But int is already signed (as Sieghard pointed out) so sint is bad.
==
That signedness isn't more than just a old (and not obvious) assumption.
--
November Webinars for C, C++, Fortran Developers
Accelerate
Use the watch window if you want to see processed
=
Very not handy :)
--
November Webinars for C, C++, Fortran Developers
Accelerate application performance with scalable programming models. Explore
Hmm, it only happens with forms having tgroupbox widget(s) on them on
disabling these groupboxes. 100% reproducible in the PODPISKA.
--
November Webinars for C, C++, Fortran Developers
Accelerate application performance
Is it called because the form looses input focus?
==
There's no user interaction (where the form may loss input focus)
between OnEventLoopStarted and disabling widget in code. And why does
it only happens on disabling groupboxes ?
Check amodalresult for mr_canclose in
You require theater seat numbers to start at 0 for each row as well?
Sine we'll have bit counts etc starting from 0 then it's more
consistent to count everything numeric from 0 as well :)
--
November
The ALGOL syntax looks as a complication seems to offer no benefits :)
For advanced syntax it worhs to look at Python (no begin end source
text garbage etc).
--
November Webinars for C, C++, Fortran Developers
Accelerate
Just have the compiler shift the (imaginary) base address
=
It may break connections with numbering in real-world task. What for
to do it if keeping this connection doesn't slow down performance ?
--
November
So card32 instead of cardinal32
=
Sure, all shortened types should have similar length shortened parts :
integer = int
cardinal = card
word = (no need to shorten)
--
November Webinars for C, C++, Fortran
See the attachment.
attachment: stack_wrong_nonlatin.png--
November Webinars for C, C++, Fortran Developers
Accelerate application performance with scalable programming models. Explore
techniques for threading, error
BTW, shortcuts for range check might be very handy :
---
if i in 1..10 then
---
or
---
if i between 1 and 10 then
---
less cumbersome than
-
if (i = 1) and (i =10) than
-
/lib/common/i18n/mseucs2ru.pas :
---
const
cp866_2: array[$2550..$256C] of byte = (
$cd,$ba,$d5,$d6,$c9,$b8,$b7,$bb,$d4,$d3,$c8,$be,$bd,$bc,$c6,
$c7,$cc,$b5,$b6,$b9,$d1,$d2,$cb,$cf,$d0,$ca,$d8,$d7,$ce
);
cpUCS2_1: array[$b5..$be] of longword = (
Reduce the language concepts one has to learn to the minimum
==
Typedefs aren't language complications - they just use the language.
Shortcuts are very handy for instance @ instaed of addr()/
--
November
Hmm, arbitrary array beginning presents one of specific Pascal
features sometimes very handy for expressing entities if the real
world. It's usually well optimized (acccess via base register,..).
While accessing may be emulated by index shift but it's definetely an
workaround.
may be emulated by index shift but it's definetely an workaround.
==
And calculating this shift on each access eats some CPU ticks :)
--
November Webinars for C, C++, Fortran Developers
A strange idea :) Why not to combine ranged ingtegers, enums sets
in a single type ?
--
November Webinars for C, C++, Fortran Developers
Accelerate application performance with scalable programming models. Explore
Ranging for integers may also help to build not yet hardware suported types,
like :
-
var
i1: integer from 0 to 2^^256; // 2prw256
i2: integer:256;
---
Sure, with emulated access to them.
--
Why do you suggest word for unsigned number?
==
word - longword(=dword = cardinal) - qword..
--
November Webinars for C, C++, Fortran Developers
Accelerate application performance with scalable
Cardinal is 32 bits only as long as word expresses unsiged types
of 2,4,8 so on bytes length.
That's currently in Pascal : *int=signed and *word=unsigned (with
aliaces shortint=int8,integer=int16, byte=1/2word,
cardinal=dword).
integer = sint32;
What for ? How about longest harware supported type (found by compiler
automatically for target architcture ) ?
--
November Webinars for C, C++, Fortran Developers
Accelerate application
Why do you want that integer has different sizes on different platforms?
==
The BASE (no user attention about range) integer. Not ranged one.
