Not a blog; Windows 7 vs. Ion; displays, laptops, etc.
Since I put up a new site with an archive of my software, a few have asked me to return the non-blog online as well. I don't think that will happen. I can't be arsed to start blogging again; it's a waste of time. I shortly entertained the idea of combining the good/essential material into a longer “Goodbuy, suckers!” anthology, but... even pissing on Linux (desktop *nix) is a waste of time. Ignorance is a bliss, and I'm much happier ignoring it all and sticking to Windows. Windows 7 is, in fact, not that bad at all. Yeah, I have been using Windows 7 for some weeks now. It's surprisingly good, far more keyboard-usable than previous releases of Windows.. let alone popular Linux DEs, which have always been inferior to Windows. Of course, I'm looking at everything through the Trackpoint-glasses; I probably would find W7 far less usable if I only had a normal mouse or, *shudder*, or had to rely on the shotgun err.. touchpad. One reason for the improved keyboard-usability is the integration of search into Windows 7. No more browsing of endless menu and dialog hierarchies (esp. control panel and startmenu), just search. No need for third-party launchers, which never worked very well on XP; just hit the Windows key and start typing. In a way, this is a partial return to the command line interface. Maybe there's still a trace of sanity left in the world. Windows 7 even supports proto-tabbing and proto-tiling. By pinning an application/shortcut to the taskbar, you can use Win+number to cycle through instances of that application. (The visual indication of the chosen instance when cycling is too weak, though; the currently focussed instance is much more strongly indicated.) And with Win+left/right arrow, you can throw a window to the left or right side of the screen, which is almost the only layout I ever used in Ion, aside from full-screen (and Ion4, if I had started working on it, would have been more based on restricted layouts like this). I just wish you could adjust (maybe you can) the widths of windows thrown this way, as in the split-float tiling mode of Ion, where frames could partially overlap each other. On a small laptop display, most windows are simply too narrow without some degree of overlapping. In any case, the ideas introduced in PWM and Ion seem to be catching on, at least in diluted forms. My first gripe is that it still doesn't have a sensible safely remove feature. Still a fly-crap-sized system tray icon... which is even hidden by default, an otherwise excellent feature for all that annoying crap that everyone wants to put there. That or opening explorer (which is the only way by default to even mount anything on the POS OpenSUSE at the office...) And even with all the other improvements on keyboard-usability (after you enable the display of accelerator keys, which is idiotically disabled by default) the fact remains that the choice of keybindings suck, a lot of things depending on poorly-accessible cursor keys etc., instead of something closer to the typing position. Fonts... difficult to say, because the display on the Thinkpad X201 is so awful -- the worst I've ever come across -- that the constant shifting of everything on the display in reaction to minute movements of the head is more annoying than the blurring that W7 almost insists upon. (You need to do a lot of work to replace the semi-hard-coded Segoe UI everywhere, to get rid of the blurring.) That said, the blurring on W7 seems better than on any other OS... depending on the font of course, and the fact that the display has so crappy contrast. I should try to get around to trying it on a semi-decent display. But all desktop displays have so horrible DPI that any kind of blurring immediately induces uncontrollable vomiting. It really sucks that there are no decent displays these days: desktop displays have shitty DPI, although you can get IPS, and laptop displays are too shallow and generally el cheapo TN-film crap, although the DPI is somewhat more reasonable. (But most smartphones have even better DPI... and have higher-quality displays too, with far better viewing angles than laptops.) Yeah... I finally gave up on trying to run Modern Bloatware(tm) on the aging T43 (with its splendid although a bit dim IPS panel), and got a new laptop. I'd have preferred a desktop computer with an IPS panel, but since a desktop display is difficult to lug when moving around the world, that's presently no option for me. And the low DPI would make my eyes bleed anyway. I was also not going to buy a big laptop with a shallow display, so a netbook or an ultraportable was the only option. Preferrably a cheap netbook, as I wouldn't want to pay much for the crap that you get as laptop displays. But netbooks are not significantly more powerful than the T43... so X20[01] it was, as I need the pointing stick. Fortunately you can get a little-used demo machine on Ebay for a somewhat more reasonable price than the list price. For idling use - random web
Archives
In case someone still cares, an archive of all the code and stuff is online at http://iki.fi/tuomov/software/ .
Re: so, what now?
On 2010-04-06 18:10 -0400, Benjamin R. Haskell wrote: Guess I'll go back to lurking. (I'm only still here for the occasional Tuomo rant.) I use Windows. I have only been in Finland three months during the last year. I'm running out of material. (Although I could rant about software bloat and ever diminishing laptop screen heights, etc., which make it impossible to work around bloatware the brute force way of adding ever more hardware. But I can't be arsed. Oh, how fscking difficult it's to buy a summer jacket these days: almost everything is full of plastic. I hate plastic noise. I hate plastic against skin.) -- Tuomo
Re: disabling mouse focus
On 2010-03-19 14:33 +0200, M. Rawash wrote: hi, is there a simple way to disable 'mousefocus' on certain clients (a on-screen keyboard, for example)? There's the 'passive' winprop. There should be examples in the list archives, maybe even something in the scripts repository. -- Tuomo
Re: Moving workspaces
On 2010-01-14 17:35 -0500, Timandahaf wrote: I want to be able to move an entire workspace from one screen to another. I tried (using the context menu), Workspace-Toggle Tag, and then went to a new workspace in the other screen, and did Workspace-Attach Tagged. What did this is: it created a new floating workspace, put all my tiled, tabbed windows from my original workspace into this new WGroupWS, and then attached this WGroupWS to the other screen. There are no tiled workspaces anymore. Workspaces are floating always, but may contain tilings. You moved the entire workspace, so you got some mess. You can do shit better with the primitives, instead of trying to extend the tagging functionality beyond its design (as a simple component of the default simple Ion way of doing things). -- Tuomo
Re: Youtube fullscreen not working
On 2009-11-27 00:54 -0500, Timandahaf wrote: I tried to do this, but I still have the exact same problem. I see what your kludge is doing, but it doesn't seem to fix the problem. Regardless of whether or not I use your script, when I press the full-screen button inside the youtube player, the video disappears, and the browser window gets surrounded by a thin read line (which I assume is the video window). Any clue what else the problem could be? The red line is probably Ion's marker for the source frame of a detached window. You can detach/reattach windows with Mod1+K D, but too big transients also get detached automatically... is flash purporting its fullscreen document window to be a transient now? And then it somehow fails when Ion's detach code keeps moving the window around? Maybe it even gives the window size hints that force it to be a single pixel, and hence basically invisible with the minimal transient decorations. (Ion always applies size hints; modern idiot-applications often request sizes that do not fall into the range of their size hints.) You may be able to find the window id with xprop -tree -root or something, if it sets any identifying information at all (probably won't, being modern crapware), and then find the size hints and the TRANSIENT_FOR property with xprop -id the_id. I'm not running Linux or even X anymore, so that's all I can help. -- Tuomo
Re: bindings to move mouse cursor?
