[ADS] - Extended Systems Inc. does not care about VO!
ilias
it's userbase.
The reason?
ESI's Advantage Database Server (ADS) and the "OrderKeyNo-problem".
The "OrderKeyNo-Problem", the latest thread on this:
http://groups.google.com/groups?selm=3f4bef65$0$49111$e4fe...@news.xs4all.nl
-
The "OrderKeyNo-problem" is known since long time and discussed in
public vo-newsgroup:
http://groups.google.com/groups?q=ads+orderkeyno
Additionally it was published in ESI's news-servers (2001-01-14 ):
"[...] Today it's almost impossible to use that class [editors remark:
DataListView] on large dbf's, with ADS, as it will be too slow. [...]"
news://devzone.advantagedatabase.com:119/3b78...@solutions.advantagedatabase.com
Even some tech-suggestions were given:
news://devzone.advantagedatabase.com:119/3b78...@solutions.advantagedatabase.com
-
"[...] Because of the slow OrderKeyno performance, some of my Visual
Objects applications run slower than before! [...]"
news://devzone.advantagedatabase.com:119/3f4c...@solutions.advantagedatabase.com
This should be enough for ESI to set this issue to a top-priority.
What did ESI do / say?
"[...] If customers find bugs, and don't report them to us, we have no
way of finding out said bugs exist. [...]"
news://devzone.advantagedatabase.com:119/3F4BAD93...@extendedsystems.com
Of course there is a way: monitoring newsgroups (public VO newsgroup
and ESI's ADS newsgroup).
"[...]
Like I said earlier, we've seen this issue before with our Delphi
client and giving a approximation for the key number was an acceptable
solution. To give exact values with performance similar to the VO RDD
would require more effort than possible at this time. With enough
customer motivation it might become a priority task for a future
release, but that is management's decision, not mine.
[...]"
news://devzone.advantagedatabase.com:119/3f4a...@solutions.advantagedatabase.com
For Delphi (much higher userbase) an approximation was an acceptable
solution.
Bad luck for VO-users, that (most) Delphi-users don't need a precise
and fast result, cause...
For VO: "...It might become a priority task for a future
release...management's decision..."
-
So, what does all this mean?
If a VO-user migrates to ADS he'll have in some points a notable
performance loss in comparision to the native VO-RDD (VO-RDD is 100%,
200%, 300% and more faster).
=> Fixing this is *not* a priority task for ESI !
And this is of course unbelievable, as the VO-user must switch back to
the native VO-RDD (how to explain a customer this speed-loss?).
ESI must provide an *immediate* fix of this critical speed issue.
-
After this fix, ADS could become the no. 1 database-choice for
VO-users, independent if a local/multiuser or a client/server version
is needed..
Especially now, as FORTRESS is effectively out-of-business (as
Loadstone / David Kuechler is out of business) [1].
-
Thoughts:
GrafX should drop FORTRESS to free up resources for the VO27
developement.
ESI could lower the prices for the lower-user-count (10 to 30), which
GrafX mainly wanted to serve with the FORTRESS solution:
http://groups.google.com/groups?threadm=5YbC8.52049%24tT1.54598695%40e3500-atl1.usenetserver.com
GrafX could make a present to the VO-community and OpenSource
FORTRESS (as e.g. Borland has done with Interbase).
http://info.borland.com/devsupport/interbase/opensource/
-
Whatever happens, both companies (ESI and GrafX) should ensure, that
ADS7.0 (including 2 user developer server with ADS-internet-server)
will be included on the VO2.7 distribution.
"[...] Maybe some day I can talk Brian into distributing a 2-user
version of Advantage with every copy of Clipper or VO that is shipped.
This way we would not have to worry about match up the license. [...]"
http://groups.google.com/groups?threadm=bd7isi%24qc5%240%40198.60.252.254
Go on !
Cooperation is guaranteed::
"[...] I look forward to continuing the long standing relationship
that we have had as "friendly competitors" We will of course provide
you with what ever you need in the way of continuing the ADS Client
for Visual Objects. We welcome your active participation in the Visual
Objects forum, and hope that as a result of "friendly competition"
that we both end up with better products for our end users.s [...]"
http://groups.google.com/groups?threadm=a_8C8.51675%24tT1.53505020%40e3500-atl1.usenetserver.com
"...provide you with what ever you need..."
