[gentoo-user] grub doesn't autoselect a kernel
I have grub running on many Gentoo machines but on one of them it sits on the kernel selection screen and doesn't autoboot even though the menu says: "The highlighted entry will be booted automatically in 2 seconds." Nothing happens for any length of time but pressing Enter boots the kernel as usual. Any ideas? - Grant
Re: [gentoo-user] Help with obscure syntax in gsignal.h?
On 07/08/2015 08:26 PM, walt wrote:
I'm trying to debug a gtk+ app so I'm trying to learn some basic gtk+
and failing :(
Can anyone splain to me what these lines mean:
typedef enum
{
G_CONNECT_AFTER = 1 << 0,
G_CONNECT_SWAPPED = 1 << 1
} GConnectFlags;
In particular I don't understand what the << operator is doing.
When I was young << meant bitwise shift to the left. When I got old
and tried to learn c++ it was a way to print a string to the terminal.
Any clues would be welcome.
Bit shift. gtk+ is written in C, so '<<' is always bit shifting. Enums
are basically constants, and they're hardcoding the values so they can
be OR'd together.
Alec
Re: [gentoo-user] Help with obscure syntax in gsignal.h?
In this context I believe it's a bit shift operator. So 1 << 0 is 1 and 1
<< 1 is 2.
As for the << operator in C++...
The operator appears in C++ as the bitshift operator too. However, C++
supports operator overloading. This means that for custom data types (I.e.
classes), one can define how certain operators (including arithmetic and
logical operators) behave. This is done in a similar way to defining member
functions of a class. For example, one could overload the multiplication
operator * for a class 'Matrix' to implement correct matrix multiplication.
One operator overload which pervades the C++ standard library is the
'stream insertion operator'. For input and output streams such as the
standard output cout, the bitshift operators (<<,>>) have been redefined to
allow you to pass data to/from the stream.
This is arguably an abuse of notation and in general such confusing use of
operator overloading would be discouraged. However, in this case it is
perhaps justified since it provides a very convenient and intuitive way of
passing large numbers of objects of varying types to a stream such as cout.
Hope this helps clear things up.
Alex
On 9 Jul 2015 1:27 am, "walt" wrote:
> I'm trying to debug a gtk+ app so I'm trying to learn some basic gtk+
> and failing :(
>
> Can anyone splain to me what these lines mean:
>
> typedef enum
> {
> G_CONNECT_AFTER = 1 << 0,
> G_CONNECT_SWAPPED = 1 << 1
> } GConnectFlags;
>
> In particular I don't understand what the << operator is doing.
>
> When I was young << meant bitwise shift to the left. When I got old
> and tried to learn c++ it was a way to print a string to the terminal.
>
> Any clues would be welcome.
>
>
>
>
[gentoo-user] Help with obscure syntax in gsignal.h?
I'm trying to debug a gtk+ app so I'm trying to learn some basic gtk+
and failing :(
Can anyone splain to me what these lines mean:
typedef enum
{
G_CONNECT_AFTER = 1 << 0,
G_CONNECT_SWAPPED = 1 << 1
} GConnectFlags;
In particular I don't understand what the << operator is doing.
When I was young << meant bitwise shift to the left. When I got old
and tried to learn c++ it was a way to print a string to the terminal.
Any clues would be welcome.
Re: [gentoo-user] Why nepomuk suddenly?
On Thursday 09 July 2015 01:10:29 Alan McKinnon wrote: > Bitter hard experience with the same mistake has taught me to always do > this when faced with such issues: > > egrep -r "nepomuk|semantic|akonadi" /etc/portage/ Indeed. It's just a matter of remembering to do that at the right time. [sigh] -- Rgds Peter
Re: [gentoo-user] KMail-2 duplicates
On Thursday 09 July 2015 01:04:13 Alan McKinnon wrote: > kmail-1 was awesome. > kmail-2 is a complete joke. > > The best recommendation I can give you is to salvage what you can from > your mailboxes and switch to something that works. Thunderbird, claws, > mutt all work fine and one of those surely fits your needs? I did survey the available clients but didn't find anything to compare with what KMail ought to be. Seriously though, after following Mick's advice, I have what seems to be a stable e-mail system now, even if it does throw up the occasional duplicate. I can live with that if it doesn't regress further. -- Rgds Peter
Re: [gentoo-user] Why nepomuk suddenly?
