From: Kazutoshi S. <k_s...@f2...> - 2009-04-25 15:58:42
|
Hi Matthieu, >From the recent thread "does anyone use view-scope buffersets?": Shlomy Reinstein wrote: (and I wrote the quoted part) >> I'm also tempted to drop the option to select the contents of new >> BufferSet. I think copied is good for new EditPane (splitting), empty is >> good for new View. Selecting one for both looks wrong and users hardly >> know when a new BufferSet is created. > > I can't give any feedback here. I think it was Matthieu who > implemented these things in the first place, I strongly recommend to > get his opinion on this before removing any such feature. I don't > think he added these features without a good reason, as they required > quite a lot of changes and maintenance. What do you think about dropping the option to select the contents of new BufferSet? -- k_satoda |
From: Alan E. <ala...@gm...> - 2009-04-25 17:10:27
|
I like the option of deciding what goes in a new bufferset. But if you think an additional option of "new for Views, copy for EditPanes", that is a good idea. And we could make that the default. I really don't like the idea of taking away functionality from the BufferSets API. We can hide complexity from the user through the creative use of options and defaults. On Sat, Apr 25, 2009 at 8:58 AM, Kazutoshi Satoda <k_s...@f2...>wrote: > Hi Matthieu, > > >From the recent thread "does anyone use view-scope buffersets?": > Shlomy Reinstein wrote: (and I wrote the quoted part) > >> I'm also tempted to drop the option to select the contents of new > >> BufferSet. I think copied is good for new EditPane (splitting), empty is > >> good for new View. Selecting one for both looks wrong and users hardly > >> know when a new BufferSet is created. > > > > I can't give any feedback here. I think it was Matthieu who > > implemented these things in the first place, I strongly recommend to > > get his opinion on this before removing any such feature. I don't > > think he added these features without a good reason, as they required > > quite a lot of changes and maintenance. > > What do you think about dropping the option to select the contents of new > BufferSet? > > -- > k_satoda > > > ------------------------------------------------------------------------------ > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > -- > ----------------------------------------------- > jEdit Users' List > jEd...@li... > https://lists.sourceforge.net/lists/listinfo/jedit-users > |
From: Kazutoshi S. <k_s...@f2...> - 2009-04-25 17:58:26
|
Alan Ezust wrote: > I like the option of deciding what goes in a new bufferset. > But if you think an additional option of "new for Views, copy for > EditPanes", that is a good idea. And we could make that the default. Please let me know what is the actual good of having the option. Just saying "I like" doesn't make sense against the proposal from a person who dislike it for a concrete reason. > I really don't like the idea of taking away functionality from the > BufferSets API. We can hide complexity from the user through the creative > use of options and defaults. Again, please let me know the reason why you don't like the idea. What is actually bad? I think having or even *creating* unnecessary complexity is obviously bad. We, as software developers, are all fighting against complexity, don't you? This is why "creeping featurism" are said as a disease, not as a evolution. -- k_satoda |
From: Shlomy R. <sre...@gm...> - 2009-04-25 18:09:16
|
On Sat, Apr 25, 2009 at 7:58 PM, Kazutoshi Satoda <k_s...@f2...> wrote: > Alan Ezust wrote: >> I like the option of deciding what goes in a new bufferset. >> But if you think an additional option of "new for Views, copy for >> EditPanes", that is a good idea. And we could make that the default. > > Please let me know what is the actual good of having the option. > > Just saying "I like" doesn't make sense against the proposal from a > person who dislike it for a concrete reason. > >> I really don't like the idea of taking away functionality from the >> BufferSets API. We can hide complexity from the user through the creative >> use of options and defaults. > > Again, please let me know the reason why you don't like the idea. What > is actually bad? > > I think having or even *creating* unnecessary complexity is obviously > bad. We, as software developers, are all fighting against complexity, > don't you? This is why "creeping featurism" are said as a disease, not > as a evolution. While this is true, it is MUCH better mentioned before the implementation, and not after the features have been implementation and undergone multiple phases of design & bug fixing... Of course, if this is really redundant complexity, then it should be removed even after a lot of effort has been put into it. But chances are, once it's released, that it can no longer be considered 'redundant'. > > -- > k_satoda > > ------------------------------------------------------------------------------ > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > -- > ----------------------------------------------- > jEdit Users' List > jEd...@li... > https://lists.sourceforge.net/lists/listinfo/jedit-users > |
