| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Comments:
<0> the current system only has one callback (here's the file"
<0> )
<0> and that's proven to be not quite sufficient
<1> I hope you won't be asked for adjusting the search order. That could result in a mess.
<0> well, this started as a let's some other callbacks about 8 years ago...got stalled and has been slooooooowly being moved from C (originally on the Amiga) to C++. I'm using the boost filesystem stuff right now
<0> I "could" easily alphabetize the files (and the order in which I do the directories for that matter) but currently it's in whatever order the boost dir iterator gives them to me
<2> Is there a way to stop destruction of an object in the destructor?%
<1> std::abort()
<0> exit()
<2> like it sees it is used somewhere else (in a reference table) and if someone tries to delete it, it cancels destruction if references exist)
<0> but why on earth would you want to DO that
<1> RoG: seriously, the answer is no.
<0> RoG what problem are you trying to solve?
<2> well, I wanna avoid destruction of an object still used in another thread
<0> and you're NOT using smart pointers?
<2> eee, i guess not
<0> you might want to consider boost::shared_ptr<T>
<1> It should not be the object's responsibility to manage its own lifetime.
<2> well if another function calls "delete object", there would a check in the destructor if the manager still needs that object elsewhere. If so, it doesnt destruct
<0> shared_ptr<T> has even been adopted (w/ minor changes) into TR1
<0> why would they be calling delete object?
<1> RoG: we understood that. It's the wrong approach.
<3> RoG - you shouldn't call delete if it shouldn't be deleted
<3> but yeah, smart pointers solve this incredibly elegantly
<2> smarts pointers are that boost stuff?
<0> no
<0> there are some smart pointers in boost
<2> ok
<0> it sounds like you want shared_ptr<> from your brief description
<2> could be yeah, ill have a look at that
<2> thanks
<4> is a DWORD unsigned ?
<2> typedef unsigned long DWORD;
<4> thanks
<0> Joey__ playing with legacy code, eh?
<4> yeah trying to implement a taskbar toolbar, needing to convert some code
<0> Joey__ note that it's really an unsigned 32bit int
<0> if you switch compilers, long may not be the "right" thing
<5> hi
<0> lo
<4> yeah I'm using UInt32 for it (.NET)
<5> is .NET worth learning ?
<3> what are you worried about?
<6> virtually everything is worth learning
<7> Sure.. .net is the future of windows
<7> WinFX ( the next API ) is managed only
<7> Any news about stl.net?
<6> Not that I've read
<7> It's bad, i'm waiting for that to start using c++/cli
<6> a touch of googling shows a total lack of information
<6> the latest bit of news I can find is from an MVP on MSs newsgroups, saying "Stay tuned, we may have more information in a week". That was over a week ago
<6> Though quite frankly, microsoft get their release schedules and feature lists from a random generator anyway
<3> still waiting on the 'quality' release for VS2005
<7> aha
<6> First, it was mildly annoying, then it was almost funny, now I just ignore microsoft entirely
<3> has VS launched internationally yet? (i'd ***ume so?)
<6> still not transitioned to VS2005. Actually, the last 5 programs of any size I wrote were either platform agnostic C++ or done entirely on my mac
<6> They've made themselves irrelevent, to me
<6> Ah, check the super-up-to-date UK Developers site
<3> I haven't used anything but 2005 since december
<6> Where you're encouraged to download VS2005 Beta 2!
<3> or november, rather
<6> http://www.microsoft.com/uk/visualstudio/default.mspx
<6> And my software supplier still has no ship date for it - been "On order only" forever now
<6> It's got to be a joke, and a stupid one for them to play. I couldn't count the amount of code I wrote specifically for windows just 2 to 3 years ago, so much there was. In the last year, I surely can count it - less than 1000 lines, unreleased
<3> even though I really don't write for anything but windows, I try to stay away from windows-specific code as much as I can, or at least contain the module as much as possible
<6> No matter. If they want to keep carping on about "TEH FUTRE! VISTA IS THE BEST LOLZ - WAIT FOR US!", that's up to them
<6> I've moved on, and plenty of others seem to be getting the same idea
<3> i do have some code that needs to fetch/save things to a database... i use ado.net in a nice c++/cli library, and that's the only part of the project not written in standard c++
<3> well, .net will really start pushing once vista is out... it's already spreading pretty good
<6> heh. You still can't write software specifically for XP, let alone Vista
<0> it would certainly help if they could get it working
<6> It will be 2012, *minimum* before it makes sense to ship anything dependent on the Vista tech
<6> Otherwise, you're basically cutting off 30% of your market
<0> and standing still
<6> Bespoke software/private deployments, sure. But general release, not a chance
<3> I guess that depends on how quickly everyone upgrades to vista
<3> 2012 seems like a stretch, though
<3> but then i'm young, and 6 years seems like a long time
<6> XP was released 2001, IIRC, and still doesn't have ubiquitous deployment
<6> and we're moving into a situation where people don't upgrade their machines as often as they used to - there's been no "big leap" for years
<6> When it comes down to it, Since Windows 2000, windows has been fine
<6> Unless you specifically need something that doesn't run on it, there is *no reason* to upgrade other than those artifically created by MS
<6> "Upgrade or get ownzed becase we won't patch you anymore"
<3> well, there's 64-bit processors, but that really hasn't caught the buzz that it should yet... maybe someone has to come out with something really useful and necessary that depends on the technology
<6> meh - I'm a hard core computer geek, and I've not upgraded my PC for 18 months in any way whatsoever
<3> I still see windows XP commercials on television. why aren't they pitching vista?