{$if REGISTERSIZE = 64}
==
User attention about range. Knowing underlying arch. Can be (and will
be) done by compliler
Both my large projects PODPISKA ROZNITSA don't care about
value ranges - they just rely on the default INTEGER type LONGINT (or
CARDINAL for unsigned data) fits any data of these projects.
Compile in default TP mode and you will be surprised. ;-)
What for ? It's hardly
We may use them as base types to use directly (with exact or default
ranges) or to spawn exact types (globally) so that people used to the
exact types to feel themselves in known environment.
--
November Webinars for C,
two concepts
===
If handled in same manner by compiler - two syntaxes for same thing ?
--
November Webinars for C, C++, Fortran Developers
Accelerate application performance with scalable
Do you use Pascal subrange type? Would you miss it?
=
Me use one in other languages - in BASH :
-
for i in `seq 1 10`; do
echo $i
done
-
for instance.
in SQL :
[..] where i between 1 and 10
--
PL/PgSQL
two syntaxes for same thing
Which should be avoided.
---
Why ? Just shortcuts (short/simplified notations) for user convinience :)
--
November Webinars for C, C++, Fortran Developers
Accelerate application
[index] is just an operator which can be redefined (overloaded) to a
function etc providing needed access (a compiler support issue). At
least it works in C++.
Me'm just affraid of UTF8 may become mandatory standard in 3..5 years
and MSElang may obsolete once ready (3..5 years too as claimed).
For new language, why not to leave only one size per type - of max
register length of target arch (may be controled by compiler options)
?
X86_32 X86_64: bool(=bool32), uint(=uint64), sint(=sint64),
float(=float64) ...
ARM-32 boxes: bool(=bool32), uint(=uint32), sint(=sint32), float(=float32)
That can not be done with perfect performance.
===
Yes, but there must exist (or to be invented) special techinques for
optimizations to some acceptable level.
Now approx 100% of WWW are utf8, most DB setups are UTF8, most text
editors are UTF8, most LINUX installations are
type
integer = sint64;
===
Not portable between architectures.
--
November Webinars for C, C++, Fortran Developers
Accelerate application performance with scalable programming models. Explore
techniques
String performance must be perfect, acceptable is not acceptable
Database servers handle UTF8 efficiently. Also they provide infexing (
hashing for textual data) - these special tchiniques...
Again the question, do you think working with current msestring is comfortable?
{$if CPUSIZE=64}
integer = sint64;
{$else}
integer = sint32;
{$endif}
==
It's the deal of compiler swicthes which should be done automatically
by build environment.
--
November Webinars for
Probably needs a wider discussion even flaming debates - on
FreePascal.Ru etc...
--
November Webinars for C, C++, Fortran Developers
Accelerate application performance with scalable programming models. Explore
techniques
var
i1: integer from 1 to 10; // assigned to UINT8 by compiler
i2: cardinal; // assigned to UINT(max_len) by compiler
i3: integer; // assigned to SINT(max_len) by compiler
begin
i1:= 0; // compile time error
i1:= 12; // compile time error
i1:= 3; // OK
Run-time range checking (it can slow down target program) might be
switched on/of by a compiler option or $ifdef.
--
November Webinars for C, C++, Fortran Developers
Accelerate application performance with scalable
In daily work the size of the base types should be unambiguously visible.
==
How about its replacement by explicit range defining/checking
(possibly with the mention reversal typedef for simplier syntax) ?
The same + some more (compiler) functionality.. And reading letter
by default
which provides faster CPUs
Which shouldn't be ignored avoided :)
Even UTF32 does not help here.
=
Once we contact with aliens (and their alphabets) then UTF8 or UTF16
become only choices.
I vote for adding Syntax for parallel processing
==
It can mainly be useful for parallel-aware languages like Haskell
(which prevents form creating state-machine applications badly
compatible with paralleizing ).
in utf-16 they fail for non BMP characters.
==
Then the NEW (free to do anything) compiler should overload the index
[] access to strings so that they don't fail. Otherwise we won't
have REALLY (crrently in the future) unicode environment :)
but as it eats up lots of
Or do you think MSEgui DB-apps are slow?
