- Pharo-dev - lists.pharo.org

Re: [Pharo-dev] spotter usage - a rough analysis of categories

by Esteban Lorenzano

> On 10 Jun 2016, at 10:11, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > >> On 10 Jun 2016, at 09:47, Guille Polito <guillermopolito(a)gmail.com> wrote: >> >> I use spotter a lot as well. > > Me too ! > >> It is actually my default entry point to the system. It superseeded the searches in Nautilus, and most of the searches that I used to start in the workspace. > > Indeed, same here. > >> The things I look most by name are tools (Menu entries), classes and packages. >> >> History is of great help and I use it a lot. But it poses a problem sometimes because it alters the order of the results, and you have to wait for the complete search to finish. If you do not do that, it happens what Max explained: you type, hit <down>,<enter> and suddenly the UI refreshed and you finished doing something else. >> >> Also, regarding the ordering of results, sometimes It bothers me that they are sorted by category and not by relevance of the item. I'll explain myself with a concrete case: >> - open spotter >> - type "Process" >> >> you get first two #menu items and then the #classes, containing the class Process. I would have expected here to show first the element that matched the most or with more relevance, regardless it's category. >> >> Question: Does spotter adapt to it's usage? Besides history, does it learn what are the items with more relevance from previous usages in the same session? > > Some simple learning would be super cool. problem for learning is that it consumes spaceâ¦ old algernon was cool on learning, but it was taking always like 10m memory extra. probably not something that you would careâ¦ and maybe some simple learning would not take as much, but well, I wanted to point it :) cheers, Esteban > >> Also, from the top of my head there is at least one use case that I cannot yet replace with spotter, and I have to use a normal workspace/playground to do it => browse senders or implementors. While i use senders/implementors search in spotter, sometimes I want to *browse* them to be able to look at them, and interact with the code. However, from spotter I can only observe results but not edit them. > > Yes, that is indeed something that I miss as well: browsing them all. > > Sven > >> Guille >> >> -------- Original Message -------- >>> and just another thing :) >>> >>> I think search patterns are a bit naive now (just matching begin of line). We could benefit from regex searchs and/or that search stuff made by Damienâ¦ I do not remember the name, but it has a name) that thing that you type âabcâ and it will find >>> - abc >>> - absoluteBinaryCapability >>> - AbstractBlooperContext >>> - etc >>> >>> I think it is a super functionality (and it was fast), that should be included by default. >>> >>> Esteban >>> >>>> On 10 Jun 2016, at 08:55, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: >>>> >>>> Ah, and I forget one very important: the menu! I almost do not use the older menu anymore :) >>>> >>>> Esteban >>>> >>>>> On 10 Jun 2016, at 08:54, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: >>>>> >>>>> I need to enable data collect ;) >>>>> >>>>> in my case, Iâm a happy Spotter userâ¦ I even have a plugin I made for my own that fits my way of work with repositoriesâ¦ anyway, I wanted to point: >>>>> >>>>> - I use a lot Classes/Methods/Packages... and catalog >>>>> - never used Senders/References/Pragmas because I do not find them easily (then is easier to search the class, then hit references) >>>>> - some categories seems confusing >>>>> - some other are not âfor extensive useâ but they are very useful: files, monticello package, playground >>>>> - this is more for the playground: I never understood why we need a difference between cached pages and named ones (a cached can just be named cached-1â¦ etc) >>>>> >>>>> I would be even happier is Yuriy extends his GitHub baseline plugin to scan subdirectories (most github projects keep mc data into a subdirectory) :) >>>>> >>>>> cheers! >>>>> Esteban >>>>> >>>>>> On 10 Jun 2016, at 08:42, Max Leske <maxleske(a)gmail.com> wrote: >>>>>> >>>>>> Hi Doru, >>>>>> >>>>>> I just want to point out that history may possibly be a bit of a false positive. When I open spotter, type and hit enter quickly I sometimes hit an entry from history that I didnât intend to (one annoying example of this is ProcessBrowser which I regularly hit accidentally when searching for Process). >>>>>> >>>>>> Cheers, >>>>>> Max >>>>>> >>>>>> >>>>>>> On 10 Jun 2016, at 00:56, Tudor Girba <tudor(a)tudorgirba.com> wrote: >>>>>>> >>>>>>> Hi, >>>>>>> >>>>>>> Juraj, Andrei and I did a rough analysis collected from 94 computers over the past 7 months. Of these, only 42 recorded more than 9 sessions so we only focused on these. It can be because the rest switched off the data collection in the meantime. We also excluded the computers of the GT team members. >>>>>>> >>>>>>> Of these 34 used the dive-in feature. That is, these users used at least one contextual search. >>>>>>> >>>>>>> We looked at the event of acting on an element (pressing Enter), and we collected the parent category. Acting on an item indicates that intent of search. There are 35 categories used in total, with 8 being used by 10 people (25% of the studied population) or more. Below you can see also the amount of computers using it: >>>>>>> >>>>>>> <spotter-categories-distribution.png> >>>>>>> >>>>>>> 'Classes'->40 >>>>>>> 'Implementors'->38 >>>>>>> 'History'->34 >>>>>>> 'Menu'->24 >>>>>>> 'Packages'->23 >>>>>>> 'Messages'->19 >>>>>>> 'Catalog Projects'->12 >>>>>>> 'Instance methods'->10 >>>>>>> 'Senders'->9 >>>>>>> 'Pragmas'->6 >>>>>>> 'References'->5 >>>>>>> 'Playground named pages'->5 >>>>>>> 'Playground cached pages'->4 >>>>>>> 'Help topics'->4 >>>>>>> 'Examples'->3 >>>>>>> 'Super instance methods'->3 >>>>>>> 'Selectors'->3 >>>>>>> 'ws.stfx.eu'->2 >>>>>>> 'GitHub Baselines'->2 >>>>>>> 'Dirty Monticello packages'->2 >>>>>>> 'Class methods'->2 >>>>>>> 'Global variables'->1 >>>>>>> 'All subclasses'->1 >>>>>>> 'Example Subjects'->1 >>>>>>> 'Files'->1 >>>>>>> 'Monticello Repositories'->1 >>>>>>> 'Metacello Configurations'->1 >>>>>>> 'Class instance variables'->1 >>>>>>> 'Items'->1 >>>>>>> 'Tags'->1 >>>>>>> 'Help contents'->1 >>>>>>> 'Monticello Package'->1 >>>>>>> 'Instance variables'->1 >>>>>>> 'Productions'->1 >>>>>>> 'Methods'->1 >>>>>>> >>>>>>> Also, in this analysis, some of the categories appear also at deeper levels (Senders, Implementors, References, Instance methods). >>>>>>> >>>>>>> As expected, Classes and Implementors are on top. Yet, the third is History, and it is also interesting to see that there is a high usage of a search through the World menu elements, but also of the Packages. >>>>>>> >>>>>>> We also note that there is quite a long tail, and this seems to confirm the hypothesis that different people have different needs and that these differences should be supported by the IDE. >>>>>>> >>>>>>> This analysis was carried out using the code that Juraj and Andrei put together for analyzing the data from the event recorder. >>>>>>> >>>>>>> Cheers, >>>>>>> Doru >>>>>>> >>>>>>> -- >>>>>>> www.tudorgirba.com >>>>>>> www.feenk.com >>>>>>> >>>>>>> "Value is always contextual." >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>> >>>>> >>>> >>> >> > >

