Squeak-dev June 2012

squeak-dev@lists.squeakfoundation.org

41 participants
155 discussions

The Inbox: Kernel-bf.700.mcz
by commits＠source.squeak.org 28 Jun '12

28 Jun '12

Bert Freudenberg uploaded a new version of Kernel to project The Inbox: http://source.squeak.org/inbox/Kernel-bf.700.mcz ==================== Summary ==================== Name: Kernel-bf.700 Author: bf Time: 28 June 2012, 7:34:28.085 pm UUID: 019680d3-785f-43a5-bc2a-9fed7bade3bb Ancestors: Kernel-fbs.699 Make DateAndTime use the Julian Calendar for dates earlier than 15 October 1582. =============== Diff against Kernel-fbs.699 =============== Item was changed: Magnitude subclass: #DateAndTime instanceVariableNames: 'seconds offset jdn nanos' classVariableNames: 'ClockProvider DaysSinceEpoch LastMilliSeconds LastTick LastTickSemaphore LocalTimeZone MilliSecondOffset OffsetsAreValid' poolDictionaries: 'ChronologyConstants' category: 'Kernel-Chronology'! + !DateAndTime commentStamp: 'bf 6/28/2012 19:09' prior: 0! - !DateAndTime commentStamp: 'brp 5/13/2003 08:07' prior: 0! I represent a point in UTC time as defined by ISO 8601. I have zero duration. + In contrast to ISO 8601, which uses the Gregorian Proleptic Calendar exclusively, I use the Julian Proleptic Calendar for dates prior to 15 Oct 1582. However, for seriously dealing with multiple calendars, you should use a more thorough date package (e.g. Calendrica). + My implementation uses three SmallIntegers and a Duration: - My implementation uses three SmallIntegers - and a Duration: jdn - julian day number. seconds - number of seconds since midnight. nanos - the number of nanoseconds since the second. - offset - duration from UTC. The nanosecond attribute is almost always zero but it defined for full ISO compliance and is suitable for timestamping. ! Item was added: + ----- Method: DateAndTime class>>gregorianYear:month:day:hour:minute:second:nanoSecond:offset: (in category 'squeak protocol') ----- + gregorianYear: year month: month day: day hour: hour minute: minute second: second nanoSecond: nanoCount offset: offset + "Return a DateAndTime" + + | daysInMonth p q r s julianDayNumber | + daysInMonth := Month + daysInMonth: month + forYear: year. + day < 1 ifTrue: [self error: 'day may not be zero or negative']. + day > daysInMonth ifTrue: [self error: 'day is after month ends']. + + p := (month - 14) quo: 12. + q := year + 4800 + p. + r := month - 2 - (12 * p). + s := (year + 4900 + p) quo: 100. + + julianDayNumber := + ( (1461 * q) quo: 4 ) + + ( (367 * r) quo: 12 ) - + ( (3 * s) quo: 4 ) + + ( day - 32075 ). + + ^self basicNew + setJdn: julianDayNumber + seconds: hour * 60 + minute * 60 + second + nano: nanoCount + offset: offset; + yourself! Item was added: + ----- Method: DateAndTime class>>julianYear:month:day:hour:minute:second:nanoSecond:offset: (in category 'squeak protocol') ----- + julianYear: year month: month day: day hour: hour minute: minute second: second nanoSecond: nanoCount offset: offset + "Return a DateAndTime" + + | days | + days := + (365 * (year - 1)) + "years since epoch" + ((year - 1) // 4) + "leap days since epoch" + (((367 * month) - 362) // 12) + "30/31 day months since newyear" + day. + month > 2 ifTrue: [ "we counted this February as 30 days; correct to 29 or 28" + days := days - ((Year isJulianLeapYear: year) ifTrue: [1] ifFalse: [2])]. + ^self basicNew + setJdn: 1721423 + days + seconds: hour * 60 + minute * 60 + second + nano: nanoCount + offset: offset; + yourself! Item was changed: ----- Method: DateAndTime class>>year:month:day:hour:minute:second:nanoSecond:offset: (in category 'squeak protocol') ----- year: year month: month day: day hour: hour minute: minute second: second nanoSecond: nanoCount offset: offset + "Return a DateAndTime. If date is after 15 Oct 1582, use Gregorian calendar, otherwise the Julian calendar" - "Return a DateAndTime" + | monthIndex | - | monthIndex daysInMonth p q r s julianDayNumber | - monthIndex := month isInteger ifTrue: [month] ifFalse: [Month indexOfMonth: month]. + ^(year > 1582 or: [year = 1582 and: [monthIndex > 10 or: [monthIndex = 10 and: [day >= 15]]]]) + ifTrue: [self gregorianYear: year month: monthIndex day: day hour: hour minute: minute second: second nanoSecond: nanoCount offset: offset] + ifFalse: [self julianYear: year month: monthIndex day: day hour: hour minute: minute second: second nanoSecond: nanoCount