fix select text with keyboard not working, regression of 0.99.26 (#1441, #1444, #1448)
updated Dutch (Luuk Geurts)
22 thoughts to “cherrytree 0.99.27 issued”
CHERRYTREE is an excellent software. I have used it for many years. Congratulations!! Thank you very much, Merry Christmas and an excellent new year !!
Thank you merry Christmas and happy 2021 🙂
I love Cherrytree. There’s nothing else as robust and stable at my favorite price of free.
Keep up the great work! Merry Christmas!!!
Thank you Merry Christmas 😉
The new icons have improved as well in 0.99.27!
Cherrytree keeps getting better, thank you very much
Merry Xmas under the cherry.. uh christmas tree 😉 and a healthy 2021!
😁 Thank you merry Christmas and a better 2921!
Changed from old TreePad (registered user) a year ago.
Cherry tree is much more responsible, and stable across OSs.
There is an exception to be faced, thus the following feedback:
On my 4 years old laptop the distribution feels quite light and fast,
Exception is the very useful tables implementation, heavy and slow,
much much slower than in the 8 years old desktop with 0.39.4.
How? Inserting, deleting or pasting row, takes 12 seconds.
Changing from a node to another, takes 2 seconds.
This may be not a priority… yet.
Maybe caused by being dependent from interpretation”.
Maybe caused by the use of smaller buffers than due.
Maybe caused by working a char at a time.
Maybe all of them. Thus mentioning, not to be forgotten.
CherryTree is now my favorite Information manager.
It’s coherent across OS’s, editing made a pleasure.
CherryTree is a true cherry. The nut can wait.
P.S.- The new icons are more visible, but unbalanced.
Rather the older, even if small. Change is not necessarily a progress.
The new one burden the eye, then becoming (as I see them) a bit stressful.
… Not a priority, just a small feedback.
Hi. I’m unsure if this should go here or in your github site. This morning a serious problem arose in my 0.99.27, reproducible both in Arch KDE & SparkyLinux [=~Debian 10 Testing] Xfce. A subnode i edited within the last day or so, apparently ok at the time, now causes a CT instant crash each & every time it is selected in the tree. CT appears to be its usual excellent self everywhere else, but this subnode is now broken. From launching CT in terminal & watching, this is the msg each time it crashes at that subnode:
“`
[2020-12-31 11:58:14.403] [ ] [debug] autosave: no needs to save file
[2020-12-31 11:58:46.108] [ ] [debug] Node name: To Do
Entity: line 16: parser error : PCDATA invalid Char value 27
nstant-Off
^
Entity: line 18: parser error : PCDATA invalid Char value 27
elay 4 Sec.
^
[2020-12-31 11:58:46.116] [gtk] [info]
unhandled exception (type std::exception) in signal handler:
what: Document not well-formed.
Line 18, column 8 (fatal):
PCDATA invalid Char value 27
Trace/breakpoint trap
“`
I am almost certain i know exactly which & where those incompatible characters/symbols are, & if only i could now access that subnode i could delete them… but as described, CT now crashes every time there. Oh, also, i have checked many of the archive/backup CT files, but they either have the same corruption, or are sufficiently old as to predate all this [i cannot merely fall back on these as i would lose too much of my recent data in numerous other nodes].
I know nothing about sqlite databases, but from the Debian repos i installed `sqlitebrowser:amd64 (3.12.1-2` in hope that somehow i might muddle my way to finding the suspect content to delete it. Here’s a picture: https://i.postimg.cc/W4jXJS0w/20201231-142533.png. As you can see, i found the subnode, but have no idea now how to “enter” the database there to edit my data.
Please can you advise how i should proceed now?
Can you send the document privately to me so I can have a look?
Hi again & thanks for your fast reply. Sorry but i can’t share the complete file with anyone because it contains some personal info as well as my tech stuff [the corrupt subnode is in the tech not personal section]. What i can do though, if this would be ok for you, tomorrow is to copy my file, then open the copy in CT & delete ALL nodes, except leaving intact this one subnode & its parent node. How would i send it to you? I’d need to know your email address.
