Is it time for a Delphi algorithms group?
John Herbster
Is it time for a Delphi algorithms group, yet?
(Thanks for your consideration -- JohnH)
Lucian.R
May be we talk about a extension of the actual algorithms, or may be about a
future version of them...or may be I'm wrong, I believe we're still
depending on them.
Rgds,
Lucian R.
"Charles Appel" <charle...@mindspring.com> wrote in message
news:41bd...@newsgroups.borland.com...
> "John Herbster" <herb-sci1_AT_sbcglobal.net> wrote in message
> news:41bd...@newsgroups.borland.com...
> >
> > Is it time for a Delphi algorithms group, yet?
>
> Sounds like a good idea.
>
> --
> Charles Appel
> "A generation which ignores history has no past - and no future."
> Robert Anson Heinlein
>
>
Charles Appel
news:41bd...@newsgroups.borland.com...
>
Sounds like a good idea.
Lucian.R
time I'll pay more attention.
Sorry.
Rgds,
Lucian R
"Lucian.R" <rlu...@easynet.ro> wrote in message
John Herbster
> Is it time for a Delphi algorithms group, yet?
Read here for some previous discussion on the subject:
http://groups-beta.google.com/groups?as_q=algorithms+delphi&as_epq=&as_oq=group+forum&as_eq=FAQ&as_ugroup=borland.*
Mark Vaughan
@newsgroups.borland.com:
> Is it time for a Delphi algorithms group, yet?
yes, definitely!
--
Mark Vaughan
___________
Visit the Numerical Methods in Pascal web page at
http://www-rab.larc.nasa.gov/nmp/fNMPhome.htm
Alisdair Meredith [TeamB]
> Is it time for a Delphi algorithms group, yet?
<takes off TeamB hat>
Creating new groups is never as easy as it seems, for all sorts of
reasons, so I'm definitely *not* wading in with a TeamB hat here <g>
But I'm going to answer your question with another:
Why a *Delphi* algorithms group? What is so special about the
algorithms you are thinking of that they might only be implemented in
Delphi?
At their heart, algorithms are an expression of a way of finding a
result, language issues are an implementation detail.
While I am not sure a Delphi algorithms group would generate enough
traffic, I would be very interested in a general group for algorithms,
where language specific implementation details were not off topic, but
also not the main traffic in the group either.
Of course the same could be said for OO, patterns, GUIs, ...
This is one (of many) reason creating new groups is hard, balancing
their interest to get sufficient-but-not-overwhelming traffic.
AlisdairM
<dons TeamB hat again>
Charles Appel
news:Xns95BE95B...@207.105.83.66...
> John Herbster wrote:
>
> > Is it time for a Delphi algorithms group, yet?
>
of
> limited use.
True. That could be a problem. A possible solution would be to grant
blanket permission to post the algorithms from the group to personal
Delphi web sites. Periodically, a list of sites could be posted in the
group.
Charles Appel
<alisdair.NO.SPAM...@uk.renaultf1.com> wrote in message
news:41bd...@newsgroups.borland.com...
> John Herbster wrote:
>
> > Is it time for a Delphi algorithms group, yet?
>
> <takes off TeamB hat>
> Creating new groups is never as easy as it seems, for all sorts of
> reasons, so I'm definitely *not* wading in with a TeamB hat here <g>
>
> But I'm going to answer your question with another:
> Why a *Delphi* algorithms group? What is so special about the
> algorithms you are thinking of that they might only be implemented in
> Delphi?
The problem is that it is often hard to find code in Delphi - where it
is easy to find the same algorithm in C/C++ or BASIC.
> At their heart, algorithms are an expression of a way of finding a
> result, language issues are an implementation detail.
True, but that detail can be quite a problem. Translating C/C++
to Delphi is not always the easiest of tasks - especially if you
have never studied C or haven't used it in ten years or more.
Translating BASIC is generally easier, but this still requires that
you have at least a working knowledge of BASIC.
> While I am not sure a Delphi algorithms group would generate enough
> traffic, I would be very interested in a general group for algorithms,
> where language specific implementation details were not off topic, but
> also not the main traffic in the group either.
Perhaps. But I'd still rather have the Delphi group.
Martin James
found were C-based, ie. actually written in C, or in C-style Pascal, ie p^,
s^, q^, AKA unfollowable spag.
Also, all the examples were based around records and pointers. Not a class
or object in sight.
Rgds,
Martin
Marcus F.
> Is it time for a Delphi algorithms group, yet?
What kind? Graphics? WinAPI related? Database? OO? General programming?...
Wait a second, aren't there newsgroups for these already?
John Jacobson
"John Herbster" <herb-sci1_AT_sbcglobal.net> wrote in message
news:41bd...@newsgroups.borland.com...
>
> Is it time for a Delphi algorithms group, yet?
YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
YES!!!!!!!!!1
Lucian.R
Are you talking about algorithms?
In fact what is an algorithm. In my opinion an algorithm is some steps
which must be followed to reach the best result. An algorithm is not only
an existing algorithm with a great name. In our days, things have changed.
Therefore if things have changed, then problems have changed, and the old
algorithms with great and fancy names aren't enough. I must recognize their
great importance, but we need more and more algorithms for more and more
coding situations. I believe also that in this newsgroups are a lot of
beginners who don't know what in fact is an algorithm. Everybody, yesterday,
today, tomorrow needs an alg for his BIG problem. There are a lot of
situations when he (our guy) post his problem, on a newsgroup, and someone
say to him WRONG GROUP. He despaired post his problem in all rest
newsgroups.
