[PHP-DEV] Bug #14541 Updated: strtok broken again

2001-12-16 Thread mlemos 

ID: 14541
User updated by: [EMAIL PROTECTED]
Reported By: [EMAIL PROTECTED]
Status: Bogus
Bug Type: Strings related
Operating System: Any
PHP Version: 4.1.0
New Comment:

No, jmoore just broke the behaviour of PHP strtok function that worked like that since 
PHP 3 for more than 4 years.

This means that he also broke PHP code that relied in that behaviour, thus 
discouraging people to upgrade to PHP 4.1.0 .

It does not matter what a Unix manual says because PHP users could not care less about 
Unix manuals when the behaviour of a function that they rely on changes without 
backwards compatibility.


Previous Comments:


[2001-12-16 05:44:55] [EMAIL PROTECTED]

Documention is now fixed in CVS, thx 2 derick.



[2001-12-16 04:45:08] [EMAIL PROTECTED]

jmoore did not break this, he fixed a faulty behavior that existed since 4.0.0.
The current behavior is the correct one, as outlined by every unix or windows manual 
about the topic 'strtok'.

Derick



[2001-12-16 02:40:29] [EMAIL PROTECTED]

I understand that it is very hard to make developers write tests for new software, but 
at least those that commit bug fixes should be required to submit test scripts that 
reproduce the bugs if they do not exist yet.

As for myself, I always present test cases when they are possible in the bug report 
itself, just like I did for this. So, developers have at least half of the job done.

I think that is a matter of making it a rule by adding to the CODING_STANDARDS.



[2001-12-16 02:25:53] [EMAIL PROTECTED]

While we could try to force developers to write tests 
before they commit code (heck, MySQL does it), but we 
might not have much luck. :)

I think that we should look to the QA team (and interested 
individuals such as yourself) to start writing tests.

I am working on tests for the array functions right now 
(coincidentally, before I commit a whack of changes to the 
array functions. :)

In Frankfurt, Rasmus suggested that we develop a web-based 
interface for developing tests as a way to lower the 
barrier for writing tests. We could look at doing this.




[2001-12-16 02:15:52] [EMAIL PROTECTED]

It seems that strtok function is broken again.

The following script returns:



Should output as always (at least until PHP 4.0.6 it does):

string(0) ""
string(9) "something"

But it outputs:

string(9) "something"
bool(false)

It seems that jmoore broken in when he tried to fix this bug:

http://bugs.php.net/bug.php?id=13866

I think that no developer should be allowed to fix bugs before:

1) submit a test case to leave in the tests directory
2) Verify that his fixes do not make his and others tests fail.

Until this procedure becomes mandatory, we'll keep seeing a history of illness in 
functions like strtok that seems to never end.

This is what regressive tests are for. Zak, please help here! :-)






Edit this bug report at http://bugs.php.net/?id=14541&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

Re: [PHP-DEV] Bug #14541 Updated: strtok broken again

2001-12-16 Thread Sterling Hughes 

> On Sun, Dec 16, 2001 at 09:45:08AM -, [EMAIL PROTECTED] wrote : 
> > ID: 14541
> > Updated by: derick
> > Reported By: [EMAIL PROTECTED]
> > Old Status: Open
> > Status: Bogus
> > Bug Type: Strings related
> > Operating System: Any
> > PHP Version: 4.1.0
> > New Comment:
> > 
> > jmoore did not break this, he fixed a faulty behavior that
> > existed since 4.0.0.  The current behavior is the correct one,
> > as outlined by every unix or windows manual about the topic
> > 'strtok'.
> 
> The problem is, it is NOT documented in the NEWS file (at
> least, grepping for strtok didn't reveal anything).
>
It was documented (I removed it), however, the problem was it
was documented as a new feature in 4.2, not 4.1, and no one caught
this till *after* the release.

-Sterling

 
> -- 
> Please always Cc to me when replying to me on the lists.
> 
> -- 
> PHP Development Mailing List 
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
> To contact the list administrators, e-mail: [EMAIL PROTECTED]
> 

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14541 Updated: strtok broken again

2001-12-16 Thread mfischer 

ID: 14541
Updated by: mfischer
Reported By: [EMAIL PROTECTED]
Status: Bogus
Bug Type: Strings related
Operating System: Any
PHP Version: 4.1.0
New Comment:

Documention is now fixed in CVS, thx 2 derick.

