BASeCamp Programming Blog

As many web developers are aware, there are a myriad of ways to develop websites and interactive web applications today. On the server side, one typically chooses from one of two technology stacks; the “Open Source” and the Microsoft. Disregarding political and license concerns, they really are very much equal in core capability. The Open Source stack typically consists of A Linux distribution as the Operating System, Apache as the Web Server, MySQL as the database, and PHP as the server-side language. These can also be installed to a Windows machine as well, and Apache has modules that even allow the use of .NET technologies via the Mono runtime. The Microsoft stack consists naturally of Windows as the OS, IIS as the webserver, and ASP as the core server-side script. ASP itself, of course, supports a number of languages; you can use JScript, VBScript, or a .NET language. Both of these platforms provide a myriad of tools at your disposal. LAMP allows you to leverage the wealth of Apache modules and use a large selection of programming languages like Ruby, Python, Perl, and of course PHP. The Microsoft Stack makes use of the very powerful SQL Server, and is particularly scalable to large operations. Of course there are variations of these; obviously MySQL could be swapped out for PostGreSQL, or even a remote SQL Server.

Costs

The cheapest hosting plans for most WebHosts use a Linux-based stack. This makes PHP one of the more accessible server-side languages to learn. When I started this sight, I wasn’t really sure how well it would turn out, what I would write about, or how much exposure it would give me or my various programs. After three years, I’m still not entirely sure of that; but at the time, I was working predominantly in Visual Basic 6. I was stagnating, and I didn’t even realize it. When starting the sight, I really had two choices- go with the Linux stack, or the Microsoft stack. My choice of the LAMP stack was done purely for a single reason: it was completely foreign to me. That may seem like an odd reason to choose a technology, but I’m always up to a challenge. I won’t try ot say that learning the “Linux way” of doing things was easy, but it did get easier over time. With the server itself of course I didn’t have full access anyway; just a standard CPanel, and I still do; but understanding Linux, Apache, and MySQL were very helpful in using PHP, which was the language I had to learn to get teh site off the ground beyond a few ugly static webpages.

I basically slogged through learning PHP, in an attempt to create a relatively simple CMS. After getting a basic CMS and some crappy side links started, I decided to redesign the site from scratch. I sketched how I wanted it to look on paper, and then set to work duplicating that appearance using the available web technologies, as well as through the use of Photoshop for the various images I needed. The end result is what you see on the main page; of course I’ve made changes to it since, and added features to the underlying CMS to support new functionality such as listing my youtube videos and different categories of items, but the visual appearance is much the same. I toyed recently with the idea of redesigning it, but decided that it could stay as it is now a little while longer; a redesign is a rather big undertaking, and I like how it looks now.

Since then, I’ve also learned and become quite adept (If I may say so) at C#. This has left me rather- annoyed- when I use PHP, which feels very messy in comparison to what is generally a very clean working environment. Not to mention being relegated to having to debug using echo, which I can’t say I really missed from using GW-BASIC.

I did install a MS Stack locally some time ago, and experiment with it for a short time before deciding to avoid it; I reasoned that if I was to use C# for web development start to avoid working on my PHP site even more. I’ve since changed my mind, however; I’ve decided to install a local IIS-based server and experiment some more with what .NET has to offer on the server-side. I’ve been able to make some pretty fine-looking stuff with WPF and Windows Forms, and I know WebForms as well as the Base-Class Library that is not heavily leveraged on the client side are one of the many areas where my abilities and knowledge can be expanded, so I can’t see why not.

Also, I’ve always thought it a bit weird that my site ran on PHP and I focussed mostly in Windows-based and MS technologies and languages. Though I don’t see a switch over occurring anytime soon.

430 total views, no views today

Important Note: In a system management setting, Or a corporation, this is NOT something I prescribe. managing and dealing with a PC that only you would use is one thing, handling them for others is, well, another thing entirely. In those cases a good AV is required and keeping it up to date as well. (With the possible exception of Linux/BSD, where you still need to be aware of any problems that crop up in the software being used)

Personally I do not use a “On-demand” or background scanner. I do have tools such as MBAM and the like installed which I will run when I notice odd processes in task manager, svchost hogging resources, or general “odd” behaviour from my system. I’ve never felt it was worth the processing overhead; The task of AV scanning takes time, and having it occur on nearly every file access is a rather hefty price. And of course, even the best AV application isn’t going to catch everything, so you need to be cautious anyway, means that, for me, I’m actually safer when I don’t have an AV installed.