From: Kazutoshi S. <k_s...@f2...> - 2009-04-25 18:37:45
|
Shlomy Reinstein wrote: >> I think having or even *creating* unnecessary complexity is obviously >> bad. We, as software developers, are all fighting against complexity, >> don't you? This is why "creeping featurism" are said as a disease, not >> as a evolution. > > While this is true, it is MUCH better mentioned before the > implementation, and not after the features have been implementation > and undergone multiple phases of design & bug fixing... Of course, if > this is really redundant complexity, then it should be removed even > after a lot of effort has been put into it. But chances are, once it's > released, that it can no longer be considered 'redundant'. I agree. But I could do nothing before the implementation for this option and BufferSet feature since the implementation comes before (without) design proposal or patch (, at least for me). To avoid this kind of unfortunate, I ask (and have asked some times) to show the reason (justification) of each change in log message. If one can't do that, then the one shouldn't commit the change, and should ask on the mailing list about the change, or post a patch to the tracker. -- k_satoda |
From: Andreas W. <wil...@fr...> - 2009-04-26 00:08:57
|
On Sun, 26 Apr 2009 03:37:42 +0900 Kazutoshi Satoda <k_s...@f2...> wrote: > To avoid this kind of unfortunate, I ask (and have asked some times) > to show the reason (justification) of each change in log message. If > one can't do that, then the one shouldn't commit the change, and > should ask on the mailing list about the change, or post a patch to > the tracker. Fair enough. I always thought that this uncoordinated coding will bite you sooner or later. (Sorry for mouthing off, but I saw it coming). Anyways, don't you think that this is now pretty much devel-stuff that you should resolve on the devel list? (Don't take this the wrong way, I appreciate how seriously you take these things!) /W |
From: Kazutoshi S. <k_s...@f2...> - 2009-04-26 04:11:25
|
Andreas Waldenburger wrote: > Anyways, don't you think that this is now pretty much devel-stuff that > you should resolve on the devel list? That's right. Thanks. I'll move the general issue to jedit-devel if it become divergent from the original question of this thread. -- k_satoda |
From: Matthieu C. <cho...@gm...> - 2009-04-26 09:26:22
|
Hi, I agree, in fact I have two screens so I usually don't use split view except when doing dual diff, and I use two views instead. And I prefers to have the same buffers in my new view when I create it. The problem is that it seems some people prefers new view with no buffer in their bufferSet so I suppose an option is necessary. About dropping the View BufferSet, I have no idea, I use EditPane BufferSet with multiple view but I don't split ... Matthieu 2009/4/25 Alan Ezust <ala...@gm...> > I like the option of deciding what goes in a new bufferset. > But if you think an additional option of "new for Views, copy for > EditPanes", that is a good idea. And we could make that the default. > > I really don't like the idea of taking away functionality from the > BufferSets API. We can hide complexity from the user through the creative > use of options and defaults. > > > > On Sat, Apr 25, 2009 at 8:58 AM, Kazutoshi Satoda <k_s...@f2...>wrote: > >> Hi Matthieu, >> >> >From the recent thread "does anyone use view-scope buffersets?": >> Shlomy Reinstein wrote: (and I wrote the quoted part) >> >> I'm also tempted to drop the option to select the contents of new >> >> BufferSet. I think copied is good for new EditPane (splitting), empty >> is >> >> good for new View. Selecting one for both looks wrong and users hardly >> >> know when a new BufferSet is created. >> > >> > I can't give any feedback here. I think it was Matthieu who >> > implemented these things in the first place, I strongly recommend to >> > get his opinion on this before removing any such feature. I don't >> > think he added these features without a good reason, as they required >> > quite a lot of changes and maintenance. >> >> What do you think about dropping the option to select the contents of new >> BufferSet? >> >> -- >> k_satoda >> >> >> ------------------------------------------------------------------------------ >> Crystal Reports - New Free Runtime and 30 Day Trial >> Check out the new simplified licensign option that enables unlimited >> royalty-free distribution of the report engine for externally facing >> server and web deployment. >> http://p.sf.net/sfu/businessobjects >> -- >> ----------------------------------------------- >> jEdit Users' List >> jEd...