Brian 2.6
Graham McKechnie
Your Readme in 2.6 states that you have fixed the problem with collation
see -
>>Changed ANSI (Windows) collation sequences.
However it refers to 2.0, not 2.5. Has there been any change in this area
from 2.5 to 2.6 with CDX's?
Last week I wrote here about a certain sequence that didn't work as compared
to the way Palm sorts their databases. This problem re the "-" char doesn't
appear to be addressed in 2.6. Many NG users (and I presume beta testers)
wrote back and said that this problem had been reported to CA and Grafx and
should be fixed in 2.6.
Brian can you confirm if this problem was addressed, if it was what where
the rules for the fix. In my testing with 2.6 this doesn't appear to be the
case.
Another related problem, is that I can create a dbf and then create an order
on the empty dbf and get a completely different sort sequence, as compared
to creating the dbf then appending records to it and then creating the
index. Can you please explain, why this is so? From my observations the
behaviour of 2.5 and 2.6 is the exactly the same in this respect. This is
obviously not right.
Can you tell me exactly how long your company will take to address this
problem and when we are likely to see the fix. Please refer to my previous
message for the exact strings to reproduce this.
I've just parted with $A415 for 2.6 and at this stage I can see no value in
2.6 in spending that amount. This appears to be CA's 2.5c version and very
little else. As you would already know I own the SDK and there is no
interest to me in the 3rd party products as I already own the best of them.
I'm quite happy with the concept of these dollars funding your effort to a
VO 3.0, but I have little interest in 3.0 today, when 2.6 doesn't correct
bugs that have been already reported long ago. How can I justify hanging
around long enough for 3.0 and another slug of dollars, if you can't get 2.6
right?
Many VO users may be "tickled pink" with this release of 2.6, unfortunately
from your point of view I am not. I note your comments about the dedication
of the alpha and the beta testers and all the other wonderful people
involved in this release. Brian, as a fellow VO developer, do you really
enjoy compiling the GUI class SDK's and getting all those warnings. Where is
the dedication to the product that you so vehemently portrayed months ago,
if you allow it to be as sloppy as the compiler says it is?
In my opinion, you had a big chance of getting this thing right. However,
again in my opinion you took the CA route, give 'em enough to keep the
dickheads happy. I've got to repeat it here again Brian, we aren't all
dickheads, some of us can see value for money - this is not what you have
delivered.
Why didn't you take the time?? - we gave you the bucks!!!
Again - when are the fixes due. Finally what a disappointment
Regards
Graham
Phil McGuinness
I know some if not all your comments about the "product" when put into
context are quite valid...
If Brian was indeed CA then I think he would have a lot to answer for.
But your demeanour is I am sure based on the "evils of the past" which I do
not believe were going to be solved in a 1st release.
But let us put this issue into the public arena for debate and also look at
the alternatives.
Let us consider is only 5 short months ... 22+ weeks since this whole VO2.6
was born.
In that time Grafx have had to seal the deal, get the source, get the team,
feed them with no income, run a beta and get what is left of the 3rd party
market behind him....... then deliver a product that is "worth the dough"
Also in that time the potential is that a percentage of VO'ers have moved
on, are thinking about moving on or will not pay the dough and sit on the
fence. A divide community would not be good for Vo's direction.
For me I also own the SDK the 3rd party products and VO2.5b3 works
reasonably well....... except it does have some very annoying bugs which are
now fixed in VO2.6 (ie. The menu getting corrupted, the opening of DBF/FPT
on XP to name a few) The money just fix these problems was worth it.
I understand your issues have not been fixed and that would piss me off if
somebody is asking you to part with your cash.. I am happy to run with
VO2,6.. but my judgement will be based on a 2,7 release I think within say
another 5 months that has addressed a lot of unresolved issues. That is an
even slicker, faster, bug free version of VO2.6 would be want I want.
I think by then the development team will be very familiar with the code and
making genuine improvements.
Then I would consider what VO3.0 has to offer for the future.
My 2 cents worth.
Phil McGuinness
----
"Graham McKechnie" <g...@bigpond.net.au> wrote in message
news:Bhyr9.29857$334....@news-server.bigpond.net.au...
Mark Cooper
IIRC, Brian was very explicit from the beginning about 2.6. There was no
promise of wondrous new fixes/enhancements. He was only going to release a
cleaned-up point upgrade. Most of the fixes had already been
started/completed by the CA team.
I for one will be glad to get some of the 3rd party products. Even though I
just purchased bBrowser and RP3, some of the other products look
interesting. If nothing else, I'll get to see how the 'pros' code with VO.
