Thread: [Vimprobable-users] Features for next version
Vimprobable is a lean web browser optimised for full keyboard control
Brought to you by:
hanness
From: Daniel C. <dan...@gm...> - 2011-08-13 16:44:11
|
Hi! Cause of the forthcoming release of the version 1.0 I took advantage of the situation to ask if there is a roadmap of features to implement for version 1.1 or 1.0.1. If there isn't such a roadmap, maybe we can discuss it here. Features I want to see implemented: 1) On the project site of uzbl I found a nice JavaScript that can block loading of flash objects and load them on mouse click (http://www.uzbl.org/wiki/flashblock). But at the time we have no possibility to load user specified scripts if a page was opened. I think it could be a nice feature to implemented something like the style.css for user scripts 'script.js' and a configuration variable to allow/disallow the loading of the script file. Is this an over bloated feature or would it be a nice to have for someone else too? 2) Often I find myself searching for something interesting via a searchengine and opening several results into new browser windows, to read them later. Maybe I could configure my windowmanager to open the new windows in the Background, to keep on searchengine's result page until I've opened all the pages I considered as interesting. But I believe it would be a very interesting feature, to have the ability to yank URLs into a list and to open them later - something like the 'u' command to open last closed page or the 'p' or 'P' command to load URL from clipboard. I think we could use a file to yank the URLs into it and have a command to open the first URL from it and pop it from the list. After reading the page or some related page the command can open the next short time bookmarked page. We could use a lowercase letter command to open the next URL from list into current window and the uppercase one, to open it into a new vimprobable instance. An additional command like :open could be implemented to show the entries of the list and to select a URL from it to open. The URL list could prevent users from open many pages only for the reason to mark them for reading later. Daniel |
From: Serge E. H. <se...@ha...> - 2011-08-13 18:49:16
|
Quoting Daniel Carl (dan...@gm...): > Hi! > > Cause of the forthcoming release of the version 1.0 I took advantage of the > 2) Often I find myself searching for something interesting via a searchengine > and opening several results into new browser windows, to read them later. > Maybe I could configure my windowmanager to open the new windows in the > Background, to keep on searchengine's result page until I've opened all the > pages I considered as interesting. But I believe it would be a very > interesting feature, to have the ability to yank URLs into a list and to > open them later - something like the 'u' command to open last closed page > or the 'p' or 'P' command to load URL from clipboard. I like it! Kind of like readitlater. For extra points, a third hint follow option (not sure which keystroke) which just moves the chosen link right into that list without ever opening it. 'Q<followed by f-like hinting>' maybe? -serge |
From: Hannes S. <ha...@yl...> - 2011-08-13 19:14:56
Attachments:
signature.asc
|
Hi! Daniel Carl <dan...@gm...> wrote: > 1) (http://www.uzbl.org/wiki/flashblock) As mentioned in previous discussions, I don't care at all about that - simply because I don't have Flash installed, so I don't need to block it. It seems to be a popular request, though. If it is done, my only request is that it should actually *block* these elements, not just *hide* them. > 2) to have the ability to yank URLs into a list and to open them > later - something like the 'u' command to open last closed page or > the 'p' or 'P' command to load URL from clipboard. This is an excellent idea and it should be very easy to implement. If we get the future hinting policy straightened out, it could even be possible to squeeze this into the 1.0 release. Since we're planning, I'd like to bring up one very general point: the future release/version number policy. As announced already, I would put the version 1 branch into bug-fixing mode. The question is: Should we do the same for the version 2 branch and make future development on a completely new one? The reason I'm proposing this is this: Adding more and more features obviously makes the browser more 'fat' gradually. This might go too far for some people. Continuing development on the same branch would mix up bugfixes and new features, making it "all or nothing" for the users: If they want bugfixes, they also have to take the new features they probably don't even want. On the other hand, starting new branches regularly allows people to either go on following the bleeding edge development or sticking to a trusted, feature-frozen, but nevertheless maintained version. On a related note, there is the question of version numbering. Should we have: - Vimprobable1 1.0 and Vimprobable2 1.0 or - Vimprobable 1.0 and Vimprobable 2.0? What does everybody think? Hannes |