=> a few tips for the OrderKeyNo-Speed-Issue !
=> a little space one the VO2.7 distribution !
Thus the end users get better products.
-
-
[1]
[FORTRESS] - Avoid usage! Loadstone is out of business!
http://groups.google.com/groups?threadm=ubinkvkojjvurrg3qb5ifrut0kmi2o3ab8%404ax.com
-
Some remarks:
Personally, I like ESI and ADS.
First of all, they have very nice webpages where if found allways what
i need (www.extendedsystems.com).
After that, they give a multi-user-database out for free - and i used
it for a Borland C++ Builder project. I was already out-of-budget,
thus this free tool saved me.
I remember when someone helps me in a difficult situation.
Somehow ESI has done that with ADS.
ilias
[...]
>ESI's Advantage Database Server (ADS) and the "OrderKeyNo-problem".
[...]
evaluation of the OrderKeyNo-problem and more:
www.marti.nl/ADSTest.zip
Geoff Schaller
(this is for the benefit of other newer readers here...)
1. You post to this forum like we care for your opinion. We don't. Please
go away. You have proven yourself an illiterate critic in this development
environment time and time over. Your "positive" responses are merely
repostings from others, not your own work. Mostly you carp and criticize.
Why can't you contribute positively if you feel the need to contribute? Why
have you picked on us? Can't you go and attack a VB forum... you'd have much
more fun.
2. I don't use ADS and for our apps, ADS has nothing to offer so I have
nothing to gain from my further commens. Many folks do use ADS (VO and
others) and they use it with great success. I have also seen that ADS gives
great support and certainly does care about VO users. Had you attended the
USA Devcon a few months back you could have spoken directly to ADS
technicians. Their attendance at no cost to the conference is a willing
demonstration of the commitment to one of their smaller user bases. Your
pronouncements that they "don't care" simply demonstrates your ignorance and
inability to properly analyse the circumstance.
3. (...and this is specially for our newer readers - I want to put this
"mystery" to bed...) OrderKeyCount(), as a concept, has a problem. The value
must count the number of keys in a given scope/filter condition in a DBF.
Prior to 2.5a, this method of the DBServer class worked great, except
according to Uwe Holz, it lied because it often failed to correctly consider
all elements of the scope condition. So it was 'repaired' in 2.5a.
Unfortunately, this had the side effect that it was slower. It is an RDD
issue (as I understand it) and because the standard DBF structure does not
carry a storage place for it in the physical file structure, it must be
calculated at least once for each order in a given DBF. From then on for the
life of the server object, changes to the count can be simply incremented at
runtime. For what ever reason, the RDD designers must have found it
problematic to fix because it is still in the same state. Its only real use
is for browser scrollbar displays - it almost has no other use that cannot
be coded another way so its value to the community is reasonably limited. In
SQL there is an easy replacement so it only affects DBF. To count 1 million
records is going to be tough at the best of times so why do it? Do we really
need a scrollbar indicating 1 million records???? How about lots <g>. Now
back to ADS. I suspect they have come against the same problem. The DBF
format simply doesn't have the structure to deal with this so it is a manual
computation for every first time an order is used. They can't use SQL
constructs because the underlying format is a DBF file that has to conform
to DBF specifications and unless these are changed, I don't think the
problem has a solution.
Geoff
ilias
>It looks like Extended Systems Inc. (ESI) does not care about VO and
>it's userbase.
>
>The reason?
>
>ESI's Advantage Database Server (ADS) and the "OrderKeyNo-problem".
>
>The "OrderKeyNo-Problem", the latest thread on this:
>http://groups.google.com/groups?selm=3f4bef65$0$49111$e4fe...@news.xs4all.nl
A quick presentation of the problem on the ESI news:
news://devzone.advantagedatabase.com:119/3efa...@solutions.advantagedatabase.com
Gary Stark
Geoff Schaller wrote:
> Sorry folks but I'm going to answer the troll here...
> (this is for the benefit of other newer readers here...)
>
> 1. You post to this forum like we care for your opinion. We don't. Please
We don't even not care about his opinion. He's a pinhead.
