On Thu, Jul 3, 2008 at 9:53 AM, Serge Stinckwich Serge.Stinckwich@gmail.com wrote:
There is a bug with the Method finder tool. Try for example : 1. 1. 2. It runs forever ...
Try the same thing in any base Squeak image starting from Squeak 3.8. It seems you found a long-standing bug.
Damien Cassou a écrit :
On Thu, Jul 3, 2008 at 9:53 AM, Serge Stinckwich Serge.Stinckwich@gmail.com wrote:
There is a bug with the Method finder tool. Try for example : 1. 1. 2. It runs forever ...
Try the same thing in any base Squeak image starting from Squeak 3.8. It seems you found a long-standing bug.
I usually show the method finder to Smalltalk beginners. It's the first time, i see this one ;-) There is another one. Try : 'abc'. 'ab' and choose the method found to launch an error.
-- Serge Stinckwich http://blog.doesnotunderstand.org/
I found the source of the bug.
1 . 1 . 2
will cause the MethodFinder to try:
1 perform: #printStringBase: withArguments: #(1)
i.e., 1 printStringBase: 1
This generates an infinite loop within Integer>>printStringBase:
Integer>>printStringBase: should be corrected to complain if its argument (base) is less than 2.
I tried the obvious "self assert: base > 1.", but the Method Finder does not like this since it only catches Errors, not AssertionFailures, so the fix is as follows:
Integer>>printStringBase: base | stream integer next | (base < 2) ifTrue: [Error signal: 'base must be >= 2']. ... "rest as before"
I did not find any error in the example 'abc' . 'ab' .
- on
On Jul 3, 2008, at 10:27, Serge Stinckwich wrote:
Damien Cassou a écrit :
On Thu, Jul 3, 2008 at 9:53 AM, Serge Stinckwich Serge.Stinckwich@gmail.com wrote:
There is a bug with the Method finder tool. Try for example : 1.
It runs forever ...
Try the same thing in any base Squeak image starting from Squeak 3.8. It seems you found a long-standing bug.
I usually show the method finder to Smalltalk beginners. It's the first time, i see this one ;-) There is another one. Try : 'abc'. 'ab' and choose the method found to launch an error.
-- Serge Stinckwich http://blog.doesnotunderstand.org/
Oscar Nierstrasz <oscar.nierstrasz <at> gmail.com> writes:
I found the source of the bug.
1 . 1 . 2
will cause the MethodFinder to try:
1 perform: #printStringBase: withArguments: #(1)
i.e., 1 printStringBase: 1
This generates an infinite loop within Integer>>printStringBase:
Good find Oscar but already on Mantis, see http://bugs.squeak.org/view.php?id=6724 See also http://bugs.squeak.org/view.php?id=6710 and search MethodFinder on Mantis.
Nicolas
Hi Nicolas,
there is no patch file on the Mantis page. Where can I find a patch with an "approved" version stamp, if anywhere?
On Thu, Jul 3, 2008 at 2:54 AM, nicolas cellier ncellier@ifrance.com wrote:
Oscar Nierstrasz <oscar.nierstrasz <at> gmail.com> writes:
I found the source of the bug.
1 . 1 . 2
will cause the MethodFinder to try:
1 perform: #printStringBase: withArguments: #(1)
i.e., 1 printStringBase: 1
This generates an infinite loop within Integer>>printStringBase:
Good find Oscar but already on Mantis, see http://bugs.squeak.org/view.php?id=6724 See also http://bugs.squeak.org/view.php?id=6710 and search MethodFinder on Mantis.
Nicolas
Eliot Miranda a écrit :
Hi Nicolas,
there is no patch file on the Mantis page. Where can I find a patchwith an "approved" version stamp, if anywhere?
On Thu, Jul 3, 2008 at 2:54 AM, nicolas cellier <ncellier@ifrance.com mailto:ncellier@ifrance.com> wrote:
Oscar Nierstrasz <oscar.nierstrasz <at> gmail.com <http://gmail.com>> writes:
My "approval" is neither required nor authoritative. Any other opinion counts. I suppose Oscar's patch is OK, didn't check because pipermail did scrub the attachment.
My image does not need such patch because MethodFinder for 1. 1. 2. does not fail anymore thanks to http://bugs.squeak.org/view.php?id=6887
And for the same price, i have faster integer printString.
Of course, this change should be approved by someone else but me.
Cheers
Nicolas
On Thu, Jul 3, 2008 at 12:09 PM, nicolas cellier ncellier@ifrance.com wrote:
Eliot Miranda a écrit :
Hi Nicolas,
there is no patch file on the Mantis page. Where can I find a patch with an "approved" version stamp, if anywhere?