From: Jason R. <jas...@gm...> - 2011-08-13 20:13:27
|
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 14/08/11 07:14, Hannes Schüller wrote: > > As announced already, I would put the version 1 branch into bug-fixing > mode. The question is: Should we do the same for the version 2 branch > and make future development on a completely new one? > > The reason I'm proposing this is this: Adding more and more features > obviously makes the browser more 'fat' gradually. This might go too far > for some people. Continuing development on the same branch would mix up > bugfixes and new features, making it "all or nothing" for the users: > If they want bugfixes, they also have to take the new features they > probably don't even want. On the other hand, starting new branches > regularly allows people to either go on following the bleeding edge > development or sticking to a trusted, feature-frozen, but nevertheless > maintained version. I would favour this approach: allowing those who wish to continue to use the more minimal browser to do so, and giving those who want to access the more featureful version that choice. > > On a related note, there is the question of version numbering. Should > we have: > - Vimprobable1 1.0 and Vimprobable2 1.0 or > - Vimprobable 1.0 and Vimprobable 2.0? The former: the latter makes Vimprobable 2.0 sound like the later release of Vimprobable 1.0 Cheers, /J -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) iQEcBAEBAgAGBQJORtUIAAoJEEReUuqxvU5ARgsIALrwxvkWfi95dbQvdj/r/ktl iMJX1C9m5i+XPUhDcJWE0oW2WEO/1ivGXx0T7bKNeOc174mp8t29e1bfHP3A2xvc KfhQlBxdRPyBYsgrhfgghl6NSovljyX0F+FGfWCTSbp8El66A/e7anAED7ZHwO2X MziiFf4YRSmrd2QylQ2T6kr6xLL6bhbfX0KCVKYu+nV/yGsphiQiGwhn19A96PvI CAbHJVeyaDNLsNYcBu26qun++spcrbLY4JhGSs/kHfqtMHU0D/p7lJp2R+Tm/9Kt jQUCJkkO0Ba17viOPSrA3heDhJkxhE7LJvDjs3JiHVPF1KiXtrqi0Lv6XvinfKI= =4oW8 -----END PGP SIGNATURE----- |
From: Matto F. <ma...@ma...> - 2011-08-14 17:49:57
|
Hi, On Sat, Aug 13, 2011 at 06:43:45PM +0200, Daniel Carl wrote: > 1) On the project site of uzbl I found a nice JavaScript that can block > loading of flash objects and load them on mouse click > (http://www.uzbl.org/wiki/flashblock). But at the time we have no > possibility to load user specified scripts if a page was opened. I think it > could be a nice feature to implemented something like the style.css for > user scripts 'script.js' and a configuration variable to allow/disallow the > loading of the script file. Is this an over bloated feature or would it be > a nice to have for someone else too? I am not a fan of flash and in the day-to-day use of Vimprobable I have never run into the need of something like this. But I wouldn't vote against it if it doesn't require much resources. > 2) Often I find myself searching for something interesting via a searchengine > and opening several results into new browser windows, to read them later. [ ... ] > to show the entries of the list and to select a URL from it to open. The > URL list could prevent users from open many pages only for the reason to > mark them for reading later. What I have been thinking about is a efficient way to shoot the current URL into an external script. Whatever that script does, is up to the user. Perhaps this is a more generic approach, that would allow for something you want but also allow users to come up with much more usefull things. OTOH, I do this now with the usage of the windowmanager. I do a yank of the current URL into the copy-and-paste buffer, and let ratpoison do the work (with "get-selection"). So if this could be done from within ratpoison, perhaps it could also be done from within other windowmanagers ... Cheers, Matto |
From: Daniel C. <dan...@gm...> - 2011-08-14 17:53:30
|
Hi! On Sat, Aug 13, 2011 at 09:14:25PM +0200, Hannes Schüller wrote: > As mentioned in previous discussions, I don't care at all about that - > simply because I don't have Flash installed, so I don't need to block > it. The blocking of flash wasn't the point of interest. It was only an example for running JavaScript from user for every Pageload. > It seems to be a popular request, though. If it is done, my only > request is that it should actually *block* these elements, not just > *hide* them. If I read the script right the flash is blocked. > As announced already, I would put the version 1 branch into bug-fixing > mode. The question is: Should we do the same for the version 2 branch > and make future development on a completely new one? I'm not even sure if it's a