@li... >> https://lists.sourceforge.net/lists/listinfo/jedit-users >> > > |
From: Kazutoshi S. <k_s...@f2...> - 2009-04-26 16:37:05
|
Matthieu Casanova wrote: > Hi, I agree, in fact I have two screens so I usually don't use split view > except when doing dual diff, and I use two views instead. > And I prefers to have the same buffers in my new view when I create it. > The problem is that it seems some people prefers new view with no buffer in > their bufferSet so I suppose an option is necessary. It sounds that the option was made to configure how a new BufferSet is created for new View, not for new EditPane (splitting). Right? I'm still wondering what was "current buffer only" actually for. Could you provide the purpose of this option at the time it was created? I can imagine that this is something for new EditPane (splitting). I'm now thinking about separating this option into two; one for new View which is "copy" or "empty", one for splitting which is "copy" or "current buffer only". Can we agree with this? -- k_satoda |
From: Alan E. <ala...@gm...> - 2009-04-26 16:49:57
|
Kazutoshi, if you use view scope buffersets, and have "new buffersets contain : an empty bufferset" then you get effectively, what you want which is that "new views have an empty buffersets, but split panes have a copy". My idea of "view scope" buffersets was to make that behavior possible, as well as having the buffersets of the splitpanes identical. Thinking about it some more, this might be desireable to some people. Even to me. So I do not want to remove view-scope buffersets. Which means that the Bufferset must have a member that indicates what scope it is. So kazutoshi, if you want to add an extra item to the combobox for "new buffersets contain" and have it be "empty for new views, or copy for new splits", that's fine with me. But I do not want to remove features from jEdit. I don't think it is too complicated. In fact, since I am the one who thought up the feature, I think it is elegant. Furthermore, the options can be "set once and forgotten about", so that it is not necessary to switch bufferset scopes if you don't want to, nor is it even necessary to fiddle with the options if you don't care about them. But what I like about jEdit is its configurability. Please don't take away choices for how BufferSets can work. You can add choices, and options, but it is really a waste of everyone's time to be taking away things that other developers put into the product. Your time could be better spent reworking our thread pool implementation or fixing the scoll code. That's my opinion. So kazutoshi, can you please move the "buffer set scope" member out of the EditPane class and back into the BufferSet class? I'd like to finish the implementation of my exclusive buffersets. On Sun, Apr 26, 2009 at 9:36 AM, Kazutoshi Satoda <k_s...@f2...>wrote: > Matthieu Casanova wrote: > >> Hi, I agree, in fact I have two screens so I usually don't use split view >> except when doing dual diff, and I use two views instead. >> And I prefers to have the same buffers in my new view when I create it. >> The problem is that it seems some people prefers new view with no buffer >> in >> their bufferSet so I suppose an option is necessary. >> > > It sounds that the option was made to configure how a new BufferSet is > created for new View, not for new EditPane (splitting). Right? > > I'm still wondering what was "current buffer only" actually for. Could > you provide the purpose of this option at the time it was created? I can > imagine that this is something for new EditPane (splitting). > > I'm now thinking about separating this option into two; one for new View > which is "copy" or "empty", one for splitting which is "copy" or > "current buffer only". Can we agree with this? > > -- > k_satoda > |
From: Kazutoshi S. <k_s...@f2...> - 2009-04-26 18:15:07
|
