November 2010 - Pharo-users

Association

by Benjamin

Hi guys, I have another noob question :) Why Association>>#= is not = anAssociation ^self == anAssociation or: [super = anAssociation and: [value = anAssociation value]] Because I can have strange behavior like association := #foo -> Float nan. {association} includes: association answer false :S Thanks, Ben

Nov. 15, 2010

permutations / combinations implementation

by Tudor Girba

Hi, Is there an implementation of permutations and combinations of a collection? Cheers, Doru -- www.tudorgirba.com "Presenting is storytelling."

Nov. 15, 2010

Re: [Pharo-users] Check a text - is it a date?

by Norbert Hartl

On 15.11.2010, at 07:26, Jeff Gray wrote: > > How do I check a text for holding a date value? > isNumber is handy for numbers. Is there anything like that? > No, there isn't as far as I know. Checking isNumber is rather easy because you can decide it by looking at a single character. A format like date you have to parse. > The scenario I'm looking at is making sure a value in a Seaside control has > been entered correctly. You can try to parse the date | date | [ date := DateAndTime fromString: aMightBeADate ] on: Error do: [ :err | ]. After that your date test is isNil. Either you get the parsed date or it is nil if it couldn't parse a date. Norbert

Nov. 15, 2010

Check a text - is it a date?

by Jeff Gray

How do I check a text for holding a date value? isNumber is handy for numbers. Is there anything like that? The scenario I'm looking at is making sure a value in a Seaside control has been entered correctly. -- View this message in context: http://forum.world.st/Check-a-text-is-it-a-date-tp3042556p3042556.html Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

Nov. 15, 2010

Re: [Pharo-users] [ANN] Smalltalk Labs Browser for blogs

by Christoph Budzinski

Hi James, glad you like SLBfb :) I hope it gets popular. It would be so funny if yet again other languages would copy from Smalltalk and you'd begin to see code browsers popping up on websites targeting the users of other languages ^^ It might even make it easier for people to switch to Smalltalk or Newspeak, since then they've already gotten used to the idea of a code browser instead of a flat text file for source code. But that's wishful thinking of course, not going to happen. Feel free to send me any suggestions for improvement or new features you might have. There are already a lot of things planned that I want to do with SLBfb and areas were I plan to look for a better solution, but it's always good to get more input from other people :) Cheers, Chris On Sun, Nov 14, 2010 at 10:33 PM, James Ashley <james.ashley(a)gmail.com>wrote: > > > Licensing issues aside (I don't really care much, though I understand > the concerns on both sides), this looks extremely slick and cool. >

Nov. 15, 2010

Re: [Pharo-users] [ANN] Smalltalk Labs Browser for blogs

by Christoph Budzinski

Hi StÃ©phane, The names are very much alike (Smalltalk Labs Browser / Smalltalk Labs Browser for blogs) so it's a bit confusing I guess, sorry about that :) The Apache License is actually quite understandable and not *that* long, so I would still call it simple. That's why I can't agree with you here and still think that the Apache License is a good choice. Sure, it's longer, but far from being as big a mess as the GPL for example. Just for the sake of having a shorter license taking a bigger risk just doesn't sound like a good idea to me. But you're completely fine to think about this however you want of course, I won't try to convice you of my point of view and I'm thankful for your input on this and will keep it in mind if other people should ask about a license change too. I already wrote why I believe that the MIT license might not be a valid license in Germany. That doesn't seem to really be a problem though, I don't want to spread fear or anything, the MIT license has been used in tons of projects all over the world including Germany and I haven't heard about tons of lawsuits yet (I don't know of even just one). Like I said, I used MIT myself before and continue to do so, Smalltalk Labs Browser (not for blogs) is completely licensed under the MIT license and will stay that way. I've even used self written licenses much shorter and of a much lesser quality than the MIT license in closed source software projects with hundred thousands or in one case millions of users and have not been sued once. Choosing the Apache License is my personal choice because it makes me sleep better, I'm not trying to convince anyone to use it too. I had actually planned to release SLBfb under the MIT license but since I used Apache licensed code from Google for a part of the server I had to read up on the Apache license and what I read made me decide to use the Apache License for SLBfb instead. Cheers, Chris On Sun, Nov 14, 2010 at 11:09 PM, StÃ©phane Ducasse < stephane.ducasse(a)inria.fr> wrote: > I confused it with the other browser. :) > So indeed the license for the javascript code does not impact us. >