When I did have one installed (after a nasty Virut infection on Windows XP to keep me from reinfecting the system using existing executables on my data drive) I actually found that a lot of activity I found suspicious and thought “oh no, I’m reinfected” could be traced to the AV. While their purposes are far more noble and good, I’m sort of felt that AV software is sort of like “fighting fire with fire”; rather than a Virus or malware building a huge root system in your machine, the AV software does. That’s why they all seem to need special software to fully remove. That, and a lot of the systems I’ve fixed for others that they blamed on “Viruses” were in fact caused by the Anti-virus software they were using, which if you ask me is utterly unacceptable. (I’ll say I’ve only seen those issues with one or two “Free” offerings, Mcaffee, and older versions of Norton, though.

Basically, my “protection” amounts to healthy cynicism. Almost all malware infections these days are trojans. So few infect a system by way of things like exploits and “drive-by” stuff that it’s hardly worth the effort to waste time thinking about. More importantly, the first line of defense even for those is the same. After all, in order to download a trojan you need to trust a website, and in order to visit a web page, you need to trust the link. Typically, when dealing with an unfamiliar executable, I’ll just run it. (unless it has a blatantly suspicious name). I might run process explorer and keep track of what the program does as well to make sure there is no funny business. I usually have that running in the background anyway. If the program requests administrator permission- it doesn’t get it. Not at first. This has prevented a good lot of “infections” if you ask me, since you can’t well infect a machine without administrator permissions. Naturally, software installers do need those permissions so depending on the software I will allow. Or, if I’m in a forgiving mood I might just say yes and deal with whatever happens later. When I am infected, I usually catch on rather quickly and am able to either kill the malware processes, or, if they are the type where they autorespawn each other, suspend all the suspicious processes and kill them all at once. Visit regedit and delete the offending entries (generally in the Run key), reboot, and typically everything is back to the way it was. In 7-8 years I was only infected once, and since moving to Vista/7, I’ve had hardly any problems.

An important Note: Typically, “manual” Virus removal is not something that just anybody does. It really requires a intimate knowledge of how Windows software works, the PE file format, and of course a willingness to “get your hands dirty”. At the same time, it really is only an extension of what you should do even when you have a AV installed- keeping a system clean requires constant vigilance and you need to constantly be assessing what possible security repurcussions your actions could have. “Does this application REALLY need admin permissions?” type stuff. More importantly, a single screw-up can cost you dearly. This is NOT something I recommend. Heck I don’t even recommend it to myself. I just don’t like the “cloud” surrounding Windows being inherently insecure to the extent of requiring an AV to function, so I don’t use one.

BC’s tips for AV-free operation

Sometimes your applications will crash; this is pretty much inevitable. Sometimes you’ll need to run Task Manager for various reasons. While there, glance over the processes. Again, this requires a familiarity with the type and number of processes you would typically see running on your machine, so it’s useless unless you are familiar both with your operating system of choice as well as your “normal” software configuration. Things like rundll32.exe showing up in there out of nowhere will make me reach for Process Explorer, where I can determine the “threat” posed by that process.

As I type this, my desktop machine does in fact have a rundll32.exe process running. Which piqued my curiousity. You can use the “Select columns” menu in the view menu of Windows Task Manager to enable columns such as “command line” which can give additional information on the process. you can also use various features of Process Explorer for that same task, or further investigation of a suspicious process, such as examining it’s in-memory layout, stack frames, loaded Libraries, etc. in the case of this particular rundll32.exe, it turns out to be used to launch a function called “GameUXShim” in C:\Windows\System32′ gameux.dll, which according to it’s description, is “Games Explorer”. the parameters, and function name, passed make it clear this is designed to “Shim” an older game to work in the newer Windows 7 environment. Again, familiarity with the Windows System helps here, the compatibility settings provided by the windows shell itself pale in comparison to what is actually provided “under the hood” which involves a massive network of shims, compatibility hooks, and databases on the sorts of the two needed for various games and older applications that were, for lack of a better word, written badly. In this case, it seems to be for “Halo.exe” the executable for the popular Halo game, which I bought and never played hardly until yesterday where I played for 5 minutes and got stuck. Knowing that I played the game, and was no longer, I can safely terminate this process and know it wasn’t malicious, and is required for proper gameplay. Same for a variety of other older games I have. {Edit: As I discovered previously , this gameUX.dll mess was actually caused by something else}

Obviously, this isn’t for people that just want to “do work” on their computer; it’s more useful for people who want to learn about how it works, and I don’t purport it as being safe, or even really that smart. An AV solution is only as good as it’s user, which is a nice way of telling people “you keep getting infected because your stupid, not because your AV sucks”. Usually I can get them to understand.