On 09/07/2015 01:05, Peter Humphrey wrote: > On Wednesday 08 July 2015 15:13:36 Alan McKinnon wrote: >> On 08/07/2015 14:53, Peter Humphrey wrote: >>> It's my turn to seek help understanding portage. >>> >>> Yesterday my daily update required a reinstallation of kdelibs with the >>> nepomuk flag switched on, and that pulled in nepomuk-core and >>> nepomuk-widgets. But according to equery, those three packages have only >>> a conditional dependence on nepomuk: >>> >>> prh@wstn ~ $ equery d nepomuk-core >>> >>> * These packages depend on nepomuk-core: >>> kde-base/kactivities-4.13.3 (nepomuk ? =kde-base/nepomuk-core-4.13.3:4[aqua=]) kde-base/kdelibs-4.14.8 (nepomuk >>> ? >=kde-base/nepomuk-core-4.14.3:4[aqua=]) >>> kde-base/nepomuk-widgets-4.14.3 (>=kde-base/nepomuk-core-4.14.3:4[aqua=]) >>> kde-base/plasma-workspace-4.11.19 (nepomuk ? =kde-base/nepomuk-core-4.11.19:4[aqua=]) prh@wstn ~ $ emerge --info | >>> grep nepomuk >>> prh@wstn ~ $ equery d nepomuk-widgets >>> >>> * These packages depend on nepomuk-widgets: >>> kde-base/kdelibs-4.14.8 (nepomuk ? =kde-base/nepomuk-widgets-4.14.3:4[aqua=]) >>> >>> Nepomuk doesn't appear in emerge --info, so why does portage insist that >>> kdelibs requires nepomuk? >>> >>> # emerge -cv nepomuk-widgets nepomuk-core >>> >>> Calculating dependencies done! >>> >>> kde-base/nepomuk-core-4.14.3 pulled in by: >>> kde-base/kdelibs-4.14.8 requires >>> >=kde-base/nepomuk-core-4.14.3:4[-aqua] >>> kde-base/nepomuk-widgets-4.14.3 requires >>> >=kde-base/nepomuk-core-4.14.3:4[-aqua]> >>> kde-base/nepomuk-widgets-4.14.3 pulled in by: >>> kde-base/kdelibs-4.14.8 requires >>> >=kde-base/nepomuk-widgets-4.14.3:4[-aqua] >>> >>> The version of kdelibs has not changed recently, nor have I made changes >>> to >>> make.conf or package.use. >>> >>> Something odd is going on. >> >> What state is the machine in now? Did you enable the flag and do what >> portage asks, or is that update still pending? >> >> I have a hunch the cause is not kdelibs directly, but rather another >> package that depends on kdelibs[nepomuk]. If your update is still >> pending, we can easily find out why with emerge -t > > Mea culpa. I had a stray entry in package.use. I don't remember where that > came from, but of course it's gone now. > > Apologies for the noise. > Ah, the simple cause is often the right one :-) Bitter hard experience with the same mistake has taught me to always do this when faced with such issues: egrep -r "nepomuk|semantic|akonadi" /etc/portage/ Tweak the regex to suit whatever the issue is, entries hiding in package.* don't stay hiding for long :-) -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Why nepomuk suddenly?
On Wednesday 08 July 2015 15:13:36 Alan McKinnon wrote: > On 08/07/2015 14:53, Peter Humphrey wrote: > > It's my turn to seek help understanding portage. > > > > Yesterday my daily update required a reinstallation of kdelibs with the > > nepomuk flag switched on, and that pulled in nepomuk-core and > > nepomuk-widgets. But according to equery, those three packages have only > > a conditional dependence on nepomuk: > > > > prh@wstn ~ $ equery d nepomuk-core > > > > * These packages depend on nepomuk-core: > > kde-base/kactivities-4.13.3 (nepomuk ? > > >=kde-base/nepomuk-core-4.13.3:4[aqua=]) kde-base/kdelibs-4.14.8 (nepomuk > > ? >=kde-base/nepomuk-core-4.14.3:4[aqua=]) > > kde-base/nepomuk-widgets-4.14.3 (>=kde-base/nepomuk-core-4.14.3:4[aqua=]) > > kde-base/plasma-workspace-4.11.19 (nepomuk ? > > >=kde-base/nepomuk-core-4.11.19:4[aqua=]) prh@wstn ~ $ emerge --info | > > grep nepomuk > > prh@wstn ~ $ equery d nepomuk-widgets > > > > * These packages depend on nepomuk-widgets: > > kde-base/kdelibs-4.14.8 (nepomuk ? > > >=kde-base/nepomuk-widgets-4.14.3:4[aqua=]) > > > > Nepomuk doesn't appear in emerge --info, so why does portage insist that > > kdelibs requires nepomuk? > > > > # emerge -cv nepomuk-widgets nepomuk-core > > > > Calculating dependencies done! > > > > kde-base/nepomuk-core-4.14.3 pulled in by: > > kde-base/kdelibs-4.14.8 requires > > >=kde-base/nepomuk-core-4.14.3:4[-aqua] > > kde-base/nepomuk-widgets-4.14.3 requires > > >=kde-base/nepomuk-core-4.14.3:4[-aqua]> > > kde-base/nepomuk-widgets-4.14.3 pulled in by: > > kde-base/kdelibs-4.14.8 requires > > >=kde-base/nepomuk-widgets-4.14.3:4[-aqua] > > > > The version of kdelibs has not changed recently, nor have I made changes > > to > > make.conf or package.use. > > > > Something odd is going on. > > What state is the machine in now? Did you enable the flag and do what > portage asks, or is that update still pending? > > I have a hunch the cause is not kdelibs directly, but rather another > package that depends on kdelibs[nepomuk]. If your update is still > pending, we can easily find out why with emerge -t Mea culpa. I had a stray entry in package.use. I don't remember where that came from, but of course it's gone now. Apologies for the noise. -- Rgds Peter