If that’s what we do, & if you find the corruption, as far as i can see though you’d still need to explain to me how i could also find it using `sqlitebrowse` & then fix it myself in my real file, otherwise how could i merge your fixed one with my faulty one?
Look at help–about for my email and if you remember what you made to corrupt the node tell me. I would modify cherrytree itself to at least not crash and then if possible correct the issue itself. To edit from terminal you can use sqlite3 other tools I don’t know
Just a thought. If you can save your whole CT file as XML – not SQL – then you can use an XML editor like XMLCopyEditor to inspect and edit the inner nodes.
Hi D L, thanks for your good suggestion. This possibility excited me, but sadly when i tried it, the instant i clicked on OK with that Save As XML option selected, CT crashed again, before the save could occur. This is repeatable. Though i clearly know nothing of the actual underlying code nor programming, intuitively i believe that the nature of this subnode’s data corruption is such that when ANY CT operation is called which needs to process the entire database rather than only those parts of it not involving that subnode, the crash is triggered. I implicitly further tested that hypothesis by launching CT again in a terminal, then initiating the “Find Into All the Tree Nodes Contents” search. Sadly as i expected, the instant the search reached that previously-identified [in my OP] data in this subnode, CT crashed again. This discovery that now i can no longer do ANY operation that processes the whole database, rather than only that one subnode, alarms me & makes me realise that the implications of this problem now are much more serious than previously i’d supposed.
Hi GIUSPEN, i confess to some disappointment that it seems i can’t get any guidance on how i might use some external tool to interrogate the database, to find then delete the specific incompatible character entries causing this problem. However i like your idea to try to fix CT so that if it encounters this bad character it stops crashing. If you can do that, then all i would need to do is upgrade to that future CT version & my current data would be ok again.
To help you see this specific bad data character i strongly suspect to match that terminal error message in my OP, i have this info for you. In that subnode “To Do” i have a CT Table, & into one row i pasted this following text… first i’ll show you a picture of it, then show you from where you can get it yourself.
The text that is highlighted here is what i had copied & pasted: https://i.postimg.cc/bv8FygKD/20210101-133239.png
I copied it from a pdf document i was viewing in Okular, but this original document can be downloaded from here: https://download1.gigabyte.com/Files/Manual/mb_manual_ga-z97(h97)-hd3_v2.x_e.pdf
You’ll find this text on page 30 of 40, very near the top of that page.
It is those unusual “double arrowheads” directly to the left of the lines beginning “Instant-Off” and “Delay 4 Sec” that i suspect CT does not like, & which trigger the crashes. I remember that when i had pasted them originally into my CT table, they did not render correctly, but changed to some weird complex rectangular shape.
Can you please advise me if this might be sufficient info for you to proceed, or if you need more from me?
Many thanks for all your work on CherryTree. It’s an awesome piece of software. I hope you had a good christmas and new year.
@guigirl it’s enough for now but better raise a GitHub issue next time
@Chris thanks happy new year 😉
Just to finish answering @guigirl .. I find that Hexeditor also searches node text patterns in *.ctb.
Raise issue in github for more discussion as @giuspen requests.
Oh how exasperating – this one also atm fails to d/l the pdf. I presume that Gigabyte must be having some server problem atm — grrrrr.
You told me before how to look up your email address. I shall do that now, then send you an email with my own copy of this pdf attached. I do hope this at least works 🙂
Thanks @DL, Last night i found Hexeditor in the AUR [i use ArchLinux] & installed it. Its UI & how to use it was completely unknown to me, but i worked out how to import [a copy of] my ctb file, open it, & find those two sections of the table. Unfortunately however, my attempts to delete that content, which seemed to go ok within Hexeditor itself, only resulted in a ctb that CT refused even to open, telling me that it was corrupt. Rats. I assume this means i am not using Hexeditor correctly.
Can you not just embed a partial screenshot (paragraphs containing special characters) in CT next time?
Explore tool .. scrot .. found in Linux.
And/or can fast forward be added to CT special characters?
CHERRYTREE is an excellent software. I have used it for many years. Congratulations!! Thank you very much, Merry Christmas and an excellent new year !!
