eg, Motorola), but that's about it. I'd guess our WAP take up is behide the US.
I wouldn't worry to much about the US being behind in Cellular usage. The techincal gap is pretty small, and insignificant.
The social change is pretty large, though. Now I expect all my friends to be available all the time, with their own number. It means, for instance, I can call one person in a two person household, and be sure of getting them, and not the other person. It means you can arrange meetings at anytime, anywhere. It's the social changes like those where the US may be behind - not techincally.
Borland would have sued MS again, and MS would have been forced to buy even more of them (just like a couple of years ago, when Borland sued MS for taking Anders, and some patent infringments, and MS ended up paying $100M in "licencing" for Borland's MIDAS technology, and buying 10% of Borland. The case was settled out of court)
After all, Object Pascal (in Delphi), which Anders designed is a fair attempt at a successor to Pascal, even if it does keep the Pascal name, backwards compatibility, and the ability to code in a non-OO style if you choose.
If you are hoping for people to help with development and/or bug fixes, then a late release of the developed code is okay, because during the earlier stages there probably isn't enough of an end product for users to use and get interested enough in to become involved in.
OTOH, if you are interested in getting feedback in precicely what your market wants and needs, then an early release is good, because it lets external people into your decision making process about the product.
You should note, though, that it is unlikely you will get development help from people at the start of the project unless (a) your product fills a need or interest that a lot of people who code for fun are into, or (b) you can form partnerships with other companies to have paid programmers working on it.
For example, if you decide that you are going to build an open source Just-In-Time parts management system, it is unlikely that many people will hack on it for fun if there is no code already written. However, if you write it, then open source it, many companies might be interested in paying someone to help with enhancements.
However, if you annouce your intention to produce this JIT system in industry papers, etc, you might get lots of well qualified people signing up on your mailing lists and giving you advice that is simply unattainable elsewhere.
I think it was the last of the "1" series before v2 introduced frames (and tables? - no, I think that was in v1).
I think I still have the floppys for a Windows 3.1 install for that. It ran really well on my 386DX40...
I remember the Netscape beta releases - hearing stories about how we couldn't get them in Australia because the servers were too overloaded for mirroring.
Then, when v1 came out, we figured out a way to set up the destops on the uni machines to use one of the lectures proxys and give us external access, and everyone downloaded it. Ahh.... makes me feel like I've been around for ever, and I'm only 25.
Sure, Mozilla the browser is way, way overdue. It's so far overdue, that I am begining to defend sites writing IE only HTML - IE has 90% of the market, after all. (I don't like it, though)
That's not the point, really. The point is that once Mozilla is stable we will have a truly cross platform, lightweight and fast alternative to the MS Active X architecture.
Sure, Java isn't bad, but for GUI apps... well, even Java evangelists aer concentrating on the server now.
Mozilla/XUL really is cross platform. Applications written using XUL will appear identical on all platforms (even the Mac!).
Mozilla is lightweight. Everyone complains about the "huge" download, but lasttime I checked, it was between 6 and 7 megs. That's for the browser, and the entire XUL operating environment. Try getting a JVM and class libraries that small, let alone an IE/Windows download.
Mozilla is fast. Okay.. it's not that fast, but the nightly builds have been getting better, and the HTML rendering is fast.
Anyway, Mozilla isn't dead. I doubt it will ever (with "ever" = the next year or so) suprass IE's market share, but it will begin to make an impression in the application development market very soon, now. I'd say within 6 months there will be more websites on XUL development than development in something like FLTK.
The idea is more to put the fear of being caught into the mind of the troll.
I'm really, really serious about the reward. I will pay it, and I will consider paying some/most of it for any infomation leading to getting him at least kicked of his ISP.
I was in a great mood yesterday until this happened. I'd just had my first story ever accepted by Slashdot (The Interbase one), I had an interesting job interview, and then K5 goes down.
I read K5 more than Slashdot these days, and post a lot more on there. I try and submit a story or two a week, and I have great fun there.
I had this great book review (of "The Forever War" by Jon Haldeman - great book, possibly the best '70s Hard SciFi I've ever read) half typed up. I log on, and I saw the submission queue with 25 entries. I think "Oh shit.. They are trolling K5" - sure enough, that's what it was.
Why would someone do this? I never understood people doing it on Slashdot, either. Once in a while, a good hand written troll is funny because of the reaction, but script-trolling? Why? Everyone knows you can do it - there is no challenge.