offset: offset] + ! - daysInMonth := Month - daysInMonth: monthIndex - forYear: year. - day < 1 ifTrue: [self error: 'day may not be zero or negative']. - day > daysInMonth ifTrue: [self error: 'day is after month ends']. - - p := (monthIndex - 14) quo: 12. - q := year + 4800 + p. - r := monthIndex - 2 - (12 * p). - s := (year + 4900 + p) quo: 100. - - julianDayNumber := - ( (1461 * q) quo: 4 ) + - ( (367 * r) quo: 12 ) - - ( (3 * s) quo: 4 ) + - ( day - 32075 ). - - ^self basicNew - setJdn: julianDayNumber - seconds: hour * 60 + minute * 60 + second - nano: nanoCount - offset: offset; - yourself! Item was changed: ----- Method: DateAndTime>>dayMonthYearDo: (in category 'squeak protocol') ----- dayMonthYearDo: aBlock + "Evaluate the block with three arguments: day, month, year. + For dates since 15 Oct 1582 use Gregorian calendar, before that, the Julian calendar" - "Evaluation the block with three arguments: day month, year." + ^jdn < 2299161 + ifTrue: [self julianDayMonthYearDo: aBlock] + ifFalse: [self gregorianDayMonthYearDo: aBlock]! - | l n i j dd mm yyyy | - l := jdn + 68569. - n := 4 * l // 146097. - l := l - (146097 * n + 3 // 4). - i := 4000 * (l + 1) // 1461001. - l := l - (1461 * i // 4) + 31. - j := 80 * l // 2447. - dd := l - (2447 * j // 80). - l := j // 11. - mm := j + 2 - (12 * l). - yyyy := 100 * (n - 49) + i + l. - - ^ aBlock - value: dd - value: mm - value: yyyy! Item was added: + ----- Method: DateAndTime>>gregorianDayMonthYearDo: (in category 'squeak protocol') ----- + gregorianDayMonthYearDo: aBlock + "Evaluate the block with three arguments: day, month, year." + + | l n i j dd mm yyyy | + l := jdn + 68569. + n := 4 * l // 146097. + l := l - (146097 * n + 3 // 4). + i := 4000 * (l + 1) // 1461001. + l := l - (1461 * i // 4) + 31. + j := 80 * l // 2447. + dd := l - (2447 * j // 80). + l := j // 11. + mm := j + 2 - (12 * l). + yyyy := 100 * (n - 49) + i + l. + + ^ aBlock + value: dd + value: mm + value: yyyy! Item was added: + ----- Method: DateAndTime>>julianDayMonthYearDo: (in category 'squeak protocol') ----- + julianDayMonthYearDo: aBlock + "Evaluate the block with three arguments: day, month, year" + + | totalDays day month year daysSinceNewYear fakeDaysSinceNewYear daysToFirstOfMonth fakeDaysToFirstOfMonth fakeAmount | + totalDays := jdn - 1721424. "days since epoch 1 January 1" + year := (4 * totalDays + 1464) // 1461. + daysSinceNewYear := totalDays + - (365 * (year - 1)) "years since epoch" + - ((year - 1) // 4) . "leap days since epoch" + "after Feb 28, make Feb look like it was a 30-day month" + fakeAmount := daysSinceNewYear <= 58 + ifTrue: [0] + ifFalse: [(Year isJulianLeapYear: year) ifTrue: [1] ifFalse: [2]]. + fakeDaysSinceNewYear := daysSinceNewYear + fakeAmount. + "month formulas assume 30/31 day months" + month := ((12 * fakeDaysSinceNewYear) + 373) // 367. + fakeDaysToFirstOfMonth := ((367 * month) - 362) // 12. + daysToFirstOfMonth := month > 2 + ifTrue: [fakeDaysToFirstOfMonth - fakeAmount] + ifFalse: [fakeDaysToFirstOfMonth]. + day := daysSinceNewYear - daysToFirstOfMonth + 1. + ^ aBlock + value: day + value: month + value: year! Item was changed: ----- Method: Year class>>daysInYear: (in category 'smalltalk-80') ----- daysInYear: yearInteger + ^ 365 + (self leapYear: yearInteger) - ^ 365 + ((self isLeapYear: yearInteger) ifTrue: [1] ifFalse: [0]). ! Item was added: + ----- Method: Year class>>isGregorianLeapYear: (in category 'squeak protocol') ----- + isGregorianLeapYear: aYearInteger + | adjustedYear | + adjustedYear := aYearInteger > 0 + ifTrue: [aYearInteger] + ifFalse: [(aYearInteger + 1) negated]. + "There was no year 0" + ^ ((adjustedYear \\ 4 ~= 0) or: [(adjustedYear \\ 100 = 0) and: [adjustedYear \\ 400 ~= 0]]) not! Item was added: + ----- Method: Year class>>isJulianLeapYear: (in category 'squeak protocol') ----- + isJulianLeapYear: aYearInteger + ^ aYearInteger \\ 4 = 0! Item was changed: ----- Method: Year class>>isLeapYear: (in category 'squeak protocol') ----- isLeapYear: aYearInteger + ^ aYearInteger >= 1582 + ifTrue: [self isGregorianLeapYear: aYearInteger] + ifFalse: [self isJulianLeapYear: aYearInteger] + ! - - | adjustedYear | - adjustedYear := aYearInteger > 0 - ifTrue: [aYearInteger] - ifFalse: [(aYearInteger + 1) negated]. - - "There was no year 0" - ^ ((adjustedYear \\ 4 ~= 0) or: [(adjustedYear \\ 100 = 0) and: [adjustedYear \\ 400 ~= 0]]) not!

