mIRC Home    About    Download    Register    News    Help

Print Thread
Page 1 of 2 1 2
$calc Limit bug #41328 12/08/03 09:29 PM
Joined: Dec 2002
Posts: 60
C
Cyber_Axe Offline OP
Babel fish
OP Offline
Babel fish
C
Joined: Dec 2002
Posts: 60
i seem to have found a bug in $calc which i have only just noticed when it reaches a certain amount of numbers it just displays zeros instead of the actuall number heres how i found it i found it by calculating bytes upto a yotta byte for example 1 byte tymes 1024 = 1kb 1kb times 1024 = 1mb heres the exact figures to show you the pattern that emmerges

1024
1048576
1073741824
1099511627776
1125899906842624
1152921504606846976
1180591620717411303000
1208925819614629174000000

then i times the yottabyte by 1024 as a test so proove my hypothosis

1237940039285380352000000000

as you can see all teh figures after the 19th are 0 and

the actual answers for the last 3 numbers should be

1180591620717411303424
1208925819614629174706176
1237940039285380274899124224

Re: $calc Limit bug #41329 12/08/03 11:45 PM
Joined: Feb 2003
Posts: 42
G
GregMo Offline
Ameglian cow
Offline
Ameglian cow
G
Joined: Feb 2003
Posts: 42
Yeah, I suppose you could call that a bug. What it should do is to return an error or else return it in scientific notation.

Reminds me of this bug:

//echo -e t: $asctime(2999999999)

Khaled didn't account for the outer limits of the variables in a way that's consistant with most languages. After "Mon Jan 18 22:14:06 2038" mIRC will no longer be able to tell what time it is since 2147483647 is the upper limit of the variable that's used. This is also the issue with an uptime over 49 days. Same size variable is used with both, only one's signed and the other is unsigned. Then again, what are the chances that Khaled will still be coding mIRC in 30 some years? ;-) For that matter, it's not often you see a Winblows system with over 49 days of uptime tho I have had mine go over 9 weeks ;-) That's one that'd I would love to see fixed tho as when it does happen all timers also stop and it's a PITA to have to restart them all.

Cheers,
GregMo

Re: $calc Limit bug #41330 13/08/03 03:33 AM
Joined: Aug 2003
Posts: 1
S
scrolli Offline
Mostly harmless
Offline
Mostly harmless
S
Joined: Aug 2003
Posts: 1
I noticed the same problem after my winchokes was up over 49 days for the first time.

Re: $calc Limit bug #41331 13/08/03 03:45 AM
Joined: Dec 2002
Posts: 2,809
C
codemastr Offline
Hoopy frood
Offline
Hoopy frood
C
Joined: Dec 2002
Posts: 2,809
Thats actually a limitation in Windows, not in mIRC.

Re: $calc Limit bug #41332 13/08/03 05:09 AM
Joined: Feb 2003
Posts: 42
G
GregMo Offline
Ameglian cow
Offline
Ameglian cow
G
Joined: Feb 2003
Posts: 42
Yeah, it's because the BIOS clock flips over and mIRC doesn't handle it well. I don't know it cold, but I do know that there's something in the BIOS that keeps track of this flip, it's just that mIRC isn't coded for it, and so it thinks the uptime is 0 once 49 days, 7 hours, and some odd minutes/seconds have passed. This also kills all timers since, I'm guessing, mIRC timers are based off of the BIOS time. I can't recall for certain, but if memory serves when this happens mIRC just flat out kills them all with no warning or other indication.

However, other than the uptime and timers, mIRC still runs fine after that point.

Cheers,
GregMo



Re: $calc Limit bug #41333 13/08/03 05:39 AM
Joined: Feb 2003
Posts: 42
G
GregMo Offline
Ameglian cow
Offline
Ameglian cow
G
Joined: Feb 2003
Posts: 42
Well, depends on how you look at it really... Win32 does use a 32-bit interger to track time, but it does filp it at it's max and store this filp in another variable. I was incorrect in thinking it was the BIOS, but this flipping is well known to those that track time as this issue was really big on Win9x as they didn't have this second variable to track it. There is a fix for it tho*. Most of your win32 uptime trackers are clueless of this too, but there are those that read it right. dduptime** is a great one, imho, which also happens to come with a mIRC formatted DLL for reading the time out of it.