On Thu, Jul 3, 2008 at 2:54 AM, nicolas cellier <ncellier@ifrance.commailto: ncellier@ifrance.com> wrote:
Oscar Nierstrasz <oscar.nierstrasz <at> gmail.com http://gmail.com> writes:
My "approval" is neither required nor authoritative. Any other opinion counts. I suppose Oscar's patch is OK, didn't check because pipermail did scrub the attachment.
My image does not need such patch because MethodFinder for 1. 1. 2. does not fail anymore thanks to http://bugs.squeak.org/view.php?id=6887
And for the same price, i have faster integer printString.
Of course, this change should be approved by someone else but me.
Cheers
Hi Nicolas,
I'm asking to avoid the "source same but rev changed" issue. I don;t want to fix the method myself with my own timestamp. Isn;t there a place where an original fix can be found with a timestamp we can all use?
Nicolas
Eliot Miranda a écrit :
Hi Nicolas,
I'm asking to avoid the "source same but rev changed" issue. Idon;t want to fix the method myself with my own timestamp. Isn;t there a place where an original fix can be found with a timestamp we can all use?
My own solution is:
Installer ensureFix: '6887 Faster printString for LargeInteger'.
which is currently equivalent to doing this:
"fix begin" Installer mantis ensureFix: 6512. Installer mantis bug: 6887 fix:'Integer-fastPrint-M6887-nice.5.cs'. "fix test" Installer mantis bug: 6887 fix:'Integer-fastPrint-test-M6887-nice.1.cs'. "fix end"
Or see directly http://bugs.squeak.org/view.php?id=6887.
Cheers
nicolas cellier a écrit :
Oscar Nierstrasz <oscar.nierstrasz <at> gmail.com> writes:
I found the source of the bug.
1 . 1 . 2
will cause the MethodFinder to try:
1 perform: #printStringBase: withArguments: #(1)
i.e., 1 printStringBase: 1
This generates an infinite loop within Integer>>printStringBase:
Good find Oscar but already on Mantis, see http://bugs.squeak.org/view.php?id=6724 See also http://bugs.squeak.org/view.php?id=6710 and search MethodFinder on Mantis.
With your patch, Method Finder works but :
1 printStringBase: 1 reply a divide by zero with is a bit strange ...
-- Serge Stinckwich http://blog.doesnotunderstand.org/
Serge Stinckwich a écrit :
nicolas cellier a écrit :
Good find Oscar but already on Mantis, see http://bugs.squeak.org/view.php?id=6724 See also http://bugs.squeak.org/view.php?id=6710 and search MethodFinder on Mantis.
With your patch, Method Finder works but :
1 printStringBase: 1 reply a divide by zero with is a bit strange ...
--
Yes, you are right. Oscar answers with a more explicit error message, which seems better also to me.
I asked to myself whether i should assert preconditions in my code, and decided not to do so for several reasons.
1) I generally avoid preconditions in Smalltalk but in a few protected places, because it's like locking the code against future extensions.
2) because the pre-condition is not needed to solve MethodFinder problem
3) because i cannot find a good reason to put a precondition specifically in this method and not into any other
4) because i can find good reasons to not put preconditions uniformely in the Kernel: it spoils both efficiency and extensibility - see 1) (printing numbers in tight loops is not that rare).
5) because in Smalltalk violated preconditions are rarely fatal
6) because the limitation is trivial and any user with decent culture should find it quickly with the Debugger
7) because even a kid opening the Debugger would maybe have a great lesson initiated on arithmetic with proper guiding (agree, not by himself)
8) because Oscar preconditions are not enough, he didn't handle this (15 printStringBase: Float pi)... Even if he checks for an Integer, did you try (100000000000 printStringBase: 123456789). Tell me, where to stop preconditions?
Of course, on this specific example, maybe i'm just wrong and your opinion may diverge. A system made for learners should better have explicit rules. Feel free to add a precondition in the code, that's what MIT is for! or just add a comment about base expectations... I won't change a bit.
Cheers
Nicolas
nicolas cellier a écrit :
- because Oscar preconditions are not enough, he didn't handle this (15
printStringBase: Float pi)... Even if he checks for an Integer, did you try (100000000000 printStringBase: 123456789). Tell me, where to stop preconditions?
100000000000 printStringBase: 123456789)
Ah Ah, damnit, I found another infinite loop in my printStringBase: !
I am trapped by my own lessons!
Finally, i will change some bits, thank you Serge!
Nicolas
nicolas cellier a écrit :
100000000000 printStringBase: 123456789
Ah Ah, damnit, I found another infinite loop in my printStringBase: !
I am trapped by my own lessons!
Finally, i will change some bits, thank you Serge!