good idea to separate vimprobable1 and vimprobable2 into different branches. It make the maintaining a little easier, but I think to apply bugfixex also to the version 1 could be forgotten. I'm not using version 1 and so I apply patches only for version 2. I'm not familar with the version 1, but could we use precompiler flags to let the user decide which version to compile, or would this make the sources to complicate to understand? > The reason I'm proposing this is this: Adding more and more features > obviously makes the browser more 'fat' gradually. This might go too far > for some people. Continuing development on the same branch would mix up > bugfixes and new features, making it "all or nothing" for the users: > If they want bugfixes, they also have to take the new features they > probably don't even want. On the other hand, starting new branches > regularly allows people to either go on following the bleeding edge > development or sticking to a trusted, feature-frozen, but nevertheless > maintained version. Maybe we can follow the vim, that allows the user to compile in different fat versions or to enable/disable every single feature by hand. So we don't have to apply bugfixes on several branches. > On a related note, there is the question of version numbering. Should > we have: > - Vimprobable1 1.0 and Vimprobable2 1.0 or > - Vimprobable 1.0 and Vimprobable 2.0? Could we abandon the deviding of version 1 and 2 and have only one flexible to compile vimprobable? Vimprobable1 could be a vimprobable compiled with tiny option, vimprobable2 compiled with the huge option, or so. Daniel |
From: Hannes S. <ha...@yl...> - 2011-08-14 20:26:07
Attachments:
signature.asc
|
Hi! Daniel Carl <dan...@gm...> wrote: > On Sat, Aug 13, 2011 at 09:14:25PM +0200, Hannes Schüller wrote: > > As announced already, I would put the version 1 branch into > > bug-fixing mode. The question is: Should we do the same for the > > version 2 branch and make future development on a completely new > > one? > > I'm not even sure if it's a good idea to separate vimprobable1 and > vimprobable2 into different branches. It make the maintaining a > little easier, but I think to apply bugfixex also to the version 1 > could be forgotten. I'm not using version 1 and so I apply patches > only for version 2. I'm not familar with the version 1, but could we > use precompiler flags to let the user decide which version to > compile, or would this make the sources to complicate to understand? For once, yes, it would increase source code complexity, yes. Concerning bugfixes, the opposite of what you assume is the case I'd say: different branches make more work. > > On a related note, there is the question of version numbering. > > Should we have: > > - Vimprobable1 1.0 and Vimprobable2 1.0 or > > - Vimprobable 1.0 and Vimprobable 2.0? > > Could we abandon the deviding of version 1 and 2 and have only one > flexible to compile vimprobable? Vimprobable1 could be a vimprobable > compiled with tiny option, vimprobable2 compiled with the huge > option, or so. See above, this would still increase code complexity, making it much less readable, less reliable, more prone to errors and race conditions etc. Hannes |
From: Daniel C. <dan...@gm...> - 2011-08-14 23:53:25
|
Hi! I have some feature from pentadactyl that I would like to have already in vimprobable. The possibility to hit CTRL-I within a textarea to open a configurable editor with the contents from the textarea. On Save the new content will be transfered back into textarea. Yes, I can do this handy, but in this way it's much faster. For the future we could also think about a simple help system that allows the user to open helppages on :help. Daniel |
From: Hannes S. <ha...@yl...> - 2011-08-15 06:16:39
|
Jason Ryan <jas...@gm...> wrote: > > On a related note, there is the question of version numbering. > > Should we have: > > - Vimprobable1 1.0 and Vimprobable2 1.0 or > > - Vimprobable 1.0 and Vimprobable 2.0? > > The former: the latter makes Vimprobable 2.0 sound like the later > release of Vimprobable 1.0 Well, it *is*, isn't it? It's a branch of the original code with stuff added to it. Hannes |
From: Jason R. <jas...@gm...> - 2011-08-15 06:26:11
|
On 15/08/11 at 08:13am, Hannes Schüller wrote: > Jason Ryan <jas...@gm...> wrote: > > > On a related note, there is the question of version numbering. > > > Should we have: > > > - Vimprobable1 1.0 and Vimprobable2 1.0 or > > > - Vimprobable 1.0 and Vimprobable 2.0? > > > > The former: the latter makes Vimprobable 2.0 sound like the later > > release of Vimprobable 1.0 > > Well, it *is*, isn't it? It's a branch of the original code with stuff > added to it. > Ah, I hadn't grasped the nature of the relationship: I thought they were developed in parallel. -- http://jasonwryan.com/ |