1 0

The Trunk: Kernel-eem.700.mcz
by commits＠source.squeak.org 28 Jun '12

28 Jun '12

Eliot Miranda uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-eem.700.mcz ==================== Summary ==================== Name: Kernel-eem.700 Author: eem Time: 28 June 2012, 10:30:56.246 am UUID: 18c57dcf-d75b-4c61-b5a3-67a9f8cfa4f6 Ancestors: Kernel-fbs.699 Back out of my bogus fix to runUntilErrorOrReturnFrom:. My "fix" breaks non-local return in the debugger. I can't find my original case that justified the new code (although it was that in the debugger too much stack was unwound), and so it needs to be backed-out and we need to find good test cases to fix this correctly. =============== Diff against Kernel-fbs.699 =============== Item was changed: ----- Method: ContextPart>>runUntilErrorOrReturnFrom: (in category 'controlling') ----- runUntilErrorOrReturnFrom: aSender "ASSUMES aSender is a sender of self. Execute self's stack until aSender returns or an unhandled exception is raised. Return a pair containing the new top context and a possibly nil exception. The exception is not nil if it was raised before aSender returned and it was not handled. The exception is returned rather than openning the debugger, giving the caller the choice of how to handle it." "Self is run by jumping directly to it (the active process abandons thisContext and executes self). However, before jumping to self we insert an ensure block under aSender that jumps back to thisContext when evaluated. We also insert an exception handler under aSender that jumps back to thisContext when an unhandled exception is raised. In either case, the inserted ensure and exception handler are removed once control jumps back to thisContext." + | error ctxt here topContext | - | error ctxt here topContext aSendersSender | here := thisContext. - aSendersSender := aSender sender. "Insert ensure and exception handler contexts under aSender" error := nil. ctxt := aSender insertSender: (ContextPart contextOn: UnhandledError do: [:ex | error ifNil: [ error := ex exception. topContext := thisContext. ex resumeUnchecked: here jump] ifNotNil: [ex pass] ]). ctxt := ctxt insertSender: (ContextPart contextEnsure: [error ifNil: [ topContext := thisContext. here jump] ]). self jump. "Control jumps to self" "Control resumes here once above ensure block or exception handler is executed" ^ error ifNil: [ "No error was raised, remove ensure context by stepping until popped" + [ctxt isDead] whileFalse: [topContext := topContext stepToCallee]. + {topContext. nil} - [ctxt isDead or: [aSender isDead]] whileFalse: [topContext := topContext stepToCallee]. - ^{aSender isDead - ifTrue: - [| retValue | - retValue := (ctxt method == (BlockClosure >> #ensure:) - or: [ctxt method == (BlockClosure >> #ifCurtailed:)]) ifTrue: - [ctxt tempAt: 3]. "returnValue in ensure: and result in ifCurtailed:" - aSendersSender push: retValue. - aSendersSender] - ifFalse: [topContext]. - nil} ] ifNotNil: [ "Error was raised, remove inserted above contexts then return signaler context" aSender terminateTo: ctxt sender. "remove above ensure and handler contexts" {topContext. error} ]. !

