Oh Windows...:facepalm:

It's another late night working on a Java project. While I'm sure there might be some good reasons to develop on a Windows machine I can't ever find them when I'm forced to use one.

Anybody have any horrible Windows stories when coding? A co-worker of mine lost some work today when he hit the Windows update pop-up today while working.
 
I only have one horrible story and it is called Visual C++ ;). I think my biggest gripe is compiling dependencies on Windows using Visual C++ for the following reasons:

  1. 99% of the time, the provided development binaries are for a newer or older version of VS.
  2. 99% of the time, the binaries use a different config (multithreaded DLL or "the other one").
  3. No standard system to compile the libraries (it is like the state of Linux about 20 years ago).
  4. No helper scripts work like wx-config, pkg-config etc.
  5. The Visual Studio clean function does not actually clean everything. It just leaves crap laying around.

And after all this, Mingw/GCC still provides more "native" feeling binaries because you don't need to install the whole Visual C++ runtime on the users machine.

Microsoft still hasn't improved this system because they have been sidetracked with .NET (VB.NET, C#) for the last few years but now they are focusing back on native development I am hoping things will get better :)
 
Hmm, Java :h

When it comes to programming in Java my favourite IDE (NetBeans, or to put this a little more ontopic: java/netbeans ;)) runs smoothly on both Windows as well as Linux (and FreeBSD!
devilgrin.gif
) so it shouldn't make too much of a difference.

As to Windows stories... Well, being a systems administrator I have to say that it's my experience that very often many people blame the operating system for their own doings. Not saying this is the case here, don't get me wrong, but it does happen a lot when Windows is being used.

Often quite appropriate considering that Windows was designed to make things as easy as possible on the end user (though Windows 8 does make me doubt this approach a little bit) but user mistakes also happen a lot.

I tend to compare it with upgrading your ports collection while ignoring /usr/ports/UPDATING. Very often things go smoothly and without problems. And suddenly "FreeBSD" or "this port" starts misbehaving, surely there's something horribly wrong with it?

Just my 2 cents on the matter ;)
 
I've lost more data and time working on Linux than Windows. That's because I mostly use Linux. Windows, in the last 7 years of irregular personal use and 5 years of regular professional use, never ever gave me a blue screen of death by itself. Trust me.

TPIBKAC applies.

Regards.
 
I'm a web developer. Recently, I had a contract to work for a small company that said I had to use their Windows 8 machines and their older copy of a trial(?) version of VS. I kid you not when I say I texted my wife, 30 minutes after starting, that I wanted to quit. At first, I couldn't get used to Windows 8 wanting to switch me out of the desktop to that start screen (and I couldn't figure out how to get out of it). Then, VS kept wanting to format my code that made the lead programmer for the company, Dr. Sheldon Cooper, blame me for lousy formatting. VS would try and insert bizarre code from its auto-insertion thing, the font sizes were huge, the color highlighting was hideous, and trying to view the same code side-by-side... I never bothered to figure it out.

Now, yes, I'm sure there were ways to configure VS to behave as I wanted but I can't help but think how many people use what I assume is the default. I've never wanted to make so many adjustments on any non-Windows machine in the past. (I don't know why the lead programmer was allowed to use a Mac.)

Fortunately, after three months, I was able to get out of my contract just a week ago.
 
I can give you a horrible coding story every single day.

At work I am forced to develop in Visual Basic 6.0 (yes, really...), and all the software must run in a mixed environment, spanning from Windows 2000 to Windows 7 x(. I'm slowly trying to make them understand the concept of "portability" and the power/flexibility of the open source tools (and operating systems too), but it's like talking with goats. But last month I showed them the power of Java and JasperReports running on a Linux machine and they were amazed :e, maybe something is changing...
 