I think that would be a very useful way to switch to the frame you want, just move the cursor until it's over that frame, should normally only take a couple of keystrokes. If you really want to do that, I suggest a Trackpoint keyboard instead. The difficult a foresee with key(board) control of the pointer -- based on experience with keyboard resize -- is control of the acceleration. You need some kind of acceleration for fast yet precise movement of the pointer, but with an on/off control device you have to do it based on hacks based on the time pressed, which is hacky and does not work so well. Trackpoint, on the other hand, does it by the pressure you apply. By applying little pressure, you can move the pointer very precisely; by a large amount of pressure, you're on the other side of the screen in no time. (Assuming you have sane almost maxed-out settings; the defaults tend to be crappy, and maybe one of the reason many people give up on the Trackpoint, and become ice-skaters.) well, it's a bit like buying a car navigation system and then only using it street by street ... maybe you should have a look at the commands ion provides. Isn't that what navigation systems are all about? If you actually bothered looking at the map for a moment, you wouldn't need some device to tell you where to go (wrong). -- Tuomo
Re: Lua..
On 2009-11-17 19:51 -0200, TC wrote: _any_ language he/she's used to, which I find a good basement philosophy if you ask me. Not quite. Such a generic scripting interface inherently has to stick to the lowest common denominator.. and hence it should really not be the main extension/scripting interface, I think, but a more simplified command interface.. For more thoughts on this, see http://iki.fi/tuomov/b/2007/scripting_configuration_and_api_design/ http://iki.fi/tuomov/vis/ I'm reluctant to learn yet another language just for the window manager. Baawaawaa, it's not written in My Language Of Choice, so it must be rewritten. That describes something like 49% of FOSS projects The other 49% is baawaawaa, it's not Freeⓡ: it doesn't use My License of Choice. The remaining 2% may have some originality. There really is not much to learn in Lua, and if you can't pick up a language on the go -- especially as simple one as lua, with a very short official reference manual -- you shouldn't be programming at all. It's not like you need a high level of proficiency for scripting. I've written plugins for programs written in perl, ruby, and python, and I have never taken the time to properly learn them, don't really like them (Lua is _much_ better, except for the library selection), and would not start a project using them myself, unless there was some very compelling reason (library availability or such). which I find far less versatile and readable than my language of choice. Then you haven't even bothered reading the basic description of the language, and are criticising just because it's not Your Language of Choice. Again, this may be just a matter of taste, but in any case, I do not feel like messing with lua, and I think it should not be imposed. It's not imposed. As far as I know, nobody has to use Ion, as it's a marginal alternative among a gazillion other window managers. It's not XFascistType, pukedev, Hardware Obstruction Layer (of the Year) or other core software that the only way to avoid is to not use Linux/*nix. Which is exactly what I have done. http://iki.fi/tuomov/b/2009/defection_part_3_windows/ My suggestion? Why not make ion controllable through IPC? maybe a fifo, socket, tcp, whatever, to _let the user choose_ the language to be used. Why don't you do that? Actually there is mod_ionflux, that you can use to send Lua code... I'm not saying that ion should be redesigned and reimplemented right away either. It will definitely be rewritten from scratch at some point I guess, so why not keep this in mind when the time comes? Ion is not being developed. For details on any possible future, see http://article.gmane.org/gmane.comp.window-managers.ion.general/9056 -- Tuomo
Re: Lua..
On 2009-11-18 02:17 -0800, Daniel Clemente wrote: Baawaawaa, it's not written in My Language Of Choice, so it must be rewritten. That describes something like 49% of FOSS projects The other 49% is baawaawaa, it's not Freeⓡ: it doesn't use My License of Choice. … The remaining 2% may have some originality. wmii is among that 2%: - it can be scripted in any language, also in your language of choice - it is free from restrictions; you may also sublicense it to your free license of choice Muahhah, looks more like belonging to both of the 49%. I suggest taking some language classes. And knowing the history of tabbing/tiling WMs, the 2% is reserved for a certain other WM or WMs... -- Tuomo
Re: Statusbar question
On 2009-10-30 16:54 -0400, Timandahaf wrote: The problem is, the workspace_pager that's shown on the screen1 is actually a copy of screen0's pager, and is therefore useless. How do I fix this? Check if the script provides %workspace_pager_n (n=0,1) or something like that. If not, then modify it to do so. -- Tuomo
On continuing Ion development [+footnote on fonts on 5800xm]
Since I've been asked a few times whether it's possible for someone to adopt or continue the development of Ion, now that I'm unlikely to work on it, I thought I should post my thoughts here on the list. First of all, as it always has been, it is, of course, possible for people to fork a _renamed_ project, and develop it into whatever direction. But, for various reasons, such complete forks are not entirely desirable, and it would be better if any new developments that fit under the banner Ion could keep using this name. I always intended Ion3plus as a sort of wiki branch, where most patches fitting some loose guidelines are welcome, and that still remains the case.. but I won't waste time applying them to the repositories (darcs and my scripts being broken on Cygwin, etc.), anyone taking over its maintenance would have to host it. These guidelines are there to keep the project true to the name Ion, and therefore include * Quality code * A veto for me for major changes that do not simply enhance existing features. XFascistType and Shitorama have snowball's chance in hell of getting into Ion. Ion runs on pure X11, not fugly complex extensions. It's all, of course, a bit vague, as it should be; strict catch-all terms and licenses suck anyway, and more should be based on case-by-case negotations. Therefore, initially, I would have to review all changes to see that whoever takes over, can produce good code (or filter out bad code from others)... take it as a job negotation, or trial period. If I'm satisfied with the results, then there shall be more free hands but, still, attempts at the scale of adding Shitorama or XFuglyType should be passed through me (to be rejected). As for development after Ion3plus, which shouldn't _fundamentally_ stray too far from Ion3, I have some new ideas to try for Ion4, that I'd still like to see implemented. I just doubt I'm going to do it myself... I'd really like to move to the management department already, and have someone else do the gruntwork. I'm not yet ready to just give up the ideas, for various reasons, but if there's good work on Ion3plus... or other arrangements are reached that ensure that they're implemented in Ion instead of a clone... I will consider that. I would also require at that point that Ion be ported to Windows and, why not, OS X at. Not necessarily as a full WM, but rather as an IDE managing some nicely behaving embeddable applications. If I magically found the time and interest to actually do the gruntwork on Ion4, it would most likely be a Windows program in any case. (Unless Apple somehow decided to add a Trackpoint on their laptops and support unblurred fonts - or screen resolutions quadrupled[1] - and my T43 died so I had to buy current shallowscreen crap.) --- [1] To replace my 5 year old phone with dying battery, and 6 year old mp3 player, I recently got a Nokia 5800 XpressMusic, as it was the only sub-300e touchscreen phone with standard USB and headphone sockets. (Easily loseable adapters suck. On the downside, no charging off USB, which some competitors' more expensive models would have. But you can get two two of these for the price of an iTrendClown co.) I like the touchscreen. With hand-writing recognition (which could admittedly perform even better), it makes the device vastly more usable than T9 or even a mini-keyboard. But that is not the topic of this footnote, so no more on that. The device blurry fonts. Perhaps surprisingly, they're not actually that bad in the �main interface. Partly this can be attributed to the graphical interface style, partly to usage, partly to choice of font, but mostly to the 230dpi screen, which makes sub-100dpi desktop displays and even ~120dpi laptop displays pale in comparison. By usage I mean that you don't read text for extended lengths on the phone, as on the computer. By choice of font I mean that the font in the interface is simple and thick, so the proportion of blurred pixels is small. It's not so great for reading texts, and the thinner font in the browser handles the blurring much more poorly. But, still, it's vastly more tolerable than on soddy computer displays resolutions, which are still the technological equivalent of the PC beeper, which could not play real music tolerably, only music designed for it. Still, despite the blurring being more tolerable on the phone than I expected, it confirms that 200dpi is not enough resolution in serious use and with a wide range of fonts, for blurring not to annoy. And, yet, OS makers keep pushing on the blur-fascist front. -- Windows may blow, but Linux sucks the crap in the wind.