In the overall scheme of things, 200USD comes out to somewhere between 2 to
4 hours of my time. If I pick up 4 tips that save me 1 hour each, it's paid
for itself.
FWIW,
Mark L. Cooper
Columbus, Ohio USA
Graham McKechnie
I don't think your post has stated anything new or addressed the issues I
raised. I really think you missed my point.
Graham
"Mark Cooper" <mlcooper...@aol.com> wrote in message
news:aomdmb$nskmt$1...@ID-147466.news.dfncis.de...
Graham McKechnie
> If Brian was indeed CA then I think he would have a lot to answer for.
> But your demeanour is I am sure based on the "evils of the past" which I
do
> not believe were going to be solved in a 1st release.
I'm not sure I feel the same way as you. "Evils of the Past" - give me a
break, I just wont a 2.6 that works and I paid heaps for.
Phil, it doesn't work mate. Are you happy with that?
> But let us put this issue into the public arena for debate and also look
at
> the alternatives.
Phil, I've put in the public arena in my original post.
What about your very real friends at MS, you talk to us every day about
them, you know, the ones you are going to invite to your next wedding etc.
Phil give me a break, you can the shit out of MS and now I can't say a bad
word about Grafx. Phil, Brian fucked up big time here. He had his chance and
blew it. We gave him the cash and he still blew it - I'm not happy Phil....,
just like Jan..
I've just read your last half - fuck mate what are you and Brian on, talk
about cheesy and we are mates. You really do believe the dribble you just
wrote I presume. If I read you right I've just paid for a non existant 2.7,
I guess you haven't paid for the fucked up 2.6, as your company is in the
credits.
Good luck anyway. But when is the CDX problem going to be fixed?
Regards
Graham
"Phil McGuinness" <hey...@sherlock.com.au> wrote in message
news:aomdk0$nurns$1...@ID-88745.news.dfncis.de...
Mark Cooper
>
> I don't think your post has stated anything new or addressed the issues I
> raised. I really think you missed my point.
>
I'm not saying there aren't any problems with 2.6. I don't have my copy yet
BTW. I probably won't even install it right away.
My point is that I'm one of those 'dickheads' that is glad to get a chance
to look at/play with a bunch of new toys. These new toys may even make it
easier to get something done in VO. I'm not a VO expert. I'm eager to see
how others program in VO.
I also feel I am contributing to the continued development of an already
decent product. This is a 'new' product for Brian. I would expect some
things that don't work correctly and will require a fix (or 2 <g>) in the
near future
Finally, I'm voting with my dollars for the health of a non-Microsoft
product.
I think that when 3.0 does get released, I'll be in a much better position
to use it.
Mark
Marc Verkade
'this is not what you have delivered.'!
I totally disagree here....
I see 2.6 as a start for future VO development..
Let's see what Brian says about your issues......
Marc
"Phil McGuinness" <hey...@sherlock.com.au> schreef in bericht
news:aomdk0$nurns$1...@ID-88745.news.dfncis.de...
Nathan Robeson
IMO you are doing EXACTLY what is needed. The cost is very minor comapred to
how much important a stable feature rich tool powerful tool can do for us.
Nathan
"Mark Cooper" <mlcooper...@aol.com> wrote in message
news:aomdmb$nskmt$1...@ID-147466.news.dfncis.de...
Adriano Rui Gominho
If you were a customer of mine I would be deligthed to refund your
money and move on. We're all in this for the money but some of us
don't dispense with the fun, not to mention politeness.
Adriano
Jari Sevon
> I've just parted with $A415 for 2.6 and at this stage I can see no value
in
> 2.6 in spending that amount. This appears to be CA's 2.5c version and very
> little else. As you would already know I own the SDK and there is no
> interest to me in the 3rd party products as I already own the best of
them.
I agree, if we look thing in this light, I could say I am not happy product
went to Grafx,
because CA would have given the 2.5c patch for free... ?? <G>
> I'm quite happy with the concept of these dollars funding your effort to a
> VO 3.0, but I have little interest in 3.0 today, when 2.6 doesn't correct
> bugs that have been already reported long ago. How can I justify hanging
> around long enough for 3.0 and another slug of dollars, if you can't get
2.6
> right?
Well, have to say it is extremely thought business model to collect
money from users for product that is not even planned. I know software
companies have a long history marketing products which don't exist, but
this...
- Jari -
Phil McGuinness
As I understand you have not bought the product... so keep the money until
things are fixed to your satisfaction then you have a choice do something
about it.