Not investigated on large data yet :) And the test application should
perfom a lot of transcoding.
--
November Webinars for C, C++, Fortran Developers
What is wrong to read in a letter the word sint32 instead of integer?
===
Means CPU architecture knowledge etc special skills. Specifying a
range instead is more meaningful. And the range supplies all needed
info to the compiler so that it makes proper
Every programmer (even the beginner) must know that CPU registers are
organized in multiples of bytes.
Java, Perl, Python, PHP, TCL, Haskell,..
MSElang must be able to program microcontrollers
There might be type definitions for ranges:
type
int8 =
Of course in a similar way also parallel execution of different code
snippets can be implemented, making the language a parallel-aware
language.
There's also the conception of pure function - which mean such
function doesn't use doesn't change any out-of-function
Probably there will be no integer
Why ? The vast majority of programs don't need typelength types
for defining varibales. They just specify type to express NATURE (
and behaviour ) of varibales. Me sugget it to be longest harware (of
target arch) supported by default.
type
uint8 = 0..$ff;
=
Then how to differentiate integers from floats for instance? To write
smth like :
float8 = 0.0..127.0;
uint8 = 0..127;
It looks weird :)
--
November Webinars for C,
Illegal. Range values must be integers.
===
Different rules for similar types ?
--
November Webinars for C, C++, Fortran Developers
Accelerate application performance with scalable programming models. Explore
against the bit numbers in type names.
==
Sure. Defined globally :
type
uint8 = integer 0..127; // uint8
[..]
Me just want to leave as low number of base types as possible so that
they only express nature and arbitary (not only 2pwr*) boundaries of
variables that's what
If it's possible to fix all UTF8 drawbacks won't it be better to
swicth to UTF8 in the new compliler ? This encoding can extend/change
safely.
--
November Webinars for C, C++, Fortran Developers
Accelerate application
What means swicth to UTF8 in the new compliler ?
==
msestring = utf8string.
Sure provided that the utf8 string RTL gets fast and there'll
redefined operator [i] for indexed char access at reasonable speed.
Sure, otherwise there'll be question - what is contained in item[0] ?
--
Android is increasing in popularity, but the open development platform that
developers love is also attractive to malware creators. Download this
Looks fixed :)
--
Android is increasing in popularity, but the open development platform that
developers love is also attractive to malware creators. Download this white
paper to learn more about secure code signing
function PrintFile(afile: msestring; out aprintcommand: msestring;
atimeout: integer = 0): integer;
var
print_cmd:msestring;
usegui: boolean;
print_ph: integer;
begin
print_cmd:= GetFinalCommand(afile, usegui);
aprintcommand:= print_cmd;
if print_cmd '' then begin // if(1)
if usegui
procedure tdmmainmo.f18monthlistcalcfields(DataSet: TDataSet);
begin
with dataset do begin
edit;
fields[1].asmsestring:= monthnames1[fields[0].asinteger];
post;
end;
end;
SF occurs atter multiple calls of OnCalcFileds. The stack :
#0 0040C3CE :0 fpc_initialize()
#1
edit() and post() in oncalcfields? That is forbidden. I'll add a check an
throw an exception.
==
Otherwise fkInternalCals-fields aren't editable.
--
Android is increasing in popularity, but the open
fkinternalcalc are normal fields
==
We used ones for storing data not related to SQL-data - boolean
switches for records (for group selections) etc. Two my big projects
are based on this feature.
Today, OnChanged event handler (see below) for fkInternalCalc-field
can't be
procedure tdmmainmo.f18monthlistcalcfields(DataSet: TDataSet);
begin
with dataset do begin
fields[1]{fkCalculated}.asmsestring:= monthnames1[fields[0]{regular
field}.asinteger];
end;
end;
fields[2] is fkInternalCalc.
Looks fixed :)
--
Android is increasing in popularity, but the open development platform that
developers love is also attractive to malware creators. Download this white
paper to learn more about secure code signing
the subj.
--
Android is increasing in popularity, but the open development platform that
developers love is also attractive to malware creators. Download this white
paper to learn more about secure code signing practices
Did You shift mouse position before undeleting ?