In conclusion, and my personal opinion, WE need an ALGORITHM group,
and if not us, THEY.
Regards,
Lucian R.
"Alisdair Meredith [TeamB]"
<alisdair.NO.SPAM...@uk.renaultf1.com> wrote in message
news:41bd...@newsgroups.borland.com...
Rene Tschaggelar
>John Herbster wrote:
>
>> Is it time for a Delphi algorithms group, yet?
>
>Basically it's a good idea. OTOH if the posts expire too fast it will
>be of limited use.
>
>Same for a "Frequently addressed problems" group, eg with the
>infamous "how can I start another process and wait until it's
>finished" and other often- encountered problems.
Ben,
Time to start running a news tool such as Xananews
to gather the news 365x24x7. No ?
Lets see, I have 87k messages just from delphi.non-tech
here. Xananews becomes a memory hog over the years then,
as it wants to have something in RAM.
Rene
--
Ing.Buro R.Tschaggelar http://www.ibrtses.com
Your newsgroups @ http://www.talkto.net
Rene Tschaggelar
You havent looked on my pages then.
With records though, I believe.
Peter Morris [Air Software Ltd]
anyway? They just add OOP to them and call them "patterns" :-)
--
Pete
====
Audio compression components, DIB graphics controls, FastStrings
http://www.droopyeyes.com
Read or write articles on just about anything
http://www.HowToDoThings.com
My blog
http://blogs.slcdug.org/petermorris/
John Jacobson
news:xn0dqyxy5000000i@ibrt_7...
> Time to start running a news tool such as Xananews
> to gather the news 365x24x7. No ?
I think Colin will be the first to tell you that Xananews is not designed
for such large-capacity storage.
> Lets see, I have 87k messages just from delphi.non-tech
> here. Xananews becomes a memory hog over the years then,
> as it wants to have something in RAM.
A newsreader that uses a relational database as the back end can do this
kind of thing very well. I turned off the message purging in my newsreader a
year and a half ago to test exactly this kind of functionality and now it
has something like 600,000 headers stored in it's Nexus* database. Still
quite useable on my 512MB laptop. The architecture is easily expanded to a
multi-tiered or client/server setup so that the newsreader could gather
messages 24x7x365 while a web server dishes out the web pages for newsgroup
web access. When I release my newsreader (there is a small possibility it
will be freeware by the way--haven't decided yet) remind me about this
discussion, so I remember to build a multitier version someone can use for
this purpose.
Note to the curious: I'm once again finding time to code my newsreader, so
it is moving ahead again after a hiatus of several months (moving, job
change, health-related activities, etc.). Writing my own RSS parser for it
(tried third-party XML parsers but they all had problems of various
kinds--sometimes because a lot of RSS feeds are not well-formed XML).
Note to trolls: don't even bother.
*Nexus rocks!!!!
ckd
<herb-sci1_AT_sbcglobal.net> wrote:
>
>Is it time for a Delphi algorithms group, yet?
>
Bucknall's algorithm book is going for around $175USD now. That would
seem to indicate that the topic of algorithms from a Delphi
perspective has some value to developers.
John Jacobson
message news:VA.00001f8...@blueyonder.co.not-this-bit.uk...
> That's going to be a bit tricky. :)
How so?
Derek Davidson
> That's going to be a bit tricky. :)
LOL!
--
Derek Davidson
http://www.ebsms.com
Send SMS Text messages from your PC. For FREE!
Deborah Pate (TeamB)
I think Colin will be the first to tell you that Xananews
is not designed for such large-capacity storage.
>>
That's going to be a bit tricky. :)
--
Deborah Pate (TeamB) http://delphi-jedi.org
TeamB don't see posts sent via Google or ISPs
Use the real Borland server: newsgroups.borland.com
http://www.borland.com/newsgroups/genl_faqs.html
John Jacobson
> I fear the rostrum has already been squatted.
I have no idea what that means.
Jim Gallagher
>
> I have no idea what that means.
It's a wonderful phrase, nonetheless. I think Deborah's point was that
Colin would have to move very quickly to be the first.
-Jim
Deborah Pate (TeamB)
How so?
>>
I think 'becoming the first to do something someone else
has already done' falls into the category of logically
impossible things even the omnipotent cannot do. :)
John Jacobson
message news:VA.00001f8...@blueyonder.co.not-this-bit.uk...
> How so?
>>>
>
> I think 'becoming the first to do something someone else
> has already done' falls into the category of logically
> impossible things even the omnipotent cannot do. :)
I'm more confused now than ever. Where did the "first to do something" part
come from? Others have been archiving and presenting newsgroups, including
Borland newsgroups, for ages now.
John Jacobson
news:41be0b22$1...@newsgroups.borland.com...
> I'm more confused now than ever. Where did the "first to do something"
> part come from? Others have been archiving and presenting newsgroups,
> including Borland newsgroups, for ages now.
Ah, it came from me, when I said Colin would be the first to *say* that
Xananews was not designed for heavy-duty storage, not do it. I should have
prefaced my question above with "What idiot introduced the 'first to do
something' part?", thus engaging in uncanny accuracy in self-description.
Alas, like most imbecility, my own stupidity prevented such self-awareness.
Deborah Pate (TeamB)
Where did the "first to do something" part come from?