> go away. You have proven yourself an illiterate critic in this development
He's an illiterate pinhead with nothing of any value whatsoever to contribute to
this community.
I sincerely hope that Brian refuses to sell Ilias a copy of 2.7. Or anything
else he might choose to buy from Brian.
Brian is a trader whose methods and dealings I respect. While I've not had any
dealings with ADS, they certainly - as evidenced from their positive
contributions in this forum - come across as a professional outfit, interested
and willing to address legitimate issues raised here.
The best that Ilias can manage is to, like a boil on one's backside, simply be
nothing more than an unpleasant nuisance and an irritant. At least a boil
eventually goes away. Ilias is apparantly so stupid that he cannot realise that
he and his comments are neither welcome nor wanted in this forum.
It's a fact that this forum is not moderated. More's the pity; we would be well
rid of this butt-boil already were some moderation in place.
Whilst there's a procedure we can go through through usenet to have a miscreant
user dealt with with, I think it would be far better if Ilias just decided to go
and bother some other more deserving group. Perhaps one of the test groups would
be appropriate. His drivel would be as meaningful and as well read.
I'd be willing to contribut to buying him a one-way ticket to Purgatory. Any
other participants?
All those in favour of bidding Ilias a Bon Voyage on his trip to Purgatory ...
--
g.
Gary Stark
gst...@RedbacksWeb.com
http://RedbacksWeb.com
ilias
<eut3...@RedbacksWeb.com> wrote:
>Geoff
>Geoff Schaller wrote:
further reference:
http://members.tripod.com/the_flys/
-
keep them alive:
Rob
I agree with everything you (Gary & Geoff ) say. However, the only way we
are going to get rid of this annoying, immature pedant is to ignore him.
Like the spoilt kid he is, if he doesn't get any attention he'll go off and
sulk somewhere else....
--
Rob Grattan
R&D Software Pty. Ltd.
Geoff Schaller
You are right, except that newer readers might see the 'criticism' of ADS
and without any rebuttal, accept it as truth. As I tried to point out, the
Ilias bilious was inaccurate in the least and misleading mostly. As to
OrderKeyCount() there is a valid discussion there as many people are
'complaining' about it. I wanted to point out that it is a resource that
comes at a cost. It works but it has a down side that can't be fixed away du
to its very nature. Sorry - I wouldn't normally respond to Ilias but
sometimes the record needs to be set straight. You will note that I won't
respond to his other posts.
Geoff
Gary Stark
Rob wrote:
> Guys, guys,
>
> I agree with everything you (Gary & Geoff ) say. However, the only way we
> are going to get rid of this annoying, immature pedant is to ignore him.
> Like the spoilt kid he is, if he doesn't get any attention he'll go off and
I'm beginning to think that we need to take this up with the newsgroups
heirarchy, so that they can have his ISP ban him.
I'm very close to going down that path.
Denis Mitrofanov
Agree with you... I hope you don't call me a 'troll' after my words about VO
compiler? <g>
Position 'this is completely shit' is not constructive. I know some Russian
phrases related to guys like 'ilias', but I will not translate them here,
sorry... <g>
Denis
ilias
<deni...@mail.ru> wrote:
>Hi, Geoff
>
>Agree with you... I hope you don't call me a 'troll' after my words about VO
>compiler? <g>
you just declare yourself a 'troll'.
>
>Position 'this is completely shit' is not constructive.
If this is true...
> I know some Russian
>phrases related to guys like 'ilias', but I will not translate them here,
>sorry... <g>
...why do you use it [the not constructive position]?
>
>Denis
>
Rob
Good point, well put. Maybe we should respond to the posts but not the
little weasel himself....
Anyway, I'm not going to waste any more time on posts from - who?
Gary Stark
Rob wrote:
> Geoff,
> Good point, well put. Maybe we should respond to the posts but not the
> little weasel himself....
That's not nice.
Please apologise.
To all of the weasels that you're insulting with that callous remark.
Igor Kaliniak
Thanks for the analysis, and...
> Now back to ADS. I suspect they have come against the same problem...
1) If we using only indices, there are no problems with OrderKeyNo/Count in ADS.
These functions are quite fast.
