- Pharo-dev - lists.pharo.org

Re: [Pharo-project] Fuel - class loading issue when superclass changed inst vars

by Martin Dias

Hi! Mariano, yes, I forgot to say that B is subclass of A. Henry, thanks a lot for sharing your knowledge about BOSS. Following my example, the "backwards-compatability reader method" has to be defined in A, right? Or do you specify it as a parameter to the class loader? I would like to have automatic tolerance to simple layout changes like this. Eliot, thanks for the correction. Rescan the methods seems to be a nice solution, and has not dependency with compiler. The migration service you describe makes me think, because we have recently added migration support to Fuel; it works like: aMaterializer migrateClassNamed: #OldPoint toClass: NewPoint variables: {('x' -> 'posX'). ('y' -> 'posY')}. On one hand, in simple cases it should be more efficient than creating a dictionary for each instance and then become them. On the other hand, in more complex cases it's not enough and the dictionary could be a more uniform solution. For example, if NewPoint has just a variable named "coordinates", which points to an array with x and y as elements. MartÃn On Tue, Mar 6, 2012 at 3:27 PM, Eliot Miranda <eliot.miranda(a)gmail.com>wrote: > > > On Tue, Mar 6, 2012 at 3:05 AM, Henrik Johansen < > henrik.s.johansen(a)veloxit.no> wrote: > >> >> On Mar 6, 2012, at 12:03 PM, Henrik Johansen wrote: >> >> >> >> > >> > Parcels >> > - Check a hash based on class layout vs equivalent hash for the stored >> method's class (stored with the method). >> > If different: >> > 1) check for existence of a backwards-compatability reader method, hand >> the old instance to it, and expect the old instance to be #become'd into >> something current, >> > 2) raise an error if no such method exists. >> > >> > Now, the main problem with this scheme is it's left as an exercise to >> the user to come up with a procedure to ensure said backwards-compat reader >> method stays up to date as additional changes are made. >> > >> > Is it a big problem? >> > Depends on your tests, and programmer diligence at any given day. >> > >> > Added inst vars are not a big deal, as if you forget them, there will >> usually be a nil #DNU somewhere down the line. (or you use lazy >> initialization, and everything works as expected even for existing >> instances) >> > >> > Removals/reorderings are a bigger issue, as the detection of failure >> (inst vars in wrong slots) are often far removed from the source of >> problems (forgetting to update the reader method), >> > If seldomly used, it may even go unnoticed until the instance is next >> saved, at which time you're in real trouble (especially if saved alongside >> newly created ones, in which case there is no consistency). >> > >> > In general I think it's an ok scheme, but would like to see a solution >> more resilient to user-error. >> > How to achieve that is an interesting topic, which I haven't yet found >> time to think through as thoroughly as I had wished/intended some months >> ago. :( >> >> Errr, BOSS, not Parcels. >> > > Right. What Parcels do is shape-change instances *and* rescan methods to > fix up inst var offsets. When a parcel is saved the "signature" of classes > of instances there-in are saved so that the parcel contains all the inst > var names of the class and its superclass. If on load the superclass chain > has a different set of inst vars then lost inst vars are omitted and added > inst vars are nil in the materialized instances. If a Parcel contains a > full class (not just instances) that has methods, and the superclass chains > inst vars have changed then these methods are "rescanned" (disassembled, > and reassembled, not using the compiler) and inst var offsets are fixed up, > missing inst vars getting changed into undeclared variable references (I > think; at least this is what *should* happen; its what happens when one > removes an inst var that is still referenced from methods). > > What's missing in the parcel scheme is any hook to allow the user to > process shape-changing instances with the values of lost inst vars in hand. > Presumably there will be cases when those values are essential to any > schema migration. Personally I would want to decouple schema migration and > add it as a post-processing step, which would imply that the materializer > would offer a service (materialize in a special mode) where it constructed > a dictionary from materialized and shape-changed instance to in parcel > state (e.g. an Array of inst var values as they occurred in the parcel). > Then the materialized instance could be migrated after the fact, with the > default behaviour being analogous to what the system does now on class > redefinition (values of deleted inst vars are lost, aded inst vars are nil). > > This reminds me of a bug with the ClassBuilder/RefactoringBrowser > combination. If one refactors pushing an inst var up to a superclass or > down to subclasses, the values of te inst var in instances are lost because > the class change is in fact done as two changes, a deletion followed by an > addition. This again could be fixed in a wrapper, building a dictionary > from instance to values, performing the set of class changes, and then > restoring the inst var state from the dictionary. > > > >> My bad. >> >> Cheers, >> Henry >> >> >> > > > -- > best, > Eliot > >

March 6, 2012

Re: [Pharo-project] .changes are now 37MBâ¦.