>>
Here:
<<John Jacobson:
I think Colin will be the first to tell you that Xananews
is not designed for such large-capacity storage.
>>
--
![Rudy Velthuis [TeamB]'s profile photo](http://lh3.googleusercontent.com/a-/ALV-UjWU6JIYTx6kJCxK1P9qy-0IkTLb3QUOOaGRszn1aPNSjVyUm9M=s40-c)
Rudy Velthuis [TeamB]
> "Deborah Pate (TeamB)" <d.p...@blueyonder.co.not-this-bit.uk> wrote
> in message news:VA.00001f8...@blueyonder.co.not-this-bit.uk...
> > <<John Jacobson:
> > How so?
> > > >
> >
> > I think 'becoming the first to do something someone else
> > has already done' falls into the category of logically
> > impossible things even the omnipotent cannot do. :)
>
> I'm more confused now than ever. Where did the "first to do
> something" part come from?
I had to ask for an explanation too, but this is it: You said: "Colin
will be the first to tell you that ...". But, since you already told it
in the same message, he can't be the first anymore. <g>
--
Rudy Velthuis [TeamB]
"A fast word about oral contraception. I asked a girl to go to bed
with me, she said 'no'." -- Woody Allen
John Herbster
"Marcus F." <a...@b.com> wrote
> What kind? Graphics? WinAPI related? Database? OO?
> General programming?
No!
Rather combinatorics, fitting, random number generators,
optimization, real-time physical modeling, sorting, searching,
rounding, integration, symbolic math, linear equation solving,
LP solving, interpolation, astronomy, civil engineering,
mechanical engineering, strength of materials, etc.
--JohnH
Colin Wilson
> A couple weeks ago, I was looking for Btree algorithm examples.
Martin James wrote:
> A couple weeks ago, I was looking for Btree algorithm examples.
http://www.wilsonc.demon.co.uk/delphi.htm
* TBTree is a disk-based B-Tree index which associates an integer value
with a variable length string. Supports strings up to about 2000
characters. Also allows sorted sequential access of strings and their
associated integers.
* TDataTree is a disk-based B-Tree index which associates a variable
length string value with an integer.
* TIndexTree is a disk-based B-Tree index which associates an integer
value with another integer value - effectively providing a sparse array
of integers.
* TBTreeIterator is a support class that allows you to quickly iterate
through a TBTree
* TDataTreeIterator is a support class that allows you to quickly
iterate through a TDataTree.
--
Colin - Author of XanaNews
--
Colin - Author of XanaNews
Colin Wilson
> I think Colin will be the first to tell you that Xananews is not
> designed for such large-capacity storage.
Yes. I don't keep stacks & stacks of old newspapers in my shed, either!
Instead, it's got a separate area for storing interesting old messages.
Conceptually like a shoe-box full of old newspaper cuttings.
And, of course we've got Google groups. Like having the British
Library on your desktop. (Apart from keeping books, the British
Library archives every newspaper, periodical, magazine, etc. that's
published. Apparently there's even a Porn section (again, just like
Google Groups!) - you just need to ask the Curators to unlock it for
you until you're 'finished' ;) )
Mind you, with this new PC of mine, XanaNews doesn't have many problems
with keeping 100,000 messages per group in the main threaded view.
BTW, how big (ie, how many megabytes) is your 600,000 message database?
Martin James
> http://www.wilsonc.demon.co.uk/delphi.htm
An excellent set of containers!
Unfortunately, my tree must support duplicate entries. Some bodging may be
needed to the R-B Btree :(
I may have a look at the queue too, since TobjectQueue is very slow & keeps
critical sections locked for too long.
Rgds,
Martin
Colin Wilson
> Unfortunately, my tree must support duplicate entries. Some bodging
> may be needed to the R-B Btree :(
Yes - sorry, there are some bugs with duplicates :( I'll sort them
out, or please let me know if you manage to fix them.
> I may have a look at the queue too, since TobjectQueue is very slow &
> keeps critical sections locked for too long.
TObjectQueue? Do you mean TObjectProcessor?
Colin Wilson
> the R-B Btree
An oxymoron? :)
Jan Mitrovics
I'd immediately subscribe to such a newsgroup!
Jan
Eric Grange
Patterns are to algorithms what pipes are to plumbing.
Eric
Uffe Kousgaard
news:41be16f7$1...@newsgroups.borland.com...
>
I agree with this part:
> Rather combinatorics, fitting, random number generators,
> optimization, real-time physical modeling, sorting, searching,
> rounding, integration, symbolic math, linear equation solving,
> LP solving, interpolation
Not this part:
> astronomy, civil engineering,
> mechanical engineering, strength of materials, etc.
You can count on me as subscriber to such a newsgroup.
Regards
Uffe
Hrvoje Brozovic
>
> While I am not sure a Delphi algorithms group would generate enough
> traffic, I would be very interested in a general group for algorithms,
Please don't post anti delphi messages in this group.
Alisdair Meredith [TeamB]
> Please don't post anti delphi messages in this group.
In what way way that an anti-Delphi message?
My concern was that a newsgroup dedicated to algorithms would not
generate enough traffic for any single product, but that a
product-agnostic group would be a good thing.
There is nothing language-specific when understanding an algorithm, it
is a process applied to data. Your representations may be Delphi
specific, C++ specific or even Java based, but discussion of the
algorithm itself would appeal to all communities.
If you are after a group for *implementing* ideas, sure the language
groups already cover that, once you have found the algorithm?