My comments about Grafx are balanced and based on fact... CA has had years
and years to fix the problems and essentially Brian took over a project in
"maintenance mode". We all knew it was going to be a big job and who knows
what manner of things were missing or broken or not documented or could
simply not be removed due to CA licensing.
He got version 2.6a out and I think he is customer responsive enough to go
the extra mile and close other important issues and release an update. As
there are a few installer issues then he may get onto this sooner than
later.
As I said I have a few show stoppers in VO2.5b3 that are fixed in VO2.6...
so for me the money was worth it and I am looking for even more in a
reasonable time period.
Give Brian and his team a break..... they have worked their guts out for the
last 5 months and this is the first or a few builds I would envisage.
Phil McGuinness
------------------
"Graham McKechnie" <g...@bigpond.net.au> wrote in message
news:r0Ar9.30247$334....@news-server.bigpond.net.au...
Phil McGuinness
I knew you must be joking.... We all know MS do not have any real friends...
only GPF's <G>
Remember MS have had 25 years to get security and other stuff together...
Brian has had 5 months....
So lets here it for VO2.6 SE ...........!
Phil McGuinness
-------------------
"Graham McKechnie" <g...@bigpond.net.au> wrote in message
news:r0Ar9.30247$334....@news-server.bigpond.net.au...
Graham McKechnie
> As I understand you have not bought the product... so keep the money until
> things are fixed to your satisfaction then you have a choice do something
> about it.
Thats an interesting assumption mate!!!. How do you figure that? Do you
really think I would be bitching if I was not using 2.6? Phil somehow you
missed the most obvious piece in my original post
"In my testing with 2.6 this doesn't appear to be the case".
Do I have to send you a copy of the invoice? I wish you would read stuff
properly, before you go shooting your mouth off.
> Give Brian and his team a break..... they have worked their guts out for
the
> last 5 months and this is the first or a few builds I would envisage.
So you want me to waste more time, waiting for another patch. Sorry Phil, I
gave them a chance, they new about the problem for over 6 months, I and
others gave them a reproducible example, it wasn't fixed and I've just done
my dough. I hope you are following along now!!!.
Phil, when you defend Grafx and blame CA, you sound just like a politican
blaming the previous government for some stuff up they haven't got around to
fixing. Brian's had over 12 hours to respond to my post, he hasn't so far,
so I guess he isn't going to. Have you now got a new role as Brian's new
spin doctor?
You obviously have more patience than me, I got stuff to do now, so I guess
I'll have to find another way without VO.
I would have been nice, if Brian had acknowledged the problem, I doubt it
would take Kuechler more than a day or two to fix it.
Regards
Graham
"Phil McGuinness" <hey...@sherlock.com.au> wrote in message
news:aonehp$o6j63$1...@ID-88745.news.dfncis.de...
RPSoftware
I understand you got valid points on 2.6, but with all due respect-- you're
bitching Brian for not fixing "bugs" you just got discovered last week. And
2.6 is already in the production for released.
Give them a break--I believe their will patch forthcoming. And stop
bitching, because it will solve anything.
And sorry you cannot collect your $200.00 (US? I hope so because PhP against
USd is awful).
BTW--just ordered 2.6. Maybe coming in couple weeks or months.
--
Regards,
Rene J. Pajaron
RPSoftware
Philippines
"Graham McKechnie" <g...@bigpond.net.au> wrote in message
news:Bhyr9.29857$334....@news-server.bigpond.net.au...
Phil McGuinness
value in 2.6 in spending that amount.]
I can see you have parted with it.. so this gives you two Free tickets to
bitch..... So you bought the upgrade to fix the Collation sequence and you
will not get any benefit in the other fixes. Would you have preferred that
the product was just canned and then where would we be.?
====
snip[ Sorry Phil, I gave them a chance, they new about the problem for over
6 months, I and others gave them a reproducible example, it wasn't fixed and
I've just done my dough. I hope you are following along now!!!.]
I feel like that about Windows.. gave them a chance and again it fails to
deliver in so many areas.
====
snip[I and others gave them a reproducible example, it wasn't fixed and I've
just done my dough. I]
Me too for many, many bugs and problems. But I am realistic what will happen
in a first build.
You know you have not done your dough.. You have a problem that is not
fixed.. OK so they fix it.
I did not do my dough because they did fix at least 3 critical problems and
I have been waiting a good year for these.
I am sure stuff will get fixed quicker once the backlog has been cleared.
I mean I hear they had something like 5000 compiler errors.. If they fixed
these it might make a big difference to some stability issues.
====
snip[ Have you now got a new role as Brian's new spin doctor?]