Alan Ezust wrote: > Kazutoshi, if you use view scope buffersets, and have "new buffersets > contain : an empty bufferset" then you get effectively, what you want which > is that "new views have an empty buffersets, but split panes have a copy". I know that. But I use EditPane-scope, to allow each pane to have different set of buffers. The conflict looks coming from sharing one configuration for two different things. > My idea of "view scope" buffersets was to make that behavior possible, as > well as having the buffersets of the splitpanes identical. > > Thinking about it some more, this might be desireable to some people. Even > to me. So I do not want to remove view-scope buffersets. (though this issue might have to go on the other thread) What is your purpose of keeping buffersets of the splitpanes identical? Shlomy taught me, at the other thread, he want it to access a newly opened buffer in both panes. But I think it can be reasonable to do it with a explicit way, like the "Buffer Switcher" macro. However, now I realized that having View-scope is a way to provide easy way to access all opened buffers. Did you mean this? > Which means that the Bufferset must have a member that indicates what scope > it is. I couldn't understand this part. How does the existence of view-scope buffersets mean the placement of the scope variable? > So kazutoshi, if you want to add an extra item to the combobox for "new > buffersets contain" and have it be "empty for new views, or copy for new > splits", that's fine with me. But I do not want to remove features from > jEdit. Now I'm not going to drop the option, but thinking to separate it (see my previous reply to Matthieu). Do you have any opinion against that? > So kazutoshi, can you please move the "buffer set scope" member out of the > EditPane class and back into the BufferSet class? I'd like to finish the > implementation of my exclusive buffersets. As I noted above, I don't see any problem of the current placement of the scope variable. What is the actual problem to use it? -- k_satoda |
From: Andreas W. <wil...@fr...> - 2009-04-26 18:30:34
|
On Mon, 27 Apr 2009 03:14:52 +0900 Kazutoshi Satoda <k_s...@f2...> wrote: > [snip] > As I noted above, I don't see any problem of the current placement of > the scope variable. What is the actual problem to use it? > Tss ... devel-talk. Wasn't there some sort of mailing list for that? ;) /W |
From: Kazutoshi S. <k_s...@f2...> - 2009-04-26 19:05:00
|
Andreas Waldenburger wrote: > On Mon, 27 Apr 2009 03:14:52 +0900 Kazutoshi Satoda > <k_s...@f2...> wrote: >> As I noted above, I don't see any problem of the current placement of >> the scope variable. What is the actual problem to use it? > > Tss ... devel-talk. Wasn't there some sort of mailing list for that? ;) Oops, sorry again. I'll take more care about this. -- k_satoda |
From: Alan E. <ala...@gm...> - 2009-04-26 18:56:29
|
> > So kazutoshi, if you want to add an extra item to the combobox for "new >> buffersets contain" and have it be "empty for new views, or copy for new >> splits", that's fine with me. But I do not want to remove features from >> jEdit. >> > > Now I'm not going to drop the option, but thinking to separate it (see my > previous reply to Matthieu). Do you have any opinion against that? > That's even better. New View Buffersets contain: and new EditPane buffersets contain: as 2 separate options. > > > So kazutoshi, can you please move the "buffer set scope" member out of the >> EditPane class and back into the BufferSet class? I'd like to finish the >> implementation of my exclusive buffersets. >> > > As I noted above, I don't see any problem of the current placement of the > scope variable. What is the actual problem to use it? > Because I need to know, given a bufferset of an editpane, what its scope is. It might be a View scope, or it might be Global. So either we need an enum member, or a reference back to the object (View or EditPane) that this is the scope for. |
From: Kazutoshi S. <k_s...@f2...> - 2009-04-26 19:21:21
|
Alan Ezust wrote: >> Now I'm not going to drop the option, but thinking to separate it (see my >> previous reply to Matthieu). Do you have any opinion against that? > > That's even better. New View Buffersets contain: and new EditPane buffersets > contain: > as 2 separate options. OK. Thanks. >>> So kazutoshi, can you please move the "buffer set scope" member out of the >>> EditPane class and back into the BufferSet class? This is not an issue for users, then moving to jedit-devel... -- k_satoda |
From: Vampire (jEdit) <Vam...@gm...> - 2009-04-27 22:42:27
|