2) Scopes make no differences with the first case, and also fast.
3) If we have filters, response time dependent from number of records, and may
be very slow.
Standard DbServer methods OrderKeyNo/Count has no appropriate parameters (should
we respect scopes/filters or not). Therefore ADSRDD realized 3rd case by default
(more general and more slow). Surely, that's not a problem. Everyone can write
own methods in a few minutes (and lines!) using their native API functions.
Maybe they will improve the results for optimized filters in the next releases,
but that's another story. In general case I also
> ... don't think the problem has a solution.
Igor
ilias
<rem4i...@gal.ukrpack.net> wrote:
>Geoff,
>
>Thanks for the analysis, and...
>> Now back to ADS. I suspect they have come against the same problem...
>
>1) If we using only indices, there are no problems with OrderKeyNo/Count in ADS.
>These functions are quite fast.
>2) Scopes make no differences with the first case, and also fast.
>3) If we have filters, response time dependent from number of records, and may
>be very slow.
>
>Standard DbServer methods OrderKeyNo/Count has no appropriate parameters (should
>we respect scopes/filters or not). Therefore ADSRDD realized 3rd case by default
>(more general and more slow). Surely, that's not a problem. Everyone can write
>own methods in a few minutes (and lines!) using their native API functions.
if it is that simpl, why don't you post a solution for the VO / ADS
community?
Rob
I've already told you - I'm not reading these posts any more!
Denis Mitrofanov
> you just declare yourself a 'troll'.
... you miss the word [don't]
> If this is true...
... you have used [if] youself. Using [if] assimes that there are [then] and
[else].
> ...why do you use it [the not constructive position]?
... this is my personal opinion and I don't publish it here instead of you.
PS. Those complete bullshit above just shows how to waste a time without any
reason.
ilias
<deni...@mail.ru> wrote:
[...]
>PS. Those complete bullshit above just shows how to waste a time without any
>reason.
>
So please stop wasting your time whilst writing all this 'bullshit'.
Be friendly - Stay in-topic.
Thank you!
Luis Felipe
> Mostly you carp and criticize.
> Why can't you contribute positively if you feel the need to contribute?
If you like VO first learn to use it, find yourself the good things and
advantages, use it for 6 months or more, then come here to help us. I'm sure
your help will be usefully.
If you dont like it...
>Why have you picked on us?
You are a newb and your postings really affects to another newbs.
Regards.
Luis MX
ilias
wrote:
>Hi Ilias
[...]
>Regards.
>
>Luis MX
>
This thread is not about my person.
It is about ESI.
Luis Felipe
> This thread is not about my person.
>
> It is about ESI.
then ESI rules :) and VO too.
Im talking as someone who have used VO for more than one year (experience).
Also I found interesting why are you posting here if you haven't use VO. But
if you dont wanna talk about it, its ok.
Regards,
Luis MX
ilias
wrote:
>Hi Ilias
>> This thread is not about my person.
>>
>> It is about ESI.
>
>then ESI rules :) and VO too.
>
>Im talking as someone who have used VO for more than one year (experience).
so, so.
>
>Also I found interesting why are you posting here if you haven't use VO.
you found wrong:
[COMSDK] - Bug Report/Fix: BSTR* [in,out]
http://groups.google.com/groups?threadm=35a10d4c.0306031406.7c654983%40posting.google.com
>But if you dont wanna talk about it, its ok.
No, i don't want to talk about irrelevant faulty conclusions.
>
>Regards,
>
>Luis MX
>
Luis Felipe
>
> you found wrong:
>
> [COMSDK] - Bug Report/Fix: BSTR* [in,out]
>
http://groups.google.com/groups?threadm=35a10d4c.0306031406.7c654983%40posting.google.com
2003-06-03 thats the date of your post
>
> >But if you dont wanna talk about it, its ok.
>
> No, i don't want to talk about irrelevant faulty conclusions.
>
Look who are talking, your ESI post its an 'irrelevant faulty conclusion'
Are you the only one here who can post irrelevant faulty conclusions ?? Me
too :) and some times others get wrong too, not big deal.
So, you have used VO for.... few months, Im just inviting you to be a
skilled VO programmer (im Not) then you can decide if you like it or not, if
you like it come here to help, if you dont like it...
