December 2016 - Pharo-users

Re: [Pharo-users] [Pharo-dev] consider putting stars on our nice books on github :)

by Ben Coman

Good idea. Done. Now just because I couldn't remember explicitly what "starring" meant, from GitHub help... Starring a repository allows you to keep track of projects that you find interesting, even if you aren't associated with the project. When you star a repository, you're actually performing two distinct actions: * Creating a bookmark for easier access * Showing appreciation to the repository maintainer for their work Many of GitHub's repository rankings depend on the number of stars a repository has. For example,repositories can be sorted and searched based on their star count. In addition, the Explore pageshows you popular repositories based on the number of stars they have. =Viewing your starred repositories= You can see all the repositories that you have starred by going to your stars page. For more information on interacting with your starred repositories, see "Managing your stars." =Notifications= Notifications are not affected when a repository is starred. You won't receive any information about a repository you've starred, unless you're also watching the repository. Activity from starred repositories don't show up in your dashboard feed, so you don't have to worry about missing important notifications for things you are working on.

Dec. 9, 2016

Re: [Pharo-users] Hash collision

by Ben Coman

On Fri, Dec 9, 2016 at 10:51 PM, Hilaire <hilaire(a)drgeo.eu> wrote: > Hi, > > In Dr. Geo I rely a lot on hash value for the geometric object. It > largely improves speed up to compare geometric items (to avoid > duplicated items). For example to detect when the user is building an > already created object. > > Also when describing geometric sketch with Smalltalk code, hundred or > thousand of geometric items are created. The hash value speeds up > equality check (otherwise true comparing is iterative in the geometric > sketch tree and it slow down) > > But sometime hash collisions occur. I guess in that case, DrGeo should > do the real equality check, or may be I am overusing the hash? I am > interested on advices. As for now I am doing nothing special. > > Here is an annoying hash collision with Pharo 3/ Pharo 5 I think should > not occur: > > (0.5(a)0.25) hash > 67006464 > (-0.5(a)0.25) hash > 67006464 > > > Squeak5.0 does not have this collision. > > It looks related to the Float hash implementation a bit different. > > What do you think? > More the point, this seems bad... -0.5 hash "==> 57344" 0.5 hash "==> 57344" I read page 307 of "Hashing in Smalltalk Theory and Practice" ** regarding Squeak's (Pharo's) Point>>hash method... "at this time we will not test this hash function for points with our floating point data sets because we already know that [Pharo's] hash function for floats is of bad quality." ** Errm... actually first time I've opened the book since I bought it a couple of years ago. And from page 273 reviewing Float's hash... "because of how the bitAnd: is used, the hash function will discard 32 of the original 64 bits. In other words what seems a hash function using all the availabel bytes is anything but. [Also] because of the way the two 16 bit chunks are added togather, by the time bitshift:-8 is sent we will only have, at most, 17 bits worth of hash value. Effectively all the bitshift: accomplishes is to get rid of the 8 lowest bits, which are zero anyway because of how the bitAnd: message were set up. Can these objections add up to particularly poor performance? The answer is definitely yes - since we will have only 17 bit hash values this hash function will be particularly unsuitable for any data set containing more than 2^17 = 131,072 objects" Float>>hash "Pharo" (self isFinite and: [self fractionPart = 0.0]) ifTrue: [^self truncated hash]. ^ (((self basicAt: 1) bitAnd: 16r00FFFF00) + ((self basicAt: 2) bitAnd: 16r00FFFF00)) bitShift: -8 The suggested fix depended on being willing to let go of byte ordering (which we might not want different results on different platforms) Float>>hash ^ByteArray hashBytes: self startingWith: self species hash but anyway that doesn't actually help this case... -0.5 hash "==>120484352" 0.5 hash "==>120484352" I see Squeak5's float hash has changed... Float>>hash "Squeak5" (self isFinite and: [self fractionPart = 0.0]) ifTrue: [^self truncated hash]. ^ ((self basicAt: 1) bitShift: -4) + ((self basicAt: 2) bitShift: -4) which gives... 0.5 hash "==>66977792" -0.5 hash "==>201195520" which solves the immediate problem, but I'm not capable of analysis its quality further. cheers -ben

Dec. 9, 2016

setter method

by Kateryna Aloshkina

Hi everybody! I've just joined the community:) and have a question How to write the method count: such that, when invoked on an instance of Counter, instance variable is set to the argument given to the message? *so thatCounter new count: 7* *would set the value of a new Counter instance to 7* (It is a little exercise from http://rmod-pharo-mooc.lille.inria.fr/MOOC/Exercises/Exo-Counter.pdf) Best wishes, Kateryna

Dec. 9, 2016

Re: [Pharo-users] [Pharo-dev] consider putting stars on our nice books on github :)

by Sven Van Caekenberghe

Done. > On 9 Dec 2016, at 16:11, stepharong <stepharong(a)free.fr> wrote: > > https://github.com/SquareBracketAssociates/EnterprisePharo > > https://github.com/SquareBracketAssociates/NumericalMethods > > https://github.com/SquareBracketAssociates/UpdatedPharoByExample > > https://github.com/SquareBracketAssociates/BuildingUIWithSpec > > https://github.com/SquareBracketAssociates/TinyBlogTutorial > > https://github.com/SquareBracketAssociates/ExploringVMs > > https://github.com/SquareBracketAssociates/DeepIntoPharo > > Stef > -- > Using Opera's mail client: http://www.opera.com/mail/ >