Sorry, my previous post here and to the other thread was without receiving the latest messages on the list. That was why I more or less suggested what you already agreed on. :-) But I would suggest offering for both options all three choices we have in the one choice currently. It is a valid wish to open the split with an empty buffer-set and also to open a new view with only the current buffer in the set. Regards Vampire Kazutoshi Satoda schrieb: > Alan Ezust wrote: > >>> Now I'm not going to drop the option, but thinking to separate it (see my >>> previous reply to Matthieu). Do you have any opinion against that? >>> >> That's even better. New View Buffersets contain: and new EditPane buffersets >> contain: >> as 2 separate options. >> > > OK. Thanks. > > >>>> So kazutoshi, can you please move the "buffer set scope" member out of the >>>> EditPane class and back into the BufferSet class? >>>> > > This is not an issue for users, then moving to jedit-devel... > > |
From: Kazutoshi S. <k_s...@f2...> - 2009-04-29 09:55:06
|
Vampire (jEdit) wrote: > But I would suggest offering for both options all three choices we have > in the one choice currently. It is a valid wish to open the split with > an empty buffer-set and also to open a new view with only the current > buffer in the set. Do you know some concrete motivation for each use case? It sounds that they can be wanted by an user who usually wish to have copied bufferset for split pane. It it is true, having the choice in the global option is not good. Instead, providing some separated actions may be better. For example, "split ..." for copied bufferset, and "open new EditPane" for empty bufferset. -- k_satoda |
From: Vampire (jEdit) <Vam...@gm...> - 2009-04-26 23:18:07
|
I already wrote it to the other thread, so just for reference here again: [snip] I'm also tempted to drop the option to select the contents of new BufferSet. I think copied is good for new EditPane (splitting), empty is good for new View. [/snip] You said it yourself. That is what YOU think. The "problem" is that some users might think that other options are better. The high configurability is one of the major USPs that jEdit has and I really like it. This is one of the main reasons I switched to jEdit as a user looong ago. I think we should have as much configurabilty while having sane defaults. The way you think is in my opinion a sane default. But we should give our users the option to choose what they consider more sane or practical for them. [snip] Selecting one for both looks wrong and users hardly know when a new BufferSet is created. [/snip] You are right. So I would suggest to split this option in two options, one for splitting, one for opening a new view with the defaults you suggested. Additionally make it more clear to the user when these settings are important. This could either be done through better, more meaningful, option names or maybe also tooltips that clarify each setting. Additionally this could and should be described in detail in the user's manual if it is not already. Regards Vampire Matthieu Casanova schrieb: > Hi, I agree, in fact I have two screens so I usually don't use split > view except when doing dual diff, and I use two views instead. > And I prefers to have the same buffers in my new view when I create it. > The problem is that it seems some people prefers new view with no > buffer in their bufferSet so I suppose an option is necessary. > About dropping the View BufferSet, I have no idea, I use EditPane > BufferSet with multiple view but I don't split ... > > Matthieu > > 2009/4/25 Alan Ezust <ala...@gm... <mailto:ala...@gm...>> > > I like the option of deciding what goes in a new bufferset. > But if you think an additional option of "new for Views, copy for > EditPanes", that is a good idea. And we could make that the default. > > I really don't like the idea of taking away functionality from the > BufferSets API. We can hide complexity from the user through the > creative use of options and defaults. > > > > On Sat, Apr 25, 2009 at 8:58 AM, Kazutoshi Satoda > <k_s...@f2... <mailto:k_s...@f2...