Anyway. I'm going to do something about it. I'm offering a $200 reward (that's Australian $s) in the event of someone turing the K5 troll in, and successful legal action being taken.
Sure, it's not a huge amount, but I hope a few others will do the same, and we'll see what happens. Yes, I'm serious.
It's not really a platform problem. You might have to partition the DB into multiple files to get around the 2GB file size limit on Linux (I think Sybase can do that), but I doubt there would be any other real problem.
Sybase runs on Linux, of course, so there is no problem there.
I'd ask in the Sybase newsgroups about the biggest database they have seen on Linux - they have a good reputation for quick answers. (About the onlt good thing I have to say about Sybase, but still....)
I'd be surprised if there aren't quite a few Linux DB's bigger than 30GB anyway.
I downloaded the CVS version of WinJab a couple of days after it's release (About 16-July-2000). It wouldn't compile.
You do expect a little bit of that with Delphi programs - missing components, etc, but this was crazy. I downloaded everthing I could find, and still it wouldn't compile. There was a whole set of XML stuff missing from CVS. God knows where to find that... (And yes, I registered the type library)
I code Delphi for a living, so I'd hate to think how much trouble other people would be having.
I'm a little annoyed about this, because I had a look at the code, and I saw a lot of things I could have fixed for very little effort. (Eg, they are auto-creating all the forms in the project at application start-up. That's a bad thing to do, but pretty easy to fix.) I wanted to help, but I couldn't.
Altering the price of digital content is a common tactic to capture the time-preference characteristics of consumers (e.g. movie tickets for immediate first releases down to free-to-air a couple of years later).
I'm sorry, but WTF??!? It took me two or three reads to work out what "time-preference characteristics of consumers" meant. Why do people have to dress up what they mean as some kind of buzz phase?
Delphi does have pointer arithmitic as well (I have no idea why I said it didn't), but you never use it.
As for garbage collection: I'm not too sure abotu that. Delphi does if you use Interfaces/COM, but it doesn't otherwise. Java and Python do, but I'm not at all sure about SmallTalk.
I've just bothered to start reading stuff about.Net - I thought it was just MS marketing speak. There's actually some pretty cool stuff in there once you get past all the crap. (I've written something at http://www.kuro5hin.org, but it was in the moderation queue when I posted this)
Anyway, reading this guys writing makes me even more impressed with.Net (although did anyone else get the impression he loves C++ even more than most Linux hackers?). For instance:
Session state can be maintained on an alternate machine, allowing server reboots with preserved state,and it can be cookieless via url munging. You get a call stack trace and some verbose error messages. Event log and mail are built in, but mail requires that damn mail service still. Interdev will debug across languages now. Oh, and IIS gets Apache like pre-emptive recycling. The surprise here is they admitted they borrowed it from Apache. ASP is finally in its own process, and has memory leak detection. You can even assign processes and processor affinity to web apps.
Pretty cool, hey! Especially the maintaing of session state across different machines.
Will someone please remind all these VB people that it isn't object oriented, its object based? You'd think these guys are coding in C++ or Java the way the bandie OO terms about.
This guy just doesn't listen. Even MS (now!) admits that VB is only object based. However VB7 (which is the.Net version) is a proper OO language. It has proper inheritance and encapsulation - and it even does exception handling.
Object oriented development in VB.NET...hehehe, title says it all. Alan Carter, yer a dork. I want to see pointers and operator overloading.
I don't. There is no need for pointers in most coding, and generally the parts where pointers are used are the most bug prone. Java, Delphi, Python, Smalltalk - all proper OO languages and none have pointers. (Not sure about operator overloading)
The biggest worry is, of course, this:
ASP+ performance and caching was another welcome to C# filled session, and offered little other than common sense...throughput vs. responsiveness, etc... Warning to ASP coders: A lot of these "features" are going to automatically be shoving preformated html to your users. Remember how frontpage hosed yer code? Now IIS will be doing it.
The funniest thing his his whole speil:
They held up a poor guy, Eiffel author, as proof of their party language support. His language looks similar to C#...he won't be there next year.
Now this poor guy was (I believe) none other than Bertrand Meyer who (while he doesn't know much about open source software) does know his stuff when it comes to high quality software engineering. For the author of this piece not to know who he was, and to claim His language looks similar to C#... is pretty dumb. Since this guy seems to think he is some kind of technology guru, I would have expected he would have heard of Eiffel. It's not particually similar to C#, btw - have a look at design by contract for a start.