AlisdairM(TeamB)
Deborah Pate (TeamB)
Please don't post anti delphi messages in this group.
>>
That wasn't anti-Delphi. I love Delphi, and I also have
doubts about whether there'd be enough traffic in a
delphi.algorithms newsgroup. (Of course if unsuccessful
newsgroups were culled this would not be an issue.)
Hrvoje Brozovic
>
> > Please don't post anti delphi messages in this group.
>
> In what way way that an anti-Delphi message?
> My concern was that a newsgroup dedicated to algorithms would not
> generate enough traffic for any single product, but that a
> product-agnostic group would be a good thing.
Because your post implies that including other language
products can somehow help group to survive.
From this, unaware reader can conclude that
delphi group is just one of 8 (eight) borland
programming language groups.
I should be obvious to frequent user of Borland forums,
and TeamB members are that by definition, that this is
not a case, so importance of Delphi is neglected in your post.
By neglecting its importance in Borland
product pallete, you are harming Delphi.
Post that harm Delphi are anti delphi posts.
But, just to back up my case, I did a google search
for messages not containing thisisnotaword word,
for posts in last year and alltime posts.
Delphi forums constitute 63% of all time posts,
and 77% of posts in last year alone.
Unfortunatelly, this increase is not because Borland is
pushing Delphi, but because, inspite borland treatment,
it is only one that somehow survives.
Here are results for proportional font
total posts last year posts product
8,140 51 TASM
9,160 229 TurboPascal
1,730,000 259,000 Delphi
64,300 3,120 Kylix
94,300 2,280 Cpp
580,000 58,100 CppBuilder
788 770 CSharpBuilder
265,000 13,500 JBuilder
Here are results for monospace font
8,140 51 TASM
9,160 229 TurboPascal
1,730,000 259,000 Delphi
64,300 3,120 Kylix
94,300 2,280 Cpp
580,000 58,100 CppBuilder
788 770 CSharpBuilder
265,000 13,500 JBuilder
Eric Grange
1:6.6 ratio
> 580,000 58,100 CppBuilder
1:10 ratio
> 265,000 13,500 JBuilder
1:20 ratio
Wooow, that's impressive, especially knowing that Delphi is the oldest
offering of the three, and overall web access (as in 'not Borland-specific')
has been growing at high pace all those years...
Eric
Hrvoje Brozovic
I saw that and commneted partly in my post.
I also did a analisys of last year / total ratio
compared to product age.
Assuming that last year is just another year in
product lifecycle (Borland's marketing plug)
Delphi would be 7 years old, CPB 10, and JB 20.
If you compare with actual age, you get
that last year is 35% percent increase
for delphi, and 20% and 64% decrease
for C and Java builders respectively.
(I assumed real age of 9, 8 and 7 for
Delphi, CB and JB, but know for sure only
for Delphi, since I use CB only because of
tools not included in Delphi, and don't use
JB at all for now)
"Eric Grange" <egra...@SPAMglscene.org> wrote in message news:41be...@newsgroups.borland.com...
Eric Grange
> compared to product age.
For more accurate trend analysis, is it be possible to obtain
activity figures for last 2 years, last 3 years, etc. too in google?
(haven't dabbled too much in the new interface...).
> [...] you get that last year is 35% percent increase
> for delphi, and 20% and 64% decrease for C and Java
> builders respectively.
The JBuilder plung is most impressive, especially since AFAIK,
unlike CPPBuilder, it is not a product that was placed into
suspended animations for years.
Eric
Hrvoje Brozovic
> For more accurate trend analysis, is it be possible to obtain
> activity figures for last 2 years, last 3 years, etc. too in google?
I done this only because it is you, Eric.
As a compensation for all patience with
Rudy in his numerous "I don't get it" threads.
Thousads of messages through years according to google.
There is something strange in google making sum of years
different than anytime, +29K in Cpp and -550 for Delphi.
last year <> 2004 is ok, since it includes last 15 days of 2003.
proportional
CppBuilder JBuilder Delphi Kylix
1995 - - - -
1996 - - - -
1997 12,7 5,3 29,2 -
1998 44,7 29,7 44,4 -
1999 60,7 34,4 61,1 -
2000 80,9 57,8 78,7 0,3
2001 111,0 75,8 110,0 32,3
2002 121,0 41,5 213,0 11,7
2003 122,0 18,9 399,0 5,6
2004 56,1 12,9 246,0 2,9
total 609,1 276,3 1.181,4 52,9
anytime 580,0 265,0 1.730,0 64,3
lastyear 58,1 13,5 259,0 3,1
monotype
CppBuilder JBuilder Delphi Kylix
1995 - - - -
1996 - - - -
1997 12,7 5,3 29,2 -
1998 44,7 29,7 44,4 -
1999 60,7 34,4 61,1 -
2000 80,9 57,8 78,7 0,3
2001 111,0 75,8 110,0 32,3
2002 121,0 41,5 213,0 11,7
2003 122,0 18,9 399,0 5,6
2004 56,1 12,9 246,0 2,9
total 609,1 276,3 1.181,4 52,9
anytime 580,0 265,0 1.730,0 64,3
lastyear 58,1 13,5 259,0 3,1
> (haven't dabbled too much in the new interface...).
