July 2015 - Pharo-dev - lists.pharo.org

Re: [Pharo-dev] About nextPutAll: $(

by Ben Coman

> Le 27/7/15 09:53, Peter UhnÃ¡k a Ã©crit : > > >> stream surround: aBlock between: '( ' and: ' )' > > > This may be personal preference, but I do not like having more method > arguments after a block argument... because if the block is longer it may > not be visible... +1 > e.g. > > stream > surround: [ :aStream | > self size > 100 > ifTrue: [ > aStream nextPutAll: 'size '. > self size printOn: aStream ] > ifFalse: [ > self keysSortedSafely > do: [ :key | > aStream > print: key; > nextPutAll: '->'; > print: (self at: key); > space ] ] ] > between: $( > and: $) > > vs between:and:surround: That could even be #between:and:putAll: On Sat, Jul 25, 2015 at 4:48 AM, stepharo <stepharo(a)free.fr> wrote: > HI > > while working on a lecture I saw that we have about 75 cases like: > > aStream nextPut: $(. > .... > aStream nextPut: $). > > ... we have surroundedBy: aString > > But only working on aString. Well it seems that #surroundedBy: only has one sender #surroundedBySingleQuotes which in turn seems to be not sent anywhere. So actually #surroundedBy: seems ripe for culling to be replaced by more generic... aStream between: '(' and: ')' putAll: [ ... nextLevelIn ... ]. Now for most of those 75 cases and other cases where this pattern might apply, is performance so critical that using $( rather than '(' gains a lot? Perhaps on ancient machine and before Cog it was important but now using $( now is premature optimisation? Otherwise we might also need #betweenChars:and:putAll: On Mon, Jul 27, 2015 at 3:59 PM, Denis Kudriashov <dionisiydk(a)gmail.com> wrote: > In most cases we always surround text by "()". Can we just provide method > for this and not duplicate "between: '( ' and: ' )'" everywhere? Something > like "stream>>printInBrackets:" Maybe it is also a useful pattern for markup languages: aStream between: '<HEAD>' and: '</HEAD>' putAll: [ ... nextLevelIn ... ]. On Sun, Jul 26, 2015 at 1:40 AM, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > (From my vacations) > > Doesn't this solves the problem? > > stream << $( << name << $) > > It is already there :) That is a reasonable point, but I think the having start and end tokens in visual close proximity has a slight advantage. On Mon, Jul 27, 2015 at 6:23 PM, stepharo <stepharo(a)free.fr> wrote: > May be I raised a not important issue. Of course its easier to have an opinion on simple things. cheers -ben

July 27, 2015

Re: [Pharo-dev] How to get better fonts?

by Peter UhnÃ¡k

I think you need to uncheck (disable) FreeType and then check (enable) it again for it to really load the fonts. And only then more fonts are available. (at least on linux) On Mon, Jul 27, 2015 at 5:37 PM, stepharo <stepharo(a)free.fr> wrote: > Hi > > with my new glasses (yes age) I know see all the details of characters. I > tried to set some freeType fonts using settings > and I enforce FreeType but I do not succeed to get nice fonts. > Am'I doing something wrong? > > Stef > >

July 27, 2015

Re: [Pharo-dev] about quality assistant

by Yuriy Tymchuk

> On 26 Jul 2015, at 14:46, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > >> I do not like useless. Because it is too violent and unclear. > > +1 on this. > > For example, in the class RTRoassalExample, there are many methods. Too much for the quality assistant. Finally, I did not press âuselessâ because I still think this is a useful rule, but not for that class. This is sad, and shows that I completely failed with that button. Because ideally you had to press it so I can know that this rule had a useful case for somebody. Uko > > Cheers, > Alexandre > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > >

July 27, 2015

Re: [Pharo-dev] about quality assistant

by Yuriy Tymchuk

> On 26 Jul 2015, at 10:55, stepharo <stepharo(a)free.fr> wrote: > > >>> I prefer >>> ban this rule = do not run it >>> false positive = rule is ok but wrong here >> >> Do you mean ban rule completely? On which level? For a developer (machine)? > > in the codecritcis the scope is clear. > On a class or a package. How is it different from false positive on a method or a class. Please do not misunderstand me, but this is a part of story telling. If you donât want to see a critic (of a class or a method), you mark it as false positive. Now you can also ban rule on the level of a class or package, so you will not see the critics about it. For me this two stories are intersecting. Why not simply say: âIf you donât want to see this critic anymore (for any reason), then you can ban it on the level of method, class, tag, package, or whole systemâ. This way the story is simpler. > >> Then again, skip works as former âfalse positiveâ. But in my opinion âfalse positiveâ is not a good name from the logical point of view. You have a rule that detects if you have no #yourself message in the end of a cascade. Now do people really mark this as âfalse positiveâ if the rule detects cascades that have #yourself message in the end? Or they do it if they donât want to see the critic? Then itâs not false positive, then they just want to hide or âskipâ the rule. > Sorry but I do not understand your argument. > For me this is simple: > the rule tells me something is wrong and it is not. > -> so I mark it as false positive. > >> This is fun part from statistics to name things as false positives. But if we program a rule in such way, that it is fuzzy, I think that itâs not correct to call it as false positive. Maybe itâs critics are useless more times that they are helpful, and if the critic is correct, the issue is not that important. Then probably we should just remove the rule or rewrite it. >> >> Now with useless button, the only reason is to collect data on which rules are actually useless to people [1], so we can take a look at them in a first place. It doest not affect the functionality. > > I do not like useless. Because it is too violent and unclear. > Please again have a look at the vocabulary used in the CodeCritics and the chapter that describes it. > Or you should fix everything but we should not have two different vocabularies. > There is enough confusing terms in Pharo so that we do not add more. I understand. I will re-read the vocabulary and fix discrepancies. But please read my argument bout duplicated behavior of banking and marking as false positive, I think that it may be confusing. Also âUseless" button does not do anything besides sending me an information that this critic was useless to somebody. Ideally there should be no button like this and I should be able to collect user actions that they perform. But from server/privacy/time perspective it was easier for me to make a simple button. Would be nice to have an advanced data collection, but I am not sure that I will have time now to program all that, so if it is bothering so much, I can simply remove it. Uko > > Stef > > >> >> Uko >> >> [1]: http://renraku.inf.usi.ch/rules <http://renraku.inf.usi.ch/rules> >> >> >>> >>> Stef >>> >>>> >>>> Uko >>>> >>>>> On 23 Jul 2015, at 22:29, stepharo <stepharo(a)free.fr <mailto:stepharo@free.fr>> wrote: >>>>> >>>>> HI yuriy >>>>> >>>>> it would be good that quality assistant uses the same vocabulary than codecritics. >>>>> for example >>>>> skip -> ban? >>>>> marks as false positif? >>>>> >>>>> Stef >>>>> >>>> >>>> >>> >>> >> >

July 27, 2015

Re: [Pharo-dev] Event handling insanity

by stepharo

> Hi igor > > I imagine that quite well. Terribly tedious. I was wondering what > would be a good architecture to handle it in the future: probably > the approach of guillermo > wtih one image poking another one > > Well, i had idea once, to implement a sandbox window.. > which works completely separate from main UI and all in it is err.. > sandboxed. :) > Everything comes in, nothing comes out.. > In that way, if something horrible happens in sandbox , it doesn't > affects the hosting environment.. Unfortunately it is easy to imagine > as concept, but quite hard to implement given the state of our current > system. > The main barrier for this is tons and tons of global state and > singletons everywhere. Yes. We will to address this one by one. First bloc will remove some of them. The bootstrap will also help. Stef > > Stef > > > Le 27/7/15 15:42, Igor Stasenko a Ã©crit : > > Nothing here is about things to fix, > but rather about how impossible to debug event handling in > system that runs and relies on very same events.. > I had very annoying problem with mouse move handling and > processing it correctly, and i finally solved the problem > after number of attempts to track down where problem lies in.. > > So, here is small list of problems i faced: > > - broken image (image won't start up) , because i send > unimplemented message in critical place, thus image simply > non-responsible/crashing immediately after startup. Easy to > fix (sarcasm) > > - debugger window spam on every event (because of DNU/other > exception). > A slightly friendlier problem than above :) > > - and last one, if you break some event handling/delivery, > even if image "behaves" you simply left without events doing > what is needed, means you got another form of a broken image, > which you cannot use anymore and need to start over and revert > the changes you made. > > - veeeerrrrryyy slow event handling, because of printing every > event into transcript , trying to see where the shit lies. > > - decrypting kilometers long event traces to see what goes where > > But when things finally start work properly, you feel excited :) > > > :) > > > -- > Best regards, > Igor Stasenko. > > > > > > > -- > Best regards, > Igor Stasenko.

July 27, 2015

How to get better fonts?

by stepharo

Hi with my new glasses (yes age) I know see all the details of characters. I tried to set some freeType fonts using settings and I enforce FreeType but I do not succeed to get nice fonts. Am'I doing something wrong? Stef

July 27, 2015

Re: [Pharo-dev] Event handling insanity

by Igor Stasenko

On 27 July 2015 at 16:31, stepharo <stepharo(a)free.fr> wrote: > Hi igor > > I imagine that quite well. Terribly tedious. I was wondering what would be > a good architecture to handle it in the future: probably the approach of > guillermo > wtih one image poking another one > > Well, i had idea once, to implement a sandbox window.. which works completely separate from main UI and all in it is err.. sandboxed. :) Everything comes in, nothing comes out.. In that way, if something horrible happens in sandbox , it doesn't affects the hosting environment.. Unfortunately it is easy to imagine as concept, but quite hard to implement given the state of our current system. The main barrier for this is tons and tons of global state and singletons everywhere. > Stef > > > Le 27/7/15 15:42, Igor Stasenko a Ã©crit : > > Nothing here is about things to fix, >> but rather about how impossible to debug event handling in system that >> runs and relies on very same events.. >> I had very annoying problem with mouse move handling and processing it >> correctly, and i finally solved the problem after number of attempts to >> track down where problem lies in.. >> >> So, here is small list of problems i faced: >> >> - broken image (image won't start up) , because i send unimplemented >> message in critical place, thus image simply non-responsible/crashing >> immediately after startup. Easy to fix (sarcasm) >> >> - debugger window spam on every event (because of DNU/other exception). >> A slightly friendlier problem than above :) >> >> - and last one, if you break some event handling/delivery, even if image >> "behaves" you simply left without events doing what is needed, means you >> got another form of a broken image, which you cannot use anymore and need >> to start over and revert the changes you made. >> >> - veeeerrrrryyy slow event handling, because of printing every event into >> transcript , trying to see where the shit lies. >> >> - decrypting kilometers long event traces to see what goes where >> >> But when things finally start work properly, you feel excited :) >> > > :) > > >> -- >> Best regards, >> Igor Stasenko. >> > > > -- Best regards, Igor Stasenko.