>Why have you picked on us?
Regards,
Luis MX
ilias
wrote:
>Hi Ilias,
>>
>> you found wrong:
>>
>> [COMSDK] - Bug Report/Fix: BSTR* [in,out]
>>
>http://groups.google.com/groups?threadm=35a10d4c.0306031406.7c654983%40posting.google.com
>
>2003-06-03 thats the date of your post
so
>
>>
>> >But if you dont wanna talk about it, its ok.
>>
>> No, i don't want to talk about irrelevant faulty conclusions.
>
>Look who are talking, your ESI post its an 'irrelevant faulty conclusion'
my conclusion is backuped with an argumentation-line.
>
>Are you the only one here who can post irrelevant faulty conclusions ?? Me
>too :) and some times others get wrong too, not big deal.
your conclusions are irrelevant, as they are off-topic.
you are still off-topic.
>
>So, you have used VO for.... few months, Im just inviting you to be a
>skilled VO programmer (im Not) then you can decide if you like it or not, if
>you like it come here to help, if you dont like it...
I don't need more time for VO.
It's another OO language, nothing special.
-
This topic discusses ESI, an the facts and interconnections which lead
me to express my conclusion are in essence decoupled from VO itself.
In other words: my VO expertise is irrelevant to my conclusion.
Thus my VO expertise is irrelevant to this topic.
>
>>Why have you picked on us?
destiny.
>
>Regards,
>Luis MX
>
Luis Felipe
> >
> >Look who are talking, your ESI post its an 'irrelevant faulty conclusion'
>
> my conclusion is backuped with an argumentation-line.
>
Your conclussion are from an unexperience VO programmer, and you are
confusing the newbs. As Geoff said OrderKeyNo its a DBF problem, thats why
you need experience with VO before you can post something like that.
> >
> >So, you have used VO for.... few months, Im just inviting you to be a
> >skilled VO programmer (im Not) then you can decide if you like it or not,
if
> >you like it come here to help, if you dont like it...
>
> I don't need more time for VO.
>
and for this forum, why ??? It's someone paying you for posting here???
> It's another OO language, nothing special.
>
so?
> -
>
> This topic discusses ESI, an the facts and interconnections which lead
> me to express my conclusion are in essence decoupled from VO itself.
>
> In other words: my VO expertise is irrelevant to my conclusion.
>
> Thus my VO expertise is irrelevant to this topic.
>
so?
> >
> >>Why have you picked on us?
>
> destiny.
>
so?
> >
> >Regards,
> >Luis MX
> >
>
ilias
wrote:
>Ilias,
>> >
>> >Look who are talking, your ESI post its an 'irrelevant faulty conclusion'
>>
>> my conclusion is backuped with an argumentation-line.
>
>Your conclussion are from an unexperience VO programmer, and you are
>confusing the newbs.
Newbies can analyze themselves.
They are newbies, not idiots.
>As Geoff said OrderKeyNo its a DBF problem, thats why
>you need experience with VO before you can post something like that.
faulty conclusion based on faulty facts.
-
however, it should be easy for any of the experienced VO programmers
to 'attack' the argumentation line at a specific point.
>> >So, you have used VO for.... few months, Im just inviting you to be a
>> >skilled VO programmer (im Not) then you can decide if you like it or not,
>if
>> >you like it come here to help, if you dont like it...
>>
>> I don't need more time for VO.
>
>and for this forum, why ???
[1]
>It's someone paying you for posting here???
[1]
>
>> It's another OO language, nothing special.
>>
>so?
[1]
>
>> -
>>
>> This topic discusses ESI, an the facts and interconnections which lead
>> me to express my conclusion are in essence decoupled from VO itself.
>>
>> In other words: my VO expertise is irrelevant to my conclusion.
>>
>> Thus my VO expertise is irrelevant to this topic.
>>
>so?
[1]
>> >
>> >>Why have you picked on us?
>>
>> destiny.
>>
>so?
[1]
>> >
>> >Regards,
>> >Luis MX
>> >
>>
>
[1] - how long do you think i'll take you serious and answer your
questions?
Luis Felipe
> [1] - how long do you think i'll take you serious and answer your
> questions?