From: Hannes S. <ha...@yl...> - 2011-08-15 10:07:23
|
Jason Ryan <jas...@gm...> wrote: > On 15/08/11 at 08:13am, Hannes Sch__ller wrote: > > Jason Ryan <jas...@gm...> wrote: > > > > On a related note, there is the question of version numbering. > > > > Should we have: > > > > - Vimprobable1 1.0 and Vimprobable2 1.0 or > > > > - Vimprobable 1.0 and Vimprobable 2.0? > > > > > > The former: the latter makes Vimprobable 2.0 sound like the later > > > release of Vimprobable 1.0 > > > > Well, it *is*, isn't it? It's a branch of the original code with > > stuff added to it. > > > Ah, I hadn't grasped the nature of the relationship: I thought they > were developed in parallel. Well, yes, this is also true. Vimprobable2 was branched off Vimprobable1. Since then, they have been developed in parallel. However, truely in parallel: Everything has been done in both branches at the same time save for the addition of new features depending on "2" exclusive functions. Hannes |
From: Daniel C. <dan...@gm...> - 2011-08-15 11:21:22
|
Hi! > On a related note, there is the question of version numbering. Should > we have: > - Vimprobable1 1.0 and Vimprobable2 1.0 or > - Vimprobable 1.0 and Vimprobable 2.0? In my opinion vimprobable1 and vimprobable2 are two different things and not only different version of the same programm. I think vimprobable1 should be in a seperate branch with seperate version numbers. Users that like the vimprobable1 can checkout this branch and can fetch bugfixes for it and, maybe, some new fetaures. Vimprobable2/master could be the other branch and have it's own versionnumbering. For the features, it could be useful to have fetature branches (could be prefixed with feature/*), that exists until the feature is merged into the master/vimprobable1. After merging we could remove the featurebranches, to ceep the repository clean. Like mentiones in a pervious post, I would prefer to put all features into the master branch and allo them to be enabled/disabled via precompiler options. For the features the complexity with the switchable features should acceptable. I think, to have only a basic vimprobable2 and several feature branches could make it difficould for users to merge the wished browser together. For every bugfix in featurebranch the user have to merge the sources together. Sometimes merging is hard enought for developers that knew the code well, but for a user who want to build the browser it would be harder. Daniel |
From: Hannes S. <ha...@yl...> - 2011-08-15 12:29:38
|
Hi! Daniel Carl <dan...@gm...> wrote: > > On a related note, there is the question of version numbering. > > Should we have: > > - Vimprobable1 1.0 and Vimprobable2 1.0 or > > - Vimprobable 1.0 and Vimprobable 2.0? > In my opinion vimprobable1 and vimprobable2 are two different things > and not only different version of the same programm. I think > vimprobable1 should be in a seperate branch with seperate version > numbers. Users that like the vimprobable1 can checkout this branch > and can fetch bugfixes for it and, maybe, some new fetaures. > > Vimprobable2/master could be the other branch and have it's own > versionnumbering. > > Like mentiones in a pervious post, I would prefer to put all features > into the master branch and allo them to be enabled/disabled via > precompiler options. So, if we went this way, wouldn't this be "Vimprobable3"? I.e. another new branch separat from the others? I.e.: Doesn't what you're saying actually *support* my suggestion to open up yet another branch to keep Vimprobable2 "cleaner"? Hannes |
From: Daniel C. <dan...@gm...> - 2011-08-15 15:21:25
|
Hi Hannes! On Mon, Aug 15, 2011 at 02:27:20PM +0200, Hannes Schüller wrote: > So, if we went this way, wouldn't this be "Vimprobable3"? I.e. another > new branch separat from the others? I.e.: Doesn't what you're saying > actually *support* my suggestion to open up yet another branch to keep > Vimprobable2 "cleaner"? Do you mean, we should use a seperate branch for every more featureful vimprobable? This could be a way, but we have to maintain 3 different versions, apply bugfixes on them, test them... . To seperate Vimprobable1 is usefull, because it's realy something other then the more featrueful Vimprobable2. But I think it's do hard to have to many branches with seperate versions. I think all future features should be avaiable for Vimprobable2 and be enabled/disabled via config.h Daniel |
From: Ryan M. <rm...@de...> - 2011-08-15 15:48:01
|
Hi, > On a related note, there is the question of version numbering. > Should we have: > - Vimprobable1 1.0 and Vimprobable2 1.0 or > - Vimprobable 1.0 and Vimprobable 2.0? vimprobable1 should be vimprobable-legacy vimprobable2 should be vimprobable All new development should be done as vimprobable, IMHO. Ryan |