Re: The end of the line
On 2009-09-01, Tuomo Valkonen tuo...@iki.fi wrote: It has come to this. Unless miracles happen, I don't think I will be using or working on Ion much anymore. Cygwin is a big pile of shit [1], with an even bigger pile of filth Yesterday I was angry, today I'm miserable. I'm too much alone with my thoughts; my anger and misery. I'm feeling somewhat regretful for such angry words; no I don't particularly like the guy and his strict discipline, but I do not want to be a bitter old man hating and hated by everybody, so I'll start by saying sorry here.
Re: The end of the line
On 2009-09-01 14:51 +, Kris Malfettone wrote: Vim works fine without cygwin on windows and I am sure emacs or its various derivitives have cygwin-less windows ports as well. I never liked emacs. Joe-bindings are half emacs, half wordstar, but I just can't stand emacs' own bindings, beyond the basic de facto standard unix movement bindings. Years ago, I tried to start usin emacs, but just couldn't. Too crippled. It's like an OS missing the editor. ^X combos are awful hand-twisters, and it seems ^X is hard-coded (at least in GNU emacs) in many places, so can't be remapped to something more useful (in joe: forward word). Another annoyance was, well, Escape Meta Alt Control Shift, how you needed one binding for search forward, another for search backward, third for replace forward, and so on, these not even including a case-sensitivity toggle. In joe, otoh, the single search function just queries the options. There were various other annoyances, like the syntax highlighter not having the concept of numbers, etc. So I doubt I'll be going for emacs, if I can find something else. Joe would be the best, but it doesn't run natively on Windows. And, since not running in a terminal, something not contstrained by the text terminal would be nice, for displaying automatic folds, etc. Vi, well, it's from such a different world that I don't dislike it like emacs, but never really “got it” either. -- Tuomo
Re: command line ?
On 2009-08-28, Ole Jørgen Brønner olejorg...@yahoo.no wrote: No, you're right, they don't. If someone wants to work on this, here's what I after a very quick glance over the code (that it's been ages since I last looked into), think might be the plan: - In input_do_refit, we have to rqgeom instead of window_do_fitrep. The crux is that with FULL_BOUNDS we know the available space. With the size policy set to something else, we only know the area the managing WMPlex (or other region) is willing to give us per the size policy. So we have to ask for as much as we'd like to have with rqgeom... but to calculate the wanted height for our requests, we have to assume that the width stays fixed. Thus, input_calc_size has to be modified to support calculations with infinite max_geom.h (virtually; INT_MAX or so). - We can no longer use the modified input_refit in input_fitrep (that gets alled by the managing object to do the resizing), instead an updated version of the current code has to be used there. (It's just a couple of lines.) - It should be possible to get rid of the last_fp field, modifying all code involving it, as we no longer make requests using the last known available space. - Of course, for nicer top-aligned queries, the completions should expand downwards, and this demands a bit of extra work, figuring out whether we're actually attached to a WMPlex, and asking for the effective size policy to see if it is a top-aligned one. Probably I'm missing something, but that should get anyone interested in this started. I have better things to do... actually haven't been able to use Ion lately, because Cygwin X has been in a crash every 5 minutes state for nearly a couple of months now. Typical FOSScrap. -- Off the dope since 2009; http://iki.fi/tuomov/b/archives/2009/07/21/T17_26_09/
Re: command line ?
On 2009-08-27, Piter_ x.pi...@gmail.com wrote: There is command line opened by F3 button. I wonder if its possible to move it up from the bottom of the screen. Thanks. Petro No, unless you write the necessary code. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde
Re: close frame when empty
On 2009-08-14, Javier Rojas jeroja...@devnull.li wrote: I think is possible to avoid such behaviour by checking the tiling (or the ws) to close the frame only when there are more frames in the tiling. But I don't know how to ask Ion about how many frames are in a tiling (or in a workspace). Any suggestions? One possibility is using WTiling.managed_i to go through the list of objects managed by the tiling. RTFM on details, but, e.g., the following snippet in Mod1+F3 prints the names of everything (except the somewhat transient stdisp) managed by the tiling: _:manager():managed_i(function(x) print(x:name()) return true end) _ is the frame in which Mod1+F3 is opened, _:manager() is typically the tiling object on tiled workspaces, etc. The function must return true to continue iteration, false (or nil) to stop iteration. Just replace the print with something that sets a (local) variable outside the iterated function to frame count, and stop iteration if this has become at least two. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: window titlebar
On 2009-08-05, Sam Mason ma...@f2s.com wrote: That's what you get for not using a proper operating system! :) Yes, on a proper operating system (instead of a half of one running on another proper one), I'd be stuck to the 80x25 VGA console. Random Cygwin brokenness (1.7 betas) is just the familiar situation from Linux, where you're also stuck to using unstable/experimental distributions, because the stable megadistros provide too old non-core software and do not run the newest third-party software without massive library recompilation efforts. At least on Windows there's a stable proper _core_ operating system underneath, running all the latest software. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: window titlebar
On 2009-08-06, Maximilian W. Zeller maw...@gmail.com wrote:
i dont't quite get how to give absolute x/y position?!
i am looking for something like
position = { x = 100, y = 100 }
There isn't one in Ion3. I did add a winprop to ion3plus,
but all the documentation is the changelog entry.
(See also WGroup.attach docs.)
Fri Oct 24 20:25:20 FLE Daylight Time 2008 Tuomo Valkonen tuo...@iki.fi
* * Added attach_params winprop and auto_placement option.
The first, attach_params, is a table containing the same parameters that can
be passed to WGroup.attach. It is applied when a new frame would normally be
created for the window, as attach parameters of the frame. The parameters
are
thus not applied when attaching a new window to an existing (tiled) frame,
or
for attaching the window within the newly created frame.
The second, auto_placement, is a boolean option in the attach parameters,
and can be used to control whether application-supplied position is used,
or whether Ion should itself calculate a position for the window.
--
Stop Gnomes and other pests! Purchase Windows today!
http://iki.fi/tuomov/b/archives/2009/07/21/T17_26_09/
Re: Compilation pb on FreeBSD.
On 2009-07-23, Albert Shih albert.s...@obspm.fr wrote: After I press F12 I've got Main menu -- session/sortir (I using French) and I've a No entry for 'session/sortir' Does it show up in tab-completion? Could there be some accent issue? Are the locales correct in that case? -- Off the dope since 2009; http://iki.fi/tuomov/b/archives/2009/07/21/T17_26_09/
Re: Ion wiki?
On 2009-06-09, Ole Jørgen Brønner olejorg...@yahoo.no wrote: I have very limited experience with web services and hosting, but I've set up a temporary dokuwiki [1] at http://folk.ntnu.no/bronner/ion3wiki (Temporary as in I don't really have a proper host. Maybe someone else has?) Maybe you an just use some of the hosting services? (sf.net, berlios, etc.) Ion of course has a berlios account for this mailing list, and in theory a wiki could be hosted on that account... but I don't support webshit-only wikis, and I don't think it's right for an unofficial wiki to appear at a semi-official address. -- Stop Gnomes and other pests! Purchase Windows today!
Re: Winsplit-revolution - window manager for MS-WindowsXP etc.