I dont know, do you?
ilias
<geo...@softwareXXobjectives.com.au> wrote:
>Sorry folks but I'm going to answer the troll here...
>(this is for the benefit of other newer readers here...)
>
>1. You post to this forum like we care for your opinion. We don't. Please
[...]
irrelevant and offtopic
>
>2. I don't use ADS and for our apps, ADS has nothing to offer so I have
>nothing to gain from my further commens. Many folks do use ADS (VO and
>others) and they use it with great success. I have also seen that ADS gives
>great support and certainly does care about VO users.
your opinion.
>Had you attended the
>USA Devcon a few months back you could have spoken directly to ADS
>technicians. Their attendance at no cost to the conference is a willing
>demonstration of the commitment to one of their smaller user bases. Your
>pronouncements that they "don't care" simply demonstrates your ignorance and
>inability to properly analyse the circumstance.
DevCon attendance of ESI is positive.
[although: they may go there for the nice food...]
But it's not the only index for customer care.
>
>3. (...and this is specially for our newer readers - I want to put this
>"mystery" to bed...) OrderKeyCount(), as a concept, has a problem. The value
>must count the number of keys in a given scope/filter condition in a DBF.
OrderKexCOUNT-Issue is fixed in ADS beta 7.0
>Prior to 2.5a, this method of the DBServer class worked great, except
>according to Uwe Holz, it lied because it often failed to correctly consider
>all elements of the scope condition. So it was 'repaired' in 2.5a.
[...]
I'm sure that this issue was (and is) evaluateable.
repaired or 'repaired'.
>records is going to be tough at the best of times so why do it? Do we really
>need a scrollbar indicating 1 million records????
[...]
The OrderKexNO-Speed-Issue exists already with a few thousands of
records.
>to DBF specifications and unless these are changed, I don't think the
>problem has a solution.
R&D admits that there is a possible solution.
It is a matter of resources and priority.
See the given links for further reference.
>
>Geoff
>
Luis Felipe
some like that.
ilias
wrote:
>>
>> [1] - how long do you think i'll take you serious and answer your
>> questions?
>
>I dont know, do you?
no.
>
Luis Felipe
>
> Newbies can analyze themselves.
>
> They are newbies, not idiots.
>
> >As Geoff said OrderKeyNo its a DBF problem, thats why
> >you need experience with VO before you can post something like that.
>
> faulty conclusion based on faulty facts.
>
> -
>
> however, it should be easy for any of the experienced VO programmers
> to 'attack' the argumentation line at a specific point.
>
Seriously. You are very good discussing.
I GTW, c u arround.
Luis, MX
ilias
wrote:
>Im just watching what happens if I feed the throll, may be it expolotes or
>some like that.
simply look at this discussion to see what happens:
(follow the conversation with "Rod da Silva", who fed me near to
explode)
[VOCOM] - any expieriences with creating Automation Server Interfaces?
http://groups.google.com/groups?threadm=35a10d4c.0306050744.5097c766%40posting.google.com
-
>
ilias
wrote:
>Ilias,
>
>> Newbies can analyze themselves.
>>
>> They are newbies, not idiots.
>>
>> >As Geoff said OrderKeyNo its a DBF problem, thats why
>> >you need experience with VO before you can post something like that.
>>
>> faulty conclusion based on faulty facts.
>>
>> however, it should be easy for any of the experienced VO programmers
>> to 'attack' the argumentation line at a specific point.
>
>Seriously. You are very good discussing.
And you are a very good 'trigger'.
>I GTW, c u arround.
I GTS, cu!
>Luis, MX
Geoff Schaller
> Agree with you... I hope you don't call me a 'troll' after my words about
VO
> compiler? <g>
Never! <g>
I credit you with substantial and worthwhile programing experience, even if
I don't necessarily agree with some of your research aims or outcomes. At
least your debate comes from a reasoned point and is worth taking seriously.
The same cannot be said for the Troll.
Geoff
Geoff Schaller
> Standard DbServer methods OrderKeyNo/Count has no appropriate parameters
(should
> we respect scopes/filters or not). Therefore ADSRDD realized 3rd case by
default
> (more general and more slow). Surely, that's not a problem. Everyone can
write
> own methods in a few minutes (and lines!) using their native API
functions.