June 10, 2016

Re: [Pharo-dev] Privacy sendDiagnosticsAndUsageData should be ternary, not binary

by Ben Coman

On Fri, Jun 10, 2016 at 2:37 PM, stepharo <stepharo(a)free.fr> wrote: > > Andrei for the shortcut reporter we put a number of show. Like that the user > will be reminder a number >> >> of times and the system will get calm. May be you could do the same > > > In a new image the message to send usage data is shown only once. > Indeed we could add a counter but that would have to maintain state between > images. > > > I would not. > Incremental thinking. Who'd have thought... a counter for incremental thinking? badaboom, badabing! ;) cheers -ben >>> >>> >>> So the proper behavior (imho) should be: >>> >>> ask Privacy for the settingâ¦ if the setting is not defined, then show a >>> popup. >>> If the setting is defined then respect it and do not show another popup. >> >> >> The current behaviour with showing the notification in Spotter and >> Nautilus should be a temporary one. >> Ideally we just need a single uniform way of showing this notification >> plus the option to control >> and see the data that each tool wants to record at a fined-grained level. >> >>> >>> >>> Also it would be nice to know what happens with the data. >> >> >> We store it on a server and use it for various analyses. >> >> what can you tell us on processor use in Spotter. > > > Not right at this moment. I'm way to busy with other urgent things. > Some older results were discussed here: > http://scg.unibe.ch/scgbib?query=Kube15a&display=abstract > Just for these result to carry more value we still need more people to send > data. > Right now few do. > > Cheers, > Andrei > >> >> >> It is also publicly available: GTEventTool default download. >> >>> >>> I mean my projects are open source so "sendSourceCode" shouldn't be an >>> issueâ¦ but what you can possibly learn from it? >>> Why not just analyze the content of SmalltalkHub/GitHub? >> >> >> Because if you have the setting enabled it does not always work to map >> different actions that with the code that you are working on. >> >> Cheers, >> Andrei >> >>> >>> >>> Peter >>> >> >> > >

June 10, 2016

Re: [Pharo-dev] System Browser UX analysis - double key shortcuts

by Ben Coman

> Le 10/6/16 Ã 03:33, Ben Coman a Ã©crit : > >> I happened to be back using a Squeak image to run the VM simulator and >> wanted to report my experience that it was a *joy* to again be able to >> use <ctrl-x> to delete methods from the System Browser. Over the past >>> >>> 12 months rather than complain about the double-key shortcuts, I've >> >> given myself time to adapt - but in practice I've adapted by always >> using the context menu rather than the double key shortcut. I don't >> understand the advantage of the double-key shortcuts here and why the >> same <ctrl-x> shortcut can't work in each of the four System Browser >> panes, since the "context" is provided by which pane is selected. >> >> I do recognise the value of double-keys for global shortcuts, since >> you need the first to select the tool to provide context for the >> second key. >> >> It would be nice to get some UX stastics for the System Browser about >> the predominance of shortcut versus menu usage for different people. On Fri, Jun 10, 2016 at 2:34 PM, stepharo <stepharo(a)free.fr> wrote: > Hi ben > > I never liked the double keys, but when they got introduced by camillo and > ben the goal was to be able to fully navigate wihout having to touch the mouse. If its all new code pouring out of your head with two hands permanently on the keyboard, maybe this may make sense. But most of my time is browsing to understand how to solve bugs, and my hand is usually on the mouse. > So from this goal point of view it was a nice attempt. I'm happy with the attempt. It sounded like a reasonable philosophy. It just didn't pan out for me. > Now we should be able > > to let user decide > > to have the same double keys in the other pane to avoid to create a > mode. I'm not clear on what you consider a "mode" here, but I'd distinguish it from "context". When I right-click the method pane I get one menu of actions, and when I right click on the class pane I get a different menu of action. The same action has a different response because of different context - and the shortcuts are attached to these contextual actions, so maybe its okay for the shortcut to be contextual. I would consider a "mode" getting different commands within in the same context - like hitting <esc> in a text editor like vi. On Fri, Jun 10, 2016 at 2:47 PM, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > wellâ¦ supposedly a revamp of the shorcuts mechanism will happen sometime during 6.0 development (if I find the time to work on it, or someone jumps in, of course :P) > > the idea with this is to centralise actions (using command pattern) and then users just declare what they will be usingâ¦ then *all tools* in system with same shortcuts will share actions. But when I right-click on the method pane, I don't get a first level menu [package;class;protocol;method] and select one of those first, then select my action. It *knows* I'm concerned with the method pane and I'm shown the context. Perhaps a single-key shortcut <ctrl-x> has its context ("M" in the method pane) added to convert it to a double-key shortcut before it is looked up centrally. > external packages should also have the possibility of install their actionsâ¦ this can create collisions, Maybe the single-key shortcuts should be limited. Its specifically <ctrl-x> that I miss, and this one I think has a broadly accepted meaning across domains, and would be rare for an App to redefine it. cheers -ben > but wellâ¦ (btw, same âuser install/declareâ idea should be taken for icons and others, as a way to avoid the stupid coupling of packages we have now)

