topbanner_forum
  *

avatar image

Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
  • Sunday March 22, 2026, 9:13 pm
  • Proudly celebrating 15+ years online.
  • Donate now to become a lifetime supporting member of the site and get a non-expiring license key for all of our programs.
  • donate

Recent Posts

Pages: prev1 ... 190 191 192 193 194 [195] 196 197 198 199 200 ... 364next
4851
General Software Discussion / Re: In search of ideal backup utility
« Last post by f0dder on August 21, 2008, 09:55 PM »
I still haven't found a backup application I like, and I'm still pondering whether I should roll my own. Would be a tedious job though, so I'll keep an eye on this thread.

A word of advice, though: don't even consider Acronis TrueImage for backup. It's super and great and lovely for imaging, but it's backup engine sucks sucks sucks. And it's unreliable, in case there's an error while backing up, you risk ruining your latest incremental backup.
4852
Let me guess, it uses ToolHelp32 to enumerate all processes, then does SetProcessWorkingSetSize(hProcess, -1, -1) on each one of them?#1

Basically useless then, although not as über-retarded as doing alloc-all-memory. It does still mean, however, that all processes will have their working sets trimmed, which very likely means flushing stuff to your pagefile. Including the apps you are currently using, which will then have to re-read their stuff back from the pagefile.

Windows does automatic process-trimming when it needs to, and it tries to only trim the least-recently-used pages of memory... which should generally mean a smoother ride.

If you need more free memory, buy more RAM. Don't fall for placebo tricks.

edit: #1: I was close - it uses EmptyWorkingSet(), which is the same as the SetProcessWorkingSetSize trick. The application is written in VB6, btw.
4853
General Software Discussion / Re: Name The Best UNZIPPER!
« Last post by f0dder on August 21, 2008, 09:42 PM »
Dirhael is right, and I am once again guilty of not expressing myself clearly. If I am unpacking a new version of a Nirsofer program, I will of course NOT want the unzipper to rename the exe file, but to replace the old version. But if I am extracting dozens of zip files containing both icons and readme files, I would not want it to replace the readmes, but to rename them. Of course various icons too can have similar names, but that is another problem. So in general it is the readme file only, that is important for me to rename. I am sorry I didn't point this out more clearly. I guess I should just learn to take the time to adjust the settings before extracting, but is so common to name the readme just that, readme, so I was expecting some author already had come with the feature to rename readmes only.
So... you don't mind overwriting the .ico files from one pack with those from another, but want to keep the readmes? O_o

Imho you'd be better off extraction each icon pack to it's own subfolder (very easy to do with winrar, simply select a bunch of archives, right-click and "Extract each archive to separate folder".) Yes, you then do have the "problem" that you need to navigate to subfolders to check the icons, but perhaps what you really want is a decent icon/graphics manager instead?

I've used RAR all the way back from the DOS days, and I think it was the first app I purchased when I finally got a VISA card. I personally don't have need for any other archiving tool, and I find WinRAR has the least cluttered interface and the most useful command-line switches as well. And while it's proprietary and shareware, there is full sourcecode available for extracting archives, and unrar has been ported to a lot of operating systems. It's also pretty fast and compresses pretty well :)
4854
Living Room / Re: What is a mouser? What does it mean to mouser someone?
« Last post by f0dder on August 21, 2008, 09:35 PM »
Try mentioning his progress on his school work. He will INSTANTLY leave the conversation without a word

He will INSTANTLY do a mouser (translation by Phil  :))
I'd say "He will INSTANTLY mouser you" - I don't think Phil's form is a verb? (but hey, I could be wrong, I suck at grammar :))
4855
General Software Discussion / Re: How to clone large HDDs?
« Last post by f0dder on August 21, 2008, 09:33 PM »
My Acer laptop HD (5 years old, IDE 40GB) is showing signs of wear. It seems to have problems reading some areas of the disk and gives me a "blue" screen followed with a major crash, at times. I have an 80GB IDE portable drive which I could swap (2 years old).

