1. RE: win32lib v0.57.7
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_000_01C20A98.96ABBF60
charset=iso-8859-1
Yes Dan, I did. However what is on the web site is not a "release" as such.
It is really a work-in-progress version because it can't really be used for
some apps (eg, the IDE!) due to bugs in it. I have spent some time fixing
the resource leaks but there is still an issue with the graphics routines. I
was asked to place it on the site so that I could get some others to help
debug it.
Along with no RunDemos, there are no updated docs either. I wanted to stick
to debugging it for now.
> -----Original Message-----
> From: Dan Moyer [mailto:DANIELMOYER at prodigy.net]
> Sent: Monday, 3 June 2002 10:27
> To: EUforum
> Subject: Re: win32lib v0.57.7
>
>
>
> Derek,
>
> Did you get the version of "RunDemos.exw" I sent you that was
> written to
> reside in the "base" folder of Win32Lib? It's not in your
> current 57.7
> download.
>
> Dan Moyer
>
> ----- Original Message -----
> From: "Derek Parnell" <ddparnell at bigpond.com>
> To: "EUforum" <EUforum at topica.com>
> Sent: Sunday, June 02, 2002 11:12 AM
> Subject: win32lib v0.57.7
>
>
> >
> > I've updated my web site with the current version of the
> Win32lib library.
> > There are some problems with Pixmaps and graphics routines.
> >
> > If you are interested in helping debug this version, you
> can download it
> > from :
> >
> > http://www.users.bigpond.com/ddparnell/euphoria/w320577.zip
> >
> > The demo program PRETEND.EXW is a prime example that causes
> Windows to
> crash
> > due to resource leaks.
> >
> > I might not be able to fix anything for two weeks as I'm on
> holidays next
> > weekend.
> >
> > -------------
> > Derek.
> >
> >
> >
>
>
>
==================================================================
De informatie opgenomen in dit bericht kan vertrouwelijk zijn en
is uitsluitend bestemd voor de geadresseerde. Indien u dit bericht
onterecht ontvangt wordt u verzocht de inhoud niet te gebruiken en
de afzender direct te informeren door het bericht te retourneren.
==================================================================
The information contained in this message may be confidential
and is intended to be exclusively for the addressee. Should you
receive this message unintentionally, please do not use the contents
herein and notify the sender immediately by return e-mail.
==================================================================
------_=_NextPart_000_01C20A98.96ABBF60
Content-Type: application/ms-tnef
2. RE: win32lib v0.57.7
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_000_01C20B46.C3C4DC30
charset=iso-8859-1
Thanks Thomas. You're correct. In principle, all Windows IDs, Handles and
Pointers should be atoms.
> -----Original Message-----
> From: Thomas Parslow (PatRat) [mailto:tom at almostobsolete.net]
> Sent: Tuesday, 4 June 2002 2:00
> To: EUforum
> Subject: Re: win32lib v0.57.7
>
>
>
> Hi,
>
> In releaseResourcesForID an integer is used to hold a windows resource
> id, integers can be a maximum of 31 bits in size and one of the
> resources that gets passed to it when my app runs is to big so I get a
> type check error.
>
> Changing the type of the variable "id" from integer to atom fixes the
> problem for this routine (not sure if it's a problem in other
> routines, none show up when I run my app).
>
> Thomas Parslow (PatRat)
> E-Mail/Jabber: tom at almostobsolete.net
> ICQ: 26359483
>
>
>
>
==================================================================
De informatie opgenomen in dit bericht kan vertrouwelijk zijn en
is uitsluitend bestemd voor de geadresseerde. Indien u dit bericht
onterecht ontvangt wordt u verzocht de inhoud niet te gebruiken en
de afzender direct te informeren door het bericht te retourneren.
==================================================================
The information contained in this message may be confidential
and is intended to be exclusively for the addressee. Should you
receive this message unintentionally, please do not use the contents
herein and notify the sender immediately by return e-mail.
==================================================================
------_=_NextPart_000_01C20B46.C3C4DC30
Content-Type: application/ms-tnef
3. RE: win32lib v0.57.7
Derek,
Are these 'problems' only showing on Win9x platform? I'm not seeing any
issues with pretend.exw on WinXP...
-- Brian
Derek Parnell wrote:
> I've updated my web site with the current version of the Win32lib
> library.
> There are some problems with Pixmaps and graphics routines.
>
> If you are interested in helping debug this version, you can download it
> from :
>
> http://www.users.bigpond.com/ddparnell/euphoria/w320577.zip
>
> The demo program PRETEND.EXW is a prime example that causes Windows to
> crash
> due to resource leaks.
