What is Delta??

What is Delta??

[Sathish Veluswamy]

Hi..

Can anybody explain me about Delta with an example..???

Sathish Veluswamy.

Re: What is Delta??

[Martin Krischik]

Sathish Veluswamy wrote:

> Can anybody explain me about Delta with an example..???

http://en.wikibooks.org/wiki/Ada_Programming/Types/delta

Martin

Re: What is Delta??

[Sathish Veluswamy]

Hi Martin Krischik..

Thanks for your help. I saw that link but still i find some
difficulties in understanding about DELTA. So if you dont mind can u
explain about DELTA in detail and if possible with an example.

With regards

Sathish Veluswamy

Re: What is Delta??

[Martin Krischik]

Hello Sathish,

Sathish Veluswamy wrote:

> Thanks for your help. I saw that link but still i find some
> difficulties in understanding about DELTA. So if you dont mind can u
> explain about DELTA in detail and if possible with an example.

what is there to understand? Delta declares a fixed point data type.
The same kind of fixed point data type you find in COBOL or PL/1. You
should have learned about fixed point data types at University or
Polytechnik. If not, you can read it up here:

http://en.wikipedia.org/wiki/Fixed-point_arithmetic

There was just recently a very exhaustive thread about Ada's fixed
point type under the heading "Ada decimal types" here at comp.lang.ada:


http://groups.google.com/group/comp.lang.ada/browse_thread/thread/16dbd7ecc6a3d50c

And I just remebered, I got an exapmple as well:

http://en.wikibooks.org/wiki/Ada_Programming/Algorithms/Chapter_6#No_64_bit_integers
http://svn.sourceforge.net/viewcvs.cgi/wikibook-ada/trunk/demos/Source/fibonacci_5.adb?view=markup

Hope that helps.

Martin

PS: its "you" not "u" - we are software engeneers here and not getto
kits.

Re: What is Delta??

[impslayer]

Martin Krischik skrev:

...
>
> PS: its "you" not "u" - we are software engeneers here and not getto
> kits.

I guess the pun was not intended, was it ;)

Re: What is Delta??

[Georg Bauhaus]

On Fri, 2006-05-12 at 06:49 -0700, Martin Krischik wrote:


> PS: its "you" not "u" - we are software engeneers here and not getto
> kits.

Martin,
if you could kindly consider being a little more thoughtful when
using  the word "ghetto". Not only because of its tradition in fairly
recent European history, but also because it seems you do not really
understand what you are talking about. There is a lot to know about
former Cabrini Green, the history of Chinese/Italian/Spanish/etc.
quarters in New York, the "suburbs" of Calcutta, Duisburg-Bruckhausen,
the aptness of the term "ghetto", and the various influences on
language, the probability of low budget thieving, and the distribution
of parental educational resources.
(Nothing implied here as regards the participants of this thread!)
Also there is more to say about the pride and meaning that some young
inhabitants (or not) of their quarters attach to the term "ghetto kid".

If you dislike IRC slang or whatever it is, or want to say,
"Young man, I must say I don't appreciate your way of addressing
things. What have you learned so far?", there is of course nothing
wrong with that.

All these 'u's and 'i's aren't highly and directly correlated with
being able to understand Ada, and not indicative of a sloppy way.
Proof attempt:
Technically, decades of "ghetto music" have shown various and 
sophisticated construction patterns, played by highly skilled
musicians, technical masters not only of their instruments but
also of the theory of music. A frequent finding with programmers.
(I'm sure you will be able to pick someone to whom these qualifications
do not apply.)

Conversely, knowing Ada very well is not an indicator of being a poet.
OTOH, being a very good and knowledgeable programmer is a necessary
requirement for involvement in high budget organized computer aided
fraud. 
I'm not accusing you or me or anyone, of course. Even though, by age
and "class", criminal statistics demonstrably makes finding high
rate bribery more likely among people in their mid-40s and beyond
who work, for example, in the car industries, which is, I hear,
a target of Ada advocacy.

FWIW, famous mathematicians did the very same thing, they abbreviate
a lot in their hand written letters. I invite you to have a look
at specimen written by Euler, or Gauß.



Sathish,
I have learned that in c.l.ada it is the custom to present some
proof of foregoing efforts to learn something or tackle a problem.
A good way IMHO, and may I suggest you try this approach and present
a few specific problems, written in Ada by you, that your have tried
with a compiler.