by Max Leske

If it wouldn't take so bloody long to run (approx. 9 hours for all Squeak version up to and including Pharo 1.0) and didn't require me to port back all the changes we made to FSGit to Pharogenesis I'd happily do that. I just don't think the effort is worth it (I'll change my mind if people on the list start begging for it :) ) Ah yes, and the fetchingâ¦ That would of course be missing tooâ¦ :p Cheers, Max On 06.03.2012, at 20:05, Camillo Bruni wrote: > max, > > maybe you want to create a test git repos with all the sources back to 1.0 :D > and then we should record all changes into a local git-repos :) > if you want to got back some versions you simply start fetching the missing objects from a central repos somewhere... > > so far dreaming a bit > > best > cami > > On 2012-03-06, at 20:01, Marcus Denker wrote: > >> >> On Mar 6, 2012, at 7:30 PM, Max Leske wrote: >> >>> This may be a stupid questionâ¦ (As far as I understand, condensing the changes is not the same) >>> Why don't you increment the version number of the sources file to 1.1 and use a fresh changes file? >>> >> The sources file does not contain the sources and the changes just the changes. (even though one could think so taking >> the names into account.) >> >> The sources file containes the sources of ages ago (1.0), the .changes is "on top of that". This means that the .sources >> file contains a lot of code from classes and methods that have been removed (like the changes), >> as well as old versions. >> >> The .sources/.changes mechanism merges three responsabilties: >> >> 1) be a log of all edited code in case of a crash >> 2) store the current source of the methods (when you ask a method for code, it reads either from the .sources >> (method not changed since 1.0) or from the .changes (method was changed). >> 3) provide a history for all code >> >> And all that in 2 big files that one can only append to because? No idea, actually... >> >> I personally think that it is a mistake today to merge all these into one such basic mechanism. Especialy if >> you look at how complex the code is in the image... it's actually amazing. >> >> When you do a #condenseSources (which is sadly broken right now), then it generates a .sources >> file with just the code in the current image. It would be interesting to see how large that is. >> >> Marcus >> >> -- >> Marcus Denker -- http://marcusdenker.de >> >> > >

March 6, 2012

Re: [Pharo-project] .changes are now 37MBâ¦.

by Max Leske

That's what I meantâ¦ So that won't workâ¦ Thanks for the explanation Marcus. On 06.03.2012, at 20:01, Marcus Denker wrote: > When you do a #condenseSources (which is sadly broken right now), then it generates a .sources > file with just the code in the current image. It would be interesting to see how large that is.

March 6, 2012

Re: [Pharo-project] .changes are now 37MBâ¦.

by Eliot Miranda

On Tue, Mar 6, 2012 at 11:01 AM, Marcus Denker <marcus.denker(a)inria.fr>wrote: > > On Mar 6, 2012, at 7:30 PM, Max Leske wrote: > > > This may be a stupid questionâ¦ (As far as I understand, condensing the > changes is not the same) > > Why don't you increment the version number of the sources file to 1.1 > and use a fresh changes file? > > > The sources file does not contain the sources and the changes just the > changes. (even though one could think so taking > the names into account.) > > The sources file containes the sources of ages ago (1.0), the .changes is > "on top of that". This means that the .sources > file contains a lot of code from classes and methods that have been > removed (like the changes), > as well as old versions. > > The .sources/.changes mechanism merges three responsabilties: > > 1) be a log of all edited code in case of a crash > 2) store the current source of the methods (when you ask a method for > code, it reads either from the .sources > (method not changed since 1.0) or from the .changes (method was changed). > 3) provide a history for all code > > And all that in 2 big files that one can only append to because? No idea, > actually... > > I personally think that it is a mistake today to merge all these into one > such basic mechanism. Especialy if > you look at how complex the code is in the image... it's actually amazing. > > When you do a #condenseSources (which is sadly broken right now), then it > generates a .sources > file with just the code in the current image. It would be interesting to > see how large that is. > IMO it's also interesting to see how big a condense of only the valid history is, i.e. removing history for missing classes and methods, and only retaining history that is referenced from current methods (which eliminates branches that don't lead to the current definition). After all this is the history one wants, and I already provided code to compute the per-method history that leads to the current definition. Why not have your cake and compress it too? > > Marcus > > -- > Marcus Denker -- http://marcusdenker.de > > > -- best, Eliot

March 6, 2012

Re: [Pharo-project] .changes are now 37MBâ¦.