>
> I might not be able to fix anything for two weeks as I'm on holidays
> next
> weekend.
>
> -------------
> Derek.
>
>
>
4. RE: win32lib v0.57.7
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_000_01C20C11.B3651A00
charset=iso-8859-1
Good question. I've only tested it on WinMe.
Okay, I just tested it on WinNT and it works fine.
It does seem to be a Win9X issue. Great! Now how do I get aound that! Any
ideas anyone?
> -----Original Message-----
> From: Brian Broker [mailto:bkb at cnw.com]
> Sent: Wednesday, 5 June 2002 3:54
> To: EUforum
> Subject: RE: win32lib v0.57.7
>
>
>
> Derek,
>
> Are these 'problems' only showing on Win9x platform? I'm not
> seeing any
> issues with pretend.exw on WinXP...
>
> -- Brian
>
> Derek Parnell wrote:
> > I've updated my web site with the current version of the Win32lib
> > library.
> > There are some problems with Pixmaps and graphics routines.
> >
> > If you are interested in helping debug this version, you
> can download it
> > from :
> >
> > http://www.users.bigpond.com/ddparnell/euphoria/w320577.zip
> >
> > The demo program PRETEND.EXW is a prime example that causes
> Windows to
> > crash
> > due to resource leaks.
> >
> > I might not be able to fix anything for two weeks as I'm on
> holidays
> > next
> > weekend.
> >
> > -------------
> > Derek.
> >
> >
> >
>
>
>
==================================================================
De informatie opgenomen in dit bericht kan vertrouwelijk zijn en
is uitsluitend bestemd voor de geadresseerde. Indien u dit bericht
onterecht ontvangt wordt u verzocht de inhoud niet te gebruiken en
de afzender direct te informeren door het bericht te retourneren.
==================================================================
The information contained in this message may be confidential
and is intended to be exclusively for the addressee. Should you
receive this message unintentionally, please do not use the contents
herein and notify the sender immediately by return e-mail.
==================================================================
------_=_NextPart_000_01C20C11.B3651A00
Content-Type: application/ms-tnef
5. RE: win32lib v0.57.7
I've only got one system to test on (my home system is Win98 but I'm
usually so tired after work that I don't use it much.) I'll see what I
can do about tracking this down and I'll let you know if I find
anything.
-- Brian
Derek Parnell wrote:
> Good question. I've only tested it on WinMe.
>
> Okay, I just tested it on WinNT and it works fine.
>
> It does seem to be a Win9X issue. Great! Now how do I get aound that!
> Any
> ideas anyone?
>
>
> > -----Original Message-----
> > From: Brian Broker [mailto:bkb at cnw.com]
> > Sent: Wednesday, 5 June 2002 3:54
> > To: EUforum
> > Subject: RE: win32lib v0.57.7
> >
> >
> >
> > Derek,
> >
> > Are these 'problems' only showing on Win9x platform? I'm not
> > seeing any
> > issues with pretend.exw on WinXP...
> >
> > -- Brian
> >
> > Derek Parnell wrote:
> > > I've updated my web site with the current version of the Win32lib
> > > library.
> > > There are some problems with Pixmaps and graphics routines.
> > >
> > > If you are interested in helping debug this version, you
> > can download it
> > > from :
> > >
> > > http://www.users.bigpond.com/ddparnell/euphoria/w320577.zip
> > >
> > > The demo program PRETEND.EXW is a prime example that causes
> > Windows to
> > > crash
> > > due to resource leaks.
> > >
> > > I might not be able to fix anything for two weeks as I'm on
> > holidays
> > > next
> > > weekend.
> > >
> > > -------------
> > > Derek.
> > >
> > >
> > >
> ==================================================================
>
>
> ==================================================================
>
>
>
6. RE: win32lib v0.57.7
>Good question. I've only tested it on WinMe.
>
>Okay, I just tested it on WinNT and it works fine.
>
>It does seem to be a Win9X issue. Great! Now how do I get aound that! Any
>ideas anyone?
Very strange results. I dont really know what to say.
If I run it as is, after about 1/2 seconds it starts to paint on the desktop
or another window and it freezes shortly after. Now dont ask me why, but if
I grab the window DC at the start of the timer routine it acts just fine
procedure Tick( integer timerID )
sequence s
atom hDC1
hDC1 = w32Func( xGetDC, {getHandle(Win)} )
drawBackground()
moveSprites()
moveShip()
-- draw the screen
copyBlt( Win, 0, 0, Buffer )
end procedure
onTimer[ Win ] = routine_id( "Tick" )
Heck... im not even *using* the DC I got back (or releasing it). Just added
the variable and grabbed the DC and everything clears right up. Dont really
know how this applies to the Win32Lib routines though. Since I am not
releasing it I would think it would just create even more leaks... *shrug*
7. RE: win32lib v0.57.7
Hi Don,
What version of Win are you running? Grabbing the DC does nothing to
correct the problem on my Windows 98 [Version 4.10.1998] system. But if
it's really true, it might be one heck of a clue in the right direction.