Humm Windows horror stories, I have hundreds but a few stick in mind.
  1. The mysteriously missing DLL, a computer I was building for a friend back around late 2003 developed a fault right after a fresh OS install I thought "odd but must be a crappy install", so I reinstall the OS, drivers etc., reboot after installing the drivers, then BANG: the same issue, so I start thinking it must be the HDD or a memory issue, so I strip the PC down to basics and start again, then add components back one at a time, ending with the floppy drive (it was a necessity for him) and what do you know it happens again. So I tried an old floppy from one of my systems with no issue. It turned out there was an issue with a generic driver used by Windows for the FDD that conflicted with the driver for the SATA card. The thing is I had the exact same card in a Windows system and never was able to replicate this issue.
  2. Winsock errors, enough said on this one but work for any ISP support desk that has lot of OAPs using hand me down hardware and you get REALLY good at supporting old hardware and legacy OS's.
  3. The XP system running as a odd server, an old customer of mine who had a need for large amounts of bandwidth and a complex setup called me one day and said: "Chris, my systems slowed down to a crawl, and my new support guys are useless." I was at another job but I not only liked this guy but knew how much money he was losing (commodity trader), asked for the rest of the day off and ran over to his house. So I get there and he is right it's taking an age to recognize keystrokes and mouse movement took about the same time as continental drift to move the same distance. So did the normal just to make sure reboot, Memtest, etc. By this point trading had finished for the day (the call was at 2 and I got their at 3:30). This is 6 PM and I saw his network monitor was still saying even for him high use, I thought OK this is either DDOS or malware. I load up nmap hoping to find something while I am building an XP live CD with Malware bytes, and I noticed port 8080 was open and fired up a disposable VM on my laptop and typed in the IP address into a web browser and what did I see? His whole desktop served up over his internet connection VNC view only style.

    On investigation it turns out the company that he used for support had installed some form of home brewed remote desktop software and opened the ports on his firewall, and for some reason people were watching it.

    While not a direct Windows fault they removed my pfSense box I set up for him and used Windows Defender.
Just a few that stick in my mind.
 
saxon3049 said:
Humm windows horror story's, I have hundreds but a few stick in mind.

Heh, I just deployed a pfSense box last night, m0n0 didn't have enough VPN options.:)

With that said who dumps a pfSense box. :\
 
I think @ShelLuser summed up my experience.

All of my bad experiences with Windows are mostly concentrated on pre-XP versions (95, 98, ME). Since moving to the NT kernel, I've never really had any problems. The problems I do have are typically caused by either unstable hardware (overclocking), crappy vendor drivers (Linux and FreeBSD have those, too) or my own stupidity (user error).

It's been very solid lately, too. Since using 7 and 8, I've never ever rebooted my system except for system updates occasionally.
 
Last edited by a moderator:
Well, I've wasted several hours today attempting to upgrade the enterprise KMS server to support Windows 8.1 and Server 2012 R2 clients.

Microsoft provided key will not install. Yay. Yes, I've run the KMS host update.

If I was a pirate it would just work, but oh no, we're paying to get screwed around here.
 
We had an interesting thing with compiling on XP and the Server 2008. The build environment is the same (from the same repository), the sources were the same, all the same - but the compilers did produce different code. This tends to make software integrators somewhat annoyed, even more so when you have some SPICE levels to keep.

But I can not give a complete list of my windows horror stories - I have to use it at work. Out IT department seems to have the option of not doing that. Once I had to call them up for a problem and told them that, while they could make me use Windows, they could not make me like it. The guy did not stop laughing for some time.
 
My stories have always been with the cursed file systems from Microsoft. Months ago I discovered, with the painful way, that the MS DFS does corrupt open files in every instance if one opened instance crash, so much for a distributed file system.

Other than that, thought they had already addressed the 255 character long paths, but not, I still have to educate the users to not use very long names nor descriptions in files or folders.

My current nightmare is a SAN pre-formatted with, guess what, NTFS, very limited and insecure (my plan is to put a ZFS on it this week) and the unidentified network traffic that comes from/to the Windows clients, not verified in clients but verified in router (My Alexa sensors tickling hard).
 
igorino said:
Other than that, thought they had already addressed the 255 character long paths, but not, I still have to educate the users to not use very long names nor descriptions in files or folders.