by Camillo Bruni

max, maybe you want to create a test git repos with all the sources back to 1.0 :D and then we should record all changes into a local git-repos :) if you want to got back some versions you simply start fetching the missing objects from a central repos somewhere... so far dreaming a bit best cami On 2012-03-06, at 20:01, Marcus Denker wrote: > > On Mar 6, 2012, at 7:30 PM, Max Leske wrote: > >> This may be a stupid questionâ¦ (As far as I understand, condensing the changes is not the same) >> Why don't you increment the version number of the sources file to 1.1 and use a fresh changes file? >> > The sources file does not contain the sources and the changes just the changes. (even though one could think so taking > the names into account.) > > The sources file containes the sources of ages ago (1.0), the .changes is "on top of that". This means that the .sources > file contains a lot of code from classes and methods that have been removed (like the changes), > as well as old versions. > > The .sources/.changes mechanism merges three responsabilties: > > 1) be a log of all edited code in case of a crash > 2) store the current source of the methods (when you ask a method for code, it reads either from the .sources > (method not changed since 1.0) or from the .changes (method was changed). > 3) provide a history for all code > > And all that in 2 big files that one can only append to because? No idea, actually... > > I personally think that it is a mistake today to merge all these into one such basic mechanism. Especialy if > you look at how complex the code is in the image... it's actually amazing. > > When you do a #condenseSources (which is sadly broken right now), then it generates a .sources > file with just the code in the current image. It would be interesting to see how large that is. > > Marcus > > -- > Marcus Denker -- http://marcusdenker.de > >

March 6, 2012

Re: [Pharo-project] .changes are now 37MBâ¦.

by Marcus Denker

On Mar 6, 2012, at 7:30 PM, Max Leske wrote: > This may be a stupid questionâ¦ (As far as I understand, condensing the changes is not the same) > Why don't you increment the version number of the sources file to 1.1 and use a fresh changes file? > The sources file does not contain the sources and the changes just the changes. (even though one could think so taking the names into account.) The sources file containes the sources of ages ago (1.0), the .changes is "on top of that". This means that the .sources file contains a lot of code from classes and methods that have been removed (like the changes), as well as old versions. The .sources/.changes mechanism merges three responsabilties: 1) be a log of all edited code in case of a crash 2) store the current source of the methods (when you ask a method for code, it reads either from the .sources (method not changed since 1.0) or from the .changes (method was changed). 3) provide a history for all code And all that in 2 big files that one can only append to because? No idea, actually... I personally think that it is a mistake today to merge all these into one such basic mechanism. Especialy if you look at how complex the code is in the image... it's actually amazing. When you do a #condenseSources (which is sadly broken right now), then it generates a .sources file with just the code in the current image. It would be interesting to see how large that is. Marcus -- Marcus Denker -- http://marcusdenker.de

March 6, 2012

Re: [Pharo-project] .changes are now 37MBâ¦.

by Max Leske

This may be a stupid questionâ¦ (As far as I understand, condensing the changes is not the same) Why don't you increment the version number of the sources file to 1.1 and use a fresh changes file? Max On 06.03.2012, at 18:42, Philippe Marschall wrote: > On 06.03.2012 09:44, Marcus Denker wrote: >> ... what do we do? >> >> http://code.google.com/p/pharo/issues/detail?id=4843 >> (but that will not help that much. And someone would need to look at the tracker entry and make >> it ready for integration). >> >> or just do a condense? The history is in Monticello. > > Do people actually really rely on the .changes or just say they do? I'd be fine with just a condense. > > Cheers > Philippe > >