More critical that what AV a person uses is learning how to use that AV software effectively. You can’t just install an AV and forget about it; they each have their own nuances and settings that you should configure to your unique usage scenario. Learning what causes their various “popups” to appear warning you about things and how severe they are is important, particularly since the way a lot of AV suites present their messages is using skinned messages and gaudy pop-ups with stupid images that depict “virus infections” or other images that are wholly unimportant. Installing an AV and blindly following it’s advice and getting all worked up because it says it quarantined something merely feeds the ignorance, it doesn’t absolve it. It just adds confusing terms. “Quarantine” for example, is just a silly term in a software environment. All it amounts it is a backup folder where the files are moved to. Why? Because AV software has false positives, so it moves it there so that if it turns out that “woops that wasn’t infected” it can be moved back. It’s sort of like the equivalent of a society where, if anybody is identified as “sick”, they are euthanised. (a bit harsh but that’s essentially the analogy as far as software goes). For “cleaning” and disinfecting files, basically at this point the analogy would be that they can cure your cold, but they will have to rip off all your limbs and cut off your ears. (the resulting program very rarely still works as it used to and you need to reinstall anyway). In such a scenario, false positives could be disasterous. Instead of just having a call to your house telling you that your test results were wrong and you don’t actually have the flu, the doctor would have to interrupt your funeral to say “oh, yeah turns out he wasn’t sick”, which usually means he is no longer invited to the reception. As such AV software does what might be done in such a scenario: instead of outright deleting/killing the victims, it moves them to a special holding area, where they are forgotten about and usually deleted anyway, but at least that way if the doctors/AV software balls’d up they can just release them back into society. This only outlines that AV software is far from perfect. using medical terminology like “quarantine” and “virus” and “heal” and “infection” only serves to confuse the issue, since it actually makes people think that the entire area of malware removal is a “profession” like your standard medicine on which the analogies are based. It’s not, certainly no where near the level of the field they have taken the terms from. At this point, Anti-Virus software as a “medical” field is about the equivalent of when we would drill holes in peoples heads to release their inner demons. That isn’t to say it’s useless, just that a lot of what it does is a tad drastic.

470 total views, no views today

First, a warning:

Now that that is out of the way…

One of the nice things about later versions of windows is that you don’t automatically have full control over everything. Some people try to say this is bad, because it is their computer so they should be able to do what they want, but the point they are missing is that the changes to the default security settings is not to prevent them from doing things, but to prevent nasty programs from being able to do anything they want. By definition the settings for a user control more what the programs running under that account can do; and only serve to restrict the user themselves by virtue of them not really being able to do anything that a program cannot do. (If no program can delete a file, that user cannot delete it either).

Sometimes, however, this can get in the way. Stubborn files, for example, might refuse to be deleted. Usually, running a program as administrator clears this up, but sometimes even this doesn’t work.

In particular, a failed Windows update, or an update that doesn’t clean up properly, can leave a mess of files around. Usually these are weirdly named folders in the root of the system drive. A quick search for words like that via google reveals that this is not an uncommon problem. The problem is that nothing can delete these files- you cannot run as an administrator to delete them, tools like unlocker and deleter don’t work either. The cause is that the files weren’t even created by the administrator, but rather by the LocalSystem account under which Windows update runs. (this is required so that the update can update dll files and other files that are in-use, which usually will require a reboot for a myriad of reasons that I won’t get into). The files are supposed to be deleted afterwards- they are simply temporary files- but a unexpected power loss or an error could prevent proper cleanup of these files. But since they are owned by LocalSystem, nobody else can delete them.

So the question is- how the heck do we clean-up the files?

well, if the only way to delete them is to become LocalSystem- let’s try that. After some experimenting, one of the most reliable ways I found was to create a service. You can do this by starting a Elevated Command Prompt, and entering the following command:

This creates a new service called runcmd. the /K start is necessary because the service control manager expects services you run to be… well, services. cmd is not a service, so it won’t register itself with the SCM and this SCM will kill the process after a timeout. using /K start, we can force that first spawned cmd to instantly start another one, since killing the parent process does not kill child processes, that cmd remains alive.

Running it is simple. just enter this command:

On my System, this didn’t seem to do anything, but a few moments later I received a “Interactive Services” notification:

So, CMD was running. I switched to the Interactive desktop, and was greeted (after this weird switch thing) with this:

The Interactive Services desktop.

Success! cmd was running under the LocalSystem account. This is good way to clean up files left about by services. However, while I was able to get explorer running (start menu), I wasn’t able to get explorer running (file manager). So I cheated, opened notepad, and used it’s file dialog. This method could be used to delete odorious files that refuse to be deleted on other ways.

Obviously, this should only be used when needed and the applications you run should be kept to a minimum, and you surely shouldn’t run browsers this way!

298 total views, no views today

So I was bored and decided to update my Flash plugin, a chore that I recollect stopping in it’s tracks previously, for reasons I couldn’t recall. Main reason was that my flash plugin has been nearly constantly crashing on certain sites. Mostly due to the ubiquitous use of flash for advertisements, which seems to be one of the dominant uses of the technology.