June 10, 2016

Re: [Pharo-dev] spotter usage - a rough analysis of categories

by Thierry Goubier

2016-06-10 10:11 GMT+02:00 Sven Van Caekenberghe <sven(a)stfx.eu>: > > > On 10 Jun 2016, at 09:47, Guille Polito <guillermopolito(a)gmail.com> > wrote: > > > > I use spotter a lot as well. > > Me too ! > > > It is actually my default entry point to the system. It superseeded the > searches in Nautilus, and most of the searches that I used to start in the > workspace. > > Indeed, same here. > > > The things I look most by name are tools (Menu entries), classes and > packages. > > > > History is of great help and I use it a lot. But it poses a problem > sometimes because it alters the order of the results, and you have to wait > for the complete search to finish. If you do not do that, it happens what > Max explained: you type, hit <down>,<enter> and suddenly the UI refreshed > and you finished doing something else. > > > > Also, regarding the ordering of results, sometimes It bothers me that > they are sorted by category and not by relevance of the item. I'll explain > myself with a concrete case: > > - open spotter > > - type "Process" > > > > you get first two #menu items and then the #classes, containing the > class Process. I would have expected here to show first the element that > matched the most or with more relevance, regardless it's category. > > > > Question: Does spotter adapt to it's usage? Besides history, does it > learn what are the items with more relevance from previous usages in the > same session? > > Some simple learning would be super cool. > > > Also, from the top of my head there is at least one use case that I > cannot yet replace with spotter, and I have to use a normal > workspace/playground to do it => browse senders or implementors. While i > use senders/implementors search in spotter, sometimes I want to *browse* > them to be able to look at them, and interact with the code. However, from > spotter I can only observe results but not edit them. > > Yes, that is indeed something that I miss as well: browsing them all. > I don't use Spotter much, because I do that often: code in the result of a search for senders / implementors. Another thing that doesn't work too well with Spotter: I like to restrict the scope to a single package / group of package, either before the search, or after (making it easy to turn off large subsets of the search results). Thierry > > Sven > > > Guille > > > > -------- Original Message -------- > >> and just another thing :) > >> > >> I think search patterns are a bit naive now (just matching begin of > line). We could benefit from regex searchs and/or that search stuff made by > Damienâ¦ I do not remember the name, but it has a name) that thing that you > type âabcâ and it will find > >> - abc > >> - absoluteBinaryCapability > >> - AbstractBlooperContext > >> - etc > >> > >> I think it is a super functionality (and it was fast), that should be > included by default. > >> > >> Esteban > >> > >>> On 10 Jun 2016, at 08:55, Esteban Lorenzano <estebanlm(a)gmail.com> > wrote: > >>> > >>> Ah, and I forget one very important: the menu! I almost do not use the > older menu anymore :) > >>> > >>> Esteban > >>> > >>>> On 10 Jun 2016, at 08:54, Esteban Lorenzano <estebanlm(a)gmail.com> > wrote: > >>>> > >>>> I need to enable data collect ;) > >>>> > >>>> in my case, Iâm a happy Spotter userâ¦ I even have a plugin I made for > my own that fits my way of work with repositoriesâ¦ anyway, I wanted to > point: > >>>> > >>>> - I use a lot Classes/Methods/Packages... and catalog > >>>> - never used Senders/References/Pragmas because I do not find them > easily (then is easier to search the class, then hit references) > >>>> - some categories seems confusing > >>>> - some other are not âfor extensive useâ but they are very useful: > files, monticello package, playground > >>>> - this is more for the playground: I never understood why we need a > difference between cached pages and named ones (a cached can just be named > cached-1â¦ etc) > >>>> > >>>> I would be even happier is Yuriy extends his GitHub baseline plugin > to scan subdirectories (most github projects keep mc data into a > subdirectory) :) > >>>> > >>>> cheers! > >>>> Esteban > >>>> > >>>>> On 10 Jun 2016, at 08:42, Max Leske <maxleske(a)gmail.com> wrote: > >>>>> > >>>>> Hi Doru, > >>>>> > >>>>> I just want to point out that history may possibly be a bit of a > false positive. When I open spotter, > type and hit enter quickly I sometimes hit an entry from history that I > didnât intend to (one annoying example of this is ProcessBrowser which I > regularly hit accidentally when searching for Process). > >>>>> > >>>>> Cheers, > >>>>> Max > >>>>> > >>>>> > >>>>>> On 10 Jun 2016, at 00:56, Tudor Girba <tudor(a)tudorgirba.com> wrote: > >>>>>> > >>>>>> Hi, > >>>>>> > >>>>>> Juraj, Andrei and I did a rough analysis collected from 94 > computers over the past 7 months. Of these, only 42 recorded more than 9 > sessions so we only focused on these. It can be because the rest > switched off the data collection in > the meantime. We also excluded the computers of the GT team members. > >>>>>> > >>>>>> Of these 34 used the dive-in feature. That is, these users used at > least one contextual search. > >>>>>> > >>>>>> We looked at the event of acting on an element (pressing Enter), > and we collected the parent category. Acting on an item indicates that > intent of search. There are 35 categories used in total, with 8 being used > by 10 people (25% of the studied population) or more. Below you can see > also the amount of computers using it: > >>>>>> > >>>>>> <spotter-categories-distribution.png> > >>>>>> > >>>>>> 'Classes'->40 > >>>>>> 'Implementors'->38 > >>>>>> 'History'->34 > >>>>>> 'Menu'->24 > >>>>>> 'Packages'->23 > >>>>>> 'Messages'->19 > >>>>>> 'Catalog Projects'->12 > >>>>>> 'Instance methods'->10 > >>>>>> 'Senders'->9 > >>>>>> 'Pragmas'->6 > >>>>>> 'References'->5 > >>>>>> 'Playground named pages'->5 > >>>>>> 'Playground cached pages'->4 > >>>>>> 'Help topics'->4 > >>>>>> 'Examples'->3 > >>>>>> 'Super instance methods'->3 > >>>>>> 'Selectors'->3 > >>>>>> 'ws.stfx.eu'->2 > >>>>>> 'GitHub Baselines'->2 > >>>>>> 'Dirty Monticello packages'->2 > >>>>>> 'Class methods'->2 > >>>>>> 'Global variables'->1 > >>>>>> 'All subclasses'->1 > >>>>>> 'Example Subjects'->1 > >>>>>> 'Files'->1 > >>>>>> 'Monticello Repositories'->1 > >>>>>> 'Metacello Configurations'->1 > >>>>>> 'Class instance variables'->1 > >>>>>> 'Items'->1 > >>>>>> 'Tags'->1 > >>>>>> 'Help contents'->1 > >>>>>> 'Monticello Package'->1 > >>>>>> 'Instance variables'->1 > >>>>>> 'Productions'->1 > >>>>>> 'Methods'->1 > >>>>>> > >>>>>> Also, in this analysis, some of the categories appear also at > deeper levels (Senders, Implementors, References, Instance methods). > >>>>>> > >>>>>> As expected, Classes and Implementors are on top. Yet, the third is > History, and it is also interesting to see that there is a high usage of a > search through the World menu elements, but also of the Packages. > >>>>>> > >>>>>> We also note that there is quite a long tail, and this seems to > confirm the hypothesis that different people have different needs and that > these differences should be supported by the IDE. > >>>>>> > >>>>>> This analysis was carried out using the code that Juraj and Andrei > put together for analyzing the data from the event recorder. > >>>>>> > >>>>>> Cheers, > >>>>>> Doru > >>>>>> > >>>>>> -- > >>>>>> www.tudorgirba.com > >>>>>> www.feenk.com > >>>>>> > >>>>>> "Value is always contextual." > >>>>>> > >>>>>> > >>>>>> > >>>>>> > >>>>> > >>>> > >>> > >> > > > > >