Cheers,
GregMo

* http://support.microsoft.com:80/support/kb/articles/q216/6/41.asp&NoWebContent=1
** This is a hard one to find because the author's site is down. If ya want it, should be able to find it on google. If not, give me a shout, I have it around here somewhere.

Re: $calc Limit bug #41334 13/08/03 01:59 PM
Joined: Dec 2002
Posts: 2,985
Watchdog Offline
Hoopy frood
Offline
Hoopy frood
Joined: Dec 2002
Posts: 2,985
For that matter, it's not often you see a Winblows system with over 49 days of uptime tho I have had mine go over 9 weeks ;-)

Probably because most people who run desktop computers (the bulk of Windows users) don't want to run up their electric light bills to prove a point. Money doesn't matter to me which is why I can successfully run one particular machine that has a nice quiet power supply for months on end. This machine runs IIS, Photoshop, umpteen browsers and WS FTP (web development purposes) and also hosts a mIRC bot I run. If you have problems with stability then it's a reflection on you, not your computer.

Re: $calc Limit bug #41335 13/08/03 02:10 PM
Joined: Dec 2002
Posts: 127
F
feud Offline
Vogon poet
Offline
Vogon poet
F
Joined: Dec 2002
Posts: 127
and that's watchdog with another one of his chart-topping, self-adoring worthless posts.

assuming you use power saving technology to have your monitor turn off after a reasonable amount of time, people won't be "running up their electric bills to prove a point". of course your electric bill will be higher, but honestly, if you can afford a computer, you can afford to pay the bill for the electricity it uses.

also, "if you have problems with stability then it's a reflection on you, not your computer." what are you talking about? yes, some things that people do can contribute to their systems crashing (and many times, they do), but to say that it's a reflection of the user is a little misguided. but, the point of your post was to praise yourself and your lack of money troubles, i suppose that doesn't really matter.


------
deep down, i'm really superficial.
Re: $calc Limit bug #41336 13/08/03 02:31 PM
Joined: Feb 2003
Posts: 42
G
GregMo Offline
Ameglian cow
Offline
Ameglian cow
G
Joined: Feb 2003
Posts: 42
Yeah, only took me reading two of his posts to gather that he's something of a whiner. I mean, anyone who's been around awhile knows that win9x and ME were pathetic for stability. XP is a bit better, but not perfect by any stretch. The problem isn't so much Windows itself, but rather how it handles applications, drivers and fault recovery.

That aside, for all the Windows supporters that want to argue such as Watchdog, setup two identical systems, one running off your choice of Win32, and one running FreeBSD, and configure them to get equal usage, similar tasks, etc, and for uptime the fbsd box will win out every time. For those that whine, "What's the big deal about uptime"; uptime is a direct indication of stability of the OS, plain and simple, no two ways about it.

Cheers,
GregMo

Re: $calc Limit bug #41337 13/08/03 03:00 PM
Joined: Dec 2002
Posts: 395
M
MonoSex Offline
Fjord artisan
Offline
Fjord artisan
M
Joined: Dec 2002
Posts: 395
I agree that Windows isn't perfect, but if you have some serious stability issues, it's either hardware or drivers fault. I started using XP in 2002 and have yet to see my computer crash.


Quote:
uptime is a direct indication of stability of the OS, plain and simple, no two ways about it.


I do not agree, such high uptimes indicates that the Admin is too lazy to upgrade his software after security fixes are released (yes, EVEN fbsd has to be restrted after kernel recompilation). smirk

Re: $calc Limit bug #41338 13/08/03 04:43 PM
Joined: Dec 2002
Posts: 2,985
Watchdog Offline
Hoopy frood
Offline
Hoopy frood
Joined: Dec 2002
Posts: 2,985
Pick out a perfect OS?

*Watchdog listens....

There isn't one? Okay, issue settled apart from one small detail that you have overlooked:

setup two identical systems, one running off your choice of Win32, and one running FreeBSD, and configure them to get equal usage, similar tasks, etc, and for uptime the fbsd box will win out every time.