Hmm... ok, care to start the ball rolling? My understanding is that all
scopes/filters are necessarily considered so what parameters are needed? If
it is indeed the RDD code doing the counting, how do we bypass this?
Geoff
Igor Kaliniak
> > Standard DbServer methods OrderKeyNo/Count has no appropriate parameters
> (should
> > we respect scopes/filters or not). Therefore ADSRDD realized 3rd case by
> default
> > (more general and more slow). Surely, that's not a problem. Everyone can
> write
> > own methods in a few minutes (and lines!) using their native API
> functions.
>
> Hmm... ok, care to start the ball rolling? My understanding is that all
No, no. I don't want it.
> scopes/filters are necessarily considered so what parameters are needed? If
> it is indeed the RDD code doing the counting, how do we bypass this?
ADS allows a bit more functionality than standard VO RDDs. We can retrieve
logical record number taking (or not) into account filters/scopes. They realized
3rd (general and slowest) case in ADSRDD. There is no 'autodetection' in the
RDD, and even you haven't filters/scopes at all, that function works slowly
respecting non existing filters. But again, we can write own fast OrderKeyNo()
method in a few lines only.
Igor
Geoff Schaller
> ADS allows a bit more functionality than standard VO RDDs. We can retrieve
> logical record number taking (or not) into account filters/scopes. They
realized
So does VO's RDD. Its just the same.
>But again, we can write own fast OrderKeyNo() method in a few lines only.
But how? Without walking the b-tree? If its purely a record count (skip
etc), and the order has all records in the table, this is why it is slow.
Geoff
Igor Kaliniak
>
> >But again, we can write own fast OrderKeyNo() method in a few lines only.
>
> But how? Without walking the b-tree? If its purely a record count (skip
> etc), and the order has all records in the table, this is why it is slow.
I think ESI have patented their algorithms<g>. It is record counting only if you
want retrieve logical recno with 'respect-filter'-parameter. I remember one
test: response time was about 15(or 20?) times faster for database with 500
records when we set 'ignore-filter'-parameter in their native API.
Igor
Igor Kaliniak
> So does VO's RDD. Its just the same.
No. In VO RDD, if you have scope, you always retrieve OrdKeyNo() respecting this
scope. In ADS you can get the same, and also 'pure' logical number which ignore
scopes and/or filters.
Igor
Marcos Nogueira
> It looks like Extended Systems Inc. (ESI) does not care about VO and
> it's userbase.
>
> The reason?
>
> ESI's Advantage Database Server (ADS) and the "OrderKeyNo-problem".
It seems to me that ADS's OrderKeyNo() and OrderKeyCount() don't have
performance issues when scopes are set - only with filters. I use them
extensively in my apps without problems, and my clients don't complain
at all. I think that some network/db administrators need to read the
manuals more carefully.
Regarding ESI, their support has been good so far.
Marcos Nogueira
S. Paulo - Brazil
Geoff Schaller
OrderKeyCount() is designed to obey scopes. Dont' forget the condition
clauses and Filters. Otherwsie we are just dealing with RecCount and there'd
be no need for OrderKeyCount(). No, it must obey scopes/conditions or it is
not reporting what it really is.
Geoff
"Igor Kaliniak" <rem4i...@gal.ukrpack.net> wrote in message
news:bj25dh$ecshc$1...@ID-179710.news.uni-berlin.de...
Phil McGuinness
Not a typo Geoff...... !!
Phil McGuinness - Sherlock Software
------------
"Geoff Schaller" <geo...@softwareXXobjectives.com.au> wrote in message
news:8gk5b.80085$bo1....@news-server.bigpond.net.au...
Geoff Schaller
(...hardly the same as about 4 in the one para and the key one being the
most important word for context <g>)
Igor Kaliniak
As usual<g>
Have you read my answers ever?
There is a little difference between 'No' and 'Count'.
OK, I also don't like "the ball rolling".
Igor
Geoff Schaller
> As usual<g>
Ah, that is your assessment. Have more faith in yourself <g>.
> Have you read my answers ever?
Yes <g>
> There is a little difference between 'No' and 'Count'.