Thank you merry Christmas and happy 2021 🙂
I love Cherrytree. There’s nothing else as robust and stable at my favorite price of free.
Keep up the great work! Merry Christmas!!!
Thank you Merry Christmas 😉
The new icons have improved as well in 0.99.27!
Cherrytree keeps getting better, thank you very much
Merry Xmas under the cherry.. uh christmas tree 😉 and a healthy 2021!
😁 Thank you merry Christmas and a better 2921!
Changed from old TreePad (registered user) a year ago.
Cherry tree is much more responsible, and stable across OSs.
There is an exception to be faced, thus the following feedback:
On my 4 years old laptop the distribution feels quite light and fast,
Exception is the very useful tables implementation, heavy and slow,
much much slower than in the 8 years old desktop with 0.39.4.
How? Inserting, deleting or pasting row, takes 12 seconds.
Changing from a node to another, takes 2 seconds.
This may be not a priority… yet.
Maybe caused by being dependent from interpretation”.
Maybe caused by the use of smaller buffers than due.
Maybe caused by working a char at a time.
Maybe all of them. Thus mentioning, not to be forgotten.
CherryTree is now my favorite Information manager.
It’s coherent across OS’s, editing made a pleasure.
CherryTree is a true cherry. The nut can wait.
P.S.- The new icons are more visible, but unbalanced.
Rather the older, even if small. Change is not necessarily a progress.
The new one burden the eye, then becoming (as I see them) a bit stressful.
… Not a priority, just a small feedback.
Hi. I’m unsure if this should go here or in your github site. This morning a serious problem arose in my 0.99.27, reproducible both in Arch KDE & SparkyLinux [=~Debian 10 Testing] Xfce. A subnode i edited within the last day or so, apparently ok at the time, now causes a CT instant crash each & every time it is selected in the tree. CT appears to be its usual excellent self everywhere else, but this subnode is now broken. From launching CT in terminal & watching, this is the msg each time it crashes at that subnode:
“`
[2020-12-31 11:58:14.403] [ ] [debug] autosave: no needs to save file
[2020-12-31 11:58:46.108] [ ] [debug] Node name: To Do
Entity: line 16: parser error : PCDATA invalid Char value 27
nstant-Off
^
Entity: line 18: parser error : PCDATA invalid Char value 27
elay 4 Sec.
^
[2020-12-31 11:58:46.116] [gtk] [info]
unhandled exception (type std::exception) in signal handler:
what: Document not well-formed.
Line 18, column 8 (fatal):
PCDATA invalid Char value 27
Trace/breakpoint trap
“`
I am almost certain i know exactly which & where those incompatible characters/symbols are, & if only i could now access that subnode i could delete them… but as described, CT now crashes every time there. Oh, also, i have checked many of the archive/backup CT files, but they either have the same corruption, or are sufficiently old as to predate all this [i cannot merely fall back on these as i would lose too much of my recent data in numerous other nodes].
I know nothing about sqlite databases, but from the Debian repos i installed `sqlitebrowser:amd64 (3.12.1-2` in hope that somehow i might muddle my way to finding the suspect content to delete it. Here’s a picture: https://i.postimg.cc/W4jXJS0w/20201231-142533.png. As you can see, i found the subnode, but have no idea now how to “enter” the database there to edit my data.
Please can you advise how i should proceed now?
Can you send the document privately to me so I can have a look?
Hi again & thanks for your fast reply. Sorry but i can’t share the complete file with anyone because it contains some personal info as well as my tech stuff [the corrupt subnode is in the tech not personal section]. What i can do though, if this would be ok for you, tomorrow is to copy my file, then open the copy in CT & delete ALL nodes, except leaving intact this one subnode & its parent node. How would i send it to you? I’d need to know your email address.
If that’s what we do, & if you find the corruption, as far as i can see though you’d still need to explain to me how i could also find it using `sqlitebrowse` & then fix it myself in my real file, otherwise how could i merge your fixed one with my faulty one?
Look at help–about for my email and if you remember what you made to corrupt the node tell me. I would modify cherrytree itself to at least not crash and then if possible correct the issue itself. To edit from terminal you can use sqlite3 other tools I don’t know
Just a thought. If you can save your whole CT file as XML – not SQL – then you can use an XML editor like XMLCopyEditor to inspect and edit the inner nodes.