<6> because it doesn't exist
<6> my parents still run Windows 98. The last 3 computers my brothers bought have all been Apple macs
<3> I haven't upgraded anything in a while either
<3> almost a year I think
<6> which is what I'll replace my parents machine with when it finally kicks the bucket
<6> Wack a Mac Mini onto the existing LCD, keyboard and mouse. It'll suit them fine until the hardware dies, and I won't have to keep going up and fixing the latest fubar by Outlook
<3> office 12 comes out this year, too
<6> Woop de do. Yet more ways to redo something that was essentially perfected in Office 95 (still the only copy I've ever bought)
<3> heh
<6> "Look! We moved all the menus around so your muscle memory is useless. Also, we changed the file format, so you need to get your friend to upgrade"
<6> "Please pay us ?250 for the priviledge (?400 for new users)"
<3> Asriel - I don't know... I like to think that MS spends a decent amount of time researching and testing user interface design, and changes their software to try and match the 'typical user' at the time
<6> If you're a new user, it's probably "Better"
<0> rdragon I don't
<6> But for 99% of people, it'll be more of a pain in the ***
<3> 'upgrading' is always a pain in the ***, but it's usually for the better in the end
<0> I think they're "****ing clueless" about good ui, and worse at real-time multi(threading/tasking/processing)
<6> One of the better interfaces I've seen for a word processor recently is "Pages" from Apple (yes, I know I sound like a fanboi, but I'm serious about this)
<3> aye, I could be wrong, I don't actually have any evidence of anything showing what they might do
<6> I just opened my copy. There are 8 icons on a toolbar at the top, and a big white page with a cursor
<3> yeah Apple usually is touted for having a superior user interface
<6> You type, and it appears
<6> It doesn't do everything Word does, but it does what 99.999% of users need, and it does it bloody quickly and bloody easily
<6> If it did Mail merge, it would literally do everything I've ever done in a word processor in my career
<6> And I'm an atypical user
<6> Their instance on carrying the cruft of a billion previous versions will drag them down, eventually
<6> Whilst backwards feature compatability is nice, I'm willing to bet there are parts of Office 12 that are literally used by maybe 1000 people in the world
<3> yeah, i agree
<3> I personally prefer the features to exist even if I don't use them, else risk the rare event of needing it and having it be a h***le
<6> I'd rather have an easier to use, and importantly cheaper experience
<6> Office is extortionatly expensive. A new office licence costs twice as much as a windows licence
<6> Again, the Apple pricing model is about right - I think iWork is ?60
<6> Openoffice, I just can't get along with. All the cruft of MS office behind an interface that seems to be running via RDP on a 14.4 modem
<6> I think MS make some good things, but they need to get some compentant global management in, stat
<6> Leverage more out of their research department, and less out of the "It made money before, lets stick a new chrome on it and release it" department
<6> then it'll look a lot less like they've just waiting for Google/Apple/etc to release something before rushing to clone it
<6> Let me rephrase - their "It made money before, lets stick a new chrome on it, announce it, say it kicks the crap out of our previous version, then delay it for about 3 years" department
<3> hehe
<6> hohum, no matter. unless a client specifically requests a windows linked system or I want to play around with .net, I can't see using their stuff much in the future
<8> what piece of code do I insert in a printf() error statement to show which function issued this error ?
<8> like when calling function_foo()
<6> other than to write portable systems. VS is still the best IDE around :/
<8> the printf() gets invoked
<6> Some compilers define the preprocessor macros __LINE__, __FILE__ etc
<6> I forget the exact names
<3> corstan why are you using printf?
<8> cout or printf
<8> doesn't matter
<6> this is actually one of the few times a real macro can be handy
<3> hmm
<3> a little while ago I ran across a library that provided a stack trace
<3> runtimwe
<8> ok
<3> so, look for one
<3> or make one
<8> is it that hard?
<8> or time consuming
<6> #define ERROR(message) std::cout << (message) << "in " << __FILE__ << " at " << __LINE__;
<6> or similar. some crts can also generate you a stacktrace
<3> I wouldn't recommend trying to do it yourself, corstan
<8> no
<8> neither do I :)
<6> or create a core|mini dump at will
<8> too much of a newbie
Return to
#c++
or
Go to some related
logs:
#AllNiteCafe
#AllNiteCafe
#linux
#MissKitten
#linuxhelp
imagecreate 300dpi
#java
#chatzone
#AllNiteCafe
wyldroze
|
|