June 10, 2016

Re: [Pharo-dev] spotter usage - a rough analysis of categories

by Christophe Demarey

> Le 10 juin 2016 Ã 09:47, Guille Polito <guillermopolito(a)gmail.com> a Ã©crit : > > I use spotter a lot as well. It is actually my default entry point to the system. It superseeded the searches in Nautilus, and most of the searches that I used to start in the workspace. The same for me. Spotter is the fastest path to open a tool, browse a class, etc. My top searches are: classes, implementors, packages I sometime use tools and use more history. (by the way, I often hit history entries by accident) Christophe > The things I look most by name are tools (Menu entries), classes and packages. > > History is of great help and I use it a lot. But it poses a problem sometimes because it alters the order of the results, and you have to wait for the complete search to finish. If you do not do that, it happens what Max explained: you type, hit <down>,<enter> and suddenly the UI refreshed and you finished doing something else. > > Also, regarding the ordering of results, sometimes It bothers me that they are sorted by category and not by relevance of the item. I'll explain myself with a concrete case: > - open spotter > - type "Process" > > you get first two #menu items and then the #classes, containing the class Process. I would have expected here to show first the element that matched the most or with more relevance, regardless it's category. > > Question: Does spotter adapt to it's usage? Besides history, does it learn what are the items with more relevance from previous usages in the same session? > > Also, from the top of my head there is at least one use case that I cannot yet replace with spotter, and I have to use a normal workspace/playground to do it => browse senders or implementors. While i use senders/implementors search in spotter, sometimes I want to *browse* them to be able to look at them, and interact with the code. However, from spotter I can only observe results but not edit them. > > Guille > > -------- Original Message -------- >> and just another thing :) >> >> I think search patterns are a bit naive now (just matching begin of line). We could benefit from regex searchs and/or that search stuff made by Damienâ¦ I do not remember the name, but it has a name) that thing that you type âabcâ and it will find >> - abc >> - absoluteBinaryCapability >> - AbstractBlooperContext >> - etc >> >> I think it is a super functionality (and it was fast), that should be included by default. >> >> Esteban >> >>> On 10 Jun 2016, at 08:55, Esteban Lorenzano <estebanlm(a)gmail.com <mailto:estebanlm@gmail.com>> wrote: >>> >>> Ah, and I forget one very important: the menu! I almost do not use the older menu anymore :) >>> >>> Esteban >>> >>>> On 10 Jun 2016, at 08:54, Esteban Lorenzano < <mailto:estebanlm@gmail.com>estebanlm(a)gmail.com <mailto:estebanlm@gmail.com>> wrote: >>>> >>>> I need to enable data collect ;) >>>> >>>> in my case, Iâm a happy Spotter userâ¦ I even have a plugin I made for my own that fits my way of work with repositoriesâ¦ anyway, I wanted to point: >>>> >>>> - I use a lot Classes/Methods/Packages... and catalog >>>> - never used Senders/References/Pragmas because I do not find them easily (then is easier to search the class, then hit references) >>>> - some categories seems confusing >>>> - some other are not âfor extensive useâ but they are very useful: files, monticello package, playground >>>> - this is more for the playground: I never understood why we need a difference between cached pages and named ones (a cached can just be named cached-1â¦ etc) >>>> >>>> I would be even happier is Yuriy extends his GitHub baseline plugin to scan subdirectories (most github projects keep mc data into a subdirectory) :) >>>> >>>> cheers! >>>> Esteban >>>> >>>>> On 10 Jun 2016, at 08:42, Max Leske < <mailto:maxleske@gmail.com>maxleske(a)gmail.com <mailto:maxleske@gmail.com>> wrote: >>>>> >>>>> Hi Doru, >>>>> >>>>> I just want to point out that history may possibly be a bit of a false positive. When I open spotter, type and hit enter quickly I sometimes hit an entry from history that I didnât intend to (one annoying example of this is ProcessBrowser which I regularly hit accidentally when searching for Process). >>>>> >>>>> Cheers, >>>>> Max >>>>> >>>>> >>>>>> On 10 Jun 2016, at 00:56, Tudor Girba < <mailto:tudor@tudorgirba.com>tudor(a)tudorgirba.com <mailto:tudor@tudorgirba.com>> wrote: >>>>>> >>>>>> Hi, >>>>>> >>>>>> Juraj, Andrei and I did a rough analysis collected from 94 computers over the past 7 months. Of these, only 42 recorded more than 9 sessions so we only focused on these. It can be because the rest switched off the data collection in the meantime. We also excluded the computers of the GT team members. >>>>>> >>>>>> Of these 34 used the dive-in feature. That is, these users used at least one contextual search. >>>>>> >>>>>> We looked at the event of acting on an element (pressing Enter), and we collected the parent category. Acting on an item indicates that intent of search. There are 35 categories used in total, with 8 being used by 10 people (25% of the studied population) or more. Below you can see also the amount of computers using it: >>>>>> >>>>>> <spotter-categories-distribution.png> >>>>>> >>>>>> 'Classes'->40 >>>>>> 'Implementors'->38 >>>>>> 'History'->34 >>>>>> 'Menu'->24 >>>>>> 'Packages'->23 >>>>>> 'Messages'->19 >>>>>> 'Catalog Projects'->12 >>>>>> 'Instance methods'->10 >>>>>> 'Senders'->9 >>>>>> 'Pragmas'->6 >>>>>> 'References'->5 >>>>>> 'Playground named pages'->5 >>>>>> 'Playground cached pages'->4 >>>>>> 'Help topics'->4 >>>>>> 'Examples'->3 >>>>>> 'Super instance methods'->3 >>>>>> 'Selectors'->3 >>>>>> 'ws.stfx.eu <http://ws.stfx.eu/>'->2 >>>>>> 'GitHub Baselines'->2 >>>>>> 'Dirty Monticello packages'->2 >>>>>> 'Class methods'->2 >>>>>> 'Global variables'->1 >>>>>> 'All subclasses'->1 >>>>>> 'Example Subjects'->1 >>>>>> 'Files'->1 >>>>>> 'Monticello Repositories'->1 >>>>>> 'Metacello Configurations'->1 >>>>>> 'Class instance variables'->1 >>>>>> 'Items'->1 >>>>>> 'Tags'->1 >>>>>> 'Help contents'->1 >>>>>> 'Monticello Package'->1 >>>>>> 'Instance variables'->1 >>>>>> 'Productions'->1 >>>>>> 'Methods'->1 >>>>>> >>>>>> Also, in this analysis, some of the categories appear also at deeper levels (Senders, Implementors, References, Instance methods). >>>>>> >>>>>> As expected, Classes and Implementors are on top. Yet, the third is History, and it is also interesting to see that there is a high usage of a search through the World menu elements, but also of the Packages. >>>>>> >>>>>> We also note that there is quite a long tail, and this seems to confirm the hypothesis that different people have different needs and that these differences should be supported by the IDE. >>>>>> >>>>>> This analysis was carried out using the code that Juraj and Andrei put together for analyzing the data from the event recorder. >>>>>> >>>>>> Cheers, >>>>>> Doru >>>>>> >>>>>> -- >>>>>> <http://www.tudorgirba.com/>www.tudorgirba.com <http://www.tudorgirba.com/> >>>>>> <http://www.feenk.com/>www.feenk.com <http://www.feenk.com/> >>>>>> >>>>>> "Value is always contextual." >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>> >>> >> >