Re: [gentoo-user] KMail-2 duplicates
On 08/07/2015 23:57, Peter Humphrey wrote: > On Wednesday 08 July 2015 22:03:16 Mick wrote: > >> I would thing that this is a relatively straight forward transaction to >> troubleshoot, but devs may not be using POP3 on Kmail2, or if the glacial >> progress of KDEPIM is anything to go by, then I seriously suspect they are >> NOT really using ... Kmail2! :-p > > You may have something there. I forget the detail now but one of their devs > remarked that it was a pity that POP3 still existed. I thought he was joking > at the time, but maybe he wasn't. > > I have only POP3 accounts: my ISP doesn't offer IMAP, and I don't propose to > change ISPs on that account. > kmail-1 was awesome. kmail-2 is a complete joke. The best recommendation I can give you is to salvage what you can from your mailboxes and switch to something that works. Thunderbird, claws, mutt all work fine and one of those surely fits your needs? -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Why nepomuk suddenly?
On Wednesday 08 July 2015 15:13:36 Alan McKinnon wrote: > On 08/07/2015 14:53, Peter Humphrey wrote: > > It's my turn to seek help understanding portage. > > > > Yesterday my daily update required a reinstallation of kdelibs with the > > nepomuk flag switched on, and that pulled in nepomuk-core and > > nepomuk-widgets. But according to equery, those three packages have only > > a conditional dependence on nepomuk: > > > > prh@wstn ~ $ equery d nepomuk-core > > > > * These packages depend on nepomuk-core: > > kde-base/kactivities-4.13.3 (nepomuk ? > > >=kde-base/nepomuk-core-4.13.3:4[aqua=]) kde-base/kdelibs-4.14.8 (nepomuk > > ? >=kde-base/nepomuk-core-4.14.3:4[aqua=]) > > kde-base/nepomuk-widgets-4.14.3 (>=kde-base/nepomuk-core-4.14.3:4[aqua=]) > > kde-base/plasma-workspace-4.11.19 (nepomuk ? > > >=kde-base/nepomuk-core-4.11.19:4[aqua=]) prh@wstn ~ $ emerge --info | > > grep nepomuk > > prh@wstn ~ $ equery d nepomuk-widgets > > > > * These packages depend on nepomuk-widgets: > > kde-base/kdelibs-4.14.8 (nepomuk ? > > >=kde-base/nepomuk-widgets-4.14.3:4[aqua=]) > > > > Nepomuk doesn't appear in emerge --info, so why does portage insist that > > kdelibs requires nepomuk? > > > > # emerge -cv nepomuk-widgets nepomuk-core > > > > Calculating dependencies done! > > > > kde-base/nepomuk-core-4.14.3 pulled in by: > > kde-base/kdelibs-4.14.8 requires > > >=kde-base/nepomuk-core-4.14.3:4[-aqua] > > kde-base/nepomuk-widgets-4.14.3 requires > > >=kde-base/nepomuk-core-4.14.3:4[-aqua]> > > kde-base/nepomuk-widgets-4.14.3 pulled in by: > > kde-base/kdelibs-4.14.8 requires > > >=kde-base/nepomuk-widgets-4.14.3:4[-aqua] > > > > The version of kdelibs has not changed recently, nor have I made changes > > to > > make.conf or package.use. > > > > Something odd is going on. > > What state is the machine in now? Did you enable the flag and do what > portage asks, or is that update still pending? I haven't set any flags, nor changed anything else, as I said. I took the update and only later started wondering what was going on. I did wonder whether the change had been engendered by my installation of akonadiconsole to debug my duplicate-messages problem, but removing that hasn't made any difference. > I have a hunch the cause is not kdelibs directly, but rather another > package that depends on kdelibs[nepomuk]. You may be right, but I have nothing requiring nepomuk as far as I can see. > If your update is still pending, we can easily find out why with emerge -t Too late, and I'm sure even I'd think of that if I were awake at the time :-) -- Rgds Peter
Re: [gentoo-user] KMail-2 duplicates
On Wednesday 08 July 2015 22:03:16 Mick wrote: > I would thing that this is a relatively straight forward transaction to > troubleshoot, but devs may not be using POP3 on Kmail2, or if the glacial > progress of KDEPIM is anything to go by, then I seriously suspect they are > NOT really using ... Kmail2! :-p You may have something there. I forget the detail now but one of their devs remarked that it was a pity that POP3 still existed. I thought he was joking at the time, but maybe he wasn't. I have only POP3 accounts: my ISP doesn't offer IMAP, and I don't propose to change ISPs on that account. -- Rgds Peter
Re: [gentoo-user] KMail-2 duplicates