Actually, most non-commercial component developers are encouraged to use the MPL, not the GPL, because (a) it gets around the issues the GPL has linking to (the non-GPL'ed) VCL and (b) many major component suites (eg, Winshoes) are MPLed, so it makes linking to them easier in commercial software.
For instance, both the USSR and Czechoslovakia (neither of which exist any more) still have su and cs, and yet Serbia (which is recognized by the UN) doesn't have one.
The "USA Minor Outlying Islands" is um, yet no one would even claim that is a country. Same with the British Indian Ocean Territory (io). Greenland (a territory of Denmark) has gl
As you can see, having a Country Code means nothing in terms of being a country or not.
Absolutly no doubt at all. (assuming the "close enough" licence really is like the GPL - I haven't read it)
They are not distributing the software, and there is nothing at all in the GPL that say they can't charge for the use of the software, because they aren't redistributing it.
I don't know if this is fair or not - but that's not the point. It is allowed by the GPL.
I personally believe there is room for another, new licence that is designed for online apps like that.
I've used Oracle 7, MS SQL Server 7 and Sybase Adaptive Server 11.somthing
Sybase can't do left outer joins and then filter the results with a "where" clause. That is stupid.
There's some stupid limitation about how you can't pass blobs as parameters to stored procedures (Can't remember the details, though).
I'd choose MS SQL Server 7 over Sybase anyday (even though NT bogs down pretty quick). ANSI 92 SQL is wonderful to work with, and the client side tools are great (compare MS Query analyzer with having to look at Sybase or Oracle query plans by hand).
Oracle is great, too, of course, although I did find a bug once. In v7.2 you can't use nested result sets in queries in stored procedures. It was fixed in 7.4, though.
I'm in Australia, so I feel your download pain. 6 Meg (in the case of Mozilla) isn't big at all for a major piece of software, though.
Don't forget that the Mozilla designers were (from the start) working on cross platform compatibility, so they couldn't rely on non-standard system libraries like the JPEG decompression lib you refer to in RiscOS.
Slashdot Mirror
How many open source project suffer from this type of problem? They decide to be increadibly flexible, and try and please everyone, but in the end they end up being too complicated. Commercial products have external pressurers (eg, deadlines) that help them avoid this.
Not that I'm saying flexibility is bad - just flexibility should be measured against ease of implementaion, ease of use and maintainability.
Just in case you aren't convinced, here's an edited dll usage thing (Just showing dlls I was watching for):
With IE loaded:
0x4550000 0x100000 * 5.00.2919.6307 11/5/99 12:00 AM C:\WINNT\system32\MSHTML.TLB
0x1650000 0x188000 * 8/8/00 11:09 PM C:\WINNT\Profiles\Administrator\Local Settings\Temporary Internet Files\Content.IE5\index.dat
0x17E0000 0xC000 * 8/8/00 11:09 PM C:\WINNT\Profiles\Administrator\Cookies\index.dat
0x1A60000 0x4000 * 8/3/00 3:54 PM C:\WINNT\Profiles\Administrator\Application Data\Microsoft\Internet Explorer\MSIMGSIZ.DAT
0x70C30000 0x241000 5.00.2919.6307 11/5/99 12:00 AM C:\WINNT\System32\mshtml.dll
0x70EB0000 0x3C000 5.00.2919.6307 11/5/99 12:00 AM C:\WINNT\System32\mshtmled.dll
0x711F0000 0x77000 5.01.0000.4615 11/5/99 12:00 AM C:\WINNT\System32\jscript.dll
0x400000 0x12000 5.00.2919.6304 11/5/99 12:00 AM C:\PROGRA~1\Plus!\MICROS~1\iexplore.exe
(Sorry about the formatting)
Now, with not running, guess what? Those Dll's aren't loaded!?!? Wow! Amazing, hey!
Now this tool wasn't written by MS, but maybe they rigged the API so it would hide IE dlls, right?
Okay - so what I did was close IE, and then renamed jscript.dll (which windows wouldn't let me do if it was in use). That was okay.. Lets try another... hmmm mshtml.dll! I bet it can't hide using that. Nup... that was okay too. Well... maybe they have a hidden file system with copies of those files... so I start up IE.. nup.. big crash.
