PhpWiki

Dan Frankowski schrieb:
> Unfortunately, I have not yet had time to try to fix this. I have 
> checked that they it is still in the current CVS version.

Just fixed it, thanks. (say print a better warning)

Dan,
I just gave you CVS access to sf.net.

I added now the preliminary lib/wikilens,
and changed plugin/RateIt.php (which is responsible for the action, not 
lib/rate.php) and the theme.

Please fix that to your needs and think of the general wide audience.
(not only peardb, ...)
Some iterator is currently broken. Probably more.
But please don't touch any other core files, send me the needed patches 
instead please.

I also added two more star icon sets with another prefix, but is isn't 
completed yet. (small and big red stars)

Maybe the external rating engines will be removed, but for now I let the 
interface in. You know that I was playing with a mysql version of 
suggest and the java server from your group will also be available soon. 
(hopefully)
-- 
Reini Urban
http://phpwiki.sf.net/

Reini Urban wrote:

> Dan,
> I just gave you CVS access to sf.net.


Thanks!

> I added now the preliminary lib/wikilens,
> and changed plugin/RateIt.php (which is responsible for the action, 
> not lib/rate.php) and the theme.


Okay.

> Please fix that to your needs and think of the general wide audience.


We always would like to do that, if we can.

> (not only peardb, ...)


Hmm. That is tougher. We probably will not broaden beyond PearDB for a 
long time. We just don't have the people to do it and support it. Heck, 
I'm swamped just keeping up with your changes (!), in addition to the 
other things I do.

> Some iterator is currently broken. Probably more.
> But please don't touch any other core files, send me the needed 
> patches instead please.


I definitely only want to change files that you are okay with. I won't 
change core files.

What about adding unit tests?

> I also added two more star icon sets with another prefix, but is isn't 
> completed yet. (small and big red stars)
>
> Maybe the external rating engines will be removed, but for now I let 
> the interface in. You know that I was playing with a mysql version of 
> suggest and the java server from your group will also be available 
> soon. (hopefully)


MultiLens, the recommendation engine that is used in MovieLens, is now 
open source, registered at sourceforge.

We will probably be using our built-in PHP recs for awhile, since it is 
more of a "small-world" recommender.

Dan

Dan Frankowski schrieb:
> Reini Urban wrote:
>> Dan,
>> I just gave you CVS access to sf.net.
> 
> Thanks!
> 
>> I added now the preliminary lib/wikilens,
>> and changed plugin/RateIt.php (which is responsible for the action, 
>> not lib/rate.php) and the theme.
> 
> Okay.
> 
>> Please fix that to your needs and think of the general wide audience.
> 
> We always would like to do that, if we can.
>> (not only peardb, ...)
> 
> Hmm. That is tougher. We probably will not broaden beyond PearDB for a 
> long time. We just don't have the people to do it and support it. 

dbi->query is a generic query that works for pear and adodb, and returns 
the backend specific iterator, which name is in ->iter_class.

> Heck, I'm swamped just keeping up with your changes (!), in addition to the 
> other things I do.
> 
>> Some iterator is currently broken. Probably more.

The iterator per se is not broken, but the query returns no result 
suitable to the iterator.

>> But please don't touch any other core files, send me the needed 
>> patches instead please.
> 
> 
> I definitely only want to change files that you are okay with. I won't 
> change core files.
> 
> What about adding unit tests?

Sure. I forgot these.
But note that I fixed and improved the test framework a bit. It works 
now for cli and web (and my debugger).

>> I also added two more star icon sets with another prefix, but is isn't 
>> completed yet. (small and big red stars)
>>
>> Maybe the external rating engines will be removed, but for now I let 
>> the interface in. You know that I was playing with a mysql version of 
>> suggest and the java server from your group will also be available 
>> soon. (hopefully)
> 
> MultiLens, the recommendation engine that is used in MovieLens, is now 
> open source, registered at sourceforge.

Yeah, just saw it. Great.

> We will probably be using our built-in PHP recs for awhile, since it is 
> more of a "small-world" recommender.
-- 
Reini Urban
http://phpwiki.sf.net/

Reini Urban wrote:

> Dan Frankowski schrieb:
>
>> Reini Urban wrote:
>>
>>> Dan,
>>> I just gave you CVS access to sf.net.
>>
>>
>> Thanks!
>


I am confused.

I am trying to update the unit tests to the package I sent you (make it 
more picky, add some tests). I tried to add a directory, and do not seem 
to have permissions:

../phpwiki-next/tests/unit) % cvs add testbox
? testbox/latest_ver
? testbox/links
? testbox/page_data
? testbox/ver_data
cvs [server aborted]: "add" requires write access to the repository
../phpwiki-next/tests/unit) % cat CVS/Root
:pserver:dfrankow@...

Can you give me write access?

Perhaps this could be because I am a project member, but as 
"Advisor/Mentor/Consultant" and not "Developer". Maybe change me to 
"Developer"?