Previous Comments:


[2001-12-16 04:45:08] [EMAIL PROTECTED]

jmoore did not break this, he fixed a faulty behavior that existed since 4.0.0.
The current behavior is the correct one, as outlined by every unix or windows manual 
about the topic 'strtok'.

Derick



[2001-12-16 02:40:29] [EMAIL PROTECTED]

I understand that it is very hard to make developers write tests for new software, but 
at least those that commit bug fixes should be required to submit test scripts that 
reproduce the bugs if they do not exist yet.

As for myself, I always present test cases when they are possible in the bug report 
itself, just like I did for this. So, developers have at least half of the job done.

I think that is a matter of making it a rule by adding to the CODING_STANDARDS.



[2001-12-16 02:25:53] [EMAIL PROTECTED]

While we could try to force developers to write tests 
before they commit code (heck, MySQL does it), but we 
might not have much luck. :)

I think that we should look to the QA team (and interested 
individuals such as yourself) to start writing tests.

I am working on tests for the array functions right now 
(coincidentally, before I commit a whack of changes to the 
array functions. :)

In Frankfurt, Rasmus suggested that we develop a web-based 
interface for developing tests as a way to lower the 
barrier for writing tests. We could look at doing this.




[2001-12-16 02:15:52] [EMAIL PROTECTED]

It seems that strtok function is broken again.

The following script returns:



Should output as always (at least until PHP 4.0.6 it does):

string(0) ""
string(9) "something"

But it outputs:

string(9) "something"
bool(false)

It seems that jmoore broken in when he tried to fix this bug:

http://bugs.php.net/bug.php?id=13866

I think that no developer should be allowed to fix bugs before:

1) submit a test case to leave in the tests directory
2) Verify that his fixes do not make his and others tests fail.

Until this procedure becomes mandatory, we'll keep seeing a history of illness in 
functions like strtok that seems to never end.

This is what regressive tests are for. Zak, please help here! :-)






Edit this bug report at http://bugs.php.net/?id=14541&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

Re: [PHP-DEV] Bug #14541 Updated: strtok broken again

2001-12-16 Thread derick 

On Sun, 16 Dec 2001, Markus Fischer wrote:

> > jmoore did not break this, he fixed a faulty behavior that
> > existed since 4.0.0.  The current behavior is the correct one,
> > as outlined by every unix or windows manual about the topic
> > 'strtok'.
>
> The problem is, it is NOT documented in the NEWS file (at
> least, grepping for strtok didn't reveal anything).

It is now documented in the manual.

Derick


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

Re: [PHP-DEV] Bug #14541 Updated: strtok broken again

2001-12-16 Thread Markus Fischer 

On Sun, Dec 16, 2001 at 09:45:08AM -, [EMAIL PROTECTED] wrote : 
> ID: 14541
> Updated by: derick
> Reported By: [EMAIL PROTECTED]
> Old Status: Open
> Status: Bogus
> Bug Type: Strings related
> Operating System: Any
> PHP Version: 4.1.0
> New Comment:
> 
> jmoore did not break this, he fixed a faulty behavior that
> existed since 4.0.0.  The current behavior is the correct one,
> as outlined by every unix or windows manual about the topic
> 'strtok'.

The problem is, it is NOT documented in the NEWS file (at
least, grepping for strtok didn't reveal anything).

-- 
Please always Cc to me when replying to me on the lists.

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14541 Updated: strtok broken again

2001-12-16 Thread derick 

ID: 14541
Updated by: derick
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Bogus
Bug Type: Strings related
Operating System: Any
PHP Version: 4.1.0
New Comment:

jmoore did not break this, he fixed a faulty behavior that existed since 4.0.0.
The current behavior is the correct one, as outlined by every unix or windows manual 
about the topic 'strtok'.

Derick

Previous Comments:


[2001-12-16 02:40:29] [EMAIL PROTECTED]

I understand that it is very hard to make developers write tests for new software, but 
at least those that commit bug fixes should be required to submit test scripts that 
reproduce the bugs if they do not exist yet.

As for myself, I always present test cases when they are possible in the bug report 
itself, just like I did for this. So, developers have at least half of the job done.

I think that is a matter of making it a rule by adding to the CODING_STANDARDS.



[2001-12-16 02:25:53] [EMAIL PROTECTED]

While we could try to force developers to write tests 
before they commit code (heck, MySQL does it), but we 
might not have much luck. :)

I think that we should look to the QA team (and interested 
individuals such as yourself) to start writing tests.