June 10, 2016

[pharo-project/pharo-core]

by GitHub

Branch: refs/tags/60075 Home: https://github.com/pharo-project/pharo-core

June 10, 2016

[pharo-project/pharo-core] 5fe187: 60075

by GitHub

Branch: refs/heads/6.0 Home: https://github.com/pharo-project/pharo-core Commit: 5fe1879036aaa6f2824b3062a6b8be0f6bf6d666 https://github.com/pharo-project/pharo-core/commit/5fe1879036aaa6f2824b3062… Author: Jenkins Build Server <board(a)pharo-project.org> Date: 2016-06-10 (Fri, 10 Jun 2016) Changed paths: M FileSystem-Core.package/FileReference.class/class/instance creation/newTempFilePrefix_suffix_.st R ScriptLoader60.package/ScriptLoader.class/instance/pharo - scripts/script60074.st A ScriptLoader60.package/ScriptLoader.class/instance/pharo - scripts/script60075.st R ScriptLoader60.package/ScriptLoader.class/instance/pharo - updates/update60074.st A ScriptLoader60.package/ScriptLoader.class/instance/pharo - updates/update60075.st M ScriptLoader60.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st R System-Sources.package/ProcessAndSessionLocalSourcesFileArray.class/README.md R System-Sources.package/ProcessAndSessionLocalSourcesFileArray.class/definition.st M System-Sources.package/SourceFileArray.class/definition.st M System-Sources.package/SourceFileArray.class/instance/initialization/initialize.st R System-Sources.package/SourceFileArray.class/instance/private - accessing file streams/localProcessReadStreamAtFileIndex_atPosition_ifPresent_ifAbsent_.st R System-Sources.package/SourceFileArray.class/instance/private - copying/localProcessReadOnlyCopy.st A System-Sources.package/SourceFileArray.class/instance/private/closeFileArray_.st M System-Sources.package/SourceFileArray.class/instance/public - file system operations/close.st M System-Sources.package/SourceFileArray.class/instance/public - file system operations/ensureOpen.st R System-Support.package/ProcessAndSessionLocalVariable.class/README.md R System-Support.package/ProcessAndSessionLocalVariable.class/definition.st R System-Support.package/ProcessAndSessionLocalVariable.class/instance/accessing/value.st R System-Support.package/ProcessAndSessionLocalVariable.class/instance/evaluating/value_.st A System-Support.package/SmalltalkImage.class/class/initialization/shutDown_.st M Tool-FileList.package/FileList.class/class/menu/contentMenu_.st Log Message: ----------- 60075 18462 FileReference>>newTempFilePrefix:suffix: https://pharo.fogbugz.com/f/cases/18462 18458 fall back debug menu in FileLists preview pane context menu https://pharo.fogbugz.com/f/cases/18458 18456 Add readOnlyQueue to SourceFilesArray https://pharo.fogbugz.com/f/cases/18456 http://files.pharo.org/image/60/60075.zip