1 0

The Trunk: Network-eem.131.mcz
by commits＠source.squeak.org 28 Jun '12

28 Jun '12

Eliot Miranda uploaded a new version of Network to project The Trunk: http://source.squeak.org/trunk/Network-eem.131.mcz ==================== Summary ==================== Name: Network-eem.131 Author: eem Time: 28 June 2012, 8:58:07.094 am UUID: df89a3e5-a9c1-46d0-a2fa-39d72e105cda Ancestors: Network-dtl.130 Add some Croquet URI manipulation routines used by the Cog VMMaker to the base Network package. =============== Diff against Network-dtl.130 =============== Item was added: + ----- Method: DosFileDirectory>>uriPathToPlatformPath: (in category '*network-uri') ----- + uriPathToPlatformPath: aString + "Convert a URI path (w/ forward slashes) into a platform path if necessary. + Also make sure we deal properly with shares vs. drives" + | parts | + parts := aString findTokens: '/'. + (parts first endsWith: ':') ifTrue:[ + "it's a drive - compose c:\foo\bar" + ^aString allButFirst copyReplaceAll: '/' with: '\' + ] ifFalse:[ + "it's a share - compose \\share\foo\bar" + ^'\', (aString copyReplaceAll: '/' with: '\') + ].! Item was added: + ----- Method: FileDirectory>>pathFromURI: (in category '*network-uri') ----- + pathFromURI: aString + | path | + aString class == Array ifTrue:["bullet-proofing" + ^FileDirectory pathFrom: {self pathName}, aString]. + "To support paths like foo/bar/user#2.doc, use #allButScheme instead of #path" + path := (self uri resolveRelativeURI: aString) allButScheme. + ^self uriPathToPlatformPath: path + ! Item was added: + ----- Method: FileDirectory>>uriPathToPlatformPath: (in category '*network-uri') ----- + uriPathToPlatformPath: aString + "Convert a URI path (w/ forward slashes) into a platform path if necessary" + ^aString! Item was added: + ----- Method: HierarchicalURI>>allButScheme (in category 'printing') ----- + allButScheme + "Answer the entire url except its scheme" + + ^String streamContents:[:s| + authority ifNotNil:[self authority printOn: s]. + s nextPutAll: super allButScheme. + query ifNotNil:[s nextPutAll: query]. + ].! Item was added: + ----- Method: URI>>allButScheme (in category 'printing') ----- + allButScheme + "Answer the entire url except its scheme" + + ^String streamContents:[:s| + s nextPutAll: schemeSpecificPart. + fragment ifNotNil: [ + s nextPut: $# . + s nextPutAll: self fragment] + ].!

1 0

Re: Environment(s) serving web pages
by Chris Cunnington 28 Jun '12

28 Jun '12

I didn't mean to say WebClient was limited. The truth of the situation is that I built things around the examples you put into WebServer examples, which depend on #default and #reset. You do say in #default that it's mainly their to facilitate the examples. But as I haven't looked at my own code for several months, it was easier to jump to Kom. "Not sure what you mean. Do you mean you'd like to have one WebServer instance listening on multiple ports? What would be the use case for this? This would be simple to add, I just hadn't considered it." No, I can't think of an situation, where that would be useful. Processes listening on different ports, as you showed here: "wc8080 := WebServer new listenOn: 8080. wc8888 := WebServer new listenOn: 8888." is what I meant. And clearly WebClient can do that. I think I've been using WebClient in a certain way, where all roads lead to the class variable Default, and could not see it otherwise. Thank you for the example. Chris

1 0

Environment(s) serving web pages
by Chris Cunnington 28 Jun '12

28 Jun '12

http://osrcon.ca/enviros.png http://osrcon.ca/chrome.png I wanted to have the same class serve web content from three Environments. I loaded KomHttpServer and HttpView2 and did so. (I would have used GreenNeon, but WebClient restricts one port per image. It has a singleton, etc. KomHttpServer lets you have as many ports per image as you'd like.) Why would you do this? Perhaps to save resources like RAM. Instead of having a Unix process per user you have a Smalltalk process per user. Or users could share a common class side persistence. Or to have one VNC connection for several applications. Most likely, because you cannot resist trying it out. Chris