> Eric
query is like this, you can easily loop years in
as_maxy and as_miny parameters,
number of hists is between "</b> of" and "from <b>".
http://groups-beta.google.com/groups?as_q=
&num=10
&scoring=r
&hl=en
&ie=UTF-8
&as_epq=
&as_oq=
&as_eq=thisisnotaword
&as_ugroup=borland.public.delphi.*
&as_usubject=
&as_uauthors=&as_qdr=
&as_drrb=b
&as_mind=1
&as_minm=1
&as_miny=2004
&as_maxd=31
&as_maxm=12
&as_maxy=2004
&safe=off
Eric Grange
Thanks :)
Just per chance, you wouldn't happen to have a thousand
euros handy too? ;)
> monotype
> CppBuilder JBuilder Delphi Kylix
> 1995 - - - -
> 1996 - - - -
> 1997 12,7 5,3 29,2 -
> 1998 44,7 29,7 44,4 -
> 1999 60,7 34,4 61,1 -
> 2000 80,9 57,8 78,7 0,3
> 2001 111,0 75,8 110,0 32,3
> 2002 121,0 41,5 213,0 11,7
> 2003 122,0 18,9 399,0 5,6
> 2004 56,1 12,9 246,0 2,9
> total 609,1 276,3 1.181,4 52,9
> anytime 580,0 265,0 1.730,0 64,3
> lastyear 58,1 13,5 259,0 3,1
Now that's interesting, personal analysis:
Apparently JBuilder started withering in 2001-2002, but according
to reports, financial hit of similar magnitude occured only a year
or two later. In a few years, chances are it'll have disappeared
from Borland products lineup.
The CPPB trend indicates that by not issuing new releases, Borland
pretty much slaughtered the goose as it used to follow the Delphi
curve nicely, insufficient compiler improvements probably did the rest...
wether or not the recent announce will be able to revive it is
anybody's guess.
Delphi's curve seemingly hit a wall in 2004, which broke its almost
geometric curve so far. If JBuilder behaviour repeats, Borland won't
feel the actual pain in their sales before 2005-2006, should D2005
fail to fix what was broken.
Finally Kylix never really caught, is now pretty much buried
and forgotten.
Quite instructive overall, if not that surprising for those of us
who followed the whole story. It's interesting to see that the product
Borland was almost ashamed off and hid in the bowels of their site,
with no incentives for the sales force and marketing, is the only one
with life in it. I guess more detailed statistics of the first months
of 2005 will be able to tell us if D2005 will help a rebound or just
fizzle out.
> There is something strange in google making sum of years
> different than anytime, +29K in Cpp and -550 for Delphi.
Total matches the years' totals, I'm not sure what "anytime"
means to google, maybe it's just a quick estimation of the record
count, while total and other figures are the actual count?
(IIRC google does something similar with websearches, the hit
count it gives in the first search is only an estimate).
Eric
Andrew Rybenkov
--
Andrew Rybenkov.
Hrvoje Brozovic
> Just per chance, you wouldn't happen to have a thousand
> euros handy too? ;)
I was just wondering what to do with my
Chrismass bonus I expect to receive tomorow.
>
> Total matches the years' totals,
Total is calculated by my application,
so hopefully it matches sum of years :)
> I'm not sure what "anytime" means to google
It is number of messages not concerning the dates.
So, any messages with broken dates are inside this.
This is how I see bigger number than mere sum of years for CPB.
What happened to half million of delphi messages is total mystery to me.
>It's interesting to see that the product
>Borland was almost ashamed off and hid in the bowels of their site,
>with no incentives for the sales force and marketing, is the only one
>with life in it. I
or, in other words, from my original post
Hrvoje Brozovic
> That wasn't anti-Delphi. I love Delphi, and I also have
> doubts about whether there'd be enough traffic in a
> delphi.algorithms newsgroup.
You are free to have your doubts
if delphi alone can generate enough traffic.
Anti Delphi is assumption that including
other languages can be of any help,
while knowing that all other together
are just fraction of Delphi traffic.
somebody
Eric Grange
> so hopefully it matches sum of years :)
Oh, that explains it ^_^
> It is number of messages not concerning the dates.
> So, any messages with broken dates are inside this.
> This is how I see bigger number than mere sum of years for CPB.
This looks like a lot just for broken dates...
> What happened to half million of delphi messages is total mystery to me.
Indeed.
> "Unfortunatelly, this increase is not because Borland is
> pushing Delphi, but because, inspite borland treatment,
> it is only one that somehow survives."
Yep.
In the same vein I also found appalling the implied statements
by some Borlanders in the early days of D8 that Delphi was only
for DB or simple GUI apps, with practically no chance of taking
on C++ in performance applications, that D8 was just as fast
as D7 if not faster, that web server stuff was "finally" possible, etc.
IMHO a lot of dumbing down of goals and expectations that did
a lot of disservice to Delphi.
Eric
somebody
> It's interesting to see that the product
> Borland was almost ashamed of
Hrvoje will surely excommunicate me for saying this but with D8 and maybe
D2005 to an extent, it's not hard to figure why they might feel ashamed <g>
Eric Grange
> D2005 to an extent, it's not hard to figure why they might feel ashamed <g>
IMHO they were most "ashamed" of Delphi during the days of JBuilder hype,
Inprise years and those that followed, D6 or D7 didn't receive much
advertisement/marketing attention from Borland.
For those two releases, if you didn't knew Delphi already or people that did,
it was a near impossibility to be aware that a new Delphi version was out.
Eric
Hrvoje Brozovic
>
> Hrvoje will surely excommunicate me for saying this but with D8 and maybe
> D2005 to an extent, it's not hard to figure why they might feel ashamed
<g>
>
Bad quality of D8 surelly was not reason of their behaviour.
It is trivial to prove it, since it started before D8 was in sight.
somebody
> "Marcus F." <a...@b.com> wrote
> > What kind? Graphics? WinAPI related? Database? OO?
> > General programming?
> Rather combinatorics, fitting, random number generators,
> optimization, real-time physical modeling, sorting, searching,
> rounding, integration, symbolic math, linear equation solving,
> LP solving, interpolation, astronomy, civil engineering,
> mechanical engineering, strength of materials, etc.
For some reason, I doubt that we don't see many such disciplines and
algorithms relating to those already in the existing newsgroups is that
there's not a dedicated algorithms group. In other words, it's not like
existing groups are flooded with such algorithms. If that were the case, I
would agree that a dedicated group would become necessary. But as it is, I
don't believe in the "if you build it they will come" mantra. Very few
people use Delphi for scientific/engineering/CAD type of stuff and those who
do professionally in isolated cases tend to be secretive. Such an algorithms
group would get enough traffic *only if* Delphi were the language of choice
in academia but it is not, an extremely small majority of academicians in
sciences and engineering have even heard of Delphi. Case in point, I have
been looking for a nice algorithmic (ie not just C/C++ code) treatment, not
even Pascal/Delphi code, of 2D polygon/polyarc offset using Voronoi diagrams
for a while with no success. You can even find tons of Java code but not a
single line of Pascal or even pseudo-code. The conclusion is that if one
needs to do the kind of stuff you mention, Pascal/Delphi is the absolute
worst choice and C/C++ is an absolute must.
Rudy Velthuis [TeamB]
> Here are results for proportional font
Doesn't that depend on the font used?
> Here are results for monospace font
>
> 8,140 51 TASM
> 9,160 229 TurboPascal
> 1,730,000 259,000 Delphi
> 64,300 3,120 Kylix
> 94,300 2,280 Cpp
> 580,000 58,100 CppBuilder
> 788 770 CSharpBuilder
> 265,000 13,500 JBuilder
Interesting. Now what do these numbers tell us, and which groups were
counted?
--
Rudy Velthuis [TeamB]
"Don't knock masturbation, it's sex with someone I love ."
-- Woody Allen, From 'Annie Hall' 1977.
Rudy Velthuis [TeamB]
> I done this only because it is you, Eric.
> As a compensation for all patience with
> Rudy in his numerous "I don't get it" threads.
Assuming you mean me: I don't get that one either. What are my numerous
"I don't get it" threads?
--
Rudy Velthuis [TeamB]
"Premature optimization is the root of all evil." -- Donald Knuth
John Herbster
> For some reason, I doubt that we don't see many such disciplines
> and algorithms relating to those already in the existing newsgroups
> is that there's not a dedicated algorithms group.
I do not understand this? Would somebody please explain?
> Very few people use Delphi for scientific/engineering/CAD type
> of stuff
I disagree.
> those who do professionally in isolated cases tend to be secretive.
I disagree
--JohnH (not afraid to use my name)
Brad White
>>"somebody" <some...@somewhere.com> wrote in message
news:41bf...@newsgroups.borland.com...
> --JohnH (not afraid to use my name)
8:-)
Go John!
--
Brad.
John Herbster
"Deborah Pate (TeamB)" <d.p...@blueyonder.co.not-this-bit.uk> wrote
> ... and I also have doubts about whether there'd be enough traffic
> in a delphi.algorithms newsgroup. (Of course if unsuccessful
> newsgroups were culled this would not be an issue.)
There are about 360 borland.public.* news groups
About 40 (11%) of these are empty of posts.
About 120 (33%) of these receive fewer than 1 post a week.
There are about 89 borland.public.delphi.* groups
About 16 (18%) of these are empty of posts.
About 24 (27%) of these have fewer than 1 post a week.
--JohnH
Rudy Velthuis [TeamB]
> Anti Delphi is assumption that including
> other languages can be of any help,
> while knowing that all other together
> are just fraction of Delphi traffic.
How is that anti-Delphi? Why is it so important to you to prove that
Delphi has such high traffic, compared to other products? What is wrong
with a language neutral group?
--
Rudy Velthuis [TeamB]
"Far too many development shops are run by fools who succeed despite
their many failings." -- Brion L. Webster
Hrvoje Brozovic
>> I done this only because it is you, Eric.
>> As a compensation for all patience with
>> Rudy in his numerous "I don't get it" threads.
>"Rudy Velthuis [TeamB]" wrote in message
>Assuming you mean me: I don't get that one either. What are my numerous
>"I don't get it" threads?
>>Hrvoje Brozovic wrote:
>> Anti Delphi is assumption that including
>> other languages can be of any help,
>> while knowing that all other together
>> are just fraction of Delphi traffic.
>"Rudy Velthuis [TeamB]" wrote in message
>>How is that anti-Delphi? Why is it so important to you to prove that
>>Delphi has such high traffic, compared to other products? What is wrong
>>with a language neutral group?
Well, Rudy, your second quote quite nicely answers the first one.
I'll try step by step just for fun.
1. someone asks for DELPHI group.
2. TeamB answers that delphi has not enought traffic,
so all languages group can help.
3. This TeamB response implies that Delphi traffic
is just fraction of overall language groups trafic.
4. I PROVE that opposite is true. Traffic from
all other languages combined is fraction of Delphi traffic alone.
5. TeamB are frequent forums users by definition,
so TeamB should know that assumption in 2. is false
even before I proved that it is, in other words,
TeamB posted assumption about low delphi trafic
knowingly it is not the case, with intention of denying
new DELPHI group.
6. Point 5 fits nicely in definition of FUD.
Mark Vaughan
news:41bf...@newsgroups.borland.com:
> Very few people use Delphi for scientific/engineering/CAD type
> of stuff and
very few, yes...but perhaps a few more than you think
> those who do professionally in isolated cases tend
> to be secretive.
this statement is absolute nonsense
--
Mark Vaughan
___________
Visit the Numerical Methods in Pascal web page at
http://www-rab.larc.nasa.gov/nmp/fNMPhome.htm
Rudy Velthuis [TeamB]
> 4. I PROVE that opposite is true.
Yes, so? The proposal to make a group language agnostic because we
might expect more traffic in a general group is not anti-Delphi. Or
should we see your survey as anti-C++ and anti-Java, since you "proved"
that there is less traffic for these languages?
And you have not proved how well a group which does not exist would be
visited. Fact is that one can expect *more* traffic in a group for all
languages than in a group for only one.
IOW, there is no anti or pro here. This is not a fight between
languages, but you appear to make one out of it.
And accusing TeamB of spreading FUD about their own favourite language
is really over the top, IMO.
--
Rudy Velthuis [TeamB]
"A lady came up to me on the street, pointed at my suede jacket and
said, "Don't you know a cow was murdered for that jacket?" I said
"I didn't know there were any witnesses. Now I'll have to kill you
too"." -- George Carlin
Rudy Velthuis [TeamB]
> > > Hrvoje Brozovic wrote:
>
> >> I done this only because it is you, Eric.
> >> As a compensation for all patience with
> >> Rudy in his numerous "I don't get it" threads.
>
>
> > "Rudy Velthuis [TeamB]" wrote in message
> > Assuming you mean me: I don't get that one either. What are my
> > numerous "I don't get it" threads?
>
>
> > > Hrvoje Brozovic wrote:
>
> >> Anti Delphi is assumption that including
> >> other languages can be of any help,
> >> while knowing that all other together
> >> are just fraction of Delphi traffic.
>
>
> > "Rudy Velthuis [TeamB]" wrote in message
> > > How is that anti-Delphi? Why is it so important to you to prove
> > > that Delphi has such high traffic, compared to other products?
> > > What is wrong with a language neutral group?
>
> Well, Rudy, your second quote quite nicely answers the first one.
>
> I'll try step by step just for fun.
>
> 1. someone asks for DELPHI group.
> 2. TeamB answers that delphi has not enought traffic,
> so all languages group can help.
No, that is not true. TeamB answers that _that_ _specific_ group for
Delphi _may_ not have enough traffic.
--
Rudy Velthuis [TeamB]
"Nietzsche was stupid and abnormal."
- Leo Tolstoy (1828-1910)
Wayne Niddery [TeamB]
>
> Well, Rudy, your second quote quite nicely answers the first one.
>
> I'll try step by step just for fun.
>
> 1. someone asks for DELPHI group.
> 2. TeamB answers that delphi has not enought traffic,
No-one said this AFAIK. What was said is that the *specific group being
asked for* would not get enough traffic just from Delphi.
--
Wayne Niddery - Logic Fundamentals, Inc. (www.logicfundamentals.com)
RADBooks: http://www.logicfundamentals.com/RADBooks.html
"A man is likely to mind his own business when it is worth minding,
when it is not, he takes his mind off his own meaningless affairs by
minding other people's business." - Eric Hoffer
Rudy Velthuis [TeamB]
> No, that is not true. TeamB answers that that specific group for
> Delphi may not have enough traffic.
Oops! Sorry for the overquote.
--
Rudy Velthuis [TeamB]
"Science is like sex: sometimes something useful comes out, but that is
not the reason we are doing it" -- Richard Feynman
somebody
> "somebody" <some...@somewhere.com> wrote
> > Very few people use Delphi for scientific/engineering/CAD type
> > of stuff and
> very few, yes...but perhaps a few more than you think
And perhaps less.
> > those who do professionally in isolated cases tend
> > to be secretive.
> this statement is absolute nonsense
Why? It's not about Delphi but programmers in general. It's very wrong to
get the idea that, since there's seemingly so much "stuff" out there on the
internet, majority of professional engineers and programmers are willing to
share their knowledge. From my experience, programmers and engineers who
actually run things (rather than hobbyists or academicians) have more of a
blue-collar mentality and either by choice or by ciscumstance (through lack
of time, incentive or due to company policies/interests and competitive
pressures) keep to themselves.
That's why technical and scientific material you can find on the internet is
mostly from either hobbyists or academicians (including government
researchers) and not from private enterprise, even less so from the more
successful private enterprise. It's too bad, because -no disrespect to
academicians- professionals working in the industry are the one who have a
handle on practical implementations.
So, since most academicians wouldn't be caught dead using Delphi for
publications, there's a shortage of algorithms in Pascal coming from that
channel. And by the above reasoning, professional developers don't often
share their knowledge so the contribution from them is negligible (as in any
language). In conclusion, unless the academia embraces Delphi (which won't
happen of course) there will always be a shortage of good quality Pascal
code pertaining to the disciplines of sciences and engineering. And that has
nothing to do with whether Borland hosts a dedicated algorithms group or
not.