It should be noted with laptops that if you set a 'Harddrive password' in the BIOS, it's a good idea to remove it BEFORE you clone the drive.
Shouldn't be necessary - if you can boot the system and the disk password is entered, you'll be getting the right data. AFAIK, it's not like you're going to get scrambled data if you don't have an authenticated disk; you will simply be denied access to it. ATA disk passwords != full-disk encryption :)
4856
General Software Discussion / Re: Your most used SPECIAL programs
« Last post by f0dder on August 21, 2008, 09:30 PM »
40Hz: is LilyPond, by any chance, built ontop of TeX?
4857
I've seen all of the above, regularly. It's sadly a very imperfect world, and, even if you don't have problems directly because of bad reg entries, your Windows install has to deal with any increased size. And even if you discount any delays in Windows, you can't deny the bigger it is, the longer it takes to search.
-mikiem
Searching the registry is very efficient - there's only really two registry problems that should ever need fixing. One is fragmentation, the other is with apps making "bad changes" to some of the more complicated parts of the registry (COM stuff, whatever) that can cause other applications to fail. But leftover bloat in itself really is never a problem.
4858
General Software Discussion / Re: What would your ideal Operating System be like?
« Last post by f0dder on August 21, 2008, 09:20 PM »
1 ) Pervasive metadata — the OS should provide not only solid metadata handling per file, but support *extensible* metadata mechanisms for any file. Tagging, file usage and history and discoverable information has to be a core OS feature, not something tacked on by 3rd-parties. The OS should provide a core search facility built robustly on this metadata, again not some proprietary 3rd-party. And this data should be accessible to the cloud through design.
-nontroppo
NTFS sorta has this, through alternate data streams... but it's in no means integrated or usable (well, there's a few standard things like "comments" and such, but meh).

2 ) As a consequence, folder hierarchies should lose predominance and smart folders should pervade. No OS is where I want it to be (r.e. metadata and smart folders) on this.
-nontroppo
I'm not sure if I agree on this... I find well-structured folder hierarchies easy to navigate, and they're fast and efficient. For metadata based navigation, you either need very smart indexing, very smart caching, or you will suffer abysmal speed and/or bloat. And you need to be very good at tagging your files for something like this to be useful, imho... (yeah, there's content-based search, but then you do need those huge index files).

3 ) Delta versioned file system. Again, this should be core OS territory (even as much as I love Filehamster!), configurable per file. The interface should allow simple searching for a file through time, and apps should allow version comparisons easily (i.e. the OS API should enforce this). Time Machine is the closest so far (great UI), but I want underlying filesystem support which HFS+ doesn't have.
-nontroppo
ZFS has this, iirc, and it's a good idea. But I see problems with it - people would feel that it's a substitute for backups. And while versioning is cool, you still need those pesky backups :)

5) Core support for the coming GPU revolution. I do a bunch of DV editing, and harnessing the GPU as a general purpose device would rock. I don't want a proprietary 3rd-party to do this, I want it pervasive and universally offered by the OS. Better support from multiple CPUs goes without saying, but it is depressing to see high-core machines having cores sitting idle.
-nontroppo
Get NVidia to allow people to use the CUDA interface for free, and get the other companies to use it. CUDA doesn't even need to be opensourced to do this, it's "just" the API specs (and perhaps a few internals-style things) that needs to be fully documented.

I agree more or less with other of your points, though :)
4859
Back when I used a paging file, I used to set a "reasonable lower bound", around 1gig, based on how much would ever go in there. But I let windows manage the upper bound, just in case I would get into a silly extreme situation where it was necessary. But of course with todays hard disk sizes, you might as well just make a fixed 4gig pagefile :) (or grab enough physical RAM you don't need one in the first place :) :)).
4860
OK, I've taken a cursory look at the source. The crash happens on line 151 in MainForm.cs, upon a call to StartHook(this.Handle); - my guess is that this hook is in a 32bit native DLL, which obviously can't be loaded when dotNET jits CircleDock for a 64bit target CPU. By creating a build configuration for "x86" (instead of "any CPU") and copying some files around manually, I was able to get CircleDock running (although there's some other problems, but we'll focus on those later).