Dan

Reini Urban wrote:

> Dan Frankowski schrieb:
>
>> I am trying to update the unit tests to the package I sent you (make 
>> it more picky, add some tests). I tried to add a directory, and do 
>> not seem to have permissions:
>>
>> ../phpwiki-next/tests/unit) % cvs add testbox
>> ? testbox/latest_ver
>> ? testbox/links
>> ? testbox/page_data
>> ? testbox/ver_data
>> cvs [server aborted]: "add" requires write access to the repository
>> ../phpwiki-next/tests/unit) % cat CVS/Root
>> :pserver:dfrankow@...
>>
>> Can you give me write access?
>>
>> Perhaps this could be because I am a project member, but as 
>> "Advisor/Mentor/Consultant" and not "Developer". Maybe change me to 
>> "Developer"?
>
>
> strange. done.


Thanks. I have not yet tested it, since you are uncertain about the 
directory. More below.

> But why would you like to add the textbox?
> Don't you have the latest test.php?
> I renamed it to .testbox, so that it wouldn't display in cvs.


I have the latest test.php in Sourceforge's repository. However, I also 
have a bunch of changes that I emailed to this list as a tarball and you 
haven't had time to apply. I thought I'd apply them.

One of the changes is to add an error handler (instead of just an assert 
handler). This is useful because then if there is incorrect PHP syntax, 
vars missing, etc., the tests don't run. In other words, the tests get 
more picky, which is good. When they get more picky, they want an 
existing testbox with some stuff in it (InterWikiMap, global_data, a few 
other things). Thus, I want to make a "testbox" in CVS with the right 
stuff and check it in, as I did on our local copy.

Since I do not wish to make changes that you don't want, please tell me 
that you approve. Then I'll go ahead and make the unit test changes, 
including adding a CVS-tracked "testbox" with stuff in it.

Dan

Dan Frankowski schrieb:
> One of the changes is to add an error handler (instead of just an assert 
> handler). This is useful because then if there is incorrect PHP syntax, 
> vars missing, etc., the tests don't run. In other words, the tests get 
> more picky, which is good. When they get more picky, they want an 
> existing testbox with some stuff in it (InterWikiMap, global_data, a few 
> other things). Thus, I want to make a "testbox" in CVS with the right 
> stuff and check it in, as I did on our local copy.
> 
> Since I do not wish to make changes that you don't want, please tell me 
> that you approve. Then I'll go ahead and make the unit test changes, 
> including adding a CVS-tracked "testbox" with stuff in it.

approved.

but the name "testbox" compared to out intermediate ".testbox"
might not fully understandable. can you think of a better name?

In a perl CPAN test framework the call it "t" (for the logic) and 
"output" (for the output which should be tested against).
i'm not quite accustomed to the java unittest structure.
-- 
Reini Urban
http://xarch.tu-graz.ac.at/home/rurban/

Dan Frankowski schrieb:
> If you put [A B] in your page, but with a newline, this breaks. In other 
> words:
> 
> [A
> B]
> 
> It used to say (in our 1.3.9-based code):
> 
> lib/InlineParser.php:297: Notice[8]: Undefined offset: 4
> lib/InlineParser.php:297: Notice[8]: Undefined offset: 3
> lib/InlineParser.php:297: Notice[8]: Undefined offset: 2
> lib/InlineParser.php:297: Notice[8]: Undefined offset: 1
> 
> Now it says:
> 
> lib/InlineParser.php:325: Notice[8]: Undefined offset: 4 (...repeated 4 
> times)

BTW: This new collapsed php warning is the new default behaviour of 
postponed errors and warnings. I thought this is better than the 
repeated output of "missing DEBUG, assuming 'DEBUG' ..." 20 times.
-- 
Reini Urban
http://xarch.tu-graz.ac.at/home/rurban/

Reini Urban wrote:

> Dan Frankowski schrieb:
>
>> One of the changes is to add an error handler (instead of just an 
>> assert handler). This is useful because then if there is incorrect 
>> PHP syntax, vars missing, etc., the tests don't run. In other words, 
>> the tests get more picky, which is good. When they get more picky, 
>> they want an existing testbox with some stuff in it (InterWikiMap, 
>> global_data, a few other things). Thus, I want to make a "testbox" in 
>> CVS with the right stuff and check it in, as I did on our local copy.
>>
>> Since I do not wish to make changes that you don't want, please tell 
>> me that you approve. Then I'll go ahead and make the unit test 
>> changes, including adding a CVS-tracked "testbox" with stuff in it.
>
>
> approved.


Cool. I won't add it until we have a name you like (see below).

> but the name "testbox" compared to out intermediate ".testbox"
> might not fully understandable. can you think of a better name?


I'm sorry, I didn't understand this.

I picked "testbox" to mean a "test sandbox." I agree it is not a great 
name, but I'm having a hard time thinking of better ones. A list of 
possibilities:

1. testbox
2. test_sandbox
3. sandbox
4. backend_dir
5. backend_data
6. testdb

Any preferences? I'm not sure any are very good.