March 6, 2012

Re: [Pharo-project] Fuel - class loading issue when superclass changed inst vars

by Eliot Miranda

On Tue, Mar 6, 2012 at 3:05 AM, Henrik Johansen < henrik.s.johansen(a)veloxit.no> wrote: > > On Mar 6, 2012, at 12:03 PM, Henrik Johansen wrote: > > >> > > > > Parcels > > - Check a hash based on class layout vs equivalent hash for the stored > method's class (stored with the method). > > If different: > > 1) check for existence of a backwards-compatability reader method, hand > the old instance to it, and expect the old instance to be #become'd into > something current, > > 2) raise an error if no such method exists. > > > > Now, the main problem with this scheme is it's left as an exercise to > the user to come up with a procedure to ensure said backwards-compat reader > method stays up to date as additional changes are made. > > > > Is it a big problem? > > Depends on your tests, and programmer diligence at any given day. > > > > Added inst vars are not a big deal, as if you forget them, there will > usually be a nil #DNU somewhere down the line. (or you use lazy > initialization, and everything works as expected even for existing > instances) > > > > Removals/reorderings are a bigger issue, as the detection of failure > (inst vars in wrong slots) are often far removed from the source of > problems (forgetting to update the reader method), > > If seldomly used, it may even go unnoticed until the instance is next > saved, at which time you're in real trouble (especially if saved alongside > newly created ones, in which case there is no consistency). > > > > In general I think it's an ok scheme, but would like to see a solution > more resilient to user-error. > > How to achieve that is an interesting topic, which I haven't yet found > time to think through as thoroughly as I had wished/intended some months > ago. :( > > Errr, BOSS, not Parcels. > Right. What Parcels do is shape-change instances *and* rescan methods to fix up inst var offsets. When a parcel is saved the "signature" of classes of instances there-in are saved so that the parcel contains all the inst var names of the class and its superclass. If on load the superclass chain has a different set of inst vars then lost inst vars are omitted and added inst vars are nil in the materialized instances. If a Parcel contains a full class (not just instances) that has methods, and the superclass chains inst vars have changed then these methods are "rescanned" (disassembled, and reassembled, not using the compiler) and inst var offsets are fixed up, missing inst vars getting changed into undeclared variable references (I think; at least this is what *should* happen; its what happens when one removes an inst var that is still referenced from methods). What's missing in the parcel scheme is any hook to allow the user to process shape-changing instances with the values of lost inst vars in hand. Presumably there will be cases when those values are essential to any schema migration. Personally I would want to decouple schema migration and add it as a post-processing step, which would imply that the materializer would offer a service (materialize in a special mode) where it constructed a dictionary from materialized and shape-changed instance to in parcel state (e.g. an Array of inst var values as they occurred in the parcel). Then the materialized instance could be migrated after the fact, with the default behaviour being analogous to what the system does now on class redefinition (values of deleted inst vars are lost, aded inst vars are nil). This reminds me of a bug with the ClassBuilder/RefactoringBrowser combination. If one refactors pushing an inst var up to a superclass or down to subclasses, the values of te inst var in instances are lost because the class change is in fact done as two changes, a deletion followed by an addition. This again could be fixed in a wrapper, building a dictionary from instance to values, performing the set of class changes, and then restoring the inst var state from the dictionary. > My bad. > > Cheers, > Henry > > > -- best, Eliot

March 6, 2012

Re: [Pharo-project] .changes are now 37MBâ¦.

by Philippe Marschall

On 06.03.2012 09:44, Marcus Denker wrote: > ... what do we do? > > http://code.google.com/p/pharo/issues/detail?id=4843 > (but that will not help that much. And someone would need to look at the tracker entry and make > it ready for integration). > > or just do a condense? The history is in Monticello. Do people actually really rely on the .changes or just say they do? I'd be fine with just a condense. Cheers Philippe

March 6, 2012

Re: [Pharo-project] Nautilus feedback

by Benjamin

It's strange ... Can you quickly list the version of each Nautilus package you use ? Thanks in advance, Ben On Mar 6, 2012, at 4:27 PM, Alexandre Bergel wrote: > Really, if I press cmd-b it opens the old system browser (the release of yesterday). > > Alexandre > > > On 6 Mar 2012, at 04:33, Tudor Girba wrote: > >> Indeed. The Moose image already has Nautilus with the default behavior. >> >> Cheers, >> Doru >> >> >> On 5 Mar 2012, at 22:34, Mariano Martinez Peck wrote: >> >>> >>> >>> On Mon, Mar 5, 2012 at 8:35 PM, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: >>> Why should I do this manually? >>> I tried 'Choose new default browser' and 'Register this Browser as default', it does not work for me (I have a 14341). Is there something incorrect I am doing? >>> >>> >>> If you use the latest Nautilus from the correct repository (ss3), then cmd+b should open Nautilus BY DEFAULT. But that was added by ben quite recently. So, be sure to be using the latest version. I saw a problem you were downloading from ss...so that will be outdated. >>> >>> cheers >>> >>> >>> Alexandre >>> >>> >>> On 5 Mar 2012, at 16:28, Benjamin wrote: >>> >>>> Why don't you set Nautilus as the default browser ? >>>> (And should be by default) >>>> >>>> >>>> Ben >>>> >>>> On Mar 5, 2012, at 8:16 PM, Alexandre Bergel wrote: >>>> >>>>>>> - Cmd-b does not open a Nautilus, but a standard browser >>>>>> >>>>>> It opens the default browser :) >>>>> >>>>> Is there a way to change this? This almost prevent me from using Nautilus. I use cmd-b a lot actually. >>>>> OB does it pretty well... >>>>> >>>>> Cheers, >>>>> Alexandre >>>>> >>>>> >>>>> -- >>>>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >>>>> Alexandre Bergel http://www.bergel.eu >>>>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>> >>>> >>> >>> -- >>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >>> Alexandre Bergel http://www.bergel.eu >>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >>> >>> >>> >>> >>> >>> >>> >>> >>> >>> -- >>> Mariano >>> http://marianopeck.wordpress.com >>> >> >> -- >> www.tudorgirba.com >> >> "Problem solving efficiency grows with the abstractness level of problem understanding." >> >> >> >> > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > >

March 6, 2012