On Wednesday 08 Jul 2015 08:56:46 Peter Humphrey wrote: > On Monday 06 July 2015 19:16:59 Mick wrote: > > Peter, I'm going from memory here, but I recovered from a kmail2 problem > > by clearing the akonadi cache and thereby forcing a reindexing. I > > recall opening the akonadi console from the tool tray and this offered > > me a GUI, which listed the various KDEPIM resources. I right clicked on > > the corrupted folder and selected to clear the cache; in my case it > > would not list any messages and was hanging when trying to sync with the > > mail server. > > > > This fixed it. > > > > Are you looking at the same place? > > I finally found it. The Browser tab shows a list of "collections" in the > top left pane. Right-clicking on an e-mail folder there offers "Clear > akonadi cache". I did that, waited a few seconds for disk activity to stop > and then took the Synchronise option. That took a long time - I don't know > how long as I had to go out and leave it to itself. > > Repeating the procedure for each of my dodgy folders appears to have tidied > up the worst of the mess, though I still get duplicates. At least now they > aren't phantoms. > > Thanks for your perseverance, Mick. You are welcome, but I am saddened to read that the problem (duplicate emails when using POP3) is *still* not resolved. I don't know what is causing it, but suspect that it is something to do with the UIDL command causing some hash malfunction on the Kmail2 POP3 client. I am guessing (wildly and with no evidence) that akonadi's (re)indexing is not managing the message ID reliably, potentially barfing on the message hash, recreating a duplicate. Either way, applications like this give MSOutlook with its reliably self-corrupting .pst files a good name. I would thing that this is a relatively straight forward transaction to troubleshoot, but devs may not be using POP3 on Kmail2, or if the glacial progress of KDEPIM is anything to go by, then I seriously suspect they are NOT really using ... Kmail2! :-p -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Anyone else having a problem with bash?
Am 08.07.2015 um 15:10 schrieb Todd Goodman: I haven't looked for that specifically in the ssh source code though. And it sounds like it's happening to people even without interrupting programs so it's unlikely the cause of all the problems. I didn't say that SSH is the cause to that, I just said I can reproduce the issue with SSH. And since this didn't happen back in bash-4.2_p53, I doubt that SSH has to do something with it directly, but rather some change that happened between bash-4.2_p53 and bash-4.3_p33-r2. --Flo
[gentoo-user] Re: HBM gpus
R0b0t1 gmail.com> writes: > 1) AMD cards usually have okay FOSS driver support, but all that is > available might be the mfr-supported ones for some time. > 2) It will never overheat in the stock configuration (which you would > have to do a great deal to change) unless you are retarded. Note > "stock configuration" usually assumes 25C as the ambient temperature. There all sold out, so it looks like I have plenty to time to figure out the supporting hardware, or a new rig, if needed. http://www.newegg.com/Product/ProductList.aspx?Submit=ENE&DEPA=0&Order=BESTMATCH&Description=amd+fury+with+4096+stream+processors&N=-1&isNodeId=1
[gentoo-user] Re: HBM gpus
R0b0t1 gmail.com> writes: > 1) AMD cards usually have okay FOSS driver support, but all that is > available might be the mfr-supported ones for some time. > 2) It will never overheat in the stock configuration (which you would > have to do a great deal to change) unless you are retarded. Note > "stock configuration" usually assumes 25C as the ambient temperature. In lieu of 'retarded' does 'dense' work for you? (OK with me). I do intend to experiment with gcc-5 and other compilers where the gpu memory & resources is available for generic compiling [1]. So I'm sure tight monitoring of the card's temperatures and power consumption is very important for my experiments. And no, I have no idea how to do this properly; I've only read about it, but that is the joy in smoking ~700 in hardware? Hopefully not. I'm still trying to figure out if my Gigabyte:: 990FXa-UD3 mobo can handle this card. It has an 8350 with a water cooler and dual fans on the cpu. The card also comes with a water cooler so I do have a roof mount for another fan in the Thermaltake chassis (V03000 series Level 10 GTS). This top of chasis fan mount is bit, so I might need and adaptor plate custom made. The last time I properly set up a Radeon card, it took some time to ferret out the versions for that particular gpu; as I vaguely recall there were dozens to choose from the Foss collection: (Cape Verde/Pitcairn). with: VIDEO_CARDS="radeon fbdev vesa" If I use the vendor drivers what does make.conf look like (it's been a long time since I used vendor drivers. Does that change if I intend to access the gpu resources in system compilations [2,3] ? Is this problematic from linux; I sure hope not as I dont do windows. Any tidbits or suggestions are appreciated, like repos that have beta foss drivers for a radeon config. Tia, James [1] https://gcc.gnu.org/wiki/OpenACC [2] http://www.openacc.org/node/365 [3] http://www.openacc-standard.org/
Re: [gentoo-user] Why nepomuk suddenly?