> In a perl CPAN test framework the call it "t" (for the logic) and 
> "output" (for the output which should be tested against).


Neither "t" nor "output" is appropriate here. It's more like "input". I 
guess that's another possibility.

> i'm not quite accustomed to the java unittest structure.


Java has no standard proposal for the input data upon which to base unit 
tests.

Dan

I am blocked on this until we choose a name for 'testbox'.

Comments?

Dan

Dan Frankowski wrote:

> Reini Urban wrote:
>
>> Dan Frankowski schrieb:
>>
>>> One of the changes is to add an error handler (instead of just an 
>>> assert handler). This is useful because then if there is incorrect 
>>> PHP syntax, vars missing, etc., the tests don't run. In other words, 
>>> the tests get more picky, which is good. When they get more picky, 
>>> they want an existing testbox with some stuff in it (InterWikiMap, 
>>> global_data, a few other things). Thus, I want to make a "testbox" 
>>> in CVS with the right stuff and check it in, as I did on our local 
>>> copy.
>>>
>>> Since I do not wish to make changes that you don't want, please tell 
>>> me that you approve. Then I'll go ahead and make the unit test 
>>> changes, including adding a CVS-tracked "testbox" with stuff in it.
>>
>>
>>
>> approved.
>
>
>
> Cool. I won't add it until we have a name you like (see below).
>
>> but the name "testbox" compared to out intermediate ".testbox"
>> might not fully understandable. can you think of a better name?
>
>
>
> I'm sorry, I didn't understand this.
>
> I picked "testbox" to mean a "test sandbox." I agree it is not a great 
> name, but I'm having a hard time thinking of better ones. A list of 
> possibilities:
>
> 1. testbox
> 2. test_sandbox
> 3. sandbox
> 4. backend_dir
> 5. backend_data
> 6. testdb
>
> Any preferences? I'm not sure any are very good.
>
>> In a perl CPAN test framework the call it "t" (for the logic) and 
>> "output" (for the output which should be tested against).
>
>
>
> Neither "t" nor "output" is appropriate here. It's more like "input". 
> I guess that's another possibility.
>
>> i'm not quite accustomed to the java unittest structure.
>
>
>
> Java has no standard proposal for the input data upon which to base 
> unit tests.
>
> Dan
>
>
>
> -------------------------------------------------------
> This SF.Net email sponsored by Black Hat Briefings & Training.
> Attend Black Hat Briefings & Training, Las Vegas July 24-29 - digital 
> self defense, top technical experts, no vendor pitches, unmatched 
> networking opportunities. Visit http://www.blackhat.com
> _______________________________________________
> Phpwiki-talk mailing list
> Phpwiki-talk@...
> https://lists.sourceforge.net/lists/listinfo/phpwiki-talk

Dan Frankowski wrote:
> I am blocked on this until we choose a name for 'testbox'.
> Comments?

let's just keep "testbox" then.

> Dan Frankowski wrote:
>> Reini Urban wrote:
>>> Dan Frankowski schrieb:
>>>
>>>> One of the changes is to add an error handler (instead of just an 
>>>> assert handler). This is useful because then if there is incorrect 
>>>> PHP syntax, vars missing, etc., the tests don't run. In other words, 
>>>> the tests get more picky, which is good. When they get more picky, 
>>>> they want an existing testbox with some stuff in it (InterWikiMap, 
>>>> global_data, a few other things). Thus, I want to make a "testbox" 
>>>> in CVS with the right stuff and check it in, as I did on our local 
>>>> copy.
>>>>
>>>> Since I do not wish to make changes that you don't want, please tell 
>>>> me that you approve. Then I'll go ahead and make the unit test 
>>>> changes, including adding a CVS-tracked "testbox" with stuff in it.
>>>
>>>
>>>
>>>
>>> approved.
>>
>>
>>
>>
>> Cool. I won't add it until we have a name you like (see below).
>>
>>> but the name "testbox" compared to out intermediate ".testbox"
>>> might not fully understandable. can you think of a better name?
>>
>>
>>
>>
>> I'm sorry, I didn't understand this.
>>
>> I picked "testbox" to mean a "test sandbox." I agree it is not a great 
>> name, but I'm having a hard time thinking of better ones. A list of 
>> possibilities:
>>
>> 1. testbox
>> 2. test_sandbox
>> 3. sandbox
>> 4. backend_dir
>> 5. backend_data
>> 6. testdb
>>
>> Any preferences? I'm not sure any are very good.
>>
>>> In a perl CPAN test framework the call it "t" (for the logic) and 
>>> "output" (for the output which should be tested against).
>>
>>
>>
>>
>> Neither "t" nor "output" is appropriate here. It's more like "input". 
>> I guess that's another possibility.
>>
>>> i'm not quite accustomed to the java unittest structure.
>>
>>
>>
>>
>> Java has no standard proposal for the input data upon which to base 
>> unit tests.
>>
>> Dan
-- 
Reini Urban
http://xarch.tu-graz.ac.at/home/rurban/