The WTF here comes from the good old pascal strings, the file system itself can do a lot better. There we have 256 bytes (or letters?) in a file/folder name. But some parts of Windows need to be compatible with pascal calling conventions, that is why the Microsoft compilers had the "\pHello World!" string format. The curse of compatibility - bit for bit and bug for bug. NTFS itself is not that bad. But then again, it was not developed in house.
 
To be fair, as bad as Windows can be sometimes, Windows 7 is definitely a winner. It still works better than almost everything else. I have a laptop with Windows 7 beside me and it's been behaving pretty well.
 
zspider said:
To be fair, as bad as Windows can be sometimes, Windows 7 is definitely a winner. It still works better than almost everything else. I have a laptop with Windows 7 beside me and it's been behaving pretty well.

Oh yes Windows can be a winner at times, hell it won the OS was (ver 1.0) because it had the backing of a recognised giant and a easy to understand layout, that didn't change for a long time. As any one who has worked support will tell you Windows 8 has destroyed 25 years of walking people through issues over the phone when RDC is not a option.

If you want a example, my dad a very intelligent man a Nuclear engineer. However he and computers do not get along he has always used windows and he hadto recently use a windows laptop... after 3 hours he returned the laptop to the shop and since then has used a mix os OSX and a VM of Win7.
 
The original NTFS was actually a pretty damn close copy of OS/2 Warp's old HPFS. Hell, they even use the exact same partition type ID. HPFS was a collaboration between IBM and MicroSoft, released in 1988, and was, in fact, written by mostly MicroSoft coders.

When MicroSoft decided to implement it with Windows NT, they changed it's name from "High Performance File System" to "New Technology File System", like Windows NT, aka New Technology.

Later, as Windows progressed, the filesystem did, too. But it's still basically the same old HPFS.

Excuse the history lesson, but I loved OS/2 Warp 4.0.
 
History is important - for curiosity and for remembering what not to do again.

Do you have some links to that history regarding NTFS? I was more under the impression that it came with the rest of NT from DEC, when Cuttler came over and "wrote all of the NT kernel in 2 weeks in some isolated hut in the mountains".
 
Crivens said:
History is important - for curiosity and for remembering what not to do again.

Do you have some links to that history regarding NTFS? I was more under the impression that it came with the rest of NT from DEC, when Cuttler came over and "wrote all of the NT kernel in 2 weeks in some isolated hut in the mountains".

I was thinking the same thing, especially the fact that VMS was written by Cutler and had ODS/Files-11, which is a fragmenting file system. Then he reimplemented VMS as Windows NT and of course the filesystem concepts came over too.
 
Good ol days

break19 said:
The original NTFS was actually a pretty damn close copy of OS/2 Warp's old HPFS. Hell, they even use the exact same partition type ID. HPFS was a collaboration between IBM and MicroSoft, released in 1988, and was, in fact, written by mostly MicroSoft coders.

When MicroSoft decided to implement it with Windows NT, they changed it's name from "High Performance File System" to "New Technology File System", like Windows NT, aka New Technology.

Later, as Windows progressed, the filesystem did, too. But it's still basically the same old HPFS.

Excuse the history lesson, but I loved OS/2 Warp 4.0.

I ordered a copy of OS/2 for a client. Seems like that was '92 or '93. When I received the software, I installed it for the company, but they didn't like the interface. Interestingly, they were a big time IBM go-blue mainframe shop. Instead, about a year later they asked for Microsoft NT 3.51 because they preferred its "WIndows 3.1 look and feel". Go figure.

I seem to remember NT 3.51 being a lot easier to install. I think the OS/2 distribution required CD-ROM drive types that were never conveniently on the machine. Until recently, I had an old computer set up with 3.51 and Visual C++. At least THAT version of Visual C++ didn't autonomously decide to insert "bizzare code into my project"! When does automajic become too automajic?
 
I really liked OS/2 Warp 3.0, especially how it handled memory. Doom also ran faster on it than MS-DOS ... at least on my 386SX
 
Back
Top