On 08/07/2015 14:53, Peter Humphrey wrote: > It's my turn to seek help understanding portage. > > Yesterday my daily update required a reinstallation of kdelibs with the > nepomuk flag switched on, and that pulled in nepomuk-core and nepomuk-widgets. > But according to equery, those three packages have only a conditional > dependence on nepomuk: > > prh@wstn ~ $ equery d nepomuk-core > * These packages depend on nepomuk-core: > kde-base/kactivities-4.13.3 (nepomuk ? > >=kde-base/nepomuk-core-4.13.3:4[aqua=]) > kde-base/kdelibs-4.14.8 (nepomuk ? >=kde-base/nepomuk-core-4.14.3:4[aqua=]) > kde-base/nepomuk-widgets-4.14.3 (>=kde-base/nepomuk-core-4.14.3:4[aqua=]) > kde-base/plasma-workspace-4.11.19 (nepomuk ? > >=kde-base/nepomuk-core-4.11.19:4[aqua=]) > prh@wstn ~ $ emerge --info | grep nepomuk > prh@wstn ~ $ equery d nepomuk-widgets > * These packages depend on nepomuk-widgets: > kde-base/kdelibs-4.14.8 (nepomuk ? >=kde-base/nepomuk-widgets-4.14.3:4[aqua=]) > > Nepomuk doesn't appear in emerge --info, so why does portage insist that > kdelibs requires nepomuk? > > # emerge -cv nepomuk-widgets nepomuk-core > > Calculating dependencies done! > kde-base/nepomuk-core-4.14.3 pulled in by: > kde-base/kdelibs-4.14.8 requires >=kde-base/nepomuk-core-4.14.3:4[-aqua] > kde-base/nepomuk-widgets-4.14.3 requires > >=kde-base/nepomuk-core-4.14.3:4[-aqua] > > kde-base/nepomuk-widgets-4.14.3 pulled in by: > kde-base/kdelibs-4.14.8 requires > >=kde-base/nepomuk-widgets-4.14.3:4[-aqua] > > The version of kdelibs has not changed recently, nor have I made changes to > make.conf or package.use. > > Something odd is going on. What state is the machine in now? Did you enable the flag and do what portage asks, or is that update still pending? I have a hunch the cause is not kdelibs directly, but rather another package that depends on kdelibs[nepomuk]. If your update is still pending, we can easily find out why with emerge -t -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Re: Anyone else having a problem with bash?
* Florian Gamböck [150708 03:15]: > Am 08.07.2015 um 02:48 schrieb walt: > > Next time this happens I'll include the output of stty -a. [..SNIP..] > > After a small `diff`, the following changes have been made: > lnext from ^V to ; icrnl, icanon, and echo from on to off (they > all got prefixed with a -). > > Any ideas what can cause this behavior? > > Regards, > --Flo A program can change terminal settings like this. For example, when ssh reads a password it can turn off echo so the password does not show up. If you interrupt the program (^C) and it doesn't have a handler that restores the original settings then this can happen. I haven't looked for that specifically in the ssh source code though. And it sounds like it's happening to people even without interrupting programs so it's unlikely the cause of all the problems. Todd
[gentoo-user] Why nepomuk suddenly?