Georg
Please excuse the length.

Re: What is Delta??

[Björn Persson]

Dennis Lee Bieber wrote:
> (based upon your surname, I suspect you used the spellings common to
> some other language -- Dutch, perhaps?)

No, Martin just isn't all that good at spelling. There's nothing wrong 
with that, but with that in mind I wouldn't have expected him to run 
down others' spelling.

-- 
Bj�rn Persson                              PGP key A88682FD
                    omb jor ers @sv ge.
                    r o.b n.p son eri nu

Re: What is Delta??

[Martin Krischik]

Dennis Lee Bieber wrote:

> On 12 May 2006 06:49:23 -0700, "Martin Krischik"
> <krischik@users.sourceforge.net> declaimed the following in
> comp.lang.ada:
> 
> 
>> PS: its "you" not "u" - we are software engeneers here and not getto
>> kits.
> 
> Whilst I abhor the public use of those illegible abbreviations (I
> usd t wrt nts in hgh scl by drpng vwls and dplct ltrs, but that is a
> private usage), if one is to make a habit of correcting the users of
> such, one should ensure their own comment is accurate for the language
> being used -- in English: "engineers", "ghetto"

I don't object to the spelling. I object to the general attitude of the
posting. The commanding tone, unneeded use abbreviations, missing proof of
foregoing efforts etc. pp. And it is not the only post.

But then, you and Georg are right - I have overreacted and I try will keep
my ranting to myself in future.

> (based upon your surname, I suspect you used the spellings common to
> some other language -- Dutch, perhaps?)

German but the Surname comes from further east.

Martin
-- 
mailto://krischik@users.sourceforge.net
Ada programming at: http://ada.krischik.com

Re: What is Delta??

[Martin Krischik]

Bjï¿œrn Persson wrote:

> Dennis Lee Bieber wrote:
>> (based upon your surname, I suspect you used the spellings common to
>> some other language -- Dutch, perhaps?)
> 
> No, Martin just isn't all that good at spelling. There's nothing wrong
> with that, but with that in mind I wouldn't have expected him to run
> down others' spelling.

Bjï¿œrn, if it was just spelling I would not have said a thing. But typing "u"
instead of "you" is not a spelling mistake but intentional.

Netiquette is quite clear here: you read before you post. And anybody
reading the group would have spotted that we use full sentences in our
posting. He/She might also have stotted the "Ada decimal types" thead - if
there is one thing good about google groups its the search function.

Martin
-- 
mailto://krischik@users.sourceforge.net
Ada programming at: http://ada.krischik.com

Spelling and netiquette (was: What is Delta??)

[Ludovic Brenta]

I think that Martin, Georg and Dennis all have valid points.  Now that
Martin admitted to overreacting and explained his posts, I think we
can move on.  For posterity however, I would like to provide this
link, which I encourage anyone joining this (or any other) newsgroup
to read several times before posting.

"How To Ask Questions The Smart Way", by Eric S. Raymond:

http://www.catb.org/~esr/faqs/smart-questions.html

-- 
Ludovic Brenta.

Re: What is Delta??

[Craig Carey]

On Fri, 12 May 2006 17:06:46 GMT, Dennis Lee Bieber wrote:

>On 12 May 2006 06:49:23 -0700, "Martin Krischik" declaimed the
> following in comp.lang.ada:
>
>> PS: its "you" not "u" - we are software engeneers here and not getto
>> kits.
>
>	Whilst I abhor the public use of those illegible abbreviations (I
>usd t wrt nts in hgh scl by drpng vwls and dplct ltrs, but that is a
...

The messages of the first poster are entirely consistent with an attack
being run on on (this) news:comp.lang.ada.

List owners at the groups.yahoo.com domain, have known an attacker with
over 300 (maybe over 600) different e-mail addresses, whoc wrote tiny
junk questions that might draw back a helpful response. The trashy
English is able to remind that one of his identifies can be destroyed,
probably with a concealed thoroughly undisclosed (and who wouldn't?)
aspect of unfairness. Also "India" meant USA.

Apparently Google is stripping off the normally interesting identiying
source IP numbers of posters. 

Craig Carey

Re: What is Delta??