June 10, 2016

Re: [Pharo-dev] spotter usage - a rough analysis of categories

by Sven Van Caekenberghe

> On 10 Jun 2016, at 09:47, Guille Polito <guillermopolito(a)gmail.com> wrote: > > I use spotter a lot as well. Me too ! > It is actually my default entry point to the system. It superseeded the searches in Nautilus, and most of the searches that I used to start in the workspace. Indeed, same here. > The things I look most by name are tools (Menu entries), classes and packages. > > History is of great help and I use it a lot. But it poses a problem sometimes because it alters the order of the results, and you have to wait for the complete search to finish. If you do not do that, it happens what Max explained: you type, hit <down>,<enter> and suddenly the UI refreshed and you finished doing something else. > > Also, regarding the ordering of results, sometimes It bothers me that they are sorted by category and not by relevance of the item. I'll explain myself with a concrete case: > - open spotter > - type "Process" > > you get first two #menu items and then the #classes, containing the class Process. I would have expected here to show first the element that matched the most or with more relevance, regardless it's category. > > Question: Does spotter adapt to it's usage? Besides history, does it learn what are the items with more relevance from previous usages in the same session? Some simple learning would be super cool. > Also, from the top of my head there is at least one use case that I cannot yet replace with spotter, and I have to use a normal workspace/playground to do it => browse senders or implementors. While i use senders/implementors search in spotter, sometimes I want to *browse* them to be able to look at them, and interact with the code. However, from spotter I can only observe results but not edit them. Yes, that is indeed something that I miss as well: browsing them all. Sven > Guille > > -------- Original Message -------- >> and just another thing :) >> >> I think search patterns are a bit naive now (just matching begin of line). We could benefit from regex searchs and/or that search stuff made by Damienâ¦ I do not remember the name, but it has a name) that thing that you type âabcâ and it will find >> - abc >> - absoluteBinaryCapability >> - AbstractBlooperContext >> - etc >> >> I think it is a super functionality (and it was fast), that should be included by default. >> >> Esteban >> >>> On 10 Jun 2016, at 08:55, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: >>> >>> Ah, and I forget one very important: the menu! I almost do not use the older menu anymore :) >>> >>> Esteban >>> >>>> On 10 Jun 2016, at 08:54, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: >>>> >>>> I need to enable data collect ;) >>>> >>>> in my case, Iâm a happy Spotter userâ¦ I even have a plugin I made for my own that fits my way of work with repositoriesâ¦ anyway, I wanted to point: >>>> >>>> - I use a lot Classes/Methods/Packages... and catalog >>>> - never used Senders/References/Pragmas because I do not find them easily (then is easier to search the class, then hit references) >>>> - some categories seems confusing >>>> - some other are not âfor extensive useâ but they are very useful: files, monticello package, playground >>>> - this is more for the playground: I never understood why we need a difference between cached pages and named ones (a cached can just be named cached-1â¦ etc) >>>> >>>> I would be even happier is Yuriy extends his GitHub baseline plugin to scan subdirectories (most github projects keep mc data into a subdirectory) :) >>>> >>>> cheers! >>>> Esteban >>>> >>>>> On 10 Jun 2016, at 08:42, Max Leske <maxleske(a)gmail.com> wrote: >>>>> >>>>> Hi Doru, >>>>> >>>>> I just want to point out that history may possibly be a bit of a false positive. When I open spotter, type and hit enter quickly I sometimes hit an entry from history that I didnât intend to (one annoying example of this is ProcessBrowser which I regularly hit accidentally when searching for Process). >>>>> >>>>> Cheers, >>>>> Max >>>>> >>>>> >>>>>> On 10 Jun 2016, at 00:56, Tudor Girba <tudor(a)tudorgirba.com> wrote: >>>>>> >>>>>> Hi, >>>>>> >>>>>> Juraj, Andrei and I did a rough analysis collected from 94 computers over the past 7 months. Of these, only 42 recorded more than 9 sessions so we only focused on these. It can be because the rest switched off the data collection in the meantime. We also excluded the computers of the GT team members. >>>>>> >>>>>> Of these 34 used the dive-in feature. That is, these users used at least one contextual search. >>>>>> >>>>>> We looked at the event of acting on an element (pressing Enter), and we collected the parent category. Acting on an item indicates that intent of search. There are 35 categories used in total, with 8 being used by 10 people (25% of the studied population) or more. Below you can see also the amount of computers using it: >>>>>> >>>>>> <spotter-categories-distribution.png> >>>>>> >>>>>> 'Classes'->40 >>>>>> 'Implementors'->38 >>>>>> 'History'->34 >>>>>> 'Menu'->24 >>>>>> 'Packages'->23 >>>>>> 'Messages'->19 >>>>>> 'Catalog Projects'->12 >>>>>> 'Instance methods'->10 >>>>>> 'Senders'->9 >>>>>> 'Pragmas'->6 >>>>>> 'References'->5 >>>>>> 'Playground named pages'->5 >>>>>> 'Playground cached pages'->4 >>>>>> 'Help topics'->4 >>>>>> 'Examples'->3 >>>>>> 'Super instance methods'->3 >>>>>> 'Selectors'->3 >>>>>> 'ws.stfx.eu'->2 >>>>>> 'GitHub Baselines'->2 >>>>>> 'Dirty Monticello packages'->2 >>>>>> 'Class methods'->2 >>>>>> 'Global variables'->1 >>>>>> 'All subclasses'->1 >>>>>> 'Example Subjects'->1 >>>>>> 'Files'->1 >>>>>> 'Monticello Repositories'->1 >>>>>> 'Metacello Configurations'->1 >>>>>> 'Class instance variables'->1 >>>>>> 'Items'->1 >>>>>> 'Tags'->1 >>>>>> 'Help contents'->1 >>>>>> 'Monticello Package'->1 >>>>>> 'Instance variables'->1 >>>>>> 'Productions'->1 >>>>>> 'Methods'->1 >>>>>> >>>>>> Also, in this analysis, some of the categories appear also at deeper levels (Senders, Implementors, References, Instance methods). >>>>>> >>>>>> As expected, Classes and Implementors are on top. Yet, the third is History, and it is also interesting to see that there is a high usage of a search through the World menu elements, but also of the Packages. >>>>>> >>>>>> We also note that there is quite a long tail, and this seems to confirm the hypothesis that different people have different needs and that these differences should be supported by the IDE. >>>>>> >>>>>> This analysis was carried out using the code that Juraj and Andrei put together for analyzing the data from the event recorder. >>>>>> >>>>>> Cheers, >>>>>> Doru >>>>>> >>>>>> -- >>>>>> www.tudorgirba.com >>>>>> www.feenk.com >>>>>> >>>>>> "Value is always contextual." >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>> >>> >> >