<whine>

If you set up two identical systems then they will both have the same OS as well as the same motherboard, processor, video card, memory, hard disc, etc. If you change any single one of those items then the boxes are NOT identical. RE: OS's, I've used all the popular ones, Win 9X, Win 2000/XP/NT, FreeBSD, Deadhat, WinLinux and Corel Linux, OS2, DOS and Crapple OSX. I'd lump Win 9X together with OS2 for dependability. They work for a while, die then after rebooting all is normal again. Linux, well, it's okay and more popular on the desktop than FreeBSD because there's more desktop apps for it. FreeBSD, Win2000 and WinXP have the same levels of stability and reliability. No amount of Bill Gates bashing will prove otherwise. I've seen both OS's last a long time. There are IRC servers running BSD that have been up for over a year, likewise and closer to home, where I work there are NT powered desktops running PABX management software that have been up for over 4 years - what can run for 1 year will probably run for four so there's no valid argument that one is better than the other. It comes down to personal choice.

</end of whine>

Re: $calc Limit bug #41339 13/08/03 05:17 PM
Joined: Feb 2003
Posts: 42
G
GregMo Offline
Ameglian cow
Offline
Ameglian cow
G
Joined: Feb 2003
Posts: 42
Quote:
I do not agree, such high uptimes indicates that the Admin is too lazy to upgrade his software after security fixes are released (yes, EVEN fbsd has to be restrted after kernel recompilation).


Not so, my unschooled friend. A considerably large portion of kernel updates deal with exploits that require the attacker to have local access to the machine. Your average server doesn't have many users that have local access and those that do have it, it's the admin's job to ensure that they're watched and at the same time, the kind of people that you can trust enough to not have to watch.

With FreeBSD, or any UNIX variant, you don't have to reboot it every time you upgrade something, only when rebuilding the kernel, and the kernel has nothing to do with the web server/client, the email server/client, or most any other server/client as far as rebooting is concerned.

<sarcasm>
On the other hand, with Win32, you have to reboot when upgrading notepad.
</sarcasm>


Re: $calc Limit bug #41340 13/08/03 05:36 PM
Joined: Dec 2002
Posts: 3,127
P
ParaBrat Offline
Hoopy frood
Offline
Hoopy frood
P
Joined: Dec 2002
Posts: 3,127
GregMo & feud: It really IS possible to disagree politely and without resorting to insults. Try it please.


ParaBrat @#mIRCAide DALnet
Re: $calc Limit bug #41341 13/08/03 05:37 PM
Joined: Dec 2002
Posts: 2,809
C
codemastr Offline
Hoopy frood
Offline
Hoopy frood
C
Joined: Dec 2002
Posts: 2,809
When I upgrade my webserver on my Windows machine, I don't reboot. When I upgrade my mail client, I don't reboot. When I update the Windows kernel, I do reboot. How is that different than with FreeBSD?

Re: $calc Limit bug #41342 14/08/03 06:39 PM
Joined: Jun 2003
Posts: 384
D
DekuHaze Offline
Fjord artisan
Offline
Fjord artisan
D
Joined: Jun 2003
Posts: 384
With all due respect, Watchdog, I'm gonna have to disagree with you on one or two things:

Quote:
Linux, well, it's okay and more popular on the desktop than FreeBSD because there's more desktop apps for it.


This is untrue. BSD has 90% binary compatibility with Linux, meaning it can use Linux apps as well as iits own native applications.

Quote:
FreeBSD, Win2000 and WinXP have the same levels of stability and reliability.


Also untrue. The BSD's, as well as the other Unix clones, have a reputation for being much more stable than that of Microsoft's offerings which is why they tend to be used more in mission-critical systems as well as on servers. Afterall, how many IRC servers do you know that run atop of Windows boxes? I imagine the number to be very small.

Quote:
No amount of Bill Gates bashing will prove otherwise.


You don't need to bash Gates to prove that BSD, and other Unix clones alike, are way more reliable and stable than Windows. The proof is all around you in the form of various servers, shell providers... etc...

Quote:
...what can run for 1 year will probably run for four...