Nov. 15, 2010

Re: [Pharo-users] [ANN] Smalltalk Labs Browser for blogs

by StÃ©phane Ducasse

I confused it with the other browser. :) So indeed the license for the javascript code does not impact us. Stef On Nov 14, 2010, at 3:13 PM, HaiColon wrote: > > Hi StÃ©phane, > > If it should really be a problem for people I can certainly change the > license, at least for the client since the server uses Apache licensed code > from Google already and then it would have two licenses which would make > things unnecessarily complicated. The server is a rather small part though, > I wrote the first version in about 5 minutes and the version that is in use > now in about fifteen (not including the website, I spent way more time > writing text than coding). > > Pharo / Squeak aren't completely MIT though, they use the same Apache > License as I do for SLBfb so I didn't think it would be a problem. Is it > planned to completely replace the Apache code from Pharo with MIT code over > time? > > Here's why I chose the Apache License (without giving any legal advice or > guaranteeing that my understanding of the terms are correct, mind you!): > > It's more popular than the MIT license by a few orders of magnitude. I > didn't want to believe this myself since I know way more MIT licensed > software than Apache Software, but to be fair, who reads licenses unless he > really has to. 25% of projects on Google Code Project hosting were Apache in > 2008, only about 8% MIT according to Google. > > It's pretty much the same as the MIT license, safe for a few convenient > additional clauses, like saying that if a contributor contributes code to > the project it is assumed that his contribution is under the Apache License > unless he explicitly states otherwise. This can greatly help to prevent > misunderstandings. The really big difference is that it's actually written > with modern copyright law in mind, which is one of the reasons it's so much > longer than the MIT license. The MIT license might very well not be a valid > license. For example, in Germany where I come from it is not allowed to > disclaim liability, but the MIT license does that. Limiting the liability is > allowed, the Apache license does that. Since this is also a website/server > and not only a "local", "client side" application, the license being > actually applicable to keep me from being sued becomes much more important. > The chances are very slim that someone sues a tiny project like this but why > make it easier for anyone, there are tons of law firms who's whole business > model is suing people over unlawful license terms on websites (it's > outrageous that doing this is actually legal). > > Here's an interesting article from Google on open source licenses: > http://google-opensource.blogspot.com/2008/05/standing-against-license-prol… > and an interview: > http://www.zdnet.com/blog/burnette/google-says-no-to-license-proliferation/… > > So to sum it up, I'm certainly open for discussion on this but I'd rather > not change the license unless it's really necessary (e.g. No one wants to > use it otherwise). It's not that I don't like the MIT license, it's my > favorite license, it's so short and easy to understand, I am using it for > Smalltalk Labs Browser (without the "for blogs" ^^) and used it many times > before, but it really does look like choosing the Apache License for future > projects is the right thing to do. What good is a license if it's not > applicable anyhow? > > Cheers, > Chris > > > StÃ©phane Ducasse wrote: >> >> would you mind to put the license MIT so that it is the same with Squeak >> and Pharo? >> >> Stef >> > > -- > View this message in context: http://forum.world.st/ANN-Smalltalk-Labs-Browser-for-blogs-tp3041451p304174… > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. > > _______________________________________________ > Pharo-users mailing list > Pharo-users(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-users

Nov. 14, 2010

Re: [Pharo-users] [ANN] Smalltalk Labs Browser for blogs

by James Ashley

On Sat, Nov 13, 2010 at 8:09 PM, HaiColon <haicolon(a)gmail.com> wrote: > > Hi, > > I released the first public version of "Smalltalk Labs Browser for blogs" > (or just SLBfb since that name is really long.. ^^), a JavaScript widget > that you can put on your blog (or any other kind of website) to offer your > readers a web based Smalltalk code browser for classes you mention or even > describe in an article or tutorial so that people don't need to have a > Smalltalk image at hand to understand what you're talking about. The widget > pops up when someone clicks on a classname and can be moved freely around a > website, but it tries to position itself automatically in the best position > when you don't move it by yourself and keeps doing so even when the web > browser is resized. Licensing issues aside (I don't really care much, though I understand the concerns on both sides), this looks extremely slick and cool. Thank you, James