Now, it's showing that I'm not exactly super familiar with dotNET :-[ - I have trouble seeing how the various DLL files are generated; does Orbit.Hook.dll , for instance, have corresponding source files in the project? I only really see it referenced in the source via a DllImport statement, so I'm wondering :)
4861
Can you program in C# and do you have Microsoft Visual C#? If you do, I can try setting up a special space somewhere for you to download it.
-VideoInPicture (August 18, 2008, 05:43 AM)
My experience with C# is limited, but the language syntax is similar enough to C++ that I don't have trouble reading it - and the dotNET framework seems to be reasonably easy to follow.

So yeah, I could probably help with debugging :)
4862
Hm, I don't know if there's such a thing as 32-bit compatibility mode... at least not with the right-click property stuff. And I don't know if there's a way to tell the dotNET jitter to jit for 32bit instead of 64bit.

All I can see is that the crash happens in the CircleDock.MainForm constructor, near the end of the method... debugging dotNET code without source is a bit tedious :)
4863
Weird - I still can't grab from the wikidot URL, even tried with wget to make sure I wasn't getting a cached copy of the error page. Anyway, I got the attached version from your post, and it still crashes.

But I'm sure we'll eventually be able to sort things out :)
4864
PS: The file does not exist. for the new download URL.

Fixed. It's working now.
-VideoInPicture (August 18, 2008, 03:23 AM)
I'm still getting the "File does not exist" error when trying to download...
4865
It's a good idea to always use fully-qualified paths when opening files - there's several things that can cause the current working directory to change, and then opening with relative paths break :)

PS: The file does not exist. for the new download URL.
4866
I will create a source-forge account to host the code since the whole solution is now around 40 megs with the icons I use for testing. I can't host this on wikidot because of file size limitations. I'll post a message when I get the source-forge account set up.
-VideoInPicture (August 17, 2008, 09:34 PM)
40 megs? That sounds like a lot! You should probably strip it down to only the source files etc. that's necessary to build, and perhaps a minimal set of images - keep the full image/icon set only in the binary release.

Oh, and make sure to clean out any temporary or auto-generated files, stuff like .pdb and .pch (dunno if those are used in dotNET projects?) can grow quite large.

Now I'm curious if a 64-bit Vista user will have the same problem...... Hmm. I'll have to wait and see.
-VideoInPicture (August 17, 2008, 09:34 PM)
Both versions crash on Vista64 as well.
4867
Crashes in the same way (and location?) as previous build, but the new version doesn't render any icons, only the main dock background.
4868
An unhandled exception of type 'System.BadImageFormatException' occurred in CircleDock.exe

Additional information: An attempt was made to load a program with an incorrect format. (Exception from HRESULT: 0x8007000B)

Happens in CircleDock.MainForm..ctor, after the third-last call instruction (happens at address 0000042C , but dunno how useful that is to you, since you're on 32bit and the JITer has JITed for 64bit).

Just like mouser, I get the crash after the main form has been painted. XP64/SP2, running with admin privileges, framework v3.5 installed.
4869
fSekrit / Re: Search as you type (or incremental search) in fSekrit
« Last post by f0dder on August 17, 2008, 02:31 PM »
Might be doable - I've considered it already. It will probably require a fair amount of code to implement, but I think it might be worth it.
4870
Deozaan: weird, I ran 32bit XPSP2 with 2 gigs and no pagefile for years without problems - except for one or two games where I had to temporarily re-enable the pagefile.