And that was my point. The very nature of OrderkeyCount is to include scopes
and conditions. Otherwsie it becomes some confusing hybrid of RecCount.
Geoff
ilias
[...}
>Whatever happens, both companies (ESI and GrafX) should ensure, that
>ADS7.0 (including 2 user developer server with ADS-internet-server)
>will be included on the VO2.7 distribution.
>
>"[...] Maybe some day I can talk Brian into distributing a 2-user
>version of Advantage with every copy of Clipper or VO that is shipped.
>This way we would not have to worry about match up the license. [...]"
>http://groups.google.com/groups?threadm=bd7isi%24qc5%240%40198.60.252.254
>
>Go on !
>
>Cooperation is guaranteed::
>
>"[...] I look forward to continuing the long standing relationship
>that we have had as "friendly competitors" We will of course provide
>you with what ever you need in the way of continuing the ADS Client
>for Visual Objects. We welcome your active participation in the Visual
>Objects forum, and hope that as a result of "friendly competition"
>that we both end up with better products for our end users.s [...]"
>http://groups.google.com/groups?threadm=a_8C8.51675%24tT1.53505020%40e3500-atl1.usenetserver.com
>
>"...provide you with what ever you need..."
>
>=> a few tips for the OrderKeyNo-Speed-Issue !
>
>=> a little space one the VO2.7 distribution !
>
>Thus the end users get better products.
[...]
But it look like that's not the main interest of GrafX.
"
Brian and I have discussed this issue. Brian is not going to include
Advantage as one of the third party tools with VO. It's not
beneficial to him. He is trying to create a larger market for
Fortress. Brian understands that there is a large VO customer base
using Advantage. He has assured me, he is not trying to force anyone
to Fortress.
[...]
"
http://groups.google.com/groups?selm=ab8u4k%242l9%240%40204.134.216.42
"beneficial to him"
What about the customers choice?
The customers benefit?
And what about transparency, thus the customer is informed what's
going on with the products he has invested time and money into?
-
The above quote refers to VO26.
Hopefully now, where Loadstone/David Kuechler is out of business,
GrafX will do the right step for the benefit of the VO-community and
of the new customers.
Troll Watcher
ilias is well-known on this, and other, newsgroups as an unreliable
source of information. The purpose of his postings is usually to
mis-represent some personal hang-up he has as a valid topic for this newsgroup.
Please ensure you check the veracity of any information he posts. We also strongly
recommend that you do not reply to him. Apart from the encouragement it may provide
him,invariably you will receive some bad-mannered, illogical and spiteful responses
for your efforts.
Tracy Thomas
of ADS on the VO 2.7 CD.
Brian has stated he isn't try to cut out the ADS users. I interpret Brian
actions to embrace ADS. Brian's inclusion of Fortress only broadens the VO
market and potential. This position does not limit your choice of
client/server options nor your potential benefits.
Regards,
TT
ilias
<atth...@btconline.net> wrote:
>I am a VO / ADS user.
ok.
>I am not offended or slighted by the lack of inclusion
>of ADS on the VO 2.7 CD.
of course not.
as you said, you ow ADS already.
>
>Brian has stated he isn't try to cut out the ADS users.
he has stated much other things, too.
>I interpret Brian actions to embrace ADS.
I see.
But the VO-distribution-CD's do not 'embrace' ADS.
>Brian's inclusion of Fortress only broadens the VO market and potential.
The inclusion of Fortress makes no sense and will reduce VO's the
strength and potential.
Loadstone is in essence out of business. [1]
Including Fortress is an Act Of Egoism and not an Act Of Rationality.
>This position does not limit your choice of client/server options nor your potential benefits.
Real choice for buyers this would be...
to find both products on CD.
except GrafX takes the wise decision to drop Fortress. [1]
-
>Regards,
>
>TT
[1]
[FORTRESS] - Avoid usage! Loadstone is out of business!
http://groups.google.com/groups?threadm=ubinkvkojjvurrg3qb5ifrut0kmi2o3ab8%404ax.com
Tracy Thomas
and I believe Bill from ADS has offered free 2 user development licenses for
VO developers.
Now _you_ don't need the ADS software included on the VO CD anymore than I
do. <g>
Cheers,
Tracy