Nov. 14, 2010

Re: [Pharo-users] [ANN] Smalltalk Labs Browser for blogs

by StÃ©phane Ducasse

> Hi StÃ©phane, > > If it should really be a problem for people I can certainly change the > license, at least for the client since the server uses Apache licensed code > from Google already Your code is not all in Smalltalk? > and then it would have two licenses which would make > things unnecessarily complicated. The server is a rather small part though, > I wrote the first version in about 5 minutes and the version that is in use > now in about fifteen (not including the website, I spent way more time > writing text than coding). > > Pharo / Squeak aren't completely MIT though, they use the same Apache > License as I do for SLBfb so I didn't think it would be a problem. multiply license is always a problem. > Is it planned to completely replace the Apache code from Pharo with MIT code over > time? If we could we would. Right now only the VM code is APSL (if I'm correct) and all the rest is MIT > Here's why I chose the Apache License (without giving any legal advice or > guaranteeing that my understanding of the terms are correct, mind you!): > > It's more popular than the MIT license by a few orders of magnitude. I > didn't want to believe this myself since I know way more MIT licensed > software than Apache Software, but to be fair, who reads licenses unless he > really has to. 25% of projects on Google Code Project hosting were Apache in > 2008, only about 8% MIT according to Google. this is not an argument that is really interesting for us. The world can do whether it wants. > It's pretty much the same as the MIT license, safe for a few convenient > additional clauses, like saying that if a contributor contributes code to > the project it is assumed that his contribution is under the Apache License > unless he explicitly states otherwise. This can greatly help to prevent > misunderstandings. The really big difference is that it's actually written > with modern copyright law in mind, Probably but it probably implies that mere mortals cannot read it too. > which is one of the reasons it's so much > longer than the MIT license. The MIT license might very well not be a valid > license. I would be curious why MIT would not be a valid license. Lot of projects would be strange. MIT is a simple and understandable license. > For example, in Germany where I come from it is not allowed to > disclaim liability, but the MIT license does that. Limiting the liability is > allowed, the Apache license does that. Since this is also a website/server > and not only a "local", "client side" application, the license being > actually applicable to keep me from being sued becomes much more important. > The chances are very slim that someone sues a tiny project like this but why > make it easier for anyone, there are tons of law firms who's whole business > model is suing people over unlawful license terms on websites (it's > outrageous that doing this is actually legal). > > Here's an interesting article from Google on open source licenses: > http://google-opensource.blogspot.com/2008/05/standing-against-license-prol… > and an interview: > http://www.zdnet.com/blog/burnette/google-says-no-to-license-proliferation/… > > So to sum it up, I'm certainly open for discussion on this but I'd rather > not change the license unless it's really necessary (e.g. No one wants to > use it otherwise). It's not that I don't like the MIT license, it's my > favorite license, it's so short and easy to understand, I am using it for > Smalltalk Labs Browser (without the "for blogs" ^^) and used it many times > before, but it really does look like choosing the Apache License for future > projects is the right thing to do. This is not the right things to do. Tracking license of code is a pain and a lost of time I prefer to focus on building exciting software than fighting with license issues. Now you can put your browser on APSL but my advice is don't do it. Keep it simple. > What good is a license if it's not > applicable anyhow? > > Cheers, > Chris > > > StÃ©phane Ducasse wrote: >> >> would you mind to put the license MIT so that it is the same with Squeak >> and Pharo? >> >> Stef >> > > -- > View this message in context: http://forum.world.st/ANN-Smalltalk-Labs-Browser-for-blogs-tp3041451p304174… > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. > > _______________________________________________ > Pharo-users mailing list > Pharo-users(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-users

Nov. 14, 2010

Re: [Pharo-users] [ANN] Smalltalk Labs Browser for blogs

by HaiColon

Hi StÃ©phane, If it should really be a problem for people I can certainly change the license, at least for the client since the server uses Apache licensed code from Google already and then it would have two licenses which would make things unnecessarily complicated. The server is a rather small part though, I wrote the first version in about 5 minutes and the version that is in use now in about fifteen (not including the website, I spent way more time writing text than coding). Pharo / Squeak aren't completely MIT though, they use the same Apache License as I do for SLBfb so I didn't think it would be a problem. Is it planned to completely replace the Apache code from Pharo with MIT code over time? Here's why I chose the Apache License (without giving any legal advice or guaranteeing that my understanding of the terms are correct, mind you!): It's more popular than the MIT license by a few orders of magnitude. I didn't want to believe this myself since I know way more MIT licensed software than Apache Software, but to be fair, who reads licenses unless he really has to. 25% of projects on Google Code Project hosting were Apache in 2008, only about 8% MIT according to Google. It's pretty much the same as the MIT license, safe for a few convenient additional clauses, like saying that if a contributor contributes code to the project it is assumed that his contribution is under the Apache License unless he explicitly states otherwise. This can greatly help to prevent misunderstandings. The really big difference is that it's actually written with modern copyright law in mind, which is one of the reasons it's so much longer than the MIT license. The MIT license might very well not be a valid license. For example, in Germany where I come from it is not allowed to disclaim liability, but the MIT license does that. Limiting the liability is allowed, the Apache license does that. Since this is also a website/server and not only a "local", "client side" application, the license being actually applicable to keep me from being sued becomes much more important. The chances are very slim that someone sues a tiny project like this but why make it easier for anyone, there are tons of law firms who's whole business model is suing people over unlawful license terms on websites (it's outrageous that doing this is actually legal). Here's an interesting article from Google on open source licenses: http://google-opensource.blogspot.com/2008/05/standing-against-license-prol… and an interview: http://www.zdnet.com/blog/burnette/google-says-no-to-license-proliferation/… So to sum it up, I'm certainly open for discussion on this but I'd rather not change the license unless it's really necessary (e.g. No one wants to use it otherwise). It's not that I don't like the MIT license, it's my favorite license, it's so short and easy to understand, I am using it for Smalltalk Labs Browser (without the "for blogs" ^^) and used it many times before, but it really does look like choosing the Apache License for future projects is the right thing to do. What good is a license if it's not applicable anyhow? Cheers, Chris StÃ©phane Ducasse wrote: > > would you mind to put the license MIT so that it is the same with Squeak > and Pharo? > > Stef > -- View this message in context: http://forum.world.st/ANN-Smalltalk-Labs-Browser-for-blogs-tp3041451p304174… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

Nov. 14, 2010