On 2009-06-05, Timandahaf timandahaf+...@gmail.com wrote: Has anyone seen/heard of this? http://winsplit-revolution.com/ I find that tiling w/o tabbing is rather useless. And it's the tabbing part that takes extra effort to implement under Windows. Under X, the WM can -- still as of June 2009 -- reparent application windows within its own windows, and draw the decorations in its frame windows. Under Windows, applications always create top-level windows, which are divided into client area and the remaining boundary, that is drawn by the system -- whatever part that is. You'd somehow have to replace that system to draw different kind of decorations and add different kind of controls, which are duplicated between all the windows associated to a pseudo-frame/ tabbed set. Of course, further problems arise because many applications seem to override the system, and draw their own decorations... and then there's the issue of the apparent lack of the equivalent of the ICCCM, so you can't enforce tiling... as far as I know, but I'm no WinAPI expert. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: confirm 852223f5c787b0857a42616c26c0a94208eae60a
On 2009-05-30, Ole Jørgen Brønner olejorg...@yahoo.no wrote: I'm not sure how you can easily force the number to be exacly 10, but ion3 will remember your layout, so if you create 10 workspaces manually they should still be there after a restart. If you restart/quit properly, i.e., though Ion's menus/functions, instead of simply killing X (C-A-BS). -- Stop Gnomes and other pests! Purchase Windows today!
Re: RPM spec file for ion3 for Fedora
On 2009-05-27, Timandahaf timandahaf+...@gmail.com wrote: Any chance you'd be willing to host this RPM .spec file anywhere in your darcs repo? It'd be very helpful to users who have an rpm based system. I have no public repositories that I could update. And anything that requires additional manual effort from me to update, is totally out of the question. -- Stop Gnomes and other pests! Purchase Windows today!
Patch collection
Some of the latest changes to Ion, including the scripts repository, can be downloaded from http://iki.fi/tuomov/ion/patches.html pending my VCS switch eternity project. (All version control systems suck. Only darcs developers have devoted more than two brain cells to the UI. But GHC sucks donkey balls and thus there's no cygwin version of darcs, and the native Windows version of darcs sucks. That's the biggest hurdle of my windows switch: lack of decent vcs/darcs. Otherwise I'm very happy to have taken broken away from the feces-throwing competition known as linux.) -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: Patch collection
On 2009-05-26, Ciprian Dorin, Craciun ciprian.crac...@gmail.com wrote: But I wonder: you like darcs (or at least compared with others), and you have searched for other distributed VCS's but all of them suck, what is wrong about (for example) Hg or Git? I mean what are your thoughts about this topic. (Disclaimer: I've switched all my repositories from CVS to SVN and then to Git... and until now, I'm happy... Indeed the Windows support is a little bit brittle, but under Cygwin it should work...) Actually, it doesn't. The packaged version (cygwin 1.7) doesn't do anything when you run it. And I couldn't compile it myself. (The build failed at some point, although it did build the git binary, that failed similarly to the cygwin-packaged version.) What annoys the shit out of me in hg, is its manual configuration approach to everything. Darcs is interactive. It asks you for your name/email when you first use it in a repository (unless you have globally configured them); hg used to assume usern...@hostname, which is 99% of the time crap, so you ruined your repository, and have to fix it. This days it iirc refuses to run until you configure. But at least this is an action you have to do only once on each computer. Unfortunately, the same single brain cell approach persists in default push locations. Darcs is nice, and it by defaults remembers the previous push/pull location, defaulting subsequent pushes and pulls there. And it is interactive, so unless you pass -a that just pushes everything without questions asked, you'll get to see where and what it is trying to push. Also you can't completely ruin your repository with darcs, because you can just unpull patches. Now, with hg, it only remembers where you first cloned from, and always defaults there.. if you've cloned from anywhere. push/pull don't make it remember the location. Also, push just pushes the patches to the default location, no questions asked, and with hg you can ruin the repository this way: there's no way to get rid of the patches. If the target repository was wrong, you have to figure out what was pushed, and try to fork from an earlier point. They want you to do extra work and first do 'hg outgoing' to check what would be pushed where, and only after that 'hg push'. 'darcs push', by contrast, includes the 'outgoing' functionality in its default interactive mode, so you can safely type the command. Also, to push by default somewhere else than the initial default (none unless a cloned repository), you always have manually edit a configuration file in .hg. That sucks. Bazaar has rather similar idiotic push location behaviour, but at least there's a --remember switch to push. Git.. I haven't tried a recent version, as remarked above. But a few years ago [1], its email push support was a complete joke, equivalent to just sending a plain old diff and applying it, without any inter-repository tracking. Same, I think, with bazaar. hg at least seems to try to do its best within its traditional 'kludge based on three-way diff' approach to version control, that is to be contrasted with the more advanced theories behind darcs. [1] http://iki.fi/tuomov/b/archives/2006/10/29/T16_57_00/ That is, indeed, another major problem with the VCS switch... it does not seem in general possible to convert _multiple_ interoperable darcs repositories (e.g. ion-3, ion-3plus) into a three-way-diff-kludge vcs, and maintain inter-operability. In some limited cases (if the patches in ion-3 and ion-3plus were cleanly ordered) it might be possible, but would demand a lot of manual effort. ... What sucks about the Windows darcs? Well, it's no posix program, so a lot of things won't work from cygwin shell. ^C either won't work, or will just kill it without cleaning up locks. single-key queries won't work; you have to press enter. Generally line-editing keys are broken. Time zones are printed in a non-standard format that other programs (such as thost that convert to other VCSs) fail to parse: 'W. Europe Standard Time' instead of WEDT. Also, if you set the TZ environment variable in cygwin, darcs gets totally confused __although it shouldn't see the variable, as it's no cygwin program!__ ssh won't work using cygwin-ssh and the keys stored in ssh-agent; you'd have to set up separate putty pagent setup for it... and a key setup is _required_, which is just shit. Also, at least this version of darcs I have, is just broken, 'reading pristine' *all the fucking time*, whatever you do. It's dog-slow. (Also seems to be compiled with a shitty compiler, and is generally slow.) -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Patch collection
On 2009-05-26, Tuomo Valkonen tuo...@iki.fi wrote: Time zones are printed in a non-standard format that other programs (such as thost that convert to other VCSs) fail to parse: 'W. Europe Standard Time' instead of WEDT. Um... WEST, of course. (Idiotic DST kludge. Just looked up the current zone from 'date', not being familiar with these western europe zone abbreviations, .fi being EET/EEST, and that patch was from before the DST switch.) -- Stop Gnomes and other pests! Purchase Windows today!
Re: Keypresses when closing the mod_sp window?
On 2009-05-15, Thomas Themel i...@themel.com wrote: When closing the mod_sp window, it seems that the window that is underneath also gets the keypress (in my default case, META..space) that I use to close the window. Is this a bug? It will get the release event, not the press (unless the key is pressed long enough to start repeating), and if it's a stupid app, it will act on it. Generally Ion ungrabs the keyboard in grabbed binding handler to avoid problems with the keyboard being grabbed on application startup etc... and there might have been other reasons too. This will cause subsequent release events to fall through. However, I'm not 100% sure that removing this hack would help in this case without other workaround, because the window is being hidden, and thus X should kill the grab in any case. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: VirtualBox seamless mode
On 2009-04-09, Vladimir Skuratovich skuratov...@gmail.com wrote: Hi, Is it possible to set up the VirtualBox 'seamless mode', where it is supposed to integrate with the window manager, so that the guest windows are managed by Ion? Does it create managed windows or just override-redirects? I'd expect the latter, which are likely to then get hidden under Ion's tiling. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: X is dead. Long live windows.