>> wrote: > > Hi Matthieu, > > >From the recent thread "does anyone use view-scope buffersets?": > Shlomy Reinstein wrote: (and I wrote the quoted part) > >> I'm also tempted to drop the option to select the contents > of new > >> BufferSet. I think copied is good for new EditPane > (splitting), empty is > >> good for new View. Selecting one for both looks wrong and > users hardly > >> know when a new BufferSet is created. > > > > I can't give any feedback here. I think it was Matthieu who > > implemented these things in the first place, I strongly > recommend to > > get his opinion on this before removing any such feature. I > don't > > think he added these features without a good reason, as they > required > > quite a lot of changes and maintenance. > > What do you think about dropping the option to select the > contents of new > BufferSet? > > -- > k_satoda > > ------------------------------------------------------------------------------ > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables > unlimited > royalty-free distribution of the report engine for > externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > -- > ----------------------------------------------- > jEdit Users' List > jEd...@li... > <mailto:jEd...@li...> > https://lists.sourceforge.net/lists/listinfo/jedit-users > > > > ------------------------------------------------------------------------ > > ------------------------------------------------------------------------------ > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects |
From: Shlomy R. <sre...@gm...> - 2009-04-26 16:47:42
|
On Sun, Apr 26, 2009 at 6:36 PM, Kazutoshi Satoda <k_s...@f2...> wrote: > Matthieu Casanova wrote: >> Hi, I agree, in fact I have two screens so I usually don't use split view >> except when doing dual diff, and I use two views instead. >> And I prefers to have the same buffers in my new view when I create it. >> The problem is that it seems some people prefers new view with no buffer in >> their bufferSet so I suppose an option is necessary. > > It sounds that the option was made to configure how a new BufferSet is > created for new View, not for new EditPane (splitting). Right? > > I'm still wondering what was "current buffer only" actually for. Could > you provide the purpose of this option at the time it was created? I can > imagine that this is something for new EditPane (splitting). > > I'm now thinking about separating this option into two; one for new View > which is "copy" or "empty", one for splitting which is "copy" or > "current buffer only". Can we agree with this? While I have no recommendations this way or another, what does "empty" mean? As far as I know, jEdit does not allow a situation where an edit pane has no buffers at all. If all buffers are closed, an untitled one automatically appears. > > -- > k_satoda > > ------------------------------------------------------------------------------ > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > -- > ----------------------------------------------- > jEdit Users' List > jEd...@li... > https://lists.sourceforge.net/lists/listinfo/jedit-users > |
From: Alan E. <ala...@gm...> - 2009-04-26 16:51:25
|
Empty means an untitled buffer. On Sun, Apr 26, 2009 at 9:47 AM, Shlomy Reinstein <sre...@gm...>wrote: > On Sun, Apr 26, 2009 at 6:36 PM, Kazutoshi Satoda > <k_s...@f2...> wrote: > > Matthieu Casanova wrote: > >> Hi, I agree, in fact I have two screens so I usually don't use split > view > >> except when doing dual diff, and I use two views instead. > >> And I prefers to have the same buffers in my new view when I create it. > >> The problem is that it seems some people prefers new view with no buffer > in > >> their bufferSet so I suppose an option is necessary. > > > > It sounds that the option was made to configure how a new BufferSet is > > created for new View, not for new EditPane (splitting). Right? > > > > I'm still wondering what was "current buffer only" actually for. Could > > you provide the purpose of this option at the time it was created? I can > > imagine that this is something for new EditPane (splitting). > > > > I'm now thinking about separating this option into two; one for new View > > which is "copy" or "empty", one for splitting which is "copy" or > > "current buffer only". Can we agree with this? > > While I have no recommendations this way or another, what does "empty" > mean? As far as I know, jEdit does not allow a situation where an edit > pane has no buffers at all. If all buffers are closed, an untitled one > automatically appears. > > > > > -- > > k_satoda > > > > > ------------------------------------------------------------------------------ > > Crystal Reports - New Free Runtime and 30 Day Trial > > Check out the new simplified licensign option that enables unlimited > > royalty-free distribution of the report engine for externally facing > > server and web deployment. > > http://p.sf.net/sfu/businessobjects > > -- > > ----------------------------------------------- > > jEdit Users' List > > jEd...@li... > > https://lists.sourceforge.net/lists/listinfo/jedit-users > > > |