[Jeffrey R. Carter]

Martin Krischik wrote:
> 
> I don't object to the spelling. I object to the general attitude of the
> posting. The commanding tone, unneeded use abbreviations, missing proof of
> foregoing efforts etc. pp. And it is not the only post.

I found the tone generally off-putting as well, but my reaction was 
simply not to respond.

As for such things as "u" for "you", SW engineering requires an emphasis 
on ease of reading over ease of writing, and such usages indicate the 
opposite emphasis. You may be able to get away with such things in 
languages such as C, but in Ada, using "u" for "You" will generally 
result in a compilation error.

-- 
Jeff Carter
"Sheriff murdered, crops burned, stores looted,
people stampeded, and cattle raped."
Blazing Saddles
35

Re: What is Delta??

[Craig Carey]

On 11 May 2006 01:20:26 -0700, "Sathish Veluswamy"
    <vsaathish@gmail.com> wrote:
>NNTP-Posting-Host: 203.101.96.2  [...]
>X-Trace: posting.google.com 1147335631 26609 127.0.0.1 (11 May 2006 08:20:31 GMT)
>X-Complaints-To: groups-abuse@google.com [...]
>
>Hi..
>
>Can anybody explain me about Delta with an example..???
>
>Sathish Veluswamy.


Correction: the source IP number was in the message. 

Perhaps a GUI windows Ada 95 program can be written that
processes a Usenet message and gives advice on whether the message
should be 'taken seriously'.

Sample specifications for a program making a decision on the
acceptibility of messages to comp.lang.ada

 * the program has a GUI in Windows

 * it gets input from the Windows clipboard text buffer.

 * the program pattern matching tests and tried to name nodes in
   a taxonomy as the best it can do in the way of making a judgement.

 * the taxonomy is stored in a readable text file could be located on
   a hard disk, or it could be in the Internet.

Five algebra Usenet groups have been under a major attack from one
Mr V. Bond__ of the uncontrolled cris.net, in Simferopol city, Ukraine;
and it has been like that for over a year and things are not improving.
Also alt.math.recreational is being hit with overly-faulty bug reports.

So that the program (if ever existing) does not be confined to this
Usenet group, it would be able to fail Usenet messages of persons that
do not quote text.

(I was browsing for bugs taxonomies at http://citeseer.ist.psu.edu/
and I failed to find a decent document.)

The person of "Sathish Veluswamy" might also have these two identities:

| From: "Nirmalraj Shanmugasundaram" <nirmalraj2383@gmail.com>
| Subject: Send me a sample prgm in ADA for linked list using access type
| NNTP-Posting-Date: Fri, 12 May 2006 09:34:42 +0000 (UTC)
| Injection-Info: y43g2000cwc.googlegroups.com; posting-host=203.101.96.2;

Matching since since the IP number is the same.

The proposed program would fail all e-mails from these sources:
  (a) "Ananth the Boss" <anboss@gmail.com>
  (b) "sangeet" <haisangeetyadav@gmail.com>

Samples of texts of each:

| From: "Ananth the Boss" <anboss@gmail.com>
| Newsgroups: comp.lang.ada
| Subject: ada IDE
| Date: 17 Apr 2006 23:54:33 -0700
| Injection-Info: j33g2000cwa.googlegroups.com; posting-host=203.200.153.7;
|    posting-account=OvrGOQwAAACBMesH2Qn9O80QIhYyyouK
| 
| we are involved in development of safety critical software for flight
| apps. can ny one say which is the best IDE for ADA for development of
| flight software


| From: "sangeet" <haisangeetyadav@gmail.com>
| Newsgroups: comp.lang.ada
| Subject: spec of mil-std-1773b
| Date: 25 Apr 2006 20:53:36 -0700
| X-Trace: posting.google.com 1146023624 32200 127.0.0.1 (26 Apr 2006
|    03:53:44 GMT)
| X-Complaints-To: groups-abuse@google.com
| Injection-Info: i39g2000cwa.googlegroups.com;
|   posting-host=202.65.134.27;
|   posting-account=T4BiWQ0AAACHJowGIRgOrETCiGmVAOy4
|
| hai friends i need a specifications of MIl-Std-1773
| what i studyied from net is its just an optical vertion of 1553
| but i need some clarified specs so if any budy in our group knows
| plz forword it


Once Mr Dewar said he had put someone's name on his kill filter.
I find that to be an inferior option to that of having software 
to raise the percentage-of-agreement on whether is required, or not.