On 2009-04-10, Tuomo Valkonen tuo...@iki.fi wrote: DOS as a simple and manageable system was the high point of computing. XP was a local maximum: the last OS with good fonts, and not even nearly as bad as W95, around the release of which I switched from DOS to what can now clearly be seen as the failure known as Linux/FOSS. And unlike Linux from 2001 -- which didn't suck quite as much as Linux 2009 -- Windows XP is still a widely-used and somewhat supported operating system that you can actually use and run new applications on. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: Attach client to several workspaces
On 2009-04-07, Matthieu Moy matthieu@imag.fr wrote: What you can do (and what wmii, awesome and other do) is that the resize the window on-the-fly when the workspace changes. Then the application redraws the frame at the new place, with its new size. You can manually emulate this by tagging a window with Mod1+T, switchking workspace, and attaching the window in a frame on the chosen workspace with Mod1+K A. It's not difficult to write a basic script to automate this. I can't find one in the repository, but I'd expect someone to already have done this... -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: VirtualBox seamless mode
On 2009-04-10, Klaus Umbach treibholz-...@sozial-inkompetent.de wrote: Just use VirtualBox in standard-fullscreen-mode on a separate Workspace. That's the most acceptable pain. No, that's Ion-in-cygwin-X-on-windows, after changing Mod1+Tab to Mod1+Q in Ion. Integrates much better than a fully keyboard-grabbing VM... albeit not perfectly :(. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: cannot type in run dialog
On 2009-04-08, Deepanjan Kesh deepk...@gmail.com wrote: i have installed ion3 in ubuntu 8.10. but i am not able to type in the run dialog though tab completion works. any help would be really appreciated Fix your locales settings. You need the .encoding (e.g. fi_FI.UTF-8) part in LC_CTYPE (or LANG), because Xlib is broken. -- Stop Gnomes and other pests! Purchase Windows today!
X is dead. Long live windows.
The X we all knew is dying, and it's being killed by its developers. As their latest onslaught, the Xorg dickheads have removed SaveUnders from X. Expect flicker when dragging tabs or moving windows under Ion and a new and rusy X. They want you to implement and use a heavy memory, cpu, and battery-hungry composite manager instead... and do not provide a generic one worth shit. They do not follow the original design of X and extend it; they gradually replace everything with crap and remove the old ways of doing things. Soon on their list will be core the good old fonts, as they want you to use their blur-fascist fantasies. It's impossible to keep up with the shit that the FOSScracy throws at you, unless you're one of the big corporate-sponsored projects that is already part of the FOSScracy. The lamers even banned me from posting on the list. Censorship, silencing the critique. Sure sing of an incompetent tyranny. It was a good time to switch to Windows. X as we knew it will soon no longer be. Something to laugh at: On the same tread, the fuckwits discuss how they need some way to identify displays for DPI settings etc... after they removed simple display identification by switching from the good old X multihead o Xinerama/Xrandr crap. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: Attach client to several workspaces
On 2009-04-06, Oskar Nordquist oskar.nordqu...@gmail.com wrote: I'm wondering how hard it would be to implement support for having clients visible in more than one workspace, e.g it could work like the tagging feature for moving windows, only clients could be attached to several frames and/or workspaces. A basic version shouldn't be too difficult. Just a script that attaches to one of the hooks that signal workspace change, and then moves the window around. What's more work is remembering where exactly the window was on each workspace... for that, there's the placeholder mechanism, but it isn't exported to the Lua side ATM. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Attach client to several workspaces
On 2009-04-07, Oskar Nordquist oskar.nordqu...@gmail.com wrote: Would it not be easier to just duplicate the client to refer to the same X window? I don't even know if that is possible, but if it is, that's how I'd prefer to do it. Ummm... I don't know what you really mean. The client always refers to the same window it initially created. You can move it around in Ion's window management hierarchy (cf. Mod1+A, Mod1+T Mod1+K A, tab drag), also to follow you around. Obviously you can automate this by attaching a script to suitable hooks. The difficult part is that it would always appear at the same index in a frame etc.; this demands exporting the placeholders to Lua side. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Ikiwiki replacement etc.
Ok, the site should be back in updatable condition, using webgen after all. Let me know if something is broken. Next up: VCS switch. (Suggestions about what to do about the site, the scripts repository, and other repositories are still welcome.) -- Tuomo
Re: Abandoning darcs
On 2009-03-31, Yves Rutschle i...@rutschle.net wrote: Yeah, the biggest problem is Windows users naming directories like Program Files, Last year's review, Deliveries for this and that, which means the 255 limitation actually comes up quickly. Then try to convince them to use short names and no spaces. Yeah, it sucks.. but it's coming in Linux too. I absolutely fucking hate the Desktop directory that every now and then shows up in ~. IT'S MY HOME DIRECTORY YOU GNOME FUCKTARDS! DON'T PUT ANYTHING VISIBLE THERE! I'S ***MNE***, NOT YOURS, YOU PIECE OF SHIT RETARDS!! GET IT?!?!? especially as NTFS uses Unicode, so special characters don't take many bytes as with UTF-8. Actually I think NTFS uses UTF-16 which is about twice as NTFS: 255 UTF-16 code points = 510 bytes ext2: 255 _bytes_ = less than 255 UTF-8 characters even with latin alphabet. This is what I meant. Of course, even with UTF-16 you get less than 255 glyphs when using combining characters and stuff. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: Abandoning darcs
On 2009-03-31, Adam Duck adam.ian.d...@gmail.com wrote: Yes, it is possible to _create_ such file names, but you can't use them. Here at WinXP I can create (best under Desktop) dirs close to 255 chars. Actually, I've been succesfully using file names with colons in them in Linux on NTFS (stupid NTFS-3g not failing on such names), and Windows shows them, but you can't access them. Actually, there is a low path length limit not in NTFS but in the some old versions of the _Windows API_ itself. This is probably what the FUD was about. (And if you consider an API originating for 8+3 names -- if this is the case -- 255 is a tolerable limitation for path length.) http://msdn.microsoft.com/en-us/library/aa365247.aspx In the Windows API (with some exceptions discussed in the following paragraphs), the maximum length for a path is MAX_PATH, which is defined as 260 characters. A local path is structured in the following order: drive letter, colon, backslash, components separated by backslashes, and a terminating null character. For example, the maximum path on drive D is D:\some 256 character path stringNUL where NUL represents the invisible terminating null character for the current system codepage. (The characters are used here for visual clarity and cannot be part of a valid path string.) Note File I/O functions in the Windows API convert / to \ as part of converting the name to an NT-style name, except when using the \\?\ prefix as detailed in the following sections. The Windows API has many functions that also have Unicode versions to permit an extended-length path for a maximum total path length of 32,767 characters -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: Abandoning darcs
On 2009-03-31, Daniel Clemente dcl441-b...@yahoo.com wrote: Yes, I also hate it, but I have learnt to ignore it with version control. Or better, I use it to store trash. And if I want a directory that I control myself, I create one, it's not so complex. But it's not ~, it's ~/my or /my or something long and non-automatic. I suspect that when they realise this home directory cluttering problem, instead of choosing saner names, they'll reinvent the wheel and introduce a Hidden attribute to file systems... in an additional gnome indirection layer, not the kernel. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Abandoning darcs
On 2009-03-28, Klaus Umbach treibholz-...@sozial-inkompetent.de wrote: Ah, OK. So it's a driver and a service. Anyway, it was pretty stable. Maybe I'll try it one day. No time now. Yes, but I can use both my cores and don't have to reserve one for my virus-scanner. From my point of view, Windows is still worse. Both cores.. doing fine with just one. Only really notice the scanner at login, which takes ages, just like the boot too. But with reliable standby/hibernate, it's not necessary to boot often, and you can still shut down the system for the night, etc. e.g. you can't overwrite a file, when it is executed. You can't open a file when another application writes in it (how am a I supposed to read logfiles?). This is indeed a major annoyance that you can't delete a file in use, as on *nix. At least in cygwin they can, however, be renamed out of the way. I've been told this should've been improved in Vista. A path in NTFS is not allowed to be longer than 255 chars... Path? You mean file name? It's the same on ext2/3 as far as I recall, and no practical limitation, just Linux fanboy FUD -- especially as NTFS uses Unicode, so special characters don't take many bytes as with UTF-8. And worst of all: it lacks a package-management. Worst of all: Linux has fragmented centralised package management. All the gazillion distributions doing redundant work have de facto central control over what software you can easily install on that particular distribution. On Windows I can just download software from the ISV, and easily run it. In a sandbox if I want to. And, best of all, nearly all the latest sofware still runs on Windows XP, which is from 2001 (and largely still uninfected by the current blur-fascist IT regime). Try running latest Linux binaries -- or even source, for that matter -- on Linux from 2001. FAIL. It's sick. Linux is outright against independent software vendors, and yet it's Microsoft that gets sued for any feature they add to their OS. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Abandoning darcs
On 2009-03-27, Nicolas Schier sch...@shf.de wrote: have you already decided which SCM you want to switch to? I have noticed you took a look at mercurial... I for myself am very content with it, but I know that your requirements have mostly a little higher level. Probably Mercurial. I've been using it for all new repositories anyway. It mostly sucks compared to darcs, but I don't know anything better that's portable. Stupid GHC, stupid compilers, their bootstrapping, and complicated build systems. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and local vendors and artisans have had to yield to all under one roof big box hypermarkets.