-- Brian
PS to Derek: I'd love to see the program that smokes.bmp belongs to
(tobacco's a vegetable ya know 
Don Phillips wrote:
> >Good question. I've only tested it on WinMe.
> >
> >Okay, I just tested it on WinNT and it works fine.
> >
> >It does seem to be a Win9X issue. Great! Now how do I get aound that!
> >Any
> >ideas anyone?
>
> Very strange results. I dont really know what to say.
>
> If I run it as is, after about 1/2 seconds it starts to paint on the
> desktop
> or another window and it freezes shortly after. Now dont ask me why,
> but if
> I grab the window DC at the start of the timer routine it acts just fine
>
>
> procedure Tick( integer timerID )
> sequence s
> atom hDC1
>
> hDC1 = w32Func( xGetDC, {getHandle(Win)} )
>
> drawBackground()
> moveSprites()
> moveShip()
>
> -- draw the screen
> copyBlt( Win, 0, 0, Buffer )
> end procedure
> onTimer[ Win ] = routine_id( "Tick" )
>
>
> Heck... im not even *using* the DC I got back (or releasing it). Just
> added
> the variable and grabbed the DC and everything clears right up. Dont
> really
> know how this applies to the Win32Lib routines though. Since I am not
> releasing it I would think it would just create even more leaks...
> *shrug*
>
>
>
8. RE: win32lib v0.57.7
>Hi Don,
>
>What version of Win are you running? Grabbing the DC does nothing to
>correct the problem on my Windows 98 [Version 4.10.1998] system. But if
>it's really true, it might be one heck of a clue in the right direction.
>
>-- Brian
I will check the version when I get back home. As expected, it was slowly
(but surely) eating through my memory because I was never releasing it. Of
course my Win98 machine has been hacked and modified by me so many times,
mine could very well be a fluke.
9. RE: win32lib v0.57.7
Don, Brian, Derek, et al....
I too have experienced this problem, specifically when using
drawRectangle repeatedly. I have contacted Derek about this as well.
The same code SEEMS to run fine on Win2000 Pro but me gut feeling is
that Win2K, ME and XP all have move headroom as far as resources are
concerned.
I have both Win95B and Win98 at home. I see the same problem on both
machines. I would be more than happy to help track this down if someone
could provide how to do this. I don't even know where to start.
I'm especially interested in seeing this get resolved as it is delaying
my completion of a win32lib data grid control.
Jonas
> I will check the version when I get back home. As expected, it was
> slowly
> (but surely) eating through my memory because I was never releasing it.
> Of
> course my Win98 machine has been hacked and modified by me so many
> times,
> mine could very well be a fluke.
>
>
>
10. RE: win32lib v0.57.7
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_000_01C20DA3.2F6195B0
charset=iso-8859-1
Thank you very very much Martin. I haven't looked into your research yet,
but it sounds promising.
I really only get to look at the library code on the weekends and then only
when the kids are asleep or otherwise occupied. So I haven't had a lot of
'quality' time to investigate the issue, that's why I asked for help.
I agree that the resource management needs a major overhaul, and that is now
on the TODO list.
---------
Derek.
> -----Original Message-----
> From: Martin Stachon [mailto:martin.stachon at worldonline.cz]
> Sent: Friday, 7 June 2002 3:16
> To: EUforum
> Subject: Re: win32lib v0.57.7
>
>
>
> Derek wrote:
> > The demo program PRETEND.EXW is a prime example that causes
> Windows to crash
> > due to resource leaks.
>
> This is what I think:
>
> The resource leak causes createPen() called in drawLine()
> and drawRectangle(). It creates a pen, but it is never
> released. The replaceObject()
> for this purpose doesn't work, because SelectObject doesn't
> return the previous
> pen (we created), but the default (I guess) - because we
> releasedDC() since then,
> and the DC info for this object was reseted.
>
> The error doesn't come up on Win2k, because it supports much
> larger number
> of resources.
>
> The bug similiar to the one reported a short time ago with
> wPuts(). I think
> this shows win32lib resource managment really needs to be
> improved (refcount
> for getDC/releaseDC). It's easy to say from my place, when I am not
> the maintainer - so good luck Derek !
