Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Hi Trevor, let's not invite Charles Dickens and William Shakespeare to this discussion, I guess they really don't care :) Trevor Vartanoff t...@codepuzzles.org writes: I'm glad to see an exception was made for node property, but that's only one of many, many problem cases. We will be able to examinate those problems if you state them as clearly and concisely as possible, one by one. Thanks in advance for that. Chances are that Org gave you a certain degree of freedom that you still have with Emacs in general. For example, you can still transpose lines within a paragraph with C-x C-t. Best, -- Bastien
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Nicolas, Org has its own definition for a paragraph, which, apparently, doesn't match yours. A paragraph ends either at a blank line, at the end of the buffer, or at the start of another non-paragraph element. In particular, indentation is unrelated to paragraph boundaries. If your definition of a paragraph excludes every single publication of fiction and nonfiction in the history of written language, you may want to rethink your definition. I think Charles Dickens knew what a paragraph boundary was. Actually, it's worse than that: even if you agree that everyone using a computer should now separate all paragraphs with a blank line, it still means that for any form of writing with closely packed separate lines, such as song lyrics, poetry, Shakespeare plays, or even basic lists of todo items, org-mode no longer lets you shift the lines around. I propose we implement an org-property value to decide which definition of element org-metaup should use. I'm glad to see an exception was made for node property, but that's only one of many, many problem cases. Regards, Trevor Vartanoff
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Hello Trevor, On 26 September 2012 19:18, Trevor Vartanoff t...@codepuzzles.org wrote: Nicolas, Org has its own definition for a paragraph, which, apparently, doesn't match yours. A paragraph ends either at a blank line, at the end of the buffer, or at the start of another non-paragraph element. In particular, indentation is unrelated to paragraph boundaries. If your definition of a paragraph excludes every single publication of fiction and nonfiction in the history of written language, you may want to rethink your definition. I think Charles Dickens knew what a paragraph boundary was. LaTeX uses the same paragraph definition, it splits on blank lines. So does the old exporter. By default if there is not a blank line they treat it as a continuation of the previous. If you treat it otherwise how do you tell the difference between a manual break and wrapped text that went to the next line? Actually, it's worse than that: even if you agree that everyone using a computer should now separate all paragraphs with a blank line, it still means that for any form of writing with closely packed separate lines, such as song lyrics, poetry, Shakespeare plays, or even basic lists of todo items, org-mode no longer lets you shift the lines around. Basic lists of todo items will still work fine, Org treats list items as their own elements and can be moved using org-metaup/down. It will also move paragraphs/verses for poetry and theater. =transpose-lines= ( C-x C-t ) can be used to move the individual lines around individually. I propose we implement an org-property value to decide which definition of element org-metaup should use. I'm glad to see an exception was made for node property, but that's only one of many, many problem cases. Regards, Trevor Vartanoff Regards, Jon
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Jonathan, I should have been more clear on what I meant by basic, quite idiotic of me and sorry for the confusion. If you write ten words on ten lines, you can't shift things around anymore until you've put everything into an Official Orgmode Endorsed Structure of some kind. Sometimes I don't need to bother. Sometimes while I'm creating the list, I don't decide until later whether to make it headings or list, and I need to be able to shift its items around first. Org-mode successfully taught me that I can move _everything_ around. I've happily adapted to this freedom, and now it wants to take it away. So, I apologize for my whining, but to me it it is as unexpected as if I updated Emacs and the macros stopped working.
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Hello, Trevor Vartanoff t...@codepuzzles.org writes: I suppose people who only write and handle text with Internet-style line breaks between paragraphs won't notice any change. Those who don't will either get cannot messages, or they will be quite surprised when org scans down, down, down for the next line break and ends up shifting 20 paragraphs as one element. Org has its own definition for a paragraph, which, apparently, doesn't match yours. A paragraph ends either at a blank line, at the end of the buffer, or at the start of another non-paragraph element. In particular, indentation is unrelated to paragraph boundaries. Hence, there is no such thing as 20 paragraphs as one element. A paragraph is always a single element. Regards, -- Nicolas Goaziou
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Hello, Bastien b...@altern.org writes: Hi Anthony, Anthony Lander anth...@landerfamily.ca writes: I use the M-up/dn behaviour a lot to move property lines up and down. Maybe each property line could be a new element recognized as such by org-element.el. This way org-metaup/down on a property line would move it up/down. (Maybe there are some other benefits from having such an element, but I'll trust Nicolas on this.) Nicolas, what do you think? I'm fine with this change. How would that new element be named? `headline-property'? `property'? Also, do we want this change in 7.9.3? In that case, it may be wise to delay its release by a couple of days. Regards, -- Nicolas Goaziou
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Hi Nicolas, Nicolas Goaziou n.goaz...@gmail.com writes: How would that new element be named? `headline-property'? `property'? `property' is fine, unless it is semantically too close to other elements and/or objects. Also, do we want this change in 7.9.3? In that case, it may be wise to delay its release by a couple of days. Having this for Org 8.0 is okay, lets commit this in master. Thanks! Ps: the next minor release will be 7.9.2. -- Bastien
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