Re: Abandoning darcs
On 2009-03-27, Sylvain Abélard sylvain.abel...@gmail.com wrote: If only Apple wasn't so blur-fascist, and had 4:3 laptops with a nipple... And non-glossy screens for all configurations, not just for 2000⬠ones. Well, yes, but the situation is almost the same on PCs: all the affordable models have a glossy screen.. and typically crappy keyboards. And the nipple you find only on the el cheapo Thinkpad SL models. The better T-model Thinkpads are just as expensive as Macbook Pros in .fi but come with a shallowscreen and megabezels these days. (The Thinkpads tend to be about 500e cheaper in .de compared to .fi!) I'm testing this right now. That's not really a problem to keep your thumbs on the trackpad, even with the tap-mode activated. If they move a little, your mouse will flicker (thak God, it disappears when you're typing), but that's not really a big deal. Would have to try for an extended period to find out if they'd have managed to make it work tolerably, but so far I've always had to disable to trackpad to be able to use a laptop. The most painful is Window Management (missing ion so much) and that Spaces multi-desktop thing not well integrated, especially overriding some keyboard bindings (which are customisable with limited choices, all of them not satisfying). Sure you can run Ion in X in OS X? I run it in Windows XP under Cygwin/X, and half of my apps/windows there. (In Xterm, but also plan to use gv/xpdf/xdvi for TeXing when it comes time to write something again.) It works comfortably enough, switching between Windows and Ion, after I changed Mod1+Tab to Mod1+Q in Ion, instead of passing -keyhook to X, which disables the Windows key too, which I'd like to work (could then Win+tab on the taskbar). This keeps the Windows window-count tolerable; it couldn't handle the gazillion terminals comfortably without a tabbing terminal emulator or something... but that's just Ion then; I consider it my IDE. You may find yourself disappointed at their quite-non-standard keyboard layout I have to reconfigure the keyboard on any system I plan to actually use for anything, in any case. Done that on Windows too. (Caps lock = control, extra altgr between shift and z, extra altgr+key bindings [1]. No dead keys, they suck.) [1]: http://article.gmane.org/gmane.comp.window-managers.ion.general/8542 -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: Abandoning darcs
On 2009-03-27 08:19 +, Tuomo Valkonen wrote:
On 2009-03-27, Nicolas Schier sch...@shf.de wrote:
have you already decided which SCM you want to switch to? I have noticed
you took a look at mercurial... I for myself am very content with it,
but I know that your requirements have mostly a little higher level.
Probably Mercurial. I've been using it for all new repositories
anyway. It mostly sucks compared to darcs, but I don't know anything
better that's portable. Stupid GHC, stupid compilers, their
bootstrapping, and complicated build systems.
*sigh* Even in my Windows switch, FOSS crap causes the most trouble.
GHC: fail (= Darcs: fail, lua-xgettext: fail, riot: fail).
Ikiwiki: fail. And Cygwin itself poorly supports locales (i.e., UTF-8
in practise), although with some effort I've managed to make it mostly
work. Ion is one that doesn't support non-ascii... with the settings
that other stuff works with. The problem seems to be, once again, that
X doesn't quite understand the cygwin/libc locales. Cygwin bash is
also _slow_... actually wrote my first _much_ faster (still under cygwin)
python program [1] ever thanks to this.
Windows (XP) itself works quite smoothly.. far better than Modern
Linux, so far.. but doesn't natively provide all the software I'd
like to use. Maybe coLinux would work better than Cygwin, although
it's in principle an ugly solution, and seems to be poorly supported
and difficult to set up... plus you have to deal with all the current
Linux crap [2] that I want to escape from, unless there's some lightweight
coLinux-oriented distro that strips away all that useless crap.
If only Apple wasn't so blur-fascist, and had 4:3 laptops with a
nipple... Although, I did try the _huge_ touchpad on one of the newer
Apple laptops recently, and it was far better than most I've come
across: no need to be skating back and forth, actually quite comfortable
to move the cursor with it. But, still, it's in the way of typing --
can't rest your thumbs anywhere, whereas on the Trackpoint you can
comfortable rest them on the buttons, ready to press -- and you need
to move your hand to actually use it, while the Trackpoint is simply
an integral part of the keyboard.
[1]: A small tool to backup ID3 tags, eyeD3 providing a python library for
reading and copying the raw frames to dummy files. Since copying my
music collection from CDs and DVDs to a hard drive -- it's quite
incredible in how small a package 500G goes in a usb-powered 2.5
disk, compared to the pile of 3.5 disks I had with combined capacity
of just 300G... and these aren't big compared to stuff you find from
the 80's -- and switching from mocp to foobar2000, I've been adding
tags to my collection and downloading cover art, and need to back them
up separately.
[2]: http://iki.fi/tuomov/b/archives/2009/03/13/T16_46_57/
\end{another status report written in a moment of boredom and tiredness}
--
Tuomo
Re: Abandoning darcs
On 2009-03-13 12:09 +, Tuomo Valkonen wrote: The Ion darcs repositories have been taken offline, and I will switch away from darcs. Reason: no version for the most viable *nix platform of the day, Cygwin. The website will be upgraded when I manage to replace Ikiwiki, which also does not work under Cygwin. This is going to take some time. I, of course, can't even use the usual tools to convert from darcs to another format, because the non-deterministic windows-cygwin hybrid darcs does not even print dates etc. in a standard format, so parsers fail. (No, passing TZ does not work. Darcs shouldn't know about it, as it's not a cygwin program. In truth, though, setting TZ messes everything up. Must be a leak in GHC from depending on cygwin for compilation, although it can not be compiled for cygwin -- which is the reason why there's no cygwin verion of darcs.) ... Has anyone any experience of a simple coLinux setup? That wouldn't mess hibernate/standby up? With internal samba? No udev and other crap; Windows handles devices. -- Tuomo
Re: Abandoning darcs
On 2009-03-26, Tuomo Valkonen tuo...@iki.fi wrote: This is going to take some time. I, of course, can't even use the usual tools to convert from darcs to another format, because the non-deterministic windows-cygwin hybrid darcs does not even print dates etc. in a standard format, so parsers fail. Another big project is writing a non-Haskell (read: Lua) implementation of lua-xgettext. Fuck GHC and its unportability. Help appreciated. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and local vendors and artisans have had to yield to all under one roof big box hypermarkets.