Yep.. waiting on my first Cheque...... can you hurry him up for me ? <g>
====
snip[ if Brian had acknowledged the problem, I doubt it would take Kuechler
more than a day or two to fix it.]
These guys are watching.. and I am sure they are listening to the
complaints..
Phil McGuinness
------------------
"Graham McKechnie" <g...@bigpond.net.au> wrote in message
news:pDIr9.31911$334....@news-server.bigpond.net.au...
> Phil,
>
Mark Cooper
Mark
RPSoftware
>And stop
> bitching, because it will solve anything.
>should be:
And stop bitching, because it will NOT solve anything.
--
Regards,
Rene J. Pajaron
RPSoftware
Philippines
www.rpsoftware.cdo.ph
"RPSoftware" <*remove_this*rps...@mozcom.com> wrote in message
news:aonl49$np1oe$1...@ID-143532.news.dfncis.de...
RPSoftware
My 2.6 is coming maybe later this month or early next month!
I am stuck right now with dbfcdx issues, freezing issue too (my datawindow
will suddenly freeze because my server is suddenly close by some damn
childwindows--and I am not doing it in my code--phantom?).
But I dveloped an dozens of workaround and I am tired doing all the time!
--
Regards,
Rene J. Pajaron
RPSoftware
Philippines
"Mark Cooper" <mlcooper...@aol.com> wrote in message
news:aonpuv$ogng0$1...@ID-147466.news.dfncis.de...
Graham McKechnie
Wrong. I may have discovered it last week, but it was reported many months
ago - check on google.
Graham
"RPSoftware" <*remove_this*rps...@mozcom.com> wrote in message
news:aonl49$np1oe$1...@ID-143532.news.dfncis.de...
Graham McKechnie
>>These guys are watching.. and I am sure they are listening to the
>>complaints..
I hope so Phil, but then again why didn't they listen before.
Graham
"Phil McGuinness" <hey...@sherlock.com.au> wrote in message
news:aono98$ns1aj$1...@ID-88745.news.dfncis.de...
Phil McGuinness
They had a long long list.... and I guess they have to fix enough to get the
next version to market.
CA left them a "Hot potato for sure" and I dont think the comments in German
would have helped a lot.
But now that you have raised it... the will surely not ingore it.
Anybody else got some major beefs that need addressing urgently.?
Phil McGuinness
---
"Graham McKechnie" <g...@bigpond.net.au> wrote in message
news:J6Nr9.32514$334....@news-server.bigpond.net.au...
RPSoftware
Gram,
> Wrong. I may have discovered it last week, but it was reported many months
> ago - check on google.
OK, i got it.
I believe Brian and his team are doing about this, if you can two weeks
waiting and expecting it get repaired, maybe a days or week or another paid
version upgrade to make our life as VO programer a little comforting for a
while. Why not?
Or you can move on, just like many others are doing.
--
Regards,
Rene J. Pajaron
RPSoftware
Philippines
www.rpsoftware.cdo.ph
"Graham McKechnie" <g...@bigpond.net.au> wrote in message
news:J6Nr9.32513$334....@news-server.bigpond.net.au...
RPSoftware
I cut a whats new on 2.6:
I am reading this many times, because it affect mine too, and do not know
mine if this is a bug or features?
....
Rebuild all ANSI index files created by CA Visual Objects 2.0 applications
if index keys contain punctuation marks and other
non-alphanumeric characters. The following sample shows the different
behavior:
The following table shows a list of words sorted both ways:
VO 2.0 Application: VO 2.6 Application:
----------------------------------------------------------------
xx rest xx'rest
xx*rest xx-rest
xx_rest xx rest
xx+rest xx*rest
xxarest xx_rest
xxbrest xx+rest
xxcrest xxarest
xx'rest xxbrest
xx-rest xxcrest
xxrest xxrest
xxzrest xxzrest
Results of seek operations in both versions:
VO 2.0 Application: VO 2.6 Application:
----------------------------------------------------------------
DBSEEK("xx-r") DBSEEK("xx-r")
Result: .T. Result: .T.
Found : xx'rest Found : xx-rest
DBSEEK("xx-") DBSEEK("xx-")
Result: .T. Result: .T.
Found : xx rest Found : xx-rest
Remark: Index files not containing punctuation marks and other
non-alphanumeric characters or index files with non STRING
key types or OEM index files do not have to be rebuilt.
--
Regards,
Rene J. Pajaron
RPSoftware
Philippines
"Graham McKechnie" <g...@bigpond.net.au> wrote in message
news:J6Nr9.32514$334....@news-server.bigpond.net.au...
Gary Stark
>
I'm cooking a couple of trout for dinner tonight ...
<g>
--
g.
