feline_dracoform wrote:

we have now moved to version 2. in the standard commit dialog there is a keyboard shortcut to jump to every single control... except for the three buttons "OK", "Cancel" and "Help".

Enter -> does OK Esc -> does Cancel F1 -> does Help This is standard in Windows applications.

feline_dracoform wrote:

however, in deference to people like me they have put a label on the screen telling me to press CTRL-ENTER to accept the dialog... and there i was expecting to use something traditional like ALT-O...

I find it interesting that you would pick WinCvs here since it is actually going an extra mile trying to ease the multiline editing - "traditionally" you would have to Tab-out to the next control to be able to confirm the dialog using Enter. This can also be configured for people with different habbits.

/* I C++, therefore I am... */

feline_dracoform wrote:

the default button seems remarkably inefective when focus is inside a multiline text edit field *sigh*

Normally a multiline edit control traps both Enter and Ctrl+Enter keys. WinCvs 2.x however allows you to configure the behaviour so that when inside multiline-edit control you either use Enter to confirm and Ctrl+Enter to line-break or vice-versa. Go to "Admin"->"Preferences"->Command Dialogs. Use the "Ctrl+Enter to confirm" checkbox to control this function.

feline_dracoform wrote:

then again i have long claimed that wincvs is not a well designed windows program, but that could simply because i don't like it

Or simply because you have the "instant gratification" attitude. WinCvs is designed for programmers - it is assumed that you will spend some time trying to understand the tool and the way it works. Or at least read the entry labels - the behaviour is described on the very dialog right above the edit control. It will tell you what commits and what line-breaks depending on the current settings. Oh, and there is a source code for you so you can "improve" it any way you want...

/* I C++, therefore I am... */

There is an extensive explanation on the subject in September 1996 "C++ Q & A" by Paul DiLascia: http://www.microsoft.com/msj/archive/S202C.aspx[^] You have to scroll down a little bit to the chapter starting "I've been teaching myself C++"

/* I C++, therefore I am... */

Chris Maunder wrote: The whole issue of Terms of Use is something we have been grappling with for (believe it or not) 18 months. That issue should never happen if Code Project was simply enforcing that any article contributed makes clear as to under what license it's distributed. For example at Source Forge you have to choose from available approved Open Source licenses or wait for approval if your license differs. Code Project is not making that part clear and it can only get worse with time. There are some GPL articles here and many people believes that if it's published on CP then GPL no longer applies. Its a ticking time bomb that might cause a serious problems in the future. The whole issue can be resolved instantly by enforcing contributors to choose from approved by CP licenses or to declare their own license that will be reviewed and either approved or rejected. As a side effect the license enforcement would probably reduce the flow of crappy artciles we are observing lately if articles were reviewed before allowing them to show up as CP articles.

/* I C++, therefore I am... */

brianwelsch wrote: so what? OS should NEVER crash running an ordinary piece of software. But since MS got people used to bugs most users just says "so what". But we are here professionals and one has to see the difference between the "perception" and the reality. The reality is that software is not getting any more stable or reliable and doesn't even work much faster inspite of growing computing power due to increasing pile of crap being produced. MS for example is wasting resource developing increasingly ugly and counter-intuitive types of menus, buttons and other GUI parts that worked perfectly before while obscure, and well-known bugs are left in the product for ages.

/* I C++, therefore I am... */

pseudonym67 wrote: Still I wonder where that term came from? I have the feeling I've seen it for the first time on JoelOnSoftware side, but I might be wrong...

/* I C++, therefore I am... */

Ian Darling wrote: You mean this: I think you can tinker the code to crash XP as well. I used to have a crash-them-all version somewhere at home, I may look it up for you. There was no loop there, and it was possible to make hung the DOS window or blue-screen or just reboot depending on the actual code. Ian Darling wrote: I'll respectfully disagree Thanks. Ian Darling wrote: Well, at least we've got some ideas for the next poll: It IS a good idea for the poll! ;)

/* I C++, therefore I am... */

Ian Darling wrote: Right, so that's why companies like Microsoft (who are now more or less a mature organisation) are releasing increasingly better and more reliable software? The software is not better or more reliable - it's just more complex and bigger, but I can still crash the latest OS with couple lines of code containing printf statement and few "\b" characters in it. I appreciate your optimism, but it's not the way it works I'm afraid :( Software is evolving into the fire-fight model. It works in typical situation and if any unexpected conditions happen it fails flat and the "rescue team" is called in to fix the problem, often by rebooting the system so it "works" again until the next crash.

/* I C++, therefore I am... */

Ian Darling wrote: A mature industry also removes the prevalence of the cowboys and promotes professional development methods - so we should start seeing significantly more reliable software across the board. In fact quite the opposite. We are starting to see more cheap, shitty software all around. There is a huge reduction on business analisys and quality assurance spending as well as workforce involved in development, it will not increase reliability but it will reduce the cost, at least in the short term.