As for system running slower after re-enabling pagefile, grab sysinternals' "contig" tool as see if the file is fragmented or not.
4871
General Software Discussion / Re: What would your ideal Operating System be like?
« Last post by f0dder on August 15, 2008, 01:14 AM »
"Remember the old style of Christmas lights where you had a big long string and if one bulb burned out the whole thing burned out and you had to go through each one and find out which single bulb failed? That's Microsoft."
-4wd
Cute analogy, but I'm afraid it doesn't really fit. Yes, you can use it as a coarse comparison between microkernels and the rest, but it isn't spot-on. If you look a recent operating systems, the lines begin to blur - NT (and linux, for that matter) are pretty monolithic kernels, but you can still have individually failing parts that are able to restart... especialy with Vista, which runs graphics as a "relatively individual part".

Makes one think pretty hard about whether to go for pure computer science ideals, or for stuff that works. Pure microkernels are lovely conceptually, but have too much overhead (imho). Striking a balance between concept and implementation is the thing to do.

The Amiga system rocked - we had multitasking way before Win3.1 and Desqview arrived (well, that's how I remember things, being in europe and pre-teen without a spending budget). Amiga had really cool hardware layout with specialized chips, which meant the relatively underpowered 68k hardwared (along with the special chips) did better than the x86 machines at the time. On the other hand, there was no proper memory isolation etc. in the lower-end 68k chips, so when something went wrong, you got a total guru meditation. Amiga was the system that taught me to <hotkey-save> after entering a sentence.

We need to shun sentimental memories (but still remember the past - good as well as bad), be as objective as we can about current affairs (drop the fanboyism and admit flaws of our favorite OS... there's a lot of them for all of them, and we won't agree on everything :)), et cetera. I feel my ideas are relatively agreeable in general, but I know that not everyone will agree - there are pretty strong opinions in this field. It's probably even worse than politics.

Anyway, I wish that people would start writing portable software (and not that autoconfig junk), and that we'd see things like ZFS and XFS ported to Windows... NTFS is cool but pretty ancient, and I'd love to see some realistic head-to-head benchmarks.
4872
Hey Sugar (oh, that sounded corny >_<),

nice to see you onboard!

I sometimes wonder what all the "quiet members" here do - we're obviously a bunch of regulars who make a lot of noise. That's pretty enjoyable and all, but I can't help thinking how many "sign up just for the benefits", and how many actually frequent the forums - lurking or active.

It's always nice seeing somebody post. Doesn't matter if a member has something to say about software or tech or something completely different, it could be a shot of your desktop, or perhaps the view outside your window, or even a single post in this thread.

The great thing about this site has always been the community thing - developers, users and people in general mixing up and having an easy time together. Oh, I might be going on a bit too much on a happy vibe, but it felt right - I hope you're doing OK :)
4873
How many Linux distros are there? :huh:
Too many :)
4874
General Software Discussion / Re: What would your ideal Operating System be like?
« Last post by f0dder on August 14, 2008, 02:47 PM »
These are just some of my thoughts. I've tried to strike some balance between idealism and realism. I wouldn't mind getting rid of all legacy (hardware as well as software), and it'd be just wonderful if you didn't have to pay for anything and everything had source code available etc... but I do live in the real world. Sorta :)

POLITICS
  • Doesn't necessarily have to be (beer) free, but price needs to be right. Considering that this should be the OS running "everywhere", price could be low. Money helps the world go round.
  • I'd prefer kernel and drivers to be opensource. At the same time, I appreciate that R&D costs means that parts of especially graphics drivers might have to be closed.
  • ONE distribution, a sense of direction, and no great egos on the core team.