Gary Stark
gst...@RedbacksWeb.com
http://RedbacksWeb.com
Thomas Venus
a little to much 'fuck' / 'shit' in your posting.
I don't think that this is the right way to talk with eachother
Thomas
"Graham McKechnie" <g...@bigpond.net.au> schrieb im Newsbeitrag
news:r0Ar9.30247$334....@news-server.bigpond.net.au...
TSDing
Regards
Ding
Graham McKechnie
Sorry about the language - but hell I was talking to Phil, who is another
Aussie. You'll just have to get used to way we talk to each other. Don't
forget it was the first word I ever heard
Graham
"Thomas Venus" <ve...@wm-schmuck.de> wrote in message
news:aooik3$j1e$02$1...@news.t-online.com...
Amilcar A. Camargo
On Fri, 18 Oct 2002 05:37:45 GMT, "Graham McKechnie"
<g...@bigpond.net.au> wrote:
>Rene,
>
>Wrong. I may have discovered it last week, but it was reported many months
>ago - check on google.
There is a full topic about ANSI-OEM conversion an collation sorting
inside the What's new document. It even displays a comparision table
for the results. Maybe you can check to see if now it works as
expected (the table include punctuation chars).
Regards,
Amilcar A. Camargo F.
Sand, S. A.
Guatemala, C. A.
Graham McKechnie
> There is a full topic about ANSI-OEM conversion an collation sorting
> inside the What's new document. It even displays a comparision table
> for the results. Maybe you can check to see if now it works as
> expected (the table include punctuation chars).
Yes I've read it, however, it still dosen't match the way Palm sorts, so no
joy. With some help from a colleague, I have now been able to force VO to
index the records in the order I want it, so that once the data is on the
Palm device, the Palm's DmFindSortPosition or any binarysearch routine
returns the same index # as you would expect to get from VO.
However, I'm more concerned with the following from my post
<<
Another related problem, is that I can create a dbf and then create an order
on the empty dbf and get a completely different sort sequence, as compared
to creating the dbf then appending records to it and then creating the
index. Can you please explain, why this is so? From my observations the
behaviour of 2.5 and 2.6 is exactly the same in this respect. This is
obviously not right.
<<
That is going to affect a lot more people, than just me. Most business dbfs,
start off empty with an index and then have records added as time goes by.
Well if you reindex at some later stage, you then get a different order.
Regards
Graham
"Amilcar A. Camargo" <acam...@acfsoft.com> wrote in message
news:3db0450d...@news.cis.dfn.de...
RPSoftware
I am email my partner today what date they ordered it so that I can make
estimate and tracked my order in Fedex.
--
Regards,
Rene J. Pajaron
RPSoftware
Philippines
www.rpsoftware.cdo.ph
"TSDing" <din...@pc.jaring.my> wrote in message
news:3daffe0c$1...@news.tm.net.my...
Geoff Schaller
Graham's right here.
We had the same problem, but to a lesser extent, in that clients were puting
in names with hyphens and some with spaces around the hyphens and then never
being able to "seek" them to retreive and correct the spelling. Then
suddenly, next day, they could! This floored me until Graham's
experimentation yielded the obvious answer. Our client's app was reindexing
every night. After the reindex, newly entered data appeared in the correct
collation sequence.
Ok, so there is an obvious workaround - reindexing. But this conclusion was
not really obvious before and I hope a future version will fix this so that
it is not necessary. There is no way our clients can afford to call a
reindex every time new data is entered where hyphens, spaces, etc are
involved. So, it is a problem for some people and it is one I intend to
track through the next beta series.
Geoff
"Amilcar A. Camargo" <acam...@acfsoft.com> wrote in message
news:3db0450d...@news.cis.dfn.de...
Amilcar A. Camargo
<g...@bigpond.net.au> wrote:
>Hi Amilcar,
>
>However, I'm more concerned with the following from my post
>
><<
>Another related problem, is that I can create a dbf and then create an order
>on the empty dbf and get a completely different sort sequence, as compared
>to creating the dbf then appending records to it and then creating the
>index. Can you please explain, why this is so? From my observations the
>behaviour of 2.5 and 2.6 is exactly the same in this respect. This is
>obviously not right.
><<
>
>That is going to affect a lot more people, than just me. Most business dbfs,
>start off empty with an index and then have records added as time goes by.
>Well if you reindex at some later stage, you then get a different order.
This one is new for me, and makes bad sounds. Do this happen just when
the key fields contain punctuation and the like?
Gary Stark
"work"
<g>
--
Stephane Hebert
There's a simple workaround though. Just a add a record after creation,
then create the index and finally delete the record.
I've never experienced that particular problem so I've never applied that
workaround. But that's what I would do.
Stephane
"Geoff Schaller" <geof...@bigpond.net.au> wrote in message
news:re0s9.67$DP6...@news-server.bigpond.net.au...
Graham McKechnie
Thanks for your invaluable advice.
Graham
"Adriano Rui Gominho" <a...@mail.EUnet.pt> wrote in message
news:avtuqu4b1l0arnqpj...@4ax.com...
> Graham,
>
> If you were a customer of mine I would be deligthed to refund your
> money and move on. We're all in this for the money but some of us
> don't dispense with the fun, not to mention politeness.
>
>
> Adriano
>
>
> On Thu, 17 Oct 2002 12:45:21 GMT, "Graham McKechnie"
> <g...@bigpond.net.au> wrote:
>
> >Hi Brian,
> >
> >Your Readme in 2.6 states that you have fixed the problem with collation
> >see -
> >
> >>>Changed ANSI (Windows) collation sequences.
> >
> >However it refers to 2.0, not 2.5. Has there been any change in this area
> >from 2.5 to 2.6 with CDX's?
> >
> >Last week I wrote here about a certain sequence that didn't work as
compared
> >to the way Palm sorts their databases. This problem re the "-" char
doesn't
> >appear to be addressed in 2.6. Many NG users (and I presume beta testers)
> >wrote back and said that this problem had been reported to CA and Grafx
and
> >should be fixed in 2.6.
> >
> >Brian can you confirm if this problem was addressed, if it was what where
> >the rules for the fix. In my testing with 2.6 this doesn't appear to be
the
> >case.
> >
> >Another related problem, is that I can create a dbf and then create an
order
> >on the empty dbf and get a completely different sort sequence, as
compared
> >to creating the dbf then appending records to it and then creating the
> >index. Can you please explain, why this is so? From my observations the
> >behaviour of 2.5 and 2.6 is the exactly the same in this respect. This is
> >obviously not right.
> >
> >Can you tell me exactly how long your company will take to address this
> >problem and when we are likely to see the fix. Please refer to my
previous
> >message for the exact strings to reproduce this.
> >
> >I've just parted with $A415 for 2.6 and at this stage I can see no value
in
Graham McKechnie
> This one is new for me, and makes bad sounds. Do this happen just when
> the key fields contain punctuation and the like?
Well it certainly happens when there is a hyphen in the key field. Why the
pharmaceutical industry wants hyphens is their drug names is beyond me, but
I guess I'm not going to be able to change that, but it certainly causes
problems for VO. Remember, this is not a 2.6 problem, its also there in 2.5.
I haven't done in specific tests without hyphens, but I suspect its not a
problem.
A simple test - on a Palm do the following in Addresss book. Make the
following entries
a-a
a-ab
a a
a ab
Now do the same in VO. I think you will see what I mean.
I'm going to do some more tests next week, but now I've got the data sorted
the way I want, for the weekend I'm back to writing code on the Palm. If you
guys have time to play with VO over the weekend, I'd like to hear your
results though.
Regards
Graham
"Amilcar A. Camargo" <acam...@acfsoft.com> wrote in message
news:3db091dc...@news.cis.dfn.de...
Ed Ratcliffe
VO Set ansi false ->
a a
a ab
a-a
a-ab
VO Set ansi true -> as per your list.
I do not own a Palm, so do not know how it would be sorted on one??
Ed
Phil McGuinness
Yeah.. and you lot should not have been listening in... to this private
conversation.!
Phil McGuinness
------------------
"Graham McKechnie" <g...@bigpond.net.au> wrote in message
news:LETr9.33706$334....@news-server.bigpond.net.au...
Geoff Schaller
Can't. You need exclusive access to reindex and so during the day, new
records are not available until the overnight reindex. This is a bugger
because you can't get to that record you just added! This would be even more
difficult in an app that needs 24hr availability. You just can't stop the
app to keep reindexing. This is the problem. It kind of solves it for Graham
for the moment but its still a nuisance. In other circumstances its a real
drama.
Geoff
"Stephane Hebert" <stephhebert@yourpants~videotron~.ca> wrote in message
news:Hh3s9.25857$48.22272@fe04...
Phil McGuinness
Problem solved if I understood the problem. The trick is to use the
Collation sequence.
Phil McGuinness
-------------------
FUNCTION Start()
LOCAL oDB AS dbserver
LOCAL cDataDBF := 'test_data.dbf' AS STRING
LOCAL aData := {} AS ARRAY
LOCAL cIndex := 'TmpIndex' AS STRING
SetDefault("c:\temp\")
SetCollation(#CLIPPER) // The collation sequence is the trick here to get
say 'a a' before 'a-a'
DBCREATE( cDataDBF, {{ "myData", "C", 6, 0 }}, 'DBFCDX')
//
//aData := { "a-a", "a-ab", "a a", "a ab" }
aData := { "a a", "a ab", "a-a", "a-ab" }
//
oDB := DBServer{cDataDBF, DBEXCLUSIVE}
IF oDB:Used
AEval( aData, { | aVal | oDB:Append(), oDB:FIELDPUT(#mydata , aVal),
oDB:Commit() })
//
ShowResults(oDB) // Results show before an Index the data in the DBF
matches the orginal input
//
oDB:CreateIndex( cIndex, [mydata] )
//
oDB:SetOrder( cIndex )
//
ShowResults(oDB) // Results show the results after index which match the
COLLATION sequence set.
//----------------------------
IF oDB:Append()
oDB:FIELDPUT(#mydata , 'a-b')
oDB:FIELDPUT(#mydata , 'c c')
oDB:Commit()
ENDIF
//
ShowResults(oDB) // Results show when appending items the Index is
correctly maintain matching Collation sequence.
//
oDB:Close()
ENDIF
//
WAIT
==============
FUNCTION ShowResults(oDB)
oDB:Gotop()
ShowArray(oDB:GetArray(oDB:LASTREC, #mydata) )
? '---------------- Sherlock Software -----'
Amilcar A. Camargo
On Fri, 18 Oct 2002 22:49:59 GMT, "Geoff Schaller"
<geof...@bigpond.net.au> wrote:
>Amilcar,
>
>Graham's right here.
>
>We had the same problem, but to a lesser extent, in that clients were puting
>in names with hyphens and some with spaces around the hyphens and then never
>being able to "seek" them to retreive and correct the spelling. Then
>suddenly, next day, they could! This floored me until Graham's
>experimentation yielded the obvious answer. Our client's app was reindexing
>every night. After the reindex, newly entered data appeared in the correct
>collation sequence.
What a nightmare. This could drive anyone crazy!!
Thx for sharing!!
Phil McGuinness
I think Geoff said he had to same problem.... but we could have guessed
that.
Phil
----
>
t
"Amilcar A. Camargo" <acam...@acfsoft.com> wrote in message
news:3db0e90a...@news.cis.dfn.de...
Graham McKechnie
I think you've just proved the point.
Graham
"Ed Ratcliffe" <Gryphon...@telus.net> wrote in message
news:NR3s9.37345$ER5.2...@news2.telusplanet.net...
Graham McKechnie
Agree with your test.
Well you are half way there, according to my test here. I don't think you
need to own a Palm for what I was talking about.
Now try the empy dbf test and create the index. Then append those records.
Then reindex. I haven't done it myself because I'm really busy with the C
stuff at the moment, but I will get back to it next week.
Geoff, confirms it with his post and what his users say sort of makew sense.
I'm not yet confinced as to one way or the other, but I have observed the
sort order is different and that is a big VO problem for all of us.
As I said before, it would be really neat if Brian could step in an make a
statement, one way or the other.
I really will get back to it, but for now its sorting ok for my means, its
held up my C stuff for long enough. The Palm stuff also has its problems, as
some of my beta testers have pointed out, so I'm into that right know - VO
is off my list right now, this weekend is all Palm. You just don't get
anything done, when you fight two compilers, and who maybe right.
However, please let me know when you are done with your tests.
Regards
Graham
"Ed Ratcliffe" <Gryphon...@telus.net> wrote in message
news:NR3s9.37345$ER5.2...@news2.telusplanet.net...
Graham McKechnie
> Let's see what Brian says about your issues......
Well I guess I'm still waiting, just like you. Then again I haven't seen
anything from Brian in this thread. How many days is it now?
Graham
"Marc Verkade" <ma...@martiNOSPAM.nl> wrote in message
news:aomkj4$1f1l$1...@news.kabelfoon.nl...
> Graham
>
> 'this is not what you have delivered.'!
> I totally disagree here....
>
> I see 2.6 as a start for future VO development..
>
> Let's see what Brian says about your issues......
> Marc
>
>
> "Phil McGuinness" <hey...@sherlock.com.au> schreef in bericht
> > "Graham McKechnie" <g...@bigpond.net.au> wrote in message
> > news:Bhyr9.29857$334....@news-server.bigpond.net.au...
Mark Cooper
Mark
Marc Verkade
I'm still hoping he will shine some light on your issue..