My conclusion: If IE is running on my computer when it says it isn't, then I sure want to know how, because it is one of the most impressive bits of windows programming I've seen and I bet I could make a lot of money duplicating it.
Sorry if I sound a bit over the top here... I just think the MS paranoia goes a little far sometimes.
People keep claiming this, and when you have Active Desktop running, the arguement probably has merit.
However, on my computer with NT4 SP6 there is no basis for this. Yes, some DLLs used by IE are already loaded into memory, and this does account for some of the speed up, but actual application itself isn't. I know you are going to say "The actual application isn't the major part of IE" - well, that isn't really true either. IEXPLORE.EXE on my computer (which I guess is the UI to the browser, the support functionality - bookmarks etc -, the rendering engine, and maybe some networking code) is taking up over 8Meg rendering Slashdot. If I close it, that application disapperars, and I get 8Meg back.
This is comparable to the 10Meg or so it takes Netscape to do the same task.
Therefore, I conclude than IE isn't loaded into memory, at least on my computer.
I'm a big IE5 fan. I think it is stable, fast, and reasonably standards compliant - despite the general view on Slashdot. I doubt anyone can argue that it is better than Netscape 4, anyway.
I tried Galeon the other day. If you haven't heard of this, it is a GNOME/Gtk based browser, which uses the Mozilla "Gecko" rendering engine. I was very impressed. You remember all that stuff everyone was saying about how Mozilla was going to be the fastest render on the block? Well.. it's a close call, but I'd say Galeon is pretty close to achievig that now (This with is M16 - it might be quicker now).
As for startup time (always the problem with normal Mozillas) - it is now much, much quicker than Netscape 4 on either Windows of Linux, and getting close to IE5 (which is the fastest starting browser I've ever seen, except maybe Lynx or something).
It's not quite ready for everyone to use - there are a few niggling problems (no animated "Page Loading" icon so you know when it is doing something for instance).
The cool thing is that using this doesn't stop you from using XUL as well. For instance, I used it to ftp a file, and was surprised to see a new window pop up with a XUL based download monitor.
Very Impressive. If you can't wait for the Mozilla UI to get better, give this a go. At least it will start putting a few "Mozilla 6" in web server logs around the place.
Oh yeah.. I found you needed Helix-Gnome for it to work correctly on Redhat 6.2
Of course, we have somewhat unique population distribution, so the phone companies can provide coverage along the east coast and a few other centers, and that's enough to get the 90% coverage they are talking about. OTOH, it is probably only about 20-25% land coverage, which I guess would be less than in the US. Our cellular system is mostly GSM, but you can get CDMA as well.
Has this level of phone penetarion made any difference? Socially, yes, and it has made big profits for phone service providers, but it hasn't made any difference to our technical capabilites. We have one phone manufacture. We have a few research labs (eg, Motorola), but that's about it. I'd guess our WAP take up is behide the US.
I wouldn't worry to much about the US being behind in Cellular usage. The techincal gap is pretty small, and insignificant.
The social change is pretty large, though. Now I expect all my friends to be available all the time, with their own number. It means, for instance, I can call one person in a two person household, and be sure of getting them, and not the other person. It means you can arrange meetings at anytime, anywhere. It's the social changes like those where the US may be behind - not techincally.
I mean - cable modems? DSL? I can only dream!
Borland would have sued MS again, and MS would have been forced to buy even more of them (just like a couple of years ago, when Borland sued MS for taking Anders, and some patent infringments, and MS ended up paying $100M in "licencing" for Borland's MIDAS technology, and buying 10% of Borland. The case was settled out of court)
After all, Object Pascal (in Delphi), which Anders designed is a fair attempt at a successor to Pascal, even if it does keep the Pascal name, backwards compatibility, and the ability to code in a non-OO style if you choose.
If you are hoping for people to help with development and/or bug fixes, then a late release of the developed code is okay, because during the earlier stages there probably isn't enough of an end product for users to use and get interested enough in to become involved in.
OTOH, if you are interested in getting feedback in precicely what your market wants and needs, then an early release is good, because it lets external people into your decision making process about the product.
You should note, though, that it is unlikely you will get development help from people at the start of the project unless (a) your product fills a need or interest that a lot of people who code for fun are into, or (b) you can form partnerships with other companies to have paid programmers working on it.