On 23.9.2012, at 10:25, Nicolas Goaziou wrote: Hello, Bastien b...@altern.org writes: Hi Anthony, Anthony Lander anth...@landerfamily.ca writes: I use the M-up/dn behaviour a lot to move property lines up and down. Maybe each property line could be a new element recognized as such by org-element.el. This way org-metaup/down on a property line would move it up/down. (Maybe there are some other benefits from having such an element, but I'll trust Nicolas on this.) Nicolas, what do you think? I'm fine with this change. How would that new element be named? `headline-property'? `property'? node-property is another possibility. - Carsten Also, do we want this change in 7.9.3? In that case, it may be wise to delay its release by a couple of days. Regards, -- Nicolas Goaziou
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Hello, Carsten Dominik carsten.domi...@gmail.com writes: node-property is another possibility. There's now a new element type in master: node-property. Regards, -- Nicolas Goaziou
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Nicolas Goaziou n.goaz...@gmail.com writes: There's now a new element type in master: node-property. Fantastic, thanks! -- Bastien
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Hi Bastien, On 12-Sep-22, at 5:24 AM, Bastien wrote: Hi Anthony, Anthony Lander anth...@landerfamily.ca writes: I use the M-up/dn behaviour a lot to move property lines up and down. Maybe each property line could be a new element recognized as such by org-element.el. This way org-metaup/down on a property line would move it up/down. That would be wonderful - and I see that it will be in 8.0! Thank you Nicolas for that. -Anthony
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Hi Anthony, Anthony Lander anth...@landerfamily.ca writes: I use the M-up/dn behaviour a lot to move property lines up and down. Maybe each property line could be a new element recognized as such by org-element.el. This way org-metaup/down on a property line would move it up/down. (Maybe there are some other benefits from having such an element, but I'll trust Nicolas on this.) Nicolas, what do you think? -- Bastien
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Thanks Anthony. Looks like that will be helpful if I ever update.
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Hi Trevor,
Trevor Vartanoff t...@codepuzzles.org writes:
I updated to 7.9.1 from 7.8.11 and was quite surprised to receive nasty
Cannot drag element backward messages when I tried to use org-metaup and
org-metadown to move text around.
It is hard to understand the problem without an example.
When did you get the message and why was it not appropriate?
Why was this done?
The origin of the org-meta{up/down} feature was to be able to move table
lines one by one.
By a small generalization, org-meta{up/down} also moved lines up/down
outside of tables.
The idea behind the current behavior is to think in terms of elements,
not in terms of lines -- which is good IMHO. (A line is an element of a
table, so moving line by line here is the right thing to do.)
How do I get back to one stroke functionality? I'm
afraid I'm not properly understanding the release notes or mailing list
archives.
Please explain the problem a bit more. Thanks!
--
Bastien
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Your explanation is useful. After some more experimenting, it looks like 7.8.11 accepts a return as the dividing line between elements, but 7.9.1 only recognizes them as separate if there's an empty line between them. Example: if I copy the first two lines of your email, I can shift It is hard to understand... to be after When did you get the message..., but not if I delete the empty line between them. And so, example 2, I was receiving cannot drag element forward/backward when I was in a heading with no line breaks. That is, a heading where paragraphs are separated with return + indent rather than a full line break of return + return. (Behavior is identical when I start with an empty .emacs, for the record) I suppose people who only write and handle text with Internet-style line breaks between paragraphs won't notice any change. Those who don't will either get cannot messages, or they will be quite surprised when org scans down, down, down for the next line break and ends up shifting 20 paragraphs as one element.
Re: [O] org-metaup / org-metadown nerfed in 7.9.1
Hi Trevor, On 12-Sep-19, at 10:44 AM, Trevor Vartanoff wrote: And so, example 2, I was receiving cannot drag element forward/backward when I was in a heading with no line breaks. That is, a heading where paragraphs are separated with return + indent rather than a full line break of return + return. I use the M-up/dn behaviour a lot to move property lines up and down. I added some hooks to restore that functionality since property lines aren't recognized as elements. Here's the code. You might be able to adapt it to do what you want. Hope it helps, -Anthony ; Fix M-up and M-down to move individual lines inside a property drawer (add-hook 'org-metaup-hook 'my-org-metaup-hook) (defun my-org-metaup-hook () When on a property line, use M-up to move the line up (when (org-region-active-p) (return nil)) (let ((element (car (org-element-at-point (if (eq element 'property-drawer) (let* ((position (- (point) (line-beginning-position))) (a (save-excursion (move-beginning-of-line 1) (point))) (b (save-excursion (move-end-of-line 1) (point))) (c (save-excursion (goto-char a) (move-beginning-of-line 0))) (d (save-excursion (goto-char a) (move-end-of-line 0) (point (transpose-regions a b c d) (goto-char c) (forward-char position) t) nil))) (add-hook 'org-metadown-hook 'my-org-metadown-hook) (defun my-org-metadown-hook () When on a property line, use M-down to move the line down (when (org-region-active-p) (return nil)) (let ((element (car (org-element-at-point (if (eq element 'property-drawer) (let* ((position (- (point) (line-beginning-position))) (at-end-of-line (eq (point) (line-end-position))) (a (save-excursion (move-beginning-of-line 1))) (b (save-excursion (move-end-of-line 1) (point))) (c (save-excursion (next-line) (move-beginning-of-line 1))) (d (save-excursion (next-line) (move-end-of-line 1) (point (transpose-regions a b c d) (move-beginning-of-line 1) (if at-end-of-line; Strange boundary condition at end of line (progn (next-line) ; Goes to wrong place. So instead just (move-end-of-line 1)) ; go to end of line. (forward-char position)) t) nil)))