Dec. 9, 2016

Re: [Pharo-users] consider putting stars on our nice books on github :)

by Offray Vladimir Luna CÃ¡rdenas

Done! and added to my Zotero Pharo/Smalltalk collection: https://www.zotero.org/groups/diseo_y_creacion_phd_msc_universidad_de_calda… Cheers, Offray On 09/12/16 10:11, stepharong wrote: > https://github.com/SquareBracketAssociates/EnterprisePharo > > https://github.com/SquareBracketAssociates/NumericalMethods > > https://github.com/SquareBracketAssociates/UpdatedPharoByExample > > https://github.com/SquareBracketAssociates/BuildingUIWithSpec > > https://github.com/SquareBracketAssociates/TinyBlogTutorial > > https://github.com/SquareBracketAssociates/ExploringVMs > > https://github.com/SquareBracketAssociates/DeepIntoPharo > > Stef

Dec. 9, 2016

Re: [Pharo-users] seaside javascript builders

by Siemen Baader

Thanks, I was unaware of the seaside list! cheers, Siemen On Fri, Dec 9, 2016 at 5:18 PM, Ben Coman <btc(a)openinworld.com> wrote: > > > On Fri, Dec 9, 2016 at 9:20 PM, Siemen Baader <siemenbaader(a)gmail.com> > wrote: > >> Hi all, >> >> I'm confused about the limitations and intended use of Seaside's >> javascript >> builders. I can follow that we can do eg: "(html jQuery: 'div') remove" or >> "html javascript alert: 'hello'" within the "renderContentOn: html" >> methods, >> but how do I deal with logic and variables? How do I eg. install an event >> listener and provide a browser-side function that extracts data from the >> event and then passes it to the server via "html jQuery ajax"? >> >> Is this something to be done in pure JS returned from the #script method? >> >> Thanks for any enlightenment! >> >> Siemen >> >> > Hi Siemen, > > Sorry I can't help, I'm not a seaside user - but depending on your > response here you might also like to try > http://www.seaside.st/community/mailinglist > > cheers -ben >

Dec. 9, 2016

Re: [Pharo-users] seaside javascript builders

by Ben Coman

On Fri, Dec 9, 2016 at 9:20 PM, Siemen Baader <siemenbaader(a)gmail.com> wrote: > Hi all, > > I'm confused about the limitations and intended use of Seaside's javascript > builders. I can follow that we can do eg: "(html jQuery: 'div') remove" or > "html javascript alert: 'hello'" within the "renderContentOn: html" > methods, > but how do I deal with logic and variables? How do I eg. install an event > listener and provide a browser-side function that extracts data from the > event and then passes it to the server via "html jQuery ajax"? > > Is this something to be done in pure JS returned from the #script method? > > Thanks for any enlightenment! > > Siemen > > Hi Siemen, Sorry I can't help, I'm not a seaside user - but depending on your response here you might also like to try http://www.seaside.st/community/mailinglist cheers -ben

Dec. 9, 2016

Re: [Pharo-users] Contributing code

by stepharong

Hi evan let us know if you need more guidance. It is really nice that you want to help improving Pharo. Tx a lot. > Hello, > > I wanted to ask if there was a good resource I could use to learn how to > find out about current bugs and submit patches/code/contributions to the > core Pharo classes (or other major Pharo projects). >I'd like to be able > to fix the things I encounter, I've just never gotten around to learning > how. > Ideally, it would be great if something could show me how to find and > fix some minor tweak of a comment or something like that, just so I know > the process. This seems like the kind of thing that is >either obvious, > or has been addressed somewhere, but for whatever reason, I've never > been able to find it. > > Thank you, > Evan -- Using Opera's mail client: http://www.opera.com/mail/

Dec. 9, 2016

consider putting stars on our nice books on github :)

by stepharong

https://github.com/SquareBracketAssociates/EnterprisePharo https://github.com/SquareBracketAssociates/NumericalMethods https://github.com/SquareBracketAssociates/UpdatedPharoByExample https://github.com/SquareBracketAssociates/BuildingUIWithSpec https://github.com/SquareBracketAssociates/TinyBlogTutorial https://github.com/SquareBracketAssociates/ExploringVMs https://github.com/SquareBracketAssociates/DeepIntoPharo Stef -- Using Opera's mail client: http://www.opera.com/mail/

Dec. 9, 2016

Hash collision

by Hilaire

Hi, In Dr. Geo I rely a lot on hash value for the geometric object. It largely improves speed up to compare geometric items (to avoid duplicated items). For example to detect when the user is building an already created object. Also when describing geometric sketch with Smalltalk code, hundred or thousand of geometric items are created. The hash value speeds up equality check (otherwise true comparing is iterative in the geometric sketch tree and it slow down) But sometime hash collisions occur. I guess in that case, DrGeo should do the real equality check, or may be I am overusing the hash? I am interested on advices. As for now I am doing nothing special. Here is an annoying hash collision with Pharo 3/ Pharo 5 I think should not occur: (0.5(a)0.25) hash 67006464 (-0.5(a)0.25) hash 67006464 Squeak5.0 does not have this collision. It looks related to the Float hash implementation a bit different. What do you think? Hilaire -- Dr. Geo http://drgeo.eu

Dec. 9, 2016