Craig Carey, Auckland city, New Zealand.

Re: What is Delta??

[Keith Thompson]

"Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:
> Martin Krischik wrote:
>> I don't object to the spelling. I object to the general attitude of
>> the
>> posting. The commanding tone, unneeded use abbreviations, missing proof of
>> foregoing efforts etc. pp. And it is not the only post.
>
> I found the tone generally off-putting as well, but my reaction was
> simply not to respond.
>
> As for such things as "u" for "you", SW engineering requires an
> emphasis on ease of reading over ease of writing, and such usages
> indicate the opposite emphasis. You may be able to get away with such
> things in languages such as C, but in Ada, using "u" for "You" will
> generally result in a compilation error.

For what it's worth, C will certainly flag an attempt to refer to
something declared as "You" as "u" (it will also flag an attempt to
refer to it as "you" or "YOU".

And over in comp.lang.c, we spend a significant amount of time asking
people to spell out their words.  C is terse; English doesn't have to
be.

-- 
Keith Thompson (The_Other_Keith) kst-u@mib.org  <http://www.ghoti.net/~kst>
San Diego Supercomputer Center             <*>  <http://users.sdsc.edu/~kst>
We must do something.  This is something.  Therefore, we must do this.

Re: What is Delta??

[Martin Krischik]

Craig Carey wrote:

> The person of "Sathish Veluswamy" might also have these two identities:
> 
> | From: "Nirmalraj Shanmugasundaram" <nirmalraj2383@gmail.com>
> | Subject: Send me a sample prgm in ADA for linked list using access type
> | NNTP-Posting-Date: Fri, 12 May 2006 09:34:42 +0000 (UTC)
> | Injection-Info: y43g2000cwc.googlegroups.com; posting-host=203.101.96.2;
> 
> Matching since since the IP number is the same.

Interesting. I somehow had the suspicion that both work for the same company
which has got some Ada contract without actually haven any Ada programmers.
I did not want to raise that point without proof - my suspicion was based
on similar sounding names (Indian?) - which is no proof at all, both using
gmail for E-Mail, and google for news.

But same IP - thats a different matter.

Martin
-- 
mailto://krischik@users.sourceforge.net
Ada programming at: http://ada.krischik.com

Re: What is Delta??

[Jeffrey R. Carter]

Keith Thompson wrote:
> 
> And over in comp.lang.c, we spend a significant amount of time asking
> people to spell out their words.  C is terse; English doesn't have to
> be.

Perhaps I should have added a smiley to my comments. Anyway, since you 
have experience with this from c.l.c, maybe you'd like to volunteer for 
the same role here?

-- 
Jeff Carter
"Perfidious English mouse-dropping hoarders."
Monty Python & the Holy Grail
10

Re: What is Delta??

[Martin Krischik]

Jeffrey R. Carter wrote:
> Martin Krischik wrote:
> >
> > I don't object to the spelling. I object to the general attitude of the
> > posting. The commanding tone, unneeded use abbreviations, missing proof of
> > foregoing efforts etc. pp. And it is not the only post.
>
> I found the tone generally off-putting as well, but my reaction was
> simply not to respond.

I thought about that too. But then there is a poblem: If someone else
searches - either google or google groups - and then finds a valid
question unanwered he/she might draw the wrong conclusion.

Martin

Re: What is Delta??

[Keith Thompson]

"Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:
> Keith Thompson wrote:
>> And over in comp.lang.c, we spend a significant amount of time asking
>> people to spell out their words.  C is terse; English doesn't have to
>> be.
>
> Perhaps I should have added a smiley to my comments. Anyway, since you
> have experience with this from c.l.c, maybe you'd like to volunteer
> for the same role here?

Sorry, I spend too much time on Usenet as it is.

-- 
Keith Thompson (The_Other_Keith) kst-u@mib.org  <http://www.ghoti.net/~kst>
San Diego Supercomputer Center             <*>  <http://users.sdsc.edu/~kst>
We must do something.  This is something.  Therefore, we must do this.

Re: What is Delta??

[M E Leypold]

Martin Krischik <krischik@users.sourceforge.net> writes:

> Interesting. I somehow had the suspicion that both work for the same company
> which has got some Ada contract without actually haven any Ada programmers.
> I did not want to raise that point without proof - my suspicion was based
> on similar sounding names (Indian?) - which is no proof at all, both using
> gmail for E-Mail, and google for news.
> 
> But same IP - thats a different matter.

Well - it might also be the proxy IP of a provider who uses a
NAT-Proxy for all of his Dial-In users. Providers like this exist,
i.e. Infocity in Germany.

Before I buy Craig's conspiracy theory (that this higher frequency of
unusual questions is an attack :-) I'd like to know more about how the
attacker would profit from that kind of attack. Around one newbie per
week who doesn't know how to ask questions properly doesn't seem an
exceptionally high number to me.

An alternative theory to explain the coincidence of "similar sounding"
names or the same IP (i.e. probably coming from the same NATted local
network) would be: There is university somewhere which is teaching an
SE course for which Ada is required. "Sathish Veluswamy" just asked
his questions striaght forward whereas "Ananth" realized that nobody
here would be interested overly much in doing his homework, so he
_pretented_ doing research for "his company". Seems even more probably
than a company having an Ada contract (and for a security critical
system too) w/o any Ada programmers and now trying to learn Ada on the
usenet, fast.

But all that is speculation and OT. Only wanted to hint, that IPs don't say
so much.

Regards -- Markus

Re: What is Delta??

[Ludovic Brenta]

M E Leypold wrote:
> An alternative theory to explain the coincidence of "similar sounding"
> names or the same IP (i.e. probably coming from the same NATted local
> network) would be: There is university somewhere which is teaching an
> SE course for which Ada is required. "Sathish Veluswamy" just asked
> his questions striaght forward whereas "Ananth" realized that nobody
> here would be interested overly much in doing his homework, so he
> _pretented_ doing research for "his company". Seems even more probably
> than a company having an Ada contract (and for a security critical
> system too) w/o any Ada programmers and now trying to learn Ada on the
> usenet, fast.

More specifically, my guess is that some company has subcontracted the
safety-critical avionics software to an Indian university. The
university wants to use the project to promote itself as a centre of
high-tech know-how, and the students will write interesting theses
based on the project. So far, so good. The only problem is that none of
the students know the first thing about Ada or safety-critical
software, and they don't even know how to ask on the usenet. They're
students, after all (and probably even cheaper than Indian
professionals). But I wonder if the same can be said of their teachers,
since the students seem to have received only minimal training in that
field.

Is anyone aware of a safety-critical avionics project that's just been
subcontracted off-shore to India? If it exists, that project's in
trouble.

All the above is mere speculation, of course.

-- 
Ludovic Brenta.

Re: What is Delta??

[M E Leypold]

"Ludovic Brenta" <ludovic@ludovic-brenta.org> writes:

> M E Leypold wrote:

> > _pretented_ doing research for "his company". Seems even more probably
> > than a company having an Ada contract (and for a security critical
> > system too) w/o any Ada programmers and now trying to learn Ada on the
> > usenet, fast.
> 
> More specifically, my guess is that some company has subcontracted the
> safety-critical avionics software to an Indian university. The
> university wants to use the project to promote itself as a centre of
> high-tech know-how, and the students will write interesting theses
> based on the project. So far, so good. The only problem is that none of
> the students know the first thing about Ada or safety-critical
> software, and they don't even know how to ask on the usenet. They're
> students, after all (and probably even cheaper than Indian
> professionals). 

Do things like this actually happen? I mean -- has anything of that
style happened before? I'm just wondering -- it seems so
unbelievable ...

Regards -- Markus

Re: What is Delta??

[Marc A. Criley]

M E Leypold wrote:
> "Ludovic Brenta" <ludovic@ludovic-brenta.org> writes:
> 
    <snips>

>> More specifically, my guess is that some company has subcontracted the
>> safety-critical avionics software to an Indian university. The
>> university wants to use the project to promote itself as a centre of
>> high-tech know-how, and the students will write interesting theses
>> based on the project. So far, so good. The only problem is that none of
>> the students know the first thing about Ada or safety-critical
>> software, and they don't even know how to ask on the usenet. They're
>> students, after all (and probably even cheaper than Indian
>> professionals). 
> 
> Do things like this actually happen? I mean -- has anything of that
> style happened before? I'm just wondering -- it seems so
> unbelievable ...

I can't say whether this his actually happened with regard to an Indian 
university or some other outsourcing initiative, but I do know that 
significant systems have been written in Ada in the late 90s by 
programmers in the US having little or no exposure to the language prior 
to the project.

An experienced Ada programmer knows what Ada code should look like, in 
terms of definitions, structures, and interactions.  A system that was 
dumped in my lap to port to a new OS version and compiler included the 
following:

- A set of project-standard integer types were defined for 8, 16, and 32 
bit items.  _Every_ other integer type was defined as a subtype of one 
of those.  Totally eliminated integer type conflicts--and checking.
- In several packages record type definitions were defined using the C 
idiom for struct definitions, to the point of having the term "struct" 
as part of the type name, AND defining a subtype of the record for 
public use--ala the practice of having a typedef for struct definitions.
- Handling message buffers by interfacing to the strcpy() function and 
using it to move bytes from memory buffer arrays to and from record 
components.
- Used tasks where virtually all the work of the task was performed 
during task rendezvous, _including_ making entry calls on other tasks. 
This resulted in the world's most computationally expensive procedure calls.

Clearly there were no experienced Ada hands on the project, and it showed.

-- Marc A. Criley
-- McKae Technologies
-- DTraq - XPath In Ada - XML EZ Out

Re: What is Delta??

[Dmitry A. Kazakov]

On Mon, 22 May 2006 09:55:53 -0500, Marc A. Criley wrote:

> - A set of project-standard integer types were defined for 8, 16, and 32 
> bit items.  _Every_ other integer type was defined as a subtype of one 
> of those.  Totally eliminated integer type conflicts--and checking.
> - In several packages record type definitions were defined using the C 
> idiom for struct definitions, to the point of having the term "struct" 
> as part of the type name, AND defining a subtype of the record for 
> public use--ala the practice of having a typedef for struct definitions.
> - Handling message buffers by interfacing to the strcpy() function and 
> using it to move bytes from memory buffer arrays to and from record 
> components.
> - Used tasks where virtually all the work of the task was performed 
> during task rendezvous, _including_ making entry calls on other tasks. 
> This resulted in the world's most computationally expensive procedure calls.
> 
> Clearly there were no experienced Ada hands on the project, and it showed.

I don't think that what you describe is just because of a lack of exposure
to Ada. I looks much like poor programming. Bad programmers are bad in all
languages. I am wondering what is taught in universities, because fresh
programmers we get are accustomed to use worst imaginable programming
practices. They like pointers, they use them everywhere, yet they don't
know what pointer actually is. Copying bytes while parsing texts is nothing
comparing with scanning the same text from the beginning each time a
parameter need to be read (an XML parser was written in this way). I don't
know where they get all that...

-- 
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de

Re: What is Delta??

[Jeffrey R. Carter]

Marc A. Criley wrote:
> 
> I can't say whether this his actually happened with regard to an Indian 
> university or some other outsourcing initiative, but I do know that 
> significant systems have been written in Ada in the late 90s by 
> programmers in the US having little or no exposure to the language prior 
> to the project.

Not to mention the 1980s and early 1990s.

> An experienced Ada programmer knows what Ada code should look like, in 
> terms of definitions, structures, and interactions.  A system that was 
> dumped in my lap to port to a new OS version and compiler included the 
> following:
> 
> - A set of project-standard integer types were defined for 8, 16, and 32 
> bit items.  _Every_ other integer type was defined as a subtype of one 
> of those.  Totally eliminated integer type conflicts--and checking.

Unfortunately, this is a fairly common practice. I find it less 
offensive than the design principle that most communication takes place 
through global variables. But what we're talking about is not just poor 
use of Ada; this is SW that has been designed by coders who are not 
competent to design SW. The construction workers are designing the bridges.

-- 
Jeff Carter
"Oh Lord, bless this thy hand grenade, that with it thou
mayst blow thine enemies to tiny bits, in thy mercy."
Monty Python and the Holy Grail
24

Re: What is Delta??

[Keith Thompson]

"Marc A. Criley" <mcNOSPAM@mckae.com> writes:
[...]
> An experienced Ada programmer knows what Ada code should look like, in
> terms of definitions, structures, and interactions.  A system that was
> dumped in my lap to port to a new OS version and compiler included the
> following:
>
> - A set of project-standard integer types were defined for 8, 16, and
> 32 bit items.  _Every_ other integer type was defined as a subtype of
> one of those.  Totally eliminated integer type conflicts--and checking.
> - In several packages record type definitions were defined using the C
> idiom for struct definitions, to the point of having the term "struct"
> as part of the type name, AND defining a subtype of the record for
> public use--ala the practice of having a typedef for struct
> definitions.
> - Handling message buffers by interfacing to the strcpy() function and
> using it to move bytes from memory buffer arrays to and from record
> components.
> - Used tasks where virtually all the work of the task was performed
> during task rendezvous, _including_ making entry calls on other
> tasks. This resulted in the world's most computationally expensive
> procedure calls.
>
> Clearly there were no experienced Ada hands on the project, and it showed.

Most of these sound horrendous, but one of them may not be.

C's strcpy() function copies a C-style string, terminated by a NUL
character.  If a NUL-terminated string format is imposed by some
external interface, then interfacing to strcpy() might be the most
reasonable approach.  Hand-written code to do the same thing, either
in C or in Ada, could easily be less efficient and more error-prone; a
C library's implementation of strcpy() is free to use system-specific
performance tricks and is likely to be thoroughly tested.

On the other hand, if the NUL-terminated string format *wasn't*
imposed by an external interface, imposing it on an Ada project seems
like a bad idea.

-- 
Keith Thompson (The_Other_Keith) kst-u@mib.org  <http://www.ghoti.net/~kst>
San Diego Supercomputer Center             <*>  <http://users.sdsc.edu/~kst>
We must do something.  This is something.  Therefore, we must do this.

Re: What is Delta??

[Marc A. Criley]

Keith Thompson wrote:
> "Marc A. Criley" <mcNOSPAM@mckae.com> writes:
> [...]
>> - Handling message buffers by interfacing to the strcpy() function and
>> using it to move bytes from memory buffer arrays to and from record
>> components.
[...]

> Most of these sound horrendous, but one of them may not be.
> 
> C's strcpy() function copies a C-style string, terminated by a NUL
> character.  If a NUL-terminated string format is imposed by some
> external interface, then interfacing to strcpy() might be the most
> reasonable approach.  Hand-written code to do the same thing, either
> in C or in Ada, could easily be less efficient and more error-prone; a
> C library's implementation of strcpy() is free to use system-specific
> performance tricks and is likely to be thoroughly tested.

Actually, I realized I misspoke -- it's been a few years since I did 
this.  It wasn't strcpy(), it was memmove() that was used to handle 
memory buffering, with all the accompanying 'Address and 'Size 
calculations being explicitly performed in code.

-- Marc A. Criley
-- McKae Technologies
-- DTraq - XPath In Ada - XML EZ Out

Re: What is Delta??

[Keith Thompson]

"Marc A. Criley" <mcNOSPAM@mckae.com> writes:
> Keith Thompson wrote:
>> "Marc A. Criley" <mcNOSPAM@mckae.com> writes:
>> [...]
>>> - Handling message buffers by interfacing to the strcpy() function and
>>> using it to move bytes from memory buffer arrays to and from record
>>> components.
> [...]
>
>> Most of these sound horrendous, but one of them may not be.
>> C's strcpy() function copies a C-style string, terminated by a NUL
>> character.  If a NUL-terminated string format is imposed by some
>> external interface, then interfacing to strcpy() might be the most
>> reasonable approach.  Hand-written code to do the same thing, either
>> in C or in Ada, could easily be less efficient and more error-prone; a
>> C library's implementation of strcpy() is free to use system-specific
>> performance tricks and is likely to be thoroughly tested.
>
> Actually, I realized I misspoke -- it's been a few years since I did
> this.  It wasn't strcpy(), it was memmove() that was used to handle
> memory buffering, with all the accompanying 'Address and 'Size
> calculations being explicitly performed in code.

Ok, 100% horrendous.  Congratulations!  8-)}

-- 
Keith Thompson (The_Other_Keith) kst-u@mib.org  <http://www.ghoti.net/~kst>
San Diego Supercomputer Center             <*>  <http://users.sdsc.edu/~kst>
We must do something.  This is something.  Therefore, we must do this.