--
Android is increasing in popularity, but the open development platform that
developers love is also attractive to malware creators. Download this white
paper to learn
The PODPISKA project :
//
procedure tloginfo.changedbexec(const sender: TObject);
begin
if showmodalwidget(tconnsetupfo) = mr_ok then begin // cancelled
here
self.window.modalresult:= mr_cancel;
application.terminate;
end;
- tried to load a report into IDE, since this report had
optionwidget[1] backward errors so its MFM-file was loaded as text for
manual fixing
- after fixing saving switching to form designer view (Show as
form), each SaveAll switches this file back to text view
Oops, really this input form was cancelled :
---
function checkmd5password(
const psw_digest: TMDDigest;
passwordencfunc: passwordencfuncty
): boolean;
var
fo: tstringenterfo;
res: modalresultty;
begin
repeat
result:= false;
Curious why you need an undelete option
===
Misclicked Cut Copy (placed closely).
--
Android is increasing in popularity, but the open development platform that
developers love is also attractive to
Please test.
===
What ? Whether are both Oncreate OnCreted called before OnLoaded
both for regular inherited forms ?
--
Android is increasing in popularity, but the open development platform that
1. How to cause toolbar-button to highlight its using visually (via
simulating mouse click) if activated via a shortcut ?
2 . How to know action of which button is curently active ?
There's an author's testcase on the site.
Proper LINUX non-latin setup for scalable fonts means deep knowledge.
(partially described in the PERSONS tutorial, see MSEdocumenting )
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate
Does it worth to strive ? At home everything was OK.
Though, possibly order of touching had the effect.
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore
connsetupfo.oncreate calls checkmd5password4file which in turn
creates fo:= tstringenterfo.create(nil) internally and this internal
form doesn't draw properly (no button captions etc). It means that we
can't use _any_ (even MSEgui's built-in) form-drawing GUI code in
form.OnCreate anymore, correct
Is it as designed ?
Yes.
Then it seems to be a newly introduced design limitation. Should be
mentioned in the README.
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application
All fields of qryAcnt1NodesList are ftString.
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get
Use asmsestring. widestring unicodestring on Windows.
Throughout the whole project for sure ?
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application
So we have 2 order options available - oncreate-oncreated-onloaded and
oncreate-onloaded-oncreated.
How to know if load = true to choose correct one?
--
October Webinars: Code for Performance
Free Intel webinars can help
tform.create() normally calls tform.create(load=true).
Can this (normal) behaviour be changed (to unnormal) by user options ?
Which widgets (datamodules) behave differently ?
--
October
That's with intention to do that in code not by [accidental]
MFM-options, correct ?
Do datamodules behave same ?
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application
Between
1) So the order of oncrated/onloaded depends on
the loading state, oncreate is always first.
If load=false the order is oncreate-onloaded-oncreated.
and
2) If load=true the order can be oncreate-onloaded-oncreated or
oncreate-oncreated-onloaded. tform.create() normally calls
Just for migrating from the older design's OnCreated OnLoaded, will
their swapping be correct
use oncreate.
==
What about oncreateD ?
--
October Webinars: Code for Performance
Free Intel webinars can help you
What about oncreateD ?
Seldom used.
It means that oncreate-oncreated-onloaded =
oncreate-onloaded-oncreated since oncreated isn't used in the PODPISKA
so there's nothing to swap, correct ? Are there scenarios when
OnLoaded fires before OnCreate ?
Please use the mentionend criteria (ex.
Thanx !
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel
Me needed some time to prepare the DB setup on my home machine. The results :
Now forms loads well even in the master branch, but the project
fails to compile from IDE.
-
Free Pascal Compiler version 2.6.2 [2013/02/12] for i386
Copyright (c) 1993-2012 by Florian Klaempfl
As to the slowliness while moving (even feasible on my quite videcard
powerfull home machine), me'd suggest to enlarge move shift to redraw
that's to recalculate redraw not for each pixel move.
--
October Webinars: Code
It the $TARGET macro correct for MSEgui 3+ for Windows ?