/* I C++, therefore I am... */

Chris Maunder wrote: You equate a site that will only email you if you provide your email address and don't take the 5 seconds to find and opt-out of any mailings with a company that site scrapes email addresses from newsgroups and web forums and then bombards you with stuff out of the blue and for which you have never shown any interest in. Actually I adequate to the sites that always check some options I don't really want to be checked to receive something I don't want to receive. Few years back, when I was joining the CP the times were different. If I was joining today and I saw the checkboxes selected to receive newsletters I would most definately either quit registration or simply provide fake or temporary email address just to be safe. The bottom line is, I should not have to opt-out, I should opt-in only. I do not approve anybody to opt me in for anything, and I don't buy your explanation about 5 seconds to find the opt-out options. That is 5 seconds too many. If I wanted the newsletter I would find it and select. If I don't want I should not have to do anything. Period. To me, if something looks like a trap, it is a trap. In that case it's a trap to catch unaware people to receive an email they didn't ask to receive. For whatever - good or bad - reason. The calculation here, quite transparent, is that even if people don't want the newsletter they will be too lazy to go back and unsubscribe. And that kinda leaves a bad taste if you ask me. Now, please don't push the comparisons to extreme and relate to the email-hunters and bots. But in that particular case - the selected checkboxes - there is no difference betweem spamers and CP. They both argue that I wanted something when I clearly didn't ask for it.

/* I C++, therefore I am... */

David Cunningham wrote: I think it's absolutely universally accepted that if you register on a site in any manner, you're added to the site's newsletter by default. If I don't make a consious choise then I didn't agree. Take the license agreements we are typically presented when installing any software as a reference point - they will never assume you agree to the terms, you have to click the damm checkbox and only then it's a valid agreement. You can not make an assumption that just because I double-clicked the setup executable I agree for the terms. Same applies to the websites - if I don't check the checkbox myself then I didn't agree. And when it comes to the newsletters it's particularly annoying to have all those checkboxes pre-selected because it indicates the site has created the traps for me fall into.

/* I C++, therefore I am... */

Chris Maunder wrote: Now suppose CodeProject was accused of spamming because someone get's a newsletter after joining and not bothering to check the sign up form where it says 'News;etter'. The checkbox should be unchecked initially, so I have to consiously choose to receive anything. Right now it seems to be checked by default and I have to find it and unselect, that puts Code Project right in the same cathegory as every other spammer.

/* I C++, therefore I am... */

Nowadays it would be "Productivity Oriented Development". It boils down to the instant gratification attitude and short attention span of the management, reduce QA in order to shorten time-to-market period and reduce the development team size to reduce the cost while pressuring the remaining folks to the limits and beyond. Sometimes outsourcing the job to the offshore companies is also applied. But that would be a rant on totally different topic...

/* I C++, therefore I am... */

Nick Hodapp (MSFT) wrote: I'm looking for customer feedback (evidence) that the VC.NET 2002 (and even 2003) project system exhibits a real-world problem that affects your adoption of the tool. Sure, here you have it: 1. It's slow. Really really slooooow. 2. It's buggy, at least 2002 version crashes way too often and then it just re-starts (betting I won't notice or what?). Even a simple operations like adding the class for a dialog puts it down. 3. Lack of backward-compatibility. E.g. I can convert the VC60 project to the new VC70 project, but then if I change some settings I can't export it back to VC60. There is a tool at CP that does that, but IMHO that should be build-in. And all the VC60 addins are not working any more, but it should be simple to implement a proxy interface to allow them to operate. People are like spaghetti - you need to pull them, not to push them. 4. It's lame at times. It has many fesatures of VC60 missing and so it doesn't feel like an upgrade but rather the opposite. E.g. most of the code wizards are very simplistic and look like "5 minutes of work just before the relase". Then, once you get to those wizards you may notice that even the edit boxes there are screwed - I can't get the right-click menu to copy/paste stuff!? 5. It does stupid things. One example is the source control support where it creates some temporary files, adds then to source safe, then checks them out and then removes them. That effectively fills the history of changes with lots of noise and make the history view pretty much useless. It also tends to add the "afxwin.h" include here and there for no reason. Time to stop the micro-management I tell ya... As for the messages - the general rule applies that if it goes to the output window then it's fine. If it pops-up and asks me to take some action then it's probably bad. If it pops-up just to tell me something, but it gives me only an "OK" button to confirm that it's very bad.

/* I C++, therefore I am... */

You may also enjoy that: F/A-18 Hornet passing through the sound barrier[^]

/* I C++, therefore I am... */

David Cunningham wrote: Hmmm, would you like it to be filtered even further? Perhaps. However, with C# and .NET propaganda is just like with those spam emails, add banners and popups - after a while I adapt to ignore them all together and then it's OK ;)

/* I C++, therefore I am... */

David Cunningham wrote: If you click on the MFC/C++ tab, then all the content, menus and questions you're presented with should be restricted to MFC/C++. You mean like .NET(Managed C++, SOAP and XML, C++ Web Services) and Web/Scripting (ISAPI)? X|

/* I C++, therefore I am... */

SimonS wrote: I C#, therefore I am Can't come up with something original, huh? Well, it would not be the first case of plagiarising when it comes to C#, it kinda keeps up with the spirit of the language... ;P

/* I C++, therefore I am... */

Junkie wrote: I was more interested in your Signature rather than your question. I'm sorry, but it's not for sale. ;)

/* I C++, therefore I am... */

OK, where's the new poll about using the third party tools gone, did I knock it out somehow?

/* I C++, therefore I am... */