It's my turn to seek help understanding portage. Yesterday my daily update required a reinstallation of kdelibs with the nepomuk flag switched on, and that pulled in nepomuk-core and nepomuk-widgets. But according to equery, those three packages have only a conditional dependence on nepomuk: prh@wstn ~ $ equery d nepomuk-core * These packages depend on nepomuk-core: kde-base/kactivities-4.13.3 (nepomuk ? >=kde-base/nepomuk-core-4.13.3:4[aqua=]) kde-base/kdelibs-4.14.8 (nepomuk ? >=kde-base/nepomuk-core-4.14.3:4[aqua=]) kde-base/nepomuk-widgets-4.14.3 (>=kde-base/nepomuk-core-4.14.3:4[aqua=]) kde-base/plasma-workspace-4.11.19 (nepomuk ? >=kde-base/nepomuk-core-4.11.19:4[aqua=]) prh@wstn ~ $ emerge --info | grep nepomuk prh@wstn ~ $ equery d nepomuk-widgets * These packages depend on nepomuk-widgets: kde-base/kdelibs-4.14.8 (nepomuk ? >=kde-base/nepomuk-widgets-4.14.3:4[aqua=]) Nepomuk doesn't appear in emerge --info, so why does portage insist that kdelibs requires nepomuk? # emerge -cv nepomuk-widgets nepomuk-core Calculating dependencies done! kde-base/nepomuk-core-4.14.3 pulled in by: kde-base/kdelibs-4.14.8 requires >=kde-base/nepomuk-core-4.14.3:4[-aqua] kde-base/nepomuk-widgets-4.14.3 requires >=kde-base/nepomuk-core-4.14.3:4[-aqua] kde-base/nepomuk-widgets-4.14.3 pulled in by: kde-base/kdelibs-4.14.8 requires >=kde-base/nepomuk-widgets-4.14.3:4[-aqua] The version of kdelibs has not changed recently, nor have I made changes to make.conf or package.use. Something odd is going on. -- Rgds Peter
Re: [gentoo-user] HBM gpus
1) AMD cards usually have okay FOSS driver support, but all that is available might be the mfr-supported ones for some time. 2) It will never overheat in the stock configuration (which you would have to do a great deal to change) unless you are retarded. Note "stock configuration" usually assumes 25C as the ambient temperature.
Re: [gentoo-user] Re: Anyone else having a problem with bash?
Florian Gamböck wrote: > I downloaded and compiled your archive. > > $ echo $0 > ./sh/OBJ/x86_64-linux-cc/sh > $ $0 --version > sh (Schily Bourne Shell) version 2015/06/27 a+ (x86_64-unknown-linux-gnu) > > Copyright (C) 1984-1989 AT&T > Copyright (C) 1989-2009 Sun Microsystems > Copyright (C) 1982-2015 Joerg Schilling > > > Now I cannot reproduce the no-echo issue, at least with my ssh method. > But as I said, I also couldn't do it with `bash --posix`. So this seems > somehow related. The non-POSIX Bash seems to trigger something and > doesn't reset it under certain circumstances. It seems that there is something related to my other mail: bash --posix bash-4.1$ ls MKLINKS abbrev.h bosh.mk1 defs.hfault.c hashserv.c mac.h name.cpwd.csh_policy.h sym.h version.h Makefile alias.c cmd.c dup.h func.cio.cmacro.c name.hservice.csignames.c test.c word.c Makefile.man args.cctype.c echo.cgmatch.c jobs.c main.c pfsh.1sh.1 stak.c timeout.h xec.c OBJ bltin.c ctype.h error.c hash.cjsh.1 mode.h pfsh.mk1 sh.tour.ps stak.h ulimit.c abbrev.c bosh.1defs.cexpand.c hash.hjsh.mk1 msg.c print.c sh_policy.c string.c umask.c bash-4.1$ stty -echo bash-4.1$ MKLINKS abbrev.hbosh.mk1 defs.hfault.c hashserv.c mac.hname.cpwd.csh_policy.h sym.h version.h Makefile alias.c cmd.c dup.h func.cio.cmacro.c name.hservice.csignames.c test.c word.c Makefile.man args.cctype.c echo.cgmatch.c jobs.c main.c pfsh.1sh.1 stak.c timeout.h xec.c OBJ bltin.c ctype.h error.c hash.cjsh.1 mode.h pfsh.mk1 sh.tour.ps stak.h ulimit.c abbrev.c bosh.1defs.cexpand.c hash.hjsh.mk1 msg.c print.c sh_policy.c string.c umask.c bash-4.1$ exit Note that this is on Linux and I typed "ls" after "stty -echo" Interesting: the final "exit" is printed...but this is done _after_ I typed the new-line. It seems that there is something really strange inside the command line editing code in bash. Note that The Bourne Shell does neither implement "emacs" nor "vi" mode but "ved" mode. Ved is an editor that is included as a reimplementation in the schily tools tarball. An editor with this interface and the same name first appeared on UNOS 1980. I don't know when the first ved version was written, it may be older than emacs. So if you are using the Bourne Shell, you may observe minor differences to bash. Note that an important difference is that my history editor remembers the cursor position in each command line. This makes it more convenient to change a single parameter in an existing command line. Jörg -- EMail:jo...@schily.net(home) Jörg Schilling D-13353 Berlin joerg.schill...@fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.org/private/ http://sourceforge.net/projects/schilytools/files/'
Re: [gentoo-user] Re: Anyone else having a problem with bash?