Ikiwiki replacement etc.
I will have to replace Ikiwiki as the Ion site generator, since it fails to work under Cygwin. (Excessive and ugly mangling of all input into UTF-8 breaks down in the 2007 snapshot I'd been using. *sigh*. The compiler doesn't need to know the encoding used; only the web frontend does. Newer versions don't build. No response from the author.) One thing that has been suggested, is webgen. But it seems that to do all the inlining or even listing of all the FAQ entries, I'd have to spend time learning Ruby/ERB/webgen internals, and I really have spent far too much time on the computer/OS switch [1] already. A quickdirty Lua or shell script would be faster to write. (Actually, I hate the way Ikiwiki handles intra-wiki links, with its own syntax instead of using markdown. Webgen, OTOH, doesn't have any support for easy linking. I'd like to just define ready aliases for markdown for pages on the site. There actually exists a markdown implementation in Lua...) Does anyone have any other suggestions for site generator, or what to do with the site, that wouldn't involve too much effort from me? This could also be tied to the scripts repository, distribution of extra patches etc.. I still refuse to edit pages through crappy browsers' editors, and whatever is chosen should support markdown for easy conversion (and because most other ASCII markup languages are too bloated). [1]: http://iki.fi/tuomov/b/archives/2009/03/13/T16_46_57/ -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Ikiwiki replacement etc.
On 2009-03-16, Evgeny Kurbatov evgenykurba...@yandex.ru wrote: Try AsciiDOC. It has not any web engine just text-to-html converter. I'm not looking for a markup language -- I'm sticking to markdown. I'm looking for a web engine. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Ikiwiki replacement etc.
On 2009-03-16, Evgeny Kurbatov evgenykurba...@yandex.ru wrote: Which means web engine? I meant no web engine -- no cgi, no web forms, no page editor thru web forms, just kosher html. You can manage your pages with ftp. Let the web server be engine. No, I'm not looking for dynamic content. I'm looking for something to generate static web pages; not he nicer-markup-to-html converter, which will be markdown, but the part that wraps that into templates; generates RSS; inlines multiple pages into one for news, FAQ; offers nicer wiki-like linking rather than referring to individual files; etc. -- Stop Gnomes and other pests! Purchase Windows today!
Abandoning darcs
The Ion darcs repositories have been taken offline, and I will switch away from darcs. Reason: no version for the most viable *nix platform of the day, Cygwin. The website will be upgraded when I manage to replace Ikiwiki, which also does not work under Cygwin. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: Scratchpad stuck in max size mode
On 2009-03-12, Per Johannes Schöön j.sch...@kabelmail.de wrote: I have been struck by a very curious problem. The scratchpad, from one day to the next, decided it wanted to be, and remain, as large as my screen. Have you upgraded recently? It may have wrong size policies etc. in that case. I suggest deleting the scratchpad (Mod1+C), and then reopening it (Mod1+space), when it will be recreated with correct settings. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: [Slightly-OT] On usability: managing usb devices network profiles in a ion3 environment
On 2009-03-07, Mico Filós elmico.fi...@gmail.com wrote: Thank you all for your feedback. I will play with the scripts a little... While you're at it, write ion-power-manager. Something simple and reliable to replace the awful bloated and complicated gnome and kde crap. No fucking HAL crap; just talk to /sys, acpi, whatever. Can be Thinkpad-specific. Indeed probably should be based on plugins specific to particular computer models or classes instead of over-engineered low-level abstr^Windirection. (I tried reading the gnome-power-damager code to see how it works... with the effort I put into it, I couldn't make heads and tails of how it's interacting with other HAL/DBus crap; how the XSync extension is supposed to keep track of idle times -- the documentation of the extension is poor at best -- if it is supposed to do that. You have to trace a gazillion levels of indirection. I'd probably just use the screensaver extension for idle timing, if it can be tricked into multiple timeouts. After all, you want to run the screen lock _before_ suspend/hibernate; the scripts should synchronise the locking program startup, or just include the screen lock in the power manager. The Gnome crap doesn't do this, and all the system-level scripts have finished restoring the system from suspension -- when all the sun spots are correctly aligned and it all works -- before it seems to run the locker, which is a big gleaming security black hole. But it's a lot of work to do all that, and for now Windows mostly works for me. I'm really liking _reliable_ suspend and hibernate in Windowsland, foobar2k, undervolting -- even that requires a fucking kernel patch on the piece of shit Linux -- etc. Finally even managed to get locales/UTF-8 working in Cygwin: the trick is to use LC_CTYPE=C-UTF-8.UTF-8. The biggest problem ATM is getting a cygwin version of darcs. If it doesn't work out, I may have to switch to Mercurial for Ion and my other projects.) -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and local vendors and artisans have had to yield to all under one roof big box hypermarkets.
Re: [Slightly-OT] On usability: managing usb devices network profiles in a ion3 environment
On 2009-03-03, Vladimir Skuratovich skuratov...@gmail.com wrote: With network profiles you actually have a lot of options, starting from ifupdown scripts in Debian, where you can specify a profile manually, through 'divine', a nice small program that is able to send ARP requests to specified hosts to check if they are on the network and then configure the interface, to 'guessnet' - a plug-in for ifupdown, which can do the same things as 'divine' and some more. NetworkManager (and its applet) are one of the few things that actually didn't totally suck abount Xubuntu. Could even get VPN working with less effort than Windows. (F-secure + Cisco VPN client - BSOD, if you install them in that order. The only one I've had, while Linux has locked up many times on console switches etc.) Otherwise, can't say much good about wireless in Xubuntu: all the radios blasting at full power on startup, have to manually turn them off, instead of it remembering the settings. And in 8.04 the Fn-key radio button just toggled the first device of wifi/bluetooth, while in 8.10 it cycles through the combos without indicating the current choice: bluetooth led works, but wifi led doesn't. In Windows I get a nice popup where to turn radios on/off, and it remembers the settings. (As a temporary measure, I've been just running the XFCE panel as Ion's stdisp for all the systray shit. But it sucks, because it can't align stuff to the right in the non-fullwidth mode.) -- Stop Gnomes and other pests! Purchase Windows today!
Re: [Slightly-OT] On usability: managing usb devices network profiles in a ion3 environment
On 2009-03-03 17:21 -0500, Antonio De Leon wrote:
On the question of detecting usb devices this tool can help on the
detection part.
Here's my ls-removable script too. It's in Lua, and needs
both luafilesystem and posix packages for Lua.
There are two versions. The suffix-1 version worked with
the old 2.6.18 or whateveritwas Etch. Suffix-2 is a
conversion to work the modified /sys layout in *buntu 8.10,
but it fails to work because pmount doesn't seem to have
been upgraded to to reflect that. It, however, shouldn't be
too difficult to create a suid-root counterpart based on my
code, that just calls mount -- this time with sane FAT and
NTFS mounting arguments.