For example, if you decide that you are going to build an open source Just-In-Time parts management system, it is unlikely that many people will hack on it for fun if there is no code already written. However, if you write it, then open source it, many companies might be interested in paying someone to help with enhancements.
However, if you annouce your intention to produce this JIT system in industry papers, etc, you might get lots of well qualified people signing up on your mailing lists and giving you advice that is simply unattainable elsewhere.
Does anyone else remember that browser?
I think it was the last of the "1" series before v2 introduced frames (and tables? - no, I think that was in v1).
I think I still have the floppys for a Windows 3.1 install for that. It ran really well on my 386DX40...
I remember the Netscape beta releases - hearing stories about how we couldn't get them in Australia because the servers were too overloaded for mirroring.
Then, when v1 came out, we figured out a way to set up the destops on the uni machines to use one of the lectures proxys and give us external access, and everyone downloaded it. Ahh.... makes me feel like I've been around for ever, and I'm only 25.
Sure, Mozilla the browser is way, way overdue. It's so far overdue, that I am begining to defend sites writing IE only HTML - IE has 90% of the market, after all. (I don't like it, though)
That's not the point, really. The point is that once Mozilla is stable we will have a truly cross platform, lightweight and fast alternative to the MS Active X architecture.
Sure, Java isn't bad, but for GUI apps... well, even Java evangelists aer concentrating on the server now.
Mozilla/XUL really is cross platform. Applications written using XUL will appear identical on all platforms (even the Mac!).
Mozilla is lightweight. Everyone complains about the "huge" download, but lasttime I checked, it was between 6 and 7 megs. That's for the browser, and the entire XUL operating environment. Try getting a JVM and class libraries that small, let alone an IE/Windows download.
Mozilla is fast. Okay.. it's not that fast, but the nightly builds have been getting better, and the HTML rendering is fast.
Anyway, Mozilla isn't dead. I doubt it will ever (with "ever" = the next year or so) suprass IE's market share, but it will begin to make an impression in the application development market very soon, now. I'd say within 6 months there will be more websites on XUL development than development in something like FLTK.
Interbase documentaion is available from ftp://ftp2.interbase.c om/pub/products/beta6.0/ib_b60_doc.zip
It's in PDF, from memory.
The idea is more to put the fear of being caught into the mind of the troll.
I'm really, really serious about the reward. I will pay it, and I will consider paying some/most of it for any infomation leading to getting him at least kicked of his ISP.
I was in a great mood yesterday until this happened. I'd just had my first story ever accepted by Slashdot (The Interbase one), I had an interesting job interview, and then K5 goes down.
I read K5 more than Slashdot these days, and post a lot more on there. I try and submit a story or two a week, and I have great fun there.
I had this great book review (of "The Forever War" by Jon Haldeman - great book, possibly the best '70s Hard SciFi I've ever read) half typed up. I log on, and I saw the submission queue with 25 entries. I think "Oh shit.. They are trolling K5" - sure enough, that's what it was.
Why would someone do this? I never understood people doing it on Slashdot, either. Once in a while, a good hand written troll is funny because of the reaction, but script-trolling? Why? Everyone knows you can do it - there is no challenge.
Anyway. I'm going to do something about it. I'm offering a $200 reward (that's Australian $s) in the event of someone turing the K5 troll in, and successful legal action being taken.
Sure, it's not a huge amount, but I hope a few others will do the same, and we'll see what happens. Yes, I'm serious.
It's not really a platform problem. You might have to partition the DB into multiple files to get around the 2GB file size limit on Linux (I think Sybase can do that), but I doubt there would be any other real problem.
Sybase runs on Linux, of course, so there is no problem there.
I'd ask in the Sybase newsgroups about the biggest database they have seen on Linux - they have a good reputation for quick answers. (About the onlt good thing I have to say about Sybase, but still....)
I'd be surprised if there aren't quite a few Linux DB's bigger than 30GB anyway.
I downloaded the CVS version of WinJab a couple of days after it's release (About 16-July-2000). It wouldn't compile.
You do expect a little bit of that with Delphi programs - missing components, etc, but this was crazy. I downloaded everthing I could find, and still it wouldn't compile. There was a whole set of XML stuff missing from CVS. God knows where to find that... (And yes, I registered the type library)
I code Delphi for a living, so I'd hate to think how much trouble other people would be having.
I left a message on SourceForge about it... now there are a whole lot complaining about that same thing, but no answers.