July 27, 2015

Re: [Pharo-dev] Event handling insanity

by stepharo

Hi igor I imagine that quite well. Terribly tedious. I was wondering what would be a good architecture to handle it in the future: probably the approach of guillermo wtih one image poking another one Stef Le 27/7/15 15:42, Igor Stasenko a Ã©crit : > Nothing here is about things to fix, > but rather about how impossible to debug event handling in system that > runs and relies on very same events.. > I had very annoying problem with mouse move handling and processing it > correctly, and i finally solved the problem after number of attempts > to track down where problem lies in.. > > So, here is small list of problems i faced: > > - broken image (image won't start up) , because i send unimplemented > message in critical place, thus image simply non-responsible/crashing > immediately after startup. Easy to fix (sarcasm) > > - debugger window spam on every event (because of DNU/other exception). > A slightly friendlier problem than above :) > > - and last one, if you break some event handling/delivery, even if > image "behaves" you simply left without events doing what is needed, > means you got another form of a broken image, which you cannot use > anymore and need to start over and revert the changes you made. > > - veeeerrrrryyy slow event handling, because of printing every event > into transcript , trying to see where the shit lies. > > - decrypting kilometers long event traces to see what goes where > > But when things finally start work properly, you feel excited :) :) > > -- > Best regards, > Igor Stasenko.

July 27, 2015

Event handling insanity

by Igor Stasenko

Nothing here is about things to fix, but rather about how impossible to debug event handling in system that runs and relies on very same events.. I had very annoying problem with mouse move handling and processing it correctly, and i finally solved the problem after number of attempts to track down where problem lies in.. So, here is small list of problems i faced: - broken image (image won't start up) , because i send unimplemented message in critical place, thus image simply non-responsible/crashing immediately after startup. Easy to fix (sarcasm) - debugger window spam on every event (because of DNU/other exception). A slightly friendlier problem than above :) - and last one, if you break some event handling/delivery, even if image "behaves" you simply left without events doing what is needed, means you got another form of a broken image, which you cannot use anymore and need to start over and revert the changes you made. - veeeerrrrryyy slow event handling, because of printing every event into transcript , trying to see where the shit lies. - decrypting kilometers long event traces to see what goes where But when things finally start work properly, you feel excited :) -- Best regards, Igor Stasenko.

July 27, 2015

Re: [Pharo-dev] About nextPutAll: $(

by Igor Stasenko

We all want to escape from routine, tedious and repeatable tasks.. But unfortunately there is some of them when it is going to stay no matter how you want to optimize, and implementing printing is one of good examples of such tasks. Adding 'convenience' methods won't help.. it will just balkanize new printing stuff from one that lies there for years. In thousands of projects.. because it is one of the central necessities, to see what your objects contain or to report something somewhere in readable form, and that involves *drumroll* printing :) On 27 July 2015 at 15:15, Igor Stasenko <siguctua(a)gmail.com> wrote: > > > On 27 July 2015 at 12:23, stepharo <stepharo(a)free.fr> wrote: > >> May be I raised a not important issue. >> >> My impression too. :) > > > >> Stef >> >> Le 27/7/15 09:53, Peter UhnÃ¡k a Ã©crit : >> >> >> stream surround: aBlock between: '( ' and: ' )' >>> >> >> This may be personal preference, but I do not like having more method >> arguments after a block argument... because if the block is longer it may >> not be visible... >> e.g. >> >> stream >> surround: [ :aStream | >> self size > 100 >> ifTrue: [ >> aStream nextPutAll: 'size '. >> self size printOn: aStream ] >> ifFalse: [ >> self keysSortedSafely >> do: [ :key | >> aStream >> print: key; >> nextPutAll: '->'; >> print: (self at: key); >> space ] ] ] >> between: $( >> and: $) >> >> vs between:and:surround: >> >> if you actually run this through formatter in Pharo, the between:and: >> is not even visible --- you have to scroll to see it. Of course if aBlock >> is a variable, than my argument falls. >> >> From visual perspective the current state actually isn't that bad >> because you immediately see where is beginning and where is end without >> even looking... with surround:between:and: not so much... >> >> Peter >> >> >> > > > -- > Best regards, > Igor Stasenko. > -- Best regards, Igor Stasenko.

July 27, 2015