I have a multiboot computer on which several operating systems are installed, on which my home directory and certain other directories are shared via ZFS (so chosen because it is, or can be made to be, supported on Linux, the BSD systems, and Solaris).
On some of these systems, the --geometry option to roxterm is interpreted in pixels, whereas on others, it is interpreted in characters. Because my .xinitrc startup file is shared, this creates inconsistency.
Is there a roxterm option or an X resource (not likely -- I don't see that roxterm uses the X resource database) that can be used to obtain consistent interpretation of the --geometry option among these different operating systems? Otherwise, my roxterm windows will be either too large on some operating systems, or they will be too small on other operating systems, which is bloody annoying.
If you think that this question is not of general interest, you can e-mail me at jay@m5.chicago.il.us. Thank you in advance for any and all replies.
Jay F. Shachter
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Roxterm parses the geometry itself, so it should always interpret it as characters. It's possible you're using a very old version which used a, now deprecated, parsing function built into GTK. This sourceforge project is no longer active, you should use https://github.com/realh/roxterm instead. Hopefully you'll get a more consistent experience with a newer version, unless you're stuck with a very old version of GTK/GNOME on Solaris or something.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Thank you. I see that ./mscript.py is no longer the build method. Some things have not changed, like I still had to correct all the misspellings of the word "color". This was more than just a global change with a text editor; I still had to change 8 to 7 in roxterm_opt_signal_handler() in lines 3579 and 3581 of src/roxterm.c (it is ungracious to complain about free software that someone else wrote and made available to the public, and I am not complaining, I am very grateful for your work, but if I had written the program I would not have hard-coded any constants in roxterm_opt_signal_handler(), I would have used strlen() or sizeof -- you can do sizeof of a character array and it is done at compile time, it does not slow your program down even by a nanosecond, and, like strlen(), it makes it much easier to correct misspelled words).
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
"Colour" is the correct spelling in the UK. I live with other programs and libraries using "color", but I didn't want to use foreign spelling in my own code. You could create an en_US translation for it, but all the other translators faded away, so I stopped payng much attention to the translation infrastructure, which never played well with SCM.
The config path names shouldn't be translated though, these are effectively hidden files. Translating them would make things harder if you wanted to use multiple languages on the same machine. If I had intended for them to be variable then I would have used named constants and sizeof, but that would have meant more typing, and it was nearly all written in an era when IDEs were kind of klunky and vim didn't have the sort of code assistance it does now.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Yes, I know. I'm actually very careful about such things, and have had occasion to be so, I have taught classes all over the English-speaking world (and even a few classes in French, in both Paris and Dakar). A few years ago I taught three consecutive Linux classes in London, at the end of three weeks I was almost remembering to look to the right when first crossing the street. Whenever I'm teaching in a place that requires it, and I say "suppose you have an input file named abc and an output file named xyz" I'm always careful to pronounce the output file "eks wye zed", just as, if I ever taught in Scotland, I would eliminate all consonants from my speech. The only thing I refuse to do is pronounce "either" and "neither" the way you do, because that's just wrong, that's empirically and historically wrong, everyone pronounced those words the right way until the early 20th century when your royal family died out, and you had to import some cousins from Germany to rule you (and, moreoever, not the first time you did that) who needed to learn how to speak English and had a tendency to pronounce "ei" the way it is pronounced in German, and because they were the royal family everyone started imitating their mispronunciation. Except for "either" and "neither", which I refuse to mispronounce, I am always respectful of the local dialect. And it's not just pronunciation, it's also vocabulary, I talk about being run over by a lorry and hit over the head with a spanner (I use vivid images when I teach). And also spelling. I really do know to write "colour" when that's how the locals spell it, Fowler and the OED to the contrary notwithstanding. And when I got to the chapter on shell programming, and I wanted to illustrate how to use the "read" statement to obtain information interactively from the user, I wrote little scripts that would ask you, e.g., "What is your name?" and "What is your favourite colour?", careful to misspell both "favorite" and "color".
Anyway, thank you again for writing roxterm (I just failed to build the new version on one of my machines because "target_link_directories" is supported only on cmake 3.13, which in my case I have not got, but that's not your fault), and let me know if you need someone culturally sensitive to teach system administration or programming in any country where they speak English or French.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Sorry I didn't realise it was a joke. You should have known that most programmers are autistic! Usually my sense of humour is quite good, but I tend more to dark humour (can get quite sick actually).
Interesting about the pronunciation of 'ei'. I started thinking about names ending in -stein. Both syllables of Einstein always rhyme with "pine", even in America, don't they? What about Brian Epstein (the Beatles' manager). I've only ever heard that pronounced the Germanic way. But I know -stein names often rhyme with "bean" instead. But then there's Harvey Weinstein (sorry to bring him up, but I can't think of another example who's well-known for better reasons) with both pronunciastions in the same word!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Esteemed Colleagues:
I have a multiboot computer on which several operating systems are installed, on which my home directory and certain other directories are shared via ZFS (so chosen because it is, or can be made to be, supported on Linux, the BSD systems, and Solaris).