Since the project has ${MSELIBDIR}kernel\$TARGET\ in MAKE paths where
the macro seems to equal to i386-win32 whereas unit msesysintf
resides in ${MSELIBDIR}kernel\windows\. So we have a compile error.
libpq.dll not found for everywhere places in Windows and the projct directory.
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP,
Possibly because me took them from the PG Server instalation (x86_64).
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced
Hmm some artifacts on loading the project - no text on buttons of
fo:= tstringenterfo.create(nil); in
function utils.passcheck.checkmd5password(
const psw_digest: TMDDigest;
passwordencfunc: passwordencfuncty
): boolean;
//passwcheckresult;
var
The project doesn't build with MSE 3.0 fpc 2.6.2 (the binaries-only
vanilla version).
Download the below file :
http://www.likan.uz/uploads/podpiska_reports_ref.rar
unpack run build.cmd (win32).
PS:
Also it seems that OnLoaded OnCreated are swapped, correct ? For which
components ?
Martin
Hmm, me skipped ,spacedb, when searching replacing.
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and
/utils/dmprint.pas:197 - print_ph:= execmse2(print_cmd);
The former version differs from getting print_ph used in the other
condition branch.
//
if usegui then begin // if(2)
// print_ph:= execmse2(print_cmd,nil,nil,nil,false,-1,false);
Restore the project files from git to be sure nothing has
been changed by accident. Now the forms can be updated step by step.
Since there can be 2+ same errors per MFM-file, could cursor be set to
exact error position upon closing message about that error ? It'll greatly
simplify fixing
AV on touching refs/common/refviewform_intcode or another...
Me loaded a group of forms then touched these forms one by one then closed
saved on request so on.
--
October Webinars: Code for Performance
Free Intel webinars
The problem is refs/common/refviewform - touching it 100% gives an AV.
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced
I'll check if I can reactivate the transition code in 3.0 for you. :-)
git checkout optionswidget ?
Still AV on touching refs/common/refviewform*.
--
October Webinars: Code for Performance
Free Intel webinars can help
I'll have a look.
Possibly because me have the design time DB connection working. Me'll
provide its database dump if needed.
--
October Webinars: Code for Performance
Free Intel webinars can help
Needed to recover a project last time able to built successfully those days.
(it's needed to keep both versions of project open simultaneously to
follow the way of old project to brick the new one)
Or (much better) another way of migrating projects as MSEgui API
changes significally. Trying to
git revert 8f5a3bcc4b052b0dae4007c1b38e8d28fa5bba04 (01 May 2010)
fails on
error: could not revert 8f5a3bc... + Interface support for
tpascaldesignparser.
--
October Webinars: Code for Performance
Free Intel webinars can
git reset 8f5a3bcc4b052b0dae4007c1b38e8d28fa5bba04 does the magic :)
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced
Hmm, it requires MCLASSES etc to build. Also complains on removed
optionswidget. Any hints ?
2013/10/3, Ivanko B ivankob4m...@gmail.com:
git reset 8f5a3bcc4b052b0dae4007c1b38e8d28fa5bba04 does the magic :)
--
October
Another try:
# git reset --hard 8f5a3bcc4b052b0dae4007c1b38e8d28fa5bba04
# git clean --force
It seems to remove the MCLASSES deps.
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate
What to fix in old (y2010) MSEgui files to cope the MSEide fails to
compile... finddir() doesn't match compile error with fpc 2.6.2 ?
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate
PS: It's our lovely PODPISKA :)
Still alive! :-)
Together with ROZNITSA - because they have been using every second in work
day. Now people (and my their chiefs) insist on combining these projects
in one + some fixes improvements.
I have never heard of the crystal effect before. No idea what it is, sorry.
==
That mean the effect was really exaggerated.
DELL U2317H(M)
==
Oops,sure - me did a mistake. Cheaper U2713HM (edge W-LED, 16mln
color) pro U2713H (matrix BG-LED, high-contrast, 1bln color)
They also changed screen coating - possibly because of the mentioned
crystall effect many sensitive people complain on. When selectiong a
new monitor, me also got affraid of these (as revealed today - vaste
ones) complains.
301 - 400 of 964 matches
Mail list logo