June 10, 2016

Re: [Pharo-dev] spotter usage - a rough analysis of categories

by Guille Polito

I use spotter a lot as well. It is actually my default entry point to the system. It superseeded the searches in Nautilus, and most of the searches that I used to start in the workspace. The things I look most by name are tools (Menu entries), classes and packages. History is of great help and I use it a lot. But it poses a problem sometimes because it alters the order of the results, and you have to wait for the complete search to finish. If you do not do that, it happens what Max explained: you type, hit <down>,<enter> and suddenly the UI refreshed and you finished doing something else. Also, regarding the ordering of results, sometimes It bothers me that they are sorted by category and not by relevance of the item. I'll explain myself with a concrete case: - open spotter - type "Process" you get first two #menu items and then the #classes, containing the class Process. I would have expected here to show first the element that matched the most or with more relevance, regardless it's category. Question: Does spotter adapt to it's usage? Besides history, does it learn what are the items with more relevance from previous usages in the same session? Also, from the top of my head there is at least one use case that I cannot yet replace with spotter, and I have to use a normal workspace/playground to do it => browse senders or implementors. While i use senders/implementors search in spotter, sometimes I want to *browse* them to be able to look at them, and interact with the code. However, from spotter I can only observe results but not edit them. Guille -------- Original Message -------- > and just another thing :) > > I think search patterns are a bit naive now (just matching begin of > line). We could benefit from regex searchs and/or that search stuff > made by Damienâ¦ I do not remember the name, but it has a name) that > thing that you type âabcâ and it will find > - abc > - absoluteBinaryCapability > - AbstractBlooperContext > - etc > > I think it is a super functionality (and it was fast), that should be > included by default. > > Esteban > >> On 10 Jun 2016, at 08:55, Esteban Lorenzano <estebanlm(a)gmail.com >> <mailto:estebanlm@gmail.com>> wrote: >> >> Ah, and I forget one very important: the menu! I almost do not use >> the older menu anymore :) >> >> Esteban >> >>> On 10 Jun 2016, at 08:54, Esteban Lorenzano <estebanlm(a)gmail.com >>> <mailto:estebanlm@gmail.com>> wrote: >>> >>> I need to enable data collect ;) >>> >>> in my case, Iâm a happy Spotter userâ¦ I even have a plugin I made >>> for my own that fits my way of work with repositoriesâ¦ anyway, I >>> wanted to point: >>> >>> - I use a lot Classes/Methods/Packages... and catalog >>> - never used Senders/References/Pragmas because I do not find them >>> easily (then is easier to search the class, then hit references) >>> - some categories seems confusing >>> - some other are not âfor extensive useâ but they are very useful: >>> files, monticello package, playground >>> - this is more for the playground: I never understood why we need a >>> difference between cached pages and named ones (a cached can just be >>> named cached-1â¦ etc) >>> >>> I would be even happier is Yuriy extends his GitHub baseline plugin >>> to scan subdirectories (most github projects keep mc data into a >>> subdirectory) :) >>> >>> cheers! >>> Esteban >>> >>>> On 10 Jun 2016, at 08:42, Max Leske <maxleske(a)gmail.com >>>> <mailto:maxleske@gmail.com>> wrote: >>>> >>>> Hi Doru, >>>> >>>> I just want to point out that history may possibly be a bit of a >>>> false positive. When I open spotter, type and hit enter quickly I >>>> sometimes hit an entry from history that I didnât intend to (one >>>> annoying example of this is ProcessBrowser which I regularly hit >>>> accidentally when searching for Process). >>>> >>>> Cheers, >>>> Max >>>> >>>> >>>>> On 10 Jun 2016, at 00:56, Tudor Girba <tudor(a)tudorgirba.com >>>>> <mailto:tudor@tudorgirba.com>> wrote: >>>>> >>>>> Hi, >>>>> >>>>> Juraj, Andrei and I did a rough analysis collected from 94 >>>>> computers over the past 7 months. Of these, only 42 recorded >>>>> more than 9 sessions so we only focused on these. It can be >>>>> because the rest switched off the data collection in the meantime. >>>>> We also excluded the computers of the GT team members. >>>>> >>>>> Of these 34 used the dive-in feature. That is, these users used at >>>>> least one contextual search. >>>>> >>>>> We looked at the event of acting on an element (pressing Enter), >>>>> and we collected the parent category. Acting on an item indicates >>>>> that intent of search. There are 35 categories used in total, with >>>>> 8 being used by 10 people (25% of the studied population) or more. >>>>> Below you can see also the amount of computers using it: >>>>> >>>>> <spotter-categories-distribution.png> >>>>> >>>>> 'Classes'->40 >>>>> 'Implementors'->38 >>>>> 'History'->34 >>>>> 'Menu'->24 >>>>> 'Packages'->23 >>>>> 'Messages'->19 >>>>> 'Catalog Projects'->12 >>>>> 'Instance methods'->10 >>>>> 'Senders'->9 >>>>> 'Pragmas'->6 >>>>> 'References'->5 >>>>> 'Playground named pages'->5 >>>>> 'Playground cached pages'->4 >>>>> 'Help topics'->4 >>>>> 'Examples'->3 >>>>> 'Super instance methods'->3 >>>>> 'Selectors'->3 >>>>> 'ws.stfx.eu <http://ws.stfx.eu/>'->2 >>>>> 'GitHub Baselines'->2 >>>>> 'Dirty Monticello packages'->2 >>>>> 'Class methods'->2 >>>>> 'Global variables'->1 >>>>> 'All subclasses'->1 >>>>> 'Example Subjects'->1 >>>>> 'Files'->1 >>>>> 'Monticello Repositories'->1 >>>>> 'Metacello Configurations'->1 >>>>> 'Class instance variables'->1 >>>>> 'Items'->1 >>>>> 'Tags'->1 >>>>> 'Help contents'->1 >>>>> 'Monticello Package'->1 >>>>> 'Instance variables'->1 >>>>> 'Productions'->1 >>>>> 'Methods'->1 >>>>> >>>>> Also, in this analysis, some of the categories appear also at >>>>> deeper levels (Senders, Implementors, References, Instance methods). >>>>> >>>>> As expected, Classes and Implementors are on top. Yet, the third >>>>> is History, and it is also interesting to see that there is a high >>>>> usage of a search through the World menu elements, but also of the >>>>> Packages. >>>>> >>>>> We also note that there is quite a long tail, and this seems to >>>>> confirm the hypothesis that different people have different needs >>>>> and that these differences should be supported by the IDE. >>>>> >>>>> This analysis was carried out using the code that Juraj and Andrei >>>>> put together for analyzing the data from the event recorder. >>>>> >>>>> Cheers, >>>>> Doru >>>>> >>>>> -- >>>>> www.tudorgirba.com <http://www.tudorgirba.com/> >>>>> www.feenk.com <http://www.feenk.com/> >>>>> >>>>> "Value is always contextual." >>>>> >>>>> >>>>> >>>>> >>>> >>> >> >