>
> Martin
>
>
>
>
==================================================================
De informatie opgenomen in dit bericht kan vertrouwelijk zijn en
is uitsluitend bestemd voor de geadresseerde. Indien u dit bericht
onterecht ontvangt wordt u verzocht de inhoud niet te gebruiken en
de afzender direct te informeren door het bericht te retourneren.
==================================================================
The information contained in this message may be confidential
and is intended to be exclusively for the addressee. Should you
receive this message unintentionally, please do not use the contents
herein and notify the sender immediately by return e-mail.
==================================================================
------_=_NextPart_000_01C20DA3.2F6195B0
Content-Type: application/ms-tnef
11. RE: win32lib v0.57.7
Martin Stachon wrote:
> I think this shows win32lib resource managment really needs to be
> improved (refcount for getDC/releaseDC). It's easy to say from my
> place, when I am not the maintainer - so good luck Derek !
>
> Martin
I'd have to agree here. I might admit defeat on this problem because
frankly I'm quite lost in your resource tracking methodology. I just
hope 'improved' implies 'simplified' otherwise I won't be able to
provide much support down the road...
-- Brian
PS: following 'grabbedHDC' is giving me problems at the moment (and just
might be the cause of this issue???) When did 'replaceObject' go from 2
lines to the current chunk of code?
12. RE: win32lib v0.57.7
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_000_01C210CB.E3D46BF0
charset=iso-8859-1
> -----Original Message-----
> From: Wolf [mailto:wolfritz at KING.IGS.NET]
> Sent: Saturday, 8 June 2002 8:20
> To: EUforum
> Subject: Re: win32lib v0.57.7
>
>
>
> > I'd have to agree here. I might admit defeat on this
> problem because
> > frankly I'm quite lost in your resource tracking
> methodology. I just
> > hope 'improved' implies 'simplified' otherwise I won't be able to
> > provide much support down the road...
>
> Sad, but I'd have to agree here. The 'rush' to add more and
> more 'exotic' and seldom used features is putting
> reliability and adequate documentation further and further
> into the background. I've almost given up on win32lib, using
> API directly instead, because its easier to find my 'bugs',
> than try to deal with win32lib's.
> ... or, to put it another way, my simple-mindedness can't
> follow the entangled win32lib web anymore...
>
> ( jeepers, now I'm starting to 'nudge'... like Euman... :( )
>
Thanks for this direction. I'm a big fan of simplification.
How about the idea of stopping "enhancement" until it stabilises, then only
adding new things when they are put to a vote and passed. That is, giving
new ideas a real user-influenced priority system.
---------
Derek.
==================================================================
De informatie opgenomen in dit bericht kan vertrouwelijk zijn en
is uitsluitend bestemd voor de geadresseerde. Indien u dit bericht
onterecht ontvangt wordt u verzocht de inhoud niet te gebruiken en
de afzender direct te informeren door het bericht te retourneren.
==================================================================
The information contained in this message may be confidential
and is intended to be exclusively for the addressee. Should you
receive this message unintentionally, please do not use the contents
herein and notify the sender immediately by return e-mail.
==================================================================
------_=_NextPart_000_01C210CB.E3D46BF0
Content-Type: application/ms-tnef
13. RE: win32lib v0.57.7
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_000_01C210D6.C953BFF0
charset=us-ascii
> -----Original Message-----
> From: Wolf [mailto:wolfritz at KING.IGS.NET]
> Sent: Tuesday, 11 June 2002 9:18
> To: EUforum
> Subject: Re: win32lib v0.57.7
>
>
Yes, and thank you very much.
> Not sure I'd take that route too seriously,
> 'Committee's' always seem to kill things,
Not in every case. I have actually had some positive experiences from well
run 'committees'. The trick is to maintain openness and focus.
> it's just... you know...
> ... we hate to see it 'broken' for so long !
Me too!!!! The reason it is taking so long is that I can't work on it during
the week, only at weekends and then only for a small portion of time. I'm
very sorry that this is how is has to be.
> I am, however, almost temped to sponsor a 'contest' to see
> who can break that *huge* HTML into 5 or 6 nicely hyperlinked
> chunks with a nifty little parser, asap.
Is the size of the HTML docs an issue? I don't recall anyone mentioning
that. I can break it up very easily if that is the case. Anyone else want
his done?
--------
Derek.
==================================================================
De informatie opgenomen in dit bericht kan vertrouwelijk zijn en
is uitsluitend bestemd voor de geadresseerde. Indien u dit bericht
onterecht ontvangt wordt u verzocht de inhoud niet te gebruiken en
de afzender direct te informeren door het bericht te retourneren.