KERNEL
  • Modular, with ability to load/unload modules as necessary. Keep core kernel LEAN_AND_MEAN, without ability to compile modules directly into the kernel (enforce modularity). Design this to have minimal overhead.
  • Keep ABI stable through major versions. Ability to use new drivers/modules without having to reboot for a kernel upgrade rocks.
  • Kernel drivers for basic video functions (accelerated 2D, mode switching, etc.) to allow for graphical applications without the need for a full windowing system (think embedded systems, games, ...). Windowing system would obviously also use this interface, allowing windowing system and other graphical apps to coexist.
  • Binary interface for getting/setting system info, instead of generating/parsing textual /proc-style pseudofiles.
  • Fine-grained ACL-based security.
  • Decent native API that's powerful yet comfortable to program for.
  • Legacy sucks, but we can't avoid it. Support different subsystems (POSIX, WIN32?) that builds ontop of native API.

FILESYSTEM LAYOUT
  • Well thought out and uncluttered. No dumping of a zillion files in monolithic /bin, /etc. Each "major" app will get it's own subfolder(s) - still split binaries and config files in separate folders for ease of backing up, though.
  • Registry. Binary format. Not a single monolithic registry, though - there will be a registry (or "hive") for system settings, per-application global registry, and per-user registries (not per-user per-app registries though). This means backing up config is easy, you don't have to deal with a zillion different config formats, and you have a single efficient API for getting/setting config information. Not to mention secure, atomic, transactional config writes.
  • One layout to rule them all. No /usr, /usr/local, /var, et cetera. /bin (and per-app subfolders) for programs and static data files, /cfg for configuration (subfolders probably not necessary, since there should be one registry/hive file per app), /home with per-user subfolders. /core (or whatever name) with core OS files, possibly subfolders for drivers, boot files, et cetera. System/kernel config would still be registry/hive file in /etc.

OPERATING SYSTEM
  • "Roles" or "profiles" at install-time - minimal, desktop/user, desktop/developer, server, manual package selection. Manual selection should be able to start with one of the other roles for default selections so you don't have to go through everything manually. Imho one base OS image can fit everybody, if modular enough.
  • Powerful package manager - one core API, with possibility for both graphical and cmdline clients. Distributed repositories to balance server load, cryptographic package signing. Package system should handle both binary and source packages, so you can either grab vanilla builds or customize to your needs, but using one interface for everything, and never ending up with a screwed-up system.
  • ONE desktop environment to rule them all, but make it customizable enough that most people will be happy. Less headache for developers, less headache if you need to use somebody else's machine.
  • Default distribution media needs to come with an assortment of app binary packages. Not everybody has internet connectivity, and not everybody who does have broadband.
  • Keep the thing user friendly, but without dumbing-down that pisses power users off. Imho OS X dumbs things down, while linux distros still require a bit too much tinkering for "normal people" use - some are pretty friendly now, but nobody should have to drop to a shell or edit config files to get a printer, multi-monitor setup etc. working.
  • Integration. Everything should feel like "it belongs together" and interoperates properly. This does mean standard choice of browser/file manager, but there should be hooks so power-users can perfectly integrate their own choice of browser etc.

And as a separate point: documentation. Proper, maintained, indexed, structured, professional, up-to-date documentation for everything - kernel, libraries, applications, et cetera.

EDIT: a thing I forgot - portability. But not over-the-edge "run frigging everywhere" mentality... limit to "sane" hardware, probably setting the bar at at least 32bit architectures with proper memory-management/protection hardware. Possibly forget about supporting "big iron" machines, I'm not sure one OS can really fit everything comfortably. And portability also shouldn't mean you can't take advantage of CPU-specific optimizations where they make sense.

Oh, and another thing: designed ground-up with hypervisor and virtualization support. This can be done without much overhead, and would mean internet-facing applications (browsers, servers, you name it) can run in very secure sandboxes, without having to code a boatload of per-application sandbox code.
4875
Living Room / Re: Skimp or splurge?
« Last post by f0dder on August 14, 2008, 12:40 AM »
Mmmh, gaffer's tape. And it's companion, ether!
Pages: prev1 ... 190 191 192 193 194 [195] 196 197 198 199 200 ... 364next