Hi D L, thanks for your good suggestion. This possibility excited me, but sadly when i tried it, the instant i clicked on OK with that Save As XML option selected, CT crashed again, before the save could occur. This is repeatable. Though i clearly know nothing of the actual underlying code nor programming, intuitively i believe that the nature of this subnode’s data corruption is such that when ANY CT operation is called which needs to process the entire database rather than only those parts of it not involving that subnode, the crash is triggered. I implicitly further tested that hypothesis by launching CT again in a terminal, then initiating the “Find Into All the Tree Nodes Contents” search. Sadly as i expected, the instant the search reached that previously-identified [in my OP] data in this subnode, CT crashed again. This discovery that now i can no longer do ANY operation that processes the whole database, rather than only that one subnode, alarms me & makes me realise that the implications of this problem now are much more serious than previously i’d supposed.
Hi GIUSPEN, i confess to some disappointment that it seems i can’t get any guidance on how i might use some external tool to interrogate the database, to find then delete the specific incompatible character entries causing this problem. However i like your idea to try to fix CT so that if it encounters this bad character it stops crashing. If you can do that, then all i would need to do is upgrade to that future CT version & my current data would be ok again.
To help you see this specific bad data character i strongly suspect to match that terminal error message in my OP, i have this info for you. In that subnode “To Do” i have a CT Table, & into one row i pasted this following text… first i’ll show you a picture of it, then show you from where you can get it yourself.
The text that is highlighted here is what i had copied & pasted: https://i.postimg.cc/bv8FygKD/20210101-133239.png
I copied it from a pdf document i was viewing in Okular, but this original document can be downloaded from here: https://download1.gigabyte.com/Files/Manual/mb_manual_ga-z97(h97)-hd3_v2.x_e.pdf
You’ll find this text on page 30 of 40, very near the top of that page.
It is those unusual “double arrowheads” directly to the left of the lines beginning “Instant-Off” and “Delay 4 Sec” that i suspect CT does not like, & which trigger the crashes. I remember that when i had pasted them originally into my CT table, they did not render correctly, but changed to some weird complex rectangular shape.
Can you please advise me if this might be sufficient info for you to proceed, or if you need more from me?
Many thanks for all your work on CherryTree. It’s an awesome piece of software. I hope you had a good christmas and new year.
@guigirl it’s enough for now but better raise a GitHub issue next time
@Chris thanks happy new year 😉
Just to finish answering @guigirl .. I find that Hexeditor also searches node text patterns in *.ctb.
Raise issue in github for more discussion as @giuspen requests.
@guigirl I’m failing to download that pdf https://download1.gigabyte.com/Files/Manual/mb_manual_ga-z97(h97)-hd3_v2.x_e.pdf
My apologies @giuspen, that link’s now also broken for me. Here’s one level higher up: https://www.gigabyte.com/Motherboard/GA-Z97-HD3-rev-21/support#support-manual — i selected the English one.
Oh how exasperating – this one also atm fails to d/l the pdf. I presume that Gigabyte must be having some server problem atm — grrrrr.
You told me before how to look up your email address. I shall do that now, then send you an email with my own copy of this pdf attached. I do hope this at least works 🙂
Thanks @DL, Last night i found Hexeditor in the AUR [i use ArchLinux] & installed it. Its UI & how to use it was completely unknown to me, but i worked out how to import [a copy of] my ctb file, open it, & find those two sections of the table. Unfortunately however, my attempts to delete that content, which seemed to go ok within Hexeditor itself, only resulted in a ctb that CT refused even to open, telling me that it was corrupt. Rats. I assume this means i am not using Hexeditor correctly.
Seems to be choking on “fast forward” special characters .. >>
https://www.fileformat.info/info/unicode/char/23e9/index.htm
Can you not just embed a partial screenshot (paragraphs containing special characters) in CT next time?
Explore tool .. scrot .. found in Linux.
And/or can fast forward be added to CT special characters?
Best discussed further as github issue.
I created https://github.com/giuspen/cherrytree/issues/1465.