Nicolas
OK, i updated http://bugs.squeak.org/view.php?id=6887
Oscar Nierstrasz a écrit :
I found the source of the bug.
1 . 1 . 2
will cause the MethodFinder to try:
1 perform: #printStringBase: withArguments: #(1)
i.e., 1 printStringBase: 1
This generates an infinite loop within Integer>>printStringBase:
Integer>>printStringBase: should be corrected to complain if its argument (base) is less than 2.
I tried the obvious "self assert: base > 1.", but the Method Finder does not like this since it only catches Errors, not AssertionFailures, so the fix is as follows:
Integer>>printStringBase: base | stream integer next | (base < 2) ifTrue: [Error signal: 'base must be >= 2']. ... "rest as before"
Ok, thank you Oscar. To follow, on the bug trackers : - for Squeak : http://bugs.squeak.org/view.php?id=7107 - for Pharo : http://code.google.com/p/pharo/issues/detail?id=113
I did not find any error in the example 'abc' . 'ab' .
I have this error with the last image : sq3.10.2-7179dev08.07.1
-- Serge Stinckwich http://blog.doesnotunderstand.org/
On Thu, Jul 3, 2008 at 10:27 AM, Serge Stinckwich Serge.Stinckwich@gmail.com wrote:
Damien Cassou a écrit :
On Thu, Jul 3, 2008 at 9:53 AM, Serge Stinckwich Serge.Stinckwich@gmail.com wrote:
There is a bug with the Method finder tool. Try for example : 1. 1. 2. It runs forever ...
Try the same thing in any base Squeak image starting from Squeak 3.8. It seems you found a long-standing bug.
I usually show the method finder to Smalltalk beginners. It's the first time, i see this one ;-) There is another one. Try : 'abc'. 'ab' and choose the method found to launch an error.
I think this one is related to 3.10. I told them at least 5 times to include the existing fix, but nobody did it.
Damien Cassou <damien.cassou <at> gmail.com> writes:
I usually show the method finder to Smalltalk beginners. It's the first time, i see this one There is another one. Try : 'abc'. 'ab' and choose the method found to launch an error.
I think this one is related to 3.10. I told them at least 5 times to include the existing fix, but nobody did it.
I think Edgar was confused by too many different reports/fixes
http://bugs.squeak.org/view.php?id=5641 http://bugs.squeak.org/view.php?id=6402 http://bugs.squeak.org/view.php?id=6418 http://bugs.squeak.org/view.php?id=6691
Nicolas
El 7/3/08 6:30 AM, "Damien Cassou" damien.cassou@gmail.com escribió:
I think this one is related to 3.10. I told at least 5 times to include the existing fix, but nobody did it.
them ? whose them ? Remember me the Mantis bug report, please.
I wish put all relevant fixes and enhancements to 3.10, but Ralph goes and I was displaced.
Edgar
Damien Cassou a écrit :
On Thu, Jul 3, 2008 at 10:27 AM, Serge Stinckwich Serge.Stinckwich@gmail.com wrote:
Damien Cassou a écrit :
On Thu, Jul 3, 2008 at 9:53 AM, Serge Stinckwich Serge.Stinckwich@gmail.com wrote:
There is a bug with the Method finder tool. Try for example : 1. 1. 2. It runs forever ...
Try the same thing in any base Squeak image starting from Squeak 3.8. It seems you found a long-standing bug.
I usually show the method finder to Smalltalk beginners. It's the first time, i see this one ;-) There is another one. Try : 'abc'. 'ab' and choose the method found to launch an error.
I think this one is related to 3.10. I told them at least 5 times to include the existing fix, but nobody did it.
Ok and where is the fix, Damien ?
-- Serge Stinckwich http://blog.doesnotunderstand.org/
Serge Stinckwich a écrit :
Damien Cassou a écrit :
I think this one is related to 3.10. I told them at least 5 times to include the existing fix, but nobody did it.
Ok and where is the fix, Damien ?
--
Serge, are you black listing me?
http://bugs.squeak.org/view.php?id=5641 http://bugs.squeak.org/view.php?id=6402 http://bugs.squeak.org/view.php?id=6418 http://bugs.squeak.org/view.php?id=6691
My image is patched with my own: Installer mantis bug: 6691 fix: 'StringVersion-Patch-M6691.2.cs'.
But, maybe Keith did correct other things too... Since he pushed his own corrections for his own bug, so you should trust him and try: Installer ensureFix: 6402.
Nicolas
squeak-dev@lists.squeakfoundation.org
-
Damien Cassou -
Edgar J. De Cleene -
Eliot Miranda -
nicolas cellier -
Oscar Nierstrasz -
Serge Stinckwich