June 10, 2016

Re: [Pharo-dev] spotter usage - a rough analysis of categories

by Esteban Lorenzano

and just another thing :) I think search patterns are a bit naive now (just matching begin of line). We could benefit from regex searchs and/or that search stuff made by Damienâ¦ I do not remember the name, but it has a name) that thing that you type âabcâ and it will find - abc - absoluteBinaryCapability - AbstractBlooperContext - etc I think it is a super functionality (and it was fast), that should be included by default. Esteban > On 10 Jun 2016, at 08:55, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > > Ah, and I forget one very important: the menu! I almost do not use the older menu anymore :) > > Esteban > >> On 10 Jun 2016, at 08:54, Esteban Lorenzano <estebanlm(a)gmail.com <mailto:estebanlm@gmail.com>> wrote: >> >> I need to enable data collect ;) >> >> in my case, Iâm a happy Spotter userâ¦ I even have a plugin I made for my own that fits my way of work with repositoriesâ¦ anyway, I wanted to point: >> >> - I use a lot Classes/Methods/Packages... and catalog >> - never used Senders/References/Pragmas because I do not find them easily (then is easier to search the class, then hit references) >> - some categories seems confusing >> - some other are not âfor extensive useâ but they are very useful: files, monticello package, playground >> - this is more for the playground: I never understood why we need a difference between cached pages and named ones (a cached can just be named cached-1â¦ etc) >> >> I would be even happier is Yuriy extends his GitHub baseline plugin to scan subdirectories (most github projects keep mc data into a subdirectory) :) >> >> cheers! >> Esteban >> >>> On 10 Jun 2016, at 08:42, Max Leske <maxleske(a)gmail.com <mailto:maxleske@gmail.com>> wrote: >>> >>> Hi Doru, >>> >>> I just want to point out that history may possibly be a bit of a false positive. When I open spotter, type and hit enter quickly I sometimes hit an entry from history that I didnât intend to (one annoying example of this is ProcessBrowser which I regularly hit accidentally when searching for Process). >>> >>> Cheers, >>> Max >>> >>> >>>> On 10 Jun 2016, at 00:56, Tudor Girba <tudor(a)tudorgirba.com <mailto:tudor@tudorgirba.com>> wrote: >>>> >>>> Hi, >>>> >>>> Juraj, Andrei and I did a rough analysis collected from 94 computers over the past 7 months. Of these, only 42 recorded more than 9 sessions so we only focused on these. It can be because the rest switched off the data collection in the meantime. We also excluded the computers of the GT team members. >>>> >>>> Of these 34 used the dive-in feature. That is, these users used at least one contextual search. >>>> >>>> We looked at the event of acting on an element (pressing Enter), and we collected the parent category. Acting on an item indicates that intent of search. There are 35 categories used in total, with 8 being used by 10 people (25% of the studied population) or more. Below you can see also the amount of computers using it: >>>> >>>> <spotter-categories-distribution.png> >>>> >>>> 'Classes'->40 >>>> 'Implementors'->38 >>>> 'History'->34 >>>> 'Menu'->24 >>>> 'Packages'->23 >>>> 'Messages'->19 >>>> 'Catalog Projects'->12 >>>> 'Instance methods'->10 >>>> 'Senders'->9 >>>> 'Pragmas'->6 >>>> 'References'->5 >>>> 'Playground named pages'->5 >>>> 'Playground cached pages'->4 >>>> 'Help topics'->4 >>>> 'Examples'->3 >>>> 'Super instance methods'->3 >>>> 'Selectors'->3 >>>> 'ws.stfx.eu <http://ws.stfx.eu/>'->2 >>>> 'GitHub Baselines'->2 >>>> 'Dirty Monticello packages'->2 >>>> 'Class methods'->2 >>>> 'Global variables'->1 >>>> 'All subclasses'->1 >>>> 'Example Subjects'->1 >>>> 'Files'->1 >>>> 'Monticello Repositories'->1 >>>> 'Metacello Configurations'->1 >>>> 'Class instance variables'->1 >>>> 'Items'->1 >>>> 'Tags'->1 >>>> 'Help contents'->1 >>>> 'Monticello Package'->1 >>>> 'Instance variables'->1 >>>> 'Productions'->1 >>>> 'Methods'->1 >>>> >>>> Also, in this analysis, some of the categories appear also at deeper levels (Senders, Implementors, References, Instance methods). >>>> >>>> As expected, Classes and Implementors are on top. Yet, the third is History, and it is also interesting to see that there is a high usage of a search through the World menu elements, but also of the Packages. >>>> >>>> We also note that there is quite a long tail, and this seems to confirm the hypothesis that different people have different needs and that these differences should be supported by the IDE. >>>> >>>> This analysis was carried out using the code that Juraj and Andrei put together for analyzing the data from the event recorder. >>>> >>>> Cheers, >>>> Doru >>>> >>>> -- >>>> www.tudorgirba.com <http://www.tudorgirba.com/> >>>> www.feenk.com <http://www.feenk.com/> >>>> >>>> "Value is always contextual." >>>> >>>> >>>> >>>> >>> >> >

June 10, 2016