4 3

Julian dates
by Bert Freudenberg 27 Jun '12

27 Jun '12

Squeak dates are based on the Julian Day Number: http://en.wikipedia.org/wiki/Julian_day This is defined as the number of days since January 1, 4713 BC in the Julian proleptic calendar. Now, to my surprise, Date julianDayNumber: 0 answers 24 November -4713 instead of 1 January -4712 Digging into this it is because Squeak uses the Fliegel-Van Flandern algorithm, which always results in a Gregorian date. Even for dates before the Gregorian calendar was introduced. If we were to use the algorithm by Jean Meeus, dates before 15 Nov 1582 would be converted to a Julian date. Meeus is a floating point algorithm which we may not want to use. I added a correction to our method similar to Meeus, but converted to integers. So this is a mixed one, pasted below. Would anyone be opposed to adopt this? The behavior will be unchanged for all dates more recent than 15 Nov 1582. If so, we also would need an adjustment for the reverse part in year:month:day:hour:minute:second:nanoSecond:offset:. This is already covered by a test using the year 1200 BC (DateAndTimeTest>>testFromString). Anybody up for making that work? I stumbled upon this because Etoys 5 has a new Calendar object which exposes the Julian Day Number, and setting it to 0 is a very convenient way of explaining what that means. Thanks! - Bert - dayMonthYearDo: aBlock "Evaluate the block with three arguments: day month, year. Uses the Fliegel-Van Flandern algorithm, with adjustment for dates before 15 Nov 1582 as given by Jean Meeus. See http://mathforum.org/library/drmath/view/51907.html" | a l n i j dd mm yyyy | a := jdn. a < 2299161 ifTrue: [ | alpha | alpha := (jdn * 4 - 7468865) // 146097. a := jdn - 1 - alpha + (alpha // 4)]. l := a + 68569. n := 4 * l // 146097. l := l - (146097 * n + 3 // 4). i := 4000 * (l + 1) // 1461001. l := l - (1461 * i // 4) + 31. j := 80 * l // 2447. dd := l - (2447 * j // 80). l := j // 11. mm := j + 2 - (12 * l). yyyy := 100 * (n - 49) + i + l. ^ aBlock value: dd value: mm value: yyyy

4 5

Environment question
by Chris Cunnington 27 Jun '12

27 Jun '12

>>"Browser fullOnClass: (self at: #HelloEnv)" Nifty! Chris

1 0

Environment question
by Chris Cunnington 27 Jun '12

27 Jun '12

http://osrcon.ca/helloenv.png I created a HelloEnv class and filed out. It has one method #sayHi which writes to the Transcript. I loaded it into an Environment and now I can reference it in a Workspace and have it print its message to the Transcript. How would I change my #sayHi method? I'm guessing this is the raw implementation and some tooling would needed, so you could change code in different environments. Chris

2 1

Environment question
by Chris Cunnington 27 Jun '12

27 Jun '12

I should read posts more closely. I think this was already answered in the affirmative. Chris

1 0

German Squeakers meeting July, 7th in Potsdam
by Rita Freudenberg 27 Jun '12

27 Jun '12

Dear Squeakers, here comes the invitation for the annual meeting of Squeak Germany e.V. on July, 7th at HPI in Potsdam. http://www.squeak.de/events-squeaktreffen-2012/squeaktreffen-2012 Greetings, Rita ------------------------------------------------------------------------------------------------------------------------------ Liebe Squeaker, auch dieses Jahr laden wir Euch wieder herzliche zum Squeak-Treffen mit Vorträgen und zur Mitgliederversammlung ein Wo? Dank Robert Hirschfeld und seinem Team wieder am Hasso Plattner Institut bei der Software-Architektur Gruppe in Potsdam. Den Raum geben wir noch bekannt. Anfahrt: http://www.hpi.uni-potsdam.de/hpi/campus/anfahrt.html S a m s t a g, 07.07.2012 Vorträge (Beginn 13:00 Uhr) Themen sind gern gesehen bzw. in Planung. Wenn Ihr Ideen habt, bitte melden! Bitte meldet Euch rechtzeitig! Vereinsversammlung (17:00 Uhr - 18:00 Uhr) - Bestimmung des Protokollführers und der Kassenprüfer - Feststellung der Beschlussfähigkeit - Vorstellung / Änderung der Tagesordnung - Aktivitäten des Vereins 2011-2012 - Bericht des Schatzmeisters und der Kassenprüfer - Künftige Aktivitäten - Sonstiges Wer noch Ideen für Tagesordnungspunkte hat, bitte satzungsgemäß innerhalb von zwei Wochen melden. Anschließend Ausklang in gemütlicher Runde in Potsdam. Bitte antwortet an uns oder an die Liste, wenn Ihr kommen könnt, so dass wir ein wenig planen können, vor allem wegen des abendlichen Treffens! Wir hoffen Euch wieder zahlreich in Potsdam treffen zu können. Viele Grüße Enno, Esther und Rita Rita Freudenberg rita.freudenberg(a)ovgu.de

1 0

Jump to page: