| Subject | Re: [IBO] Timeframe for new release? |
|---|---|
| Author | Jason Wharton |
| Post date | 2002-01-15T20:28:56Z |
Martijn,
In your case where you are dealing with a new IB 6.5 issue I understand you
expect me to just have it finished and supported by my own voluntary
efforts. To this I agree.
(I was late getting my IB 6.5 media kit but have it installed now and as
soon as I finish a deadline at the state on a new web app I'll get to it.)
just your specific case.
You nor anyone else is at fault. All that is done is appreciated. It's just
that when people want things done quickly I expect to see some sleeves
rolled up and some action taking place besides "pointing things out".
My comments are about the words you chose to use more than what you are
actually doing (which is much and is very helpful). You told me various
steps to repeat a case and it is in the queue. Because you didn't provide a
simple sample app isolating the problem it got put in a queue that has a
lower priority.
Here's a rough sketch of my priority system:
(All must contain clear and simple descriptions of the problem.)
1) IBO source files changed by the submitter who feels they are a good and
dependable fix. It's a huge bonus if they also provide their sample case but
not necessary.
2) A simple sample application that immediately demonstrates the problem
with full source and no hassles. It is preferred that the database be self
generating from a script, not a backup file or GDB directly. But, I will
take them as such as long as it is clear which version of InterBase/Firebird
they are using.
3) I begin to look at rejected source submissions and consider how they can
be cleaned up and better integrated. Some source submissions do not meet my
standards of coding or the submitter isn't aware that they are breaking
other situations. If I cannot get them cleaned up fairly quickly they get
rejected or bumped into my potential actions for future consideration.
4) A careful step-by-step description of a problem which has been well
defined and researched by the submitter. It's especially nice if it is using
one of my existing sample apps as a starting point.
5) I begin to look at rejected sample apps that I start into but for one
reason or another get frustrated with because they aren't simple enough or I
don't have a compatible working environment setup at the time.
6) I begin looking at step-by-step instructions which became obscure or
confusing at any point along the way and I decided to push them aside in
order to cover more pressing and convenient tasks.
7) This is when I am really hard-up for something to do and I begin taking
an exhaustive scan of all submitted bug cases, potential and otherwise, and
I try to make sense of them and build test cases for them. It is usually
rather unproductive but there is a lot of good which can come from doing
this so about once a month I spend a day or two doing this.
Based on what you have submitted you got a #4 ranking. The #1 and #2 buckets
still have some items in them so please be patient or consider elevating
your item's priority by raising it to those levels.
Use a TIB_Cursor if you want totally raw dataset processing.
What do you need the ROWS capability for with a buffered dataset anyway?
Seems if you are using the ROWS capability a cursor should fulfill the needs
just fine.
The changes to make will probably take less time than it will to setup the
test application to ensure their accuracy in. That is mainly what is holding
me back from completing this, other than I finally got my IB 6.5 just the
other week.
Regards,
Jason Wharton
CPS - Mesa AZ
http://www.ibobjects.com
> (no trying to sound rude or anything, having to do with users myself)I understand totally.
In your case where you are dealing with a new IB 6.5 issue I understand you
expect me to just have it finished and supported by my own voluntary
efforts. To this I agree.
(I was late getting my IB 6.5 media kit but have it installed now and as
soon as I finish a deadline at the state on a new web app I'll get to it.)
> In this case, I presented you with an easy reproducible case (preparingKeep in mind, that message was in part addressing the general audience, not
> a single SQL statement) and traced it down to the method where it went
> wrong... Then, I had no idea what the method actually did :)
just your specific case.
You nor anyone else is at fault. All that is done is appreciated. It's just
that when people want things done quickly I expect to see some sleeves
rolled up and some action taking place besides "pointing things out".
My comments are about the words you chose to use more than what you are
actually doing (which is much and is very helpful). You told me various
steps to repeat a case and it is in the queue. Because you didn't provide a
simple sample app isolating the problem it got put in a queue that has a
lower priority.
Here's a rough sketch of my priority system:
(All must contain clear and simple descriptions of the problem.)
1) IBO source files changed by the submitter who feels they are a good and
dependable fix. It's a huge bonus if they also provide their sample case but
not necessary.
2) A simple sample application that immediately demonstrates the problem
with full source and no hassles. It is preferred that the database be self
generating from a script, not a backup file or GDB directly. But, I will
take them as such as long as it is clear which version of InterBase/Firebird
they are using.
3) I begin to look at rejected source submissions and consider how they can
be cleaned up and better integrated. Some source submissions do not meet my
standards of coding or the submitter isn't aware that they are breaking
other situations. If I cannot get them cleaned up fairly quickly they get
rejected or bumped into my potential actions for future consideration.
4) A careful step-by-step description of a problem which has been well
defined and researched by the submitter. It's especially nice if it is using
one of my existing sample apps as a starting point.
5) I begin to look at rejected sample apps that I start into but for one
reason or another get frustrated with because they aren't simple enough or I
don't have a compatible working environment setup at the time.
6) I begin looking at step-by-step instructions which became obscure or
confusing at any point along the way and I decided to push them aside in
order to cover more pressing and convenient tasks.
7) This is when I am really hard-up for something to do and I begin taking
an exhaustive scan of all submitted bug cases, potential and otherwise, and
I try to make sense of them and build test cases for them. It is usually
rather unproductive but there is a lot of good which can come from doing
this so about once a month I spend a day or two doing this.
Based on what you have submitted you got a #4 ranking. The #1 and #2 buckets
still have some items in them so please be patient or consider elevating
your item's priority by raising it to those levels.
> Personally, I thinkIBO does harness a lot. That is what makes it such a great product to use.
> this is a bit too much work done by IBO and it makes it extra hard to
> keep up with new stuff, like ROWS, for example...
Use a TIB_Cursor if you want totally raw dataset processing.
What do you need the ROWS capability for with a buffered dataset anyway?
Seems if you are using the ROWS capability a cursor should fulfill the needs
just fine.
The changes to make will probably take less time than it will to setup the
test application to ensure their accuracy in. That is mainly what is holding
me back from completing this, other than I finally got my IB 6.5 just the
other week.
Regards,
Jason Wharton
CPS - Mesa AZ
http://www.ibobjects.com