John Herbster
This branch is way off-topic!
Andrew Rybenkov
> share their knowledge.
Engineers and programmers do not develop science (that is foundationd of all programs),
they are implementators by definition. Science is developed by academicians, and they share
their knowledge again by definition. Is decartes geometry (corner stone of CAD programs)
is some patented mystery?
Or let's look at landscape generating programs
Terragen (program that is actively used by Hollywood) is just combination of algorithms that are well known
and available in Internet - if you have problem to find the info - just ask in comp.graphics.algorithms.
Or other two well known professional specimens - Bryce and MojoWorld. Both are founded on works of Ken Musgrave,
who did/do not hide his developments at all.
> So, since most academicians wouldn't be caught dead using Delphi for
> publications, there's a shortage of algorithms in Pascal coming from that
> channel.
yet academicians are tend to use pseudo Pascal/Algol notations in their books.
--
Andrew Rybenkov.
John Herbster
There seems to be a lot of need and support for an
algorithms group.
How can we make an algorithms group come to pass?
--JohnH
John Kaster (Borland)
> How can we make an algorithms group come to pass?
Chocolate. Lots and lots of chocolate, in small, easy to eat pieces.
And it better be good chocolate, too.
P.S. Just kidding about the chocolate.
--
John Kaster http://blogs.borland.com/johnk
Features and bugs: http://qc.borland.com
Get source: http://cc.borland.com
What's going on? http://calendar.borland.com
David Clegg
> Chocolate. Lots and lots of chocolate, in small, easy to eat pieces.
Hmm.... Finally a chink in John's armour is detected. :-)
--
Cheers,
David Clegg
dcl...@gmail.com
John Kaster (Borland)
> Hmm.... Finally a chink in John's armour is detected. :-)
Oh, it wasn't for *me*!
Sorry to disappoint you!
John Herbster
> > How can we make an algorithms group come to pass?
"John Kaster (Borland)" <jo...@borland.com> wrote
> Chocolate. Lots and lots of chocolate, in small, easy to
> eat pieces. And it better be good chocolate, too.
John K, You should have received it by now, about 1800
little M&M pieces. I hope that it was good enough.
Now how about our group borland.public.algorithms?
What else do we need to do? Regards, JohnH
John Kaster (Borland)
> John K, You should have received it by now, about 1800
> little M&M pieces. I hope that it was good enough.
Yes, I did receive it, and we have photographic evidence. Watch for a
blog entry near you soon. All the people who needed to receive the
incentive have already. Thanks, we all got a good laugh out of it.
> Now how about our group borland.public.algorithms?
Stay tuned.
> What else do we need to do? Regards, JohnH
Stay tuned ;)
Captain Jake
<41f057ca$1...@newsgroups.borland.com>
>
> > Now how about our group borland.public.algorithms?
>
> Stay tuned.
This could be VERY dangerous. Coming up with good algorithms can be very
addictive!
--
***Free Your Mind***
Posted with JSNewsreader-BETA 0.9.4.369
Captain Jake
> Now how about our group borland.public.algorithms?
Oh, yeah. This is something I would absolutely LOVE to see. Absolutely,
positively!
David Clegg
> Oh, yeah. This is something I would absolutely LOVE to see.
> Absolutely, positively!
<AOL />
I doubt I'd be able to provide much input, but I'd love to be able to
learn from the experts. I sometimes find myself lurking in the basm
group before I have to remind myself "Small steps, Dave. Small steps"
--
Cheers,
David Clegg
dcl...@gmail.com
Vote 1 http://cc.borland.com/codecentral/ccweb.exe/listing?id=21489 :-)
Now supports Google Groups searching with Dyna-extend(tm) technology!
QualityCentral. The best way to bug Borland about bugs.
http://qc.borland.com
"Just because I don't care doesn't mean I don't understand." - Homer
Simpson
Wayne Niddery [TeamB]
>
> This could be VERY dangerous. Coming up with good algorithms can be
> very addictive!
Very true, but that may be exactly what makes such a new group worthy of
creation - criteria being whether there'd be enough traffic).
John Kaster (Borland)
> Very true, but that may be exactly what makes such a new group worthy
> of creation - criteria being whether there'd be enough traffic).
I've already talked to Fraz about it. Will take it to the private
newsgroup.
Rene Tschaggelar
>John Jacobson wrote:
>
>> I think Colin will be the first to tell you that Xananews is not
>> designed for such large-capacity storage.
>
>Yes. I don't keep stacks & stacks of old newspapers in my shed,
>either!
>
>Instead, it's got a separate area for storing interesting old
>messages. Conceptually like a shoe-box full of old newspaper
>cuttings.
>
>And, of course we've got Google groups. Like having the British
>Library on your desktop. (Apart from keeping books, the British
>Library archives every newspaper, periodical, magazine, etc. that's
>published. Apparently there's even a Porn section (again, just like
>Google Groups!) - you just need to ask the Curators to unlock it for
>you until you're 'finished' ;) )
>
>Mind you, with this new PC of mine, XanaNews doesn't have many
>problems with keeping 100,000 messages per group in the main threaded
>view.
>
>BTW, how big (ie, how many megabytes) is your 600,000 message
>database?
Sorry for the delay, it also took a while to figure out.
The whole database, more than 3.5GBytes.
Rene