==================================================================
The information contained in this message may be confidential
and is intended to be exclusively for the addressee. Should you
receive this message unintentionally, please do not use the contents
herein and notify the sender immediately by return e-mail.
==================================================================
------_=_NextPart_000_01C210D6.C953BFF0
Content-Type: application/ms-tnef
14. RE: win32lib v0.57.7
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_000_01C210DF.065AA370
charset=iso-8859-1
> -----Original Message-----
> From: Wolf [mailto:wolfritz at KING.IGS.NET]
> Sent: Tuesday, 11 June 2002 10:24
> To: EUforum
> Subject: Re: win32lib v0.57.7
>
>
>
> > Is the size of the HTML docs an issue? I don't recall
> anyone mentioning
> > that. I can break it up very easily if that is the case.
> Anyone else want
> > his done?
>
> It's been mentioned by others. On a P166/48meg/IE5.01, I can
> go for a cup of coffee before it finds the first jump-to...
>
Okay! I get the message (this time). I'll split it up along "topic" lines.
In doing so, I'll also review the topic contents and rationalise these if
required.
---------
Dere.
==================================================================
De informatie opgenomen in dit bericht kan vertrouwelijk zijn en
is uitsluitend bestemd voor de geadresseerde. Indien u dit bericht
onterecht ontvangt wordt u verzocht de inhoud niet te gebruiken en
de afzender direct te informeren door het bericht te retourneren.
==================================================================
The information contained in this message may be confidential
and is intended to be exclusively for the addressee. Should you
receive this message unintentionally, please do not use the contents
herein and notify the sender immediately by return e-mail.
==================================================================
------_=_NextPart_000_01C210DF.065AA370
Content-Type: application/ms-tnef
15. RE: win32lib v0.57.7
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_000_01C2129E.EB2F6600
charset=iso-8859-1
Yes to both questions.
> -----Original Message-----
> From: Juergen Luethje [mailto:jluethje at gmx.de]
> Sent: Thursday, 13 June 2002 15:27
> To: EUforum
> Subject: Re: win32lib v0.57.7
>
>
>
> Hello Dan,
>
> you wrote:
>
> > Juergen,
>
> > But isn't this how it's being done now?
>
> I don't know.
>
> > On the RDS site is the latest *official* version of
> Win32Lib, v 0.55.1,
>
> There is no resource leakage or other known serious bug in
> the "official"
> version?
>
> > and at Derek's site is the latest "bleeding edge" version, 0.57.7,
> > for which he asks for help debugging.
>
> And it is planned, that this will become the "official" version, only
> without serious bugs like resource leakages?
>
> > Only problem I see with this is that we find the features of 57 so
> > interesting we start to use it before it's stable & finished! :)
>
> If the answers to both questions above is "yes", than I agree
> with you.
>
> > Dan Moyer
>
> Best regards,
> Juergen
>
>
>
> > ----- Original Message -----
> > From: "Juergen Luethje" <jluethje at gmx.de>
> > To: "EUforum" <EUforum at topica.com>
> > Sent: Wednesday, June 12, 2002 8:03 AM
> > Subject: Re: win32lib v0.57.7
>
> <snip>
>
> >> How about maintaining two different versions of Win32Lib:
> >> one conservative version ("conservative" in it's best sense),
> >> let's say "Win32Lib classic", and one experimental version,
> >> let's say "Win32Lib future".
> >>
> >> "Win32Lib classic" should be as stable and reliable as possible,
> >> "Win32Lib future" could be used to experiment with cool new
> >> features and new controls.
> >> New features will only be taken over from "Win32Lib future" to
> >> "Win32Lib classic" after they didn't cause any problem for
> >> let's say 3 months.
> >>
> >> Doing it this way will be more work, but one person could maintain
> >> "Win32Lib classic", and someone else could maintain
> "Win32Lib future".
> >> In this case, both should work close together, of course.
> >>
> >> Just some ideas.
> >> Best regards,
> >> Juergen
>
>
>
>
==================================================================
De informatie opgenomen in dit bericht kan vertrouwelijk zijn en
is uitsluitend bestemd voor de geadresseerde. Indien u dit bericht
onterecht ontvangt wordt u verzocht de inhoud niet te gebruiken en
de afzender direct te informeren door het bericht te retourneren.
==================================================================
The information contained in this message may be confidential
and is intended to be exclusively for the addressee. Should you
receive this message unintentionally, please do not use the contents
herein and notify the sender immediately by return e-mail.
==================================================================
------_=_NextPart_000_01C2129E.EB2F6600
Content-Type: application/ms-tnef