Whilst Windows 2000/XP could probably manage an uptime of a year (without system updates (replace a system file and it wants to reboot. Windows Update being a prime example.)) I highly highly doubt it could manage 4 years. Maybe if the system was under little stress and performed just one task it MIGHT manage four years but I remain skeptical. BSD and other Unix clones have proven they can stay up for years at a time under varying system loads performing multiple tasks. 'Google' being an excellent example.

Anyway, the purpose of this post is not to flame (so plz don't take it to heart) but just to make a couple of points. Oh... and to help this thread stray even FURTHER off topic :P

Re: $calc Limit bug #41343 15/08/03 02:43 AM
Joined: Dec 2002
Posts: 2,985
Watchdog Offline
Hoopy frood
Offline
Hoopy frood
Joined: Dec 2002
Posts: 2,985
Afterall, how many IRC servers do you know that run atop of Windows boxes? I imagine the number to be very small.

Because Windows hosting would chew up their budgets for a start. How many IRC networks could afford the hundreds of thousands of dollars it would require to run 10 - 20 NT servers running commercial software? I would say none at all apart from those that get such hosting either subsidised by ISPs (a big ask) and/or paid for directly by subscription to an ISP (once again a big ask as you are asking a business that is responsible to shareholders to give hundreds or thousands of people a free ride at their expense).

That said, I don't give a damn about reputation. Reputation is more often than not based on urban myth rather than legend. On that note if my network can run NT boxes for 5 - 6 years with Microsoft Exchange, Chatspace and ConferenceRoom with no problems at all then, aside from affordability, I cannot see why that can't work elsewhere. It's not like a pocketfull of networks around the world are performing miracles.

The rest of what you said I don't disagree with because you just said what I did, albeit a different way.

Re: $calc Limit bug #41344 15/08/03 02:50 AM
Joined: Dec 2002
Posts: 1,527
_
_D3m0n_ Offline
Hoopy frood
Offline
Hoopy frood
_
Joined: Dec 2002
Posts: 1,527
yes and i can see the point of your post having lots to do with a $calc limit bug ..... man some ppl just love to debate and argue


D3m0nnet.com
Re: $calc Limit bug #41345 15/08/03 02:56 AM
Joined: Dec 2002
Posts: 2,985
Watchdog Offline
Hoopy frood
Offline
Hoopy frood
Joined: Dec 2002
Posts: 2,985
It all started with GregMo. It's not my problem that people object to his inaccuracies being corrected. You'll notice if you read up, his second post supports this. There's always someone ready to blame problems that lie in mIRC on some hatred that they have for operating systems instead of doing the right thing and discussing the factual technical aspects of it.

On that note that is all I have to say. Anything else would just be repeats of what has already been mentioned.

Re: $calc Limit bug #41346 15/08/03 02:59 PM
Joined: Jun 2003
Posts: 384
D
DekuHaze Offline
Fjord artisan
Offline
Fjord artisan
D
Joined: Jun 2003
Posts: 384
Quote:
yes and i can see the point of your post having lots to do with a $calc limit bug ....


Much like your own post then smile

The purpose of my previous post was merely to correct some of watchdog's corrections. Nothing more. cool

Re: $calc Limit bug #41347 15/08/03 03:20 PM
Joined: Jun 2003
Posts: 384
D
DekuHaze Offline
Fjord artisan
Offline
Fjord artisan
D
Joined: Jun 2003
Posts: 384
Quote:
How many IRC networks could afford the hundreds of thousands of dollars it would require to run 10 - 20 NT servers running commercial software?


IRC networks tend to be a collection of servers donated by different companies. There's no one single network-wide budget. The budget is dependant only on the site that is hosting the server, which, mostly likely, are running Windows boxes. So that suggests to me that they certainly HAD an option to use Windows but didn't because of the Unix-like OS's greater reliability, security and dependability. smirk

As for the $calc fault: mIRC is written in C/C++ (according to Khaled's FAQ) and C/C++ (like Turbo Pascal) has a limit on how high the value of an Integer can be. So I'd say that's a fault with the language itself, not the IRC client. smile

Page 1 of 2 1 2