Grtz, Marc
"Graham McKechnie" <g...@bigpond.net.au> schreef in bericht
news:Blns9.3057$DP6....@news-server.bigpond.net.au...
Marc
> <<
> Another related problem, is that I can create a dbf and then create an order
> on the empty dbf and get a completely different sort sequence, as compared
> to creating the dbf then appending records to it and then creating the
> index. Can you please explain, why this is so? From my observations the
> behaviour of 2.5 and 2.6 is exactly the same in this respect. This is
> obviously not right.
> <<
Although I have not seen this myself I can confirm that another VO
developer (Larry Peterson) has mentioned this to me before (w/2.5).
-Marc-
Graham McKechnie
I didn't notice this post, I thought you just sent it to me privately.
As you haven't got back to me, I hope you will excuse me, if I post my
version of the code, which more clearly demonstrates the problem.
Actually I had hoped by now, you would have got around to checking what I
sent.
It doesn't seem right, to do this in private, the others here, should see
this example with the corrections, which then demonstrates the problem.
For anyone interested enough, and you bloody well should be. First time you
run it, create the index first, second time create the index last. Don't get
too carried away with SetCollation/SetAnsi, I now don't believe that to be
the problem, I originally thought it was, but you will get a different order
depending on when you create the index, as Geoff confirmed a couple of days
ago.
I'm going to start another thread, with the rest of my comments for I don't
believe they should be here - its subject is "Reply from Brian"
Regards
Graham
PS My modified code follows, with comments prefaced by gmk, the rest are
Phils
FUNCTION Start()
LOCAL oDB AS dbserver
LOCAL cDataDBF := 'test_data.dbf' AS STRING
LOCAL aData := {} AS ARRAY
LOCAL cIndex := 'tmpIndex' AS STRING
Local cFileName As String
Local lSuccess As Logic
SetDefault("d:\McGuinness")
SetCollation(#Windows) // The collation sequence is the trick here to get
say 'a a' before 'a-a'
DBCREATE( cDataDBF, {{ "myData", "C", 6, 0 }}, 'DBFCDX')
//
//aData := { "a-a", "a-ab", "a a", "a ab" }
aData := { "a a", "a ab", "a-a", "a-ab" }
//
cFileName := GetDefault() + "tmpIndex.cdx"
If File( cFileName )
lSuccess := FErase( Psz( cFileName ) )
If ! lSuccess
? DosErrString( FError() )
Endif
Endif
// gmk -- Phil, I think you meant for a CDX, but you left the param off, so
got an ntx
oDB := DBServer{cDataDBF, DBEXCLUSIVE,DBREADWRITE, "DBFCDX"}
IF oDB:Used
// gmk 21/10/2002 Create it before you put in data in
// oDB:CreateIndex( cIndex, [mydata] )
//
oDB:SetOrder( cIndex )
AEval( aData, { | aVal | oDB:Append(), oDB:FIELDPUT(#mydata , aVal),
oDB:Commit() })
//
ShowResults(oDB) // Results show before an Index the data in the DBF
matches the orginal input
//
// oDB:CreateIndex( cIndex, [mydata] )
//
// oDB:SetOrder( cIndex )
//
ShowResults(oDB) // Results show the results after index which match the
COLLATION sequence set.
//----------------------------
IF oDB:Append()
oDB:FIELDPUT(#mydata , 'a-b')
oDB:FIELDPUT(#mydata , 'c c')
oDB:Commit()
ENDIF
//
// gmk 21/10/2002 Create it after you put the data in. Note how the order
is different, as compared to creating the index before adding the data.
oDB:CreateIndex( cIndex, [mydata] )
ShowResults(oDB) // Results show when appending items the Index is
correctly maintain matching Collation sequence.
//
oDB:Close()
ENDIF
//
WAIT
Larry Peterson
automatically strips the hyphen before updating the file. I've had to do the
same thing for my insurance carrier database and inventory items list, as the
same bug has shown up in each case. I've found that recreating the index solves
the problem until they update the index again, and eventually it's out of wack
again. I'm using Ansii collation, and I've seen it on two systems running under
Advantage. I'm going to rebuild and recreate the indexes with Clipper collation
and see if there is any change.
Larry
Geoff Schaller
loss of the hyphen may lead to some confusion looking up the data. The users
are used to the hyphen being present because visually, it helps break up
long names etc.
In our case it involves names. People don't like their letters sent to them
with the hyphens in their names. They get all uppity and fussy <g>.
Geoff
"Larry Peterson" <la...@lynxmedsoft.com> wrote in message
news:3DB69F41...@lynxmedsoft.com...