Hi Jörg, Am 08.07.2015 um 12:13 schrieb Joerg Schilling: Sorry for asking, but does Gentoo include the Bourne Shell? $ readlink /bin/sh bash I guess this means "no" ... If you like to do, the latest portable Bourne Shell is in: http://sourceforge.net/projects/schilytools/files/schily-2015-07-07.tar.bz2 I downloaded and compiled your archive. $ echo $0 ./sh/OBJ/x86_64-linux-cc/sh $ $0 --version sh (Schily Bourne Shell) version 2015/06/27 a+ (x86_64-unknown-linux-gnu) Copyright (C) 1984-1989 AT&T Copyright (C) 1989-2009 Sun Microsystems Copyright (C) 1982-2015 Joerg Schilling Now I cannot reproduce the no-echo issue, at least with my ssh method. But as I said, I also couldn't do it with `bash --posix`. So this seems somehow related. The non-POSIX Bash seems to trigger something and doesn't reset it under certain circumstances. --Flo
Re: [gentoo-user] Re: Anyone else having a problem with bash?
Nikos Chantziaras wrote: > On 06/07/15 20:01, walt wrote: > > This is the problem: occasionally bash gets in a state where it stops > > echoing the characters I type. The commands I type continue to work > > properly and I can see the output from them but I can't see the commands > > on the screen as I type them. > > I remember having this problem. Especially when aborting programs with > CTRL+C. I am interested to know how this can happen. Note that AFAIK, UNIX would not allow a suitable tty mode that both allows to implement interactive editing inside a program _and_ letting the kernel do echoing for the editor at the same time. This was possible in UNOS in 1980 (the first UNIX clone) and I used it for the first shell with comand line editor in 1982, but I could not use something similar on UNIX. Now I did just run a test on Linux using bash and it turns out that calling "stty -echo" brings bash into a mode that does not print typed characters in the command line. The shells that use my command line editor from 1982 (which is the first shell with command line editor "bsh" and since December 2006, my portable version of the Bourne Shell) always use a suitable tty mode for the command line editor - regardless what the user did set up using stty. It only restores the settings from the user when a user command is running. Is bash explicitly implementing a history editor that does not echo characters in case that "stty -echo" was run? This would be a really bad idea. Jörg -- EMail:jo...@schily.net(home) Jörg Schilling D-13353 Berlin joerg.schill...@fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.org/private/ http://sourceforge.net/projects/schilytools/files/'
[gentoo-user] Re: Anyone else having a problem with bash?
On 06/07/15 20:01, walt wrote: This is the problem: occasionally bash gets in a state where it stops echoing the characters I type. The commands I type continue to work properly and I can see the output from them but I can't see the commands on the screen as I type them. I remember having this problem. Especially when aborting programs with CTRL+C. I don't seem to have it anymore though. Not sure when it stopped, but I suspect it's when I upgraded to bash-4.3_p39 from 4.3_p33-r2? (I'm on ~amd64).
Re: [gentoo-user] Re: Anyone else having a problem with bash?
Florian Gamböck wrote: > Hi Stephan, > > Am 08.07.2015 um 11:28 schrieb Stephan Müller: > > As you can replicate it reliable, did you test it in Bourne shell? > > Maybe its not related to bash at all? > > $ echo $SHELL > /bin/bash Sorry for asking, but does Gentoo include the Bourne Shell? If you like to do, the latest portable Bourne Shell is in: http://sourceforge.net/projects/schilytools/files/schily-2015-07-07.tar.bz2 Jörg -- EMail:jo...@schily.net(home) Jörg Schilling D-13353 Berlin joerg.schill...@fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.org/private/ http://sourceforge.net/projects/schilytools/files/'
Re: [gentoo-user] Re: Anyone else having a problem with bash?
Hi Stephan, Am 08.07.2015 um 11:28 schrieb Stephan Müller: As you can replicate it reliable, did you test it in Bourne shell? Maybe its not related to bash at all? $ echo $SHELL /bin/bash $ bash --version | head -n1 GNU bash, Version 4.3.33(1)-release (x86_64-pc-linux-gnu) If I use /bin/sh (or to be accurate: /bin/bash --posix, since it is a symlink) then I do NOT run into the issue. --Flo
Re: [gentoo-user] Re: Anyone else having a problem with bash?