So, I visit adobe.com and go to download the player. First, they try to shove a McAfee scan down my throat. You know the drill. They know we just want to get the hell away from them, so they decide to helpfully fill out the “default” options for us, which just so happen to correspond with the options one would need to choose to give them the most revenue.

So I finally manage to get past that brigade of crap, and then it asks to install software. fair enough- that is what I was doing.

Much to my chagrin, however, it isn’t installing flash, it wants to install Adobe DLM, DLM I assume stands for DownLoad manager, although it could very well stand for Dingo-Llama-Mammoth for all I care.

let’s analyze the sequence of events so far:

1. I find that because of the shitty programming by Adobe their newest, most stable release of Flash, which I might point out isn’t even theirs and is just a sodomized and tortured version of what was at least a personable Macromedia Flash, it crashes nearly constantly doing routine tasks, like showing, I don’t know, pictures, as it’s designed to.
2. So, I decide to go to the vendor page (although I would prefer not to) in order to see if an upgrade is available. I believe there is. So I click to download Adobe Flash player. Before I do this I have to uncheck the “agreement” on my part to have them perform rectal scans of my computer using an AV product that probably is rather familiar with the appearance of a rectum, since that’s it’s origin. Now, remember, when I clicked the button, it said I was downloading Flash Player. And then *poof* up comes the prompt screen asking me to download this entirely unrelated “download manager” which brings me to another point.

Every single fucking program I download wants to install a god damned download manager! how many bloody download managers do I need? Am I going to need a download manager manager to manage all the download managers that all manage only the specific downloads from that specific company? Is there something wrong with the concept of downloading a program, I don’t know, using the conventional browser method? You know, like any other sane person? No, Adobe has decided to decide for me. “We won’t install Flash like you wanted, but we will install a download manager that will consume resources indefinitely for this one-time installation of Flash. Then it will sit in the background and make sure your updated, because god forbid if your version get’s out of date!”

Which brings me to another rant, Versioning. I mean, I totally understand why you might want to have the latest version of an application- it fixes bugs, adds features, and so forth. and being notified, and even having the opportunity to update with a few clicks is very convenient. I have no beef with the concept.

What I disagree with is this whole “OMG if you aren’t updated to the latest version you will get haxored!” there are people who say this about every bloody program. It’s understandable for browsers, and for a number of browser-based/web-based technologies, as well as things like the .NET framework, and of course the core of windows itself. But, seriously, the main reason you update a program is to fix bugs and add features, and hope that the bugs and security concerns that a new version adds (And they always do, unless the change is extremely minor) don’t outweigh the benefit of having the known vulnerabilities and the existing bugs eliminated.

Additionally, this very mantra is proposed on applications that have little relevance to web technologies. I mean, Microsoft Word has been relatively unchanged since version 6, with of course downlevel changes (which I’m sure took a lot of effort, I’m not downplaying that) But the fact is the entire purpose of the program is to be a word processor. The fact that it now represents a bloody programming platform should be some indication that they might have sort of lost their focus on what the program is supposed to do. It’s supposed to make it easy to edit documents, not make it easy to program spam e-mail merge programs or even be a platform from which to launch your own applications.

I don’t mean to pick on Word or Microsoft by any means- this seems to be a problem with a global scale. It’s a complex with versioning. If somebody has a problem, and they don’t have the latest version, that is automatically the cause, and truly, this attitude, or more precisely, the logic behind me, continues to elude me. They don’t understand the various downlevel changes, and half the time the release notes and changelog for said program mention nothing even remotely relevant to the various issues the person might be having.

Going almost hand-in-hand with the “download manager” syndrome is the “background updater”. Each company seems to have it’s own. You’ve got the Adobe one, the one from, say, Google, Apple, and so forth. And every single one of them is sitting in the background making sure I’m “up to date”. The problem here is that they all have to same goal but they all have very different UIs and they all act entirely different and essentially have different paradigms. This is something where Linux has the right idea; the package manager can update any package you install through either the GUI package manager or through a apt-get command in the terminal. The thing is, the environment is different; Linux programmers have no problem submitting their updates and new packages to the essentially neutral repository folks. With Windows, the best solution, which is the integrate this all into Windows Update, is owned by MS, which many of the companies who would have their software in it are competing with, which seems a bit like a conflict of interest; who knows if MS will “accidentally” forget to update users of competing products?

Back to the various “update” managers, they don’t simply update the programs you already have from their company; they also inform you of “updates” to their other products. The Apple update software makes sure you know when a new version of Safari is available, even if you only have iTunes; Google’s updater makes sure that you’re fully aware of when a new version of Picasa is released. And so on.

In conclusion, suffice it to say that currently update and download managers are wholly unnecessary (especially with the latter) and a huge pain in the ass for everybody.

534 total views, 2 views today