On some of these systems, the --geometry option to roxterm is interpreted in pixels, whereas on others, it is interpreted in characters. Because my .xinitrc startup file is shared, this creates inconsistency.
Is there a roxterm option or an X resource (not likely -- I don't see that roxterm uses the X resource database) that can be used to obtain consistent interpretation of the --geometry option among these different operating systems? Otherwise, my roxterm windows will be either too large on some operating systems, or they will be too small on other operating systems, which is bloody annoying.
If you think that this question is not of general interest, you can e-mail me at jay@m5.chicago.il.us. Thank you in advance for any and all replies.
Roxterm parses the geometry itself, so it should always interpret it as characters. It's possible you're using a very old version which used a, now deprecated, parsing function built into GTK. This sourceforge project is no longer active, you should use https://github.com/realh/roxterm instead. Hopefully you'll get a more consistent experience with a newer version, unless you're stuck with a very old version of GTK/GNOME on Solaris or something.
Thank you. I see that ./mscript.py is no longer the build method. Some things have not changed, like I still had to correct all the misspellings of the word "color". This was more than just a global change with a text editor; I still had to change 8 to 7 in roxterm_opt_signal_handler() in lines 3579 and 3581 of src/roxterm.c (it is ungracious to complain about free software that someone else wrote and made available to the public, and I am not complaining, I am very grateful for your work, but if I had written the program I would not have hard-coded any constants in roxterm_opt_signal_handler(), I would have used strlen() or sizeof -- you can do sizeof of a character array and it is done at compile time, it does not slow your program down even by a nanosecond, and, like strlen(), it makes it much easier to correct misspelled words).
"Colour" is the correct spelling in the UK. I live with other programs and libraries using "color", but I didn't want to use foreign spelling in my own code. You could create an en_US translation for it, but all the other translators faded away, so I stopped payng much attention to the translation infrastructure, which never played well with SCM.
The config path names shouldn't be translated though, these are effectively hidden files. Translating them would make things harder if you wanted to use multiple languages on the same machine. If I had intended for them to be variable then I would have used named constants and sizeof, but that would have meant more typing, and it was nearly all written in an era when IDEs were kind of klunky and vim didn't have the sort of code assistance it does now.
No one laughs at my jokes.
Yes, I know. I'm actually very careful about such things, and have had occasion to be so, I have taught classes all over the English-speaking world (and even a few classes in French, in both Paris and Dakar). A few years ago I taught three consecutive Linux classes in London, at the end of three weeks I was almost remembering to look to the right when first crossing the street. Whenever I'm teaching in a place that requires it, and I say "suppose you have an input file named abc and an output file named xyz" I'm always careful to pronounce the output file "eks wye zed", just as, if I ever taught in Scotland, I would eliminate all consonants from my speech. The only thing I refuse to do is pronounce "either" and "neither" the way you do, because that's just wrong, that's empirically and historically wrong, everyone pronounced those words the right way until the early 20th century when your royal family died out, and you had to import some cousins from Germany to rule you (and, moreoever, not the first time you did that) who needed to learn how to speak English and had a tendency to pronounce "ei" the way it is pronounced in German, and because they were the royal family everyone started imitating their mispronunciation. Except for "either" and "neither", which I refuse to mispronounce, I am always respectful of the local dialect. And it's not just pronunciation, it's also vocabulary, I talk about being run over by a lorry and hit over the head with a spanner (I use vivid images when I teach). And also spelling. I really do know to write "colour" when that's how the locals spell it, Fowler and the OED to the contrary notwithstanding. And when I got to the chapter on shell programming, and I wanted to illustrate how to use the "read" statement to obtain information interactively from the user, I wrote little scripts that would ask you, e.g., "What is your name?" and "What is your favourite colour?", careful to misspell both "favorite" and "color".
Anyway, thank you again for writing roxterm (I just failed to build the new version on one of my machines because "target_link_directories" is supported only on cmake 3.13, which in my case I have not got, but that's not your fault), and let me know if you need someone culturally sensitive to teach system administration or programming in any country where they speak English or French.
Sorry I didn't realise it was a joke. You should have known that most programmers are autistic! Usually my sense of humour is quite good, but I tend more to dark humour (can get quite sick actually).
Interesting about the pronunciation of 'ei'. I started thinking about names ending in -stein. Both syllables of Einstein always rhyme with "pine", even in America, don't they? What about Brian Epstein (the Beatles' manager). I've only ever heard that pronounced the Germanic way. But I know -stein names often rhyme with "bean" instead. But then there's Harvey Weinstein (sorry to bring him up, but I can't think of another example who's well-known for better reasons) with both pronunciastions in the same word!