Am 08.07.2015 um 11:06 schrieb Florian Gamböck: I can replicate it 100% when I ssh to another machine and immediately interrupt it with Ctrl-C. I hadn't considered this a bug, but after reading this thread I can confirm that it also happens randomly with other commands. Not sure if there is always interruption involved, though ... Perhaps some problem with forking / spawning a subshell and returning to the "status quo" after a non-zero exit code? Regards, --Flo As you can replicate it reliable, did you test it in Bourne shell? Maybe its not related to bash at all? ~frukto
Re: [gentoo-user] Re: Anyone else having a problem with bash?
Am 08.07.2015 um 09:00 schrieb Alan McKinnon: On 08/07/2015 07:00, Anton Shumskyi wrote: Same for me, but it appears only after I'm canceling some job on terminal with CTRL+C, maybe in 10% of total cases. I thought that was some side-effect of switching env back, but because job is terminated in a "bad way" haven't considered that as a bug. Same here. I always get this effect shortly after a Ctrl-key combination, not sure which one but it's one of the grouping on the left of the keyboard surrounding "S" I can replicate it 100% when I ssh to another machine and immediately interrupt it with Ctrl-C. I hadn't considered this a bug, but after reading this thread I can confirm that it also happens randomly with other commands. Not sure if there is always interruption involved, though ... Perhaps some problem with forking / spawning a subshell and returning to the "status quo" after a non-zero exit code? Regards, --Flo
Re: [gentoo-user] KMail-2 duplicates
On Monday 06 July 2015 19:16:59 Mick wrote: > Peter, I'm going from memory here, but I recovered from a kmail2 problem by > clearing the akonadi cache and thereby forcing a reindexing. I recall > opening the akonadi console from the tool tray and this offered me a GUI, > which listed the various KDEPIM resources. I right clicked on the > corrupted folder and selected to clear the cache; in my case it would not > list any messages and was hanging when trying to sync with the mail server. > This fixed it. > > Are you looking at the same place? I finally found it. The Browser tab shows a list of "collections" in the top left pane. Right-clicking on an e-mail folder there offers "Clear akonadi cache". I did that, waited a few seconds for disk activity to stop and then took the Synchronise option. That took a long time - I don't know how long as I had to go out and leave it to itself. Repeating the procedure for each of my dodgy folders appears to have tidied up the worst of the mess, though I still get duplicates. At least now they aren't phantoms. Thanks for your perseverance, Mick. -- Rgds Peter
Re: [gentoo-user] Re: Anyone else having a problem with bash?
Am 08.07.2015 um 02:48 schrieb walt: Next time this happens I'll include the output of stty -a. Since I just hit this bug I will do it for you if you don't mind. ;-) I somehow managed to reproduce this issue by typing `ssh myothermachine`, hitting Enter, and immediately hitting Ctrl-C. Phenomenon is the same as described here. `stty -a` before the incident: speed 38400 baud; rows 26; columns 190; line = 0; intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = M-^?; eol2 = M-^?; swtch = M-^?; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; min = 1; time = 0; -parenb -parodd cs8 hupcl -cstopb cread -clocal -crtscts -ignbrk brkint -ignpar -parmrk -inpck -istrip -inlcr -igncr icrnl ixon -ixoff -iuclc ixany imaxbel iutf8 opost -olcuc -ocrnl onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 ff0 isig icanon iexten echo echoe echok -echonl -noflsh -xcase -tostop -echoprt echoctl echoke `stty -a` immediately after the incident, typed blindly into the terminal: speed 38400 baud; rows 26; columns 190; line = 0; intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = M-^?; eol2 = M-^?; swtch = M-^?; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ; flush = ^O; min = 1; time = 0; -parenb -parodd cs8 hupcl -cstopb cread -clocal -crtscts -ignbrk brkint -ignpar -parmrk -inpck -istrip -inlcr -igncr -icrnl ixon -ixoff -iuclc ixany imaxbel iutf8 opost -olcuc -ocrnl onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 ff0 isig -icanon iexten -echo echoe echok -echonl -noflsh -xcase -tostop -echoprt echoctl echoke After a small `diff`, the following changes have been made: lnext from ^V to ; icrnl, icanon, and echo from on to off (they all got prefixed with a -). Any ideas what can cause this behavior? Regards, --Flo
Re: [gentoo-user] Re: Anyone else having a problem with bash?
On 08/07/2015 07:00, Anton Shumskyi wrote: > Same for me, but it appears only after I'm canceling some job on > terminal with CTRL+C, maybe in 10% of total cases. I thought that was > some side-effect of switching env back, but because job is terminated in > a "bad way" haven't considered that as a bug. Same here. I always get this effect shortly after a Ctrl-key combination, not sure which one but it's one of the grouping on the left of the keyboard surrounding "S" -- Alan McKinnon alan.mckin...@gmail.com