I'm a little annoyed about this, because I had a look at the code, and I saw a lot of things I could have fixed for very little effort. (Eg, they are auto-creating all the forms in the project at application start-up. That's a bad thing to do, but pretty easy to fix.) I wanted to help, but I couldn't.
I'm sorry, but WTF??!? It took me two or three reads to work out what "time-preference characteristics of consumers" meant. Why do people have to dress up what they mean as some kind of buzz phase?
(I was actually looking for that page when I wrote my original response, but I couldn't find it.)
Yes.
Delphi does have pointer arithmitic as well (I have no idea why I said it didn't), but you never use it.
As for garbage collection: I'm not too sure abotu that. Delphi does if you use Interfaces/COM, but it doesn't otherwise. Java and Python do, but I'm not at all sure about SmallTalk.
I've just bothered to start reading stuff about .Net - I thought it was just MS marketing speak. There's actually some pretty cool stuff in there once you get past all the crap. (I've written something at http://www.kuro5hin.org, but it was in the moderation queue when I posted this)
Anyway, reading this guys writing makes me even more impressed with .Net (although did anyone else get the impression he loves C++ even more than most Linux hackers?). For instance:
Pretty cool, hey! Especially the maintaing of session state across different machines.
This guy just doesn't listen. Even MS (now!) admits that VB is only object based. However VB7 (which is the .Net version) is a proper OO language. It has proper inheritance and encapsulation - and it even does exception handling.
I don't. There is no need for pointers in most coding, and generally the parts where pointers are used are the most bug prone. Java, Delphi, Python, Smalltalk - all proper OO languages and none have pointers. (Not sure about operator overloading)
The biggest worry is, of course, this:
The funniest thing his his whole speil:
Now this poor guy was (I believe) none other than Bertrand Meyer who (while he doesn't know much about open source software) does know his stuff when it comes to high quality software engineering. For the author of this piece not to know who he was, and to claim His language looks similar to C#... is pretty dumb. Since this guy seems to think he is some kind of technology guru, I would have expected he would have heard of Eiffel. It's not particually similar to C#, btw - have a look at design by contract for a start.
Actually, most non-commercial component developers are encouraged to use the MPL, not the GPL, because (a) it gets around the issues the GPL has linking to (the non-GPL'ed) VCL and (b) many major component suites (eg, Winshoes) are MPLed, so it makes linking to them easier in commercial software.
For instance, both the USSR and Czechoslovakia (neither of which exist any more) still have su and cs, and yet Serbia (which is recognized by the UN) doesn't have one.
The "USA Minor Outlying Islands" is um, yet no one would even claim that is a country. Same with the British Indian Ocean Territory (io). Greenland (a territory of Denmark) has gl
As you can see, having a Country Code means nothing in terms of being a country or not.
All logged in users can look in the story queue, and vote if it should be posted or not.
It's working fairly well at the moment, although there is occasionally a shortage of discussion on stories.
You can also make editorial comments when you are voteing, which are separate to the "content comments"
Absolutly no doubt at all. (assuming the "close enough" licence really is like the GPL - I haven't read it)
They are not distributing the software, and there is nothing at all in the GPL that say they can't charge for the use of the software, because they aren't redistributing it.
I don't know if this is fair or not - but that's not the point. It is allowed by the GPL.
I personally believe there is room for another, new licence that is designed for online apps like that.
I've used Oracle 7, MS SQL Server 7 and Sybase Adaptive Server 11.somthing
Sybase can't do left outer joins and then filter the results with a "where" clause. That is stupid.
There's some stupid limitation about how you can't pass blobs as parameters to stored procedures (Can't remember the details, though).
I'd choose MS SQL Server 7 over Sybase anyday (even though NT bogs down pretty quick). ANSI 92 SQL is wonderful to work with, and the client side tools are great (compare MS Query analyzer with having to look at Sybase or Oracle query plans by hand).
Oracle is great, too, of course, although I did find a bug once. In v7.2 you can't use nested result sets in queries in stored procedures. It was fixed in 7.4, though.
I think SOAP mandates the use of HTTP, though.
Other things (like ICE) would work well over BXXP, though.
I'm in Australia, so I feel your download pain. 6 Meg (in the case of Mozilla) isn't big at all for a major piece of software, though.
Don't forget that the Mozilla designers were (from the start) working on cross platform compatibility, so they couldn't rely on non-standard system libraries like the JPEG decompression lib you refer to in RiscOS.