I am working on tests for the array functions right now 
(coincidentally, before I commit a whack of changes to the 
array functions. :)

In Frankfurt, Rasmus suggested that we develop a web-based 
interface for developing tests as a way to lower the 
barrier for writing tests. We could look at doing this.




[2001-12-16 02:15:52] [EMAIL PROTECTED]

It seems that strtok function is broken again.

The following script returns:



Should output as always (at least until PHP 4.0.6 it does):

string(0) ""
string(9) "something"

But it outputs:

string(9) "something"
bool(false)

It seems that jmoore broken in when he tried to fix this bug:

http://bugs.php.net/bug.php?id=13866

I think that no developer should be allowed to fix bugs before:

1) submit a test case to leave in the tests directory
2) Verify that his fixes do not make his and others tests fail.

Until this procedure becomes mandatory, we'll keep seeing a history of illness in 
functions like strtok that seems to never end.

This is what regressive tests are for. Zak, please help here! :-)






Edit this bug report at http://bugs.php.net/?id=14541&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14541 Updated: strtok broken again

2001-12-15 Thread mlemos 

ID: 14541
User updated by: [EMAIL PROTECTED]
Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: Strings related
Operating System: Any
PHP Version: 4.1.0
New Comment:

I understand that it is very hard to make developers write tests for new software, but 
at least those that commit bug fixes should be required to submit test scripts that 
reproduce the bugs if they do not exist yet.

As for myself, I always present test cases when they are possible in the bug report 
itself, just like I did for this. So, developers have at least half of the job done.

I think that is a matter of making it a rule by adding to the CODING_STANDARDS.

Previous Comments:


[2001-12-16 02:25:53] [EMAIL PROTECTED]

While we could try to force developers to write tests 
before they commit code (heck, MySQL does it), but we 
might not have much luck. :)

I think that we should look to the QA team (and interested 
individuals such as yourself) to start writing tests.

I am working on tests for the array functions right now 
(coincidentally, before I commit a whack of changes to the 
array functions. :)

In Frankfurt, Rasmus suggested that we develop a web-based 
interface for developing tests as a way to lower the 
barrier for writing tests. We could look at doing this.




[2001-12-16 02:15:52] [EMAIL PROTECTED]

It seems that strtok function is broken again.

The following script returns:



Should output as always (at least until PHP 4.0.6 it does):

string(0) ""
string(9) "something"

But it outputs:

string(9) "something"
bool(false)

It seems that jmoore broken in when he tried to fix this bug:

http://bugs.php.net/bug.php?id=13866

I think that no developer should be allowed to fix bugs before:

1) submit a test case to leave in the tests directory
2) Verify that his fixes do not make his and others tests fail.

Until this procedure becomes mandatory, we'll keep seeing a history of illness in 
functions like strtok that seems to never end.

This is what regressive tests are for. Zak, please help here! :-)






Edit this bug report at http://bugs.php.net/?id=14541&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14541 Updated: strtok broken again

2001-12-15 Thread zak 

ID: 14541
Updated by: zak
Reported By: [EMAIL PROTECTED]
Status: Open
Old Bug Type: Unknown/Other Function
Bug Type: Strings related
Operating System: Any
PHP Version: 4.1.0
New Comment:

While we could try to force developers to write tests 
before they commit code (heck, MySQL does it), but we 
might not have much luck. :)

I think that we should look to the QA team (and interested 
individuals such as yourself) to start writing tests.

I am working on tests for the array functions right now 
(coincidentally, before I commit a whack of changes to the 
array functions. :)

In Frankfurt, Rasmus suggested that we develop a web-based 
interface for developing tests as a way to lower the 
barrier for writing tests. We could look at doing this.


Previous Comments:


[2001-12-16 02:15:52] [EMAIL PROTECTED]

It seems that strtok function is broken again.

The following script returns:



Should output as always (at least until PHP 4.0.6 it does):

string(0) ""
string(9) "something"

But it outputs:

string(9) "something"
bool(false)

It seems that jmoore broken in when he tried to fix this bug:

http://bugs.php.net/bug.php?id=13866

I think that no developer should be allowed to fix bugs before:

1) submit a test case to leave in the tests directory
2) Verify that his fixes do not make his and others tests fail.

Until this procedure becomes mandatory, we'll keep seeing a history of illness in 
functions like strtok that seems to never end.

This is what regressive tests are for. Zak, please help here! :-)






Edit this bug report at http://bugs.php.net/?id=14541&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]