(*sigh*, even the ntfs-3g in 8.10 ubuntu is old, and does
not seem to support .NTFS-3G/UserMapping. A lot of shit in
it is _old_, and yet it has just half-a-year release
cycles.)
--
[Fashion] is usually a form of ugliness so intolerable that we have
to alter it every six months. -- Oscar Wilde
The computer industry is the only industry that is more fashion-driven
than women's fashion. -- RMS
#!/usr/local/bin/lua
--
-- Copyright (c) Tuomo Valkonen 2008.
--
require('lfs')
require('posix')
local function filtered_dir(d)
local f, s_, v_ = lfs.dir(d)
local function g(s, v)
while true do
local vn=f(s, v)
if vn~='.' and vn~='..' then
return vn
end
end
end
return g, s_, v_
end
local function readfile(d)
local f, err = io.open(d, 'r')
if f then
local s=f:read('*a')
f:close()
return s
end
end
local function simplify_string(s)
if s then
return string.gsub(s, \n$, )
end
end
local function readfile_simple(d)
return simplify_string(readfile(d))
end
local function removable(d)
return readfile_simple(d../removable)==1;
end
local function removable_class(d)
-- Super-dirty hack
return d and string.match(d, /usb%d+/)
end
local function simplify_path(p)
local capts={}
local start=string.match(p, ^(/))
string.gsub(p, ([^/]+),
function(e)
if e==.. and #capts 1 then
capts[#capts]=nil
elseif e~=. then
table.insert(capts, e)
end
end)
return (start or )..table.concat(capts, /)
end
local function get_device(d)
l=posix.readlink(d../device)
if not l then
return nil
end
if string.match(l, ^.) then
return simplify_path(d../..l)
else
return simplify_path(l)
end
end
local function get_info(dev)
return {
vendor=readfile_simple(dev../vendor),
model=readfile_simple(dev../model),
serial=readfile_simple(dev../serial)
}
end
local function get_info_usb(dev)
par=string.match(dev, (.*)/[^/]+)
if not par then
return {}
else
local m=readfile_simple(par../manufacturer)
if not m then
return get_info_usb(par)
else
return {
vendor=m,
model=readfile_simple(par../product),
serial=readfile_simple(par../serial)
}
end
end
end
local function scan_volumes(bd, d)
local function g()
local found=false
for v in filtered_dir(bd..d) do
if string.match(v, ^..d..%d+$) then
coroutine.yield(v)
found=true
end
end
if not found then
coroutine.yield(d)
end
end
return coroutine.wrap(g)
end
local mounts={}
--local mounts_=io.popen(mount):read('*a')
--string.gsub(mounts_, [^\n]+, function(l) table.insert(mounts, l) end)
local f=io.open(/etc/mtab)
for l in f:lines() do
table.insert(mounts, l)
end
f:close()
local function mounted(v)
for _, l in ipairs(mounts) do
local mp=string.match(l, ^/dev/..v..%s+([^%s]+))
if mp then
return mp
end
end
return false
end
local bd='/sys/block/'
local volumes={}
for d in filtered_dir(bd) do
local dev=get_device(bd..d)
if removable(bd..d) or removable_class(dev) then
local i1=get_info(dev)
local i2=get_info_usb(dev)
if i2 and i2.vendor then
info=table.concat({i2.vendor, i2.model, i2.serial}, )
else
info=table.concat({i1.vendor, i1.model, i1.serial}, )
end
for v in scan_volumes(bd, d) do
table.insert(volumes, {
kernel=v,
info=info,
mountpoint=mounted(v),
})
end
end
end
table.sort(volumes,
function(a, b)
if a.mountpoint and not b.mountpoint then
return false
elseif not a.mountpoint and b.mountpoint then
return true
else
return (a.kernel b.kernel
Re: Keyboard focus lost in Swing applications (ex: Netbeans) after changing to window
On 2009-02-09, Daniel Clemente dcl441-b...@yahoo.com wrote: Well, it improved lots of things. It works perfectly with gtk, qt, tk, keyboard, mouse, ⦠The example code now works when I use Java 1.6.0_12-b04. But if I compile and run it with the unreleased 1.7.0-ea-b43, the focus is still buggy: You could, again, try reversing the order of the finalise_focus and sendmsg in the TAKE_FOCUS branch of the clientwin focus function. -- Stop Gnomes and other pests! Purchase Windows today!
Re: signal 11 from X 1.4.2 with nvidia driver
On 2009-02-09, Antonio De Leon aldle...@gmail.com wrote: it only happens when using ion3 from 200808 onwards. xfce and others work normally. Sounds like an Xorg bug. -- Stop Gnomes and other pests! Purchase Windows today!
Re: Keyboard focus lost in Swing applications (ex: Netbeans) after changing to window
On 2009-02-08 21:52 +0100, Daniel Clemente wrote:
Hi; thanks for the helpful explanations.
I tried doing that change and it stayed the same. I also tried doing a
sleep(2) before sending the WM_TAKE_FOCUS (with the +2s timestamp), but
without success.
What did work is this:
Try the attached patch. (Disclaimer: I haven't particularly tried it;
just quickly threw it together.)
--
Stop Gnomes and other pests! Purchase Windows today!
diff -rN -u old-ion-3/ioncore/clientwin.c new-ion-3/ioncore/clientwin.c
--- old-ion-3/ioncore/clientwin.c 2009-02-09 00:35:07.380883290 +0200
+++ new-ion-3/ioncore/clientwin.c 2009-02-09 00:35:07.488878930 +0200
@@ -1015,10 +1015,11 @@
{
if(cwin-flagsCLIENTWIN_P_WM_TAKE_FOCUS){
Time stmp=ioncore_get_timestamp();
+region_finalise_focusing((WRegion*)cwin, cwin-win, warp, stmp);
send_clientmsg(cwin-win, ioncore_g.atom_wm_take_focus, stmp);
+}else{
+region_finalise_focusing((WRegion*)cwin, cwin-win, warp, CurrentTime);
}
-
-region_finalise_focusing((WRegion*)cwin, cwin-win, warp);
XSync(ioncore_g.dpy, 0);
}
diff -rN -u old-ion-3/ioncore/focus.c new-ion-3/ioncore/focus.c
--- old-ion-3/ioncore/focus.c 2009-02-09 00:35:07.368883774 +0200
+++ new-ion-3/ioncore/focus.c 2009-02-09 00:35:07.484879092 +0200
@@ -347,7 +347,7 @@
/*Time ioncore_focus_time=CurrentTime;*/
-void region_finalise_focusing(WRegion* reg, Window win, bool warp)
+void region_finalise_focusing(WRegion* reg, Window win, bool warp, Time time)
{
if(warp)
region_do_warp(reg);
@@ -356,10 +356,7 @@
return;
region_set_await_focus(reg);
-/*xwindow_do_set_focus(win);*/
-XSetInputFocus(ioncore_g.dpy, win, RevertToParent,
- CurrentTime/*ioncore_focus_time*/);
-/*ioncore_focus_time=CurrentTime;*/
+XSetInputFocus(ioncore_g.dpy, win, RevertToParent, time);
}
diff -rN -u old-ion-3/ioncore/focus.h new-ion-3/ioncore/focus.h
--- old-ion-3/ioncore/focus.h 2009-02-09 00:35:07.368883774 +0200
+++ new-ion-3/ioncore/focus.h 2009-02-09 00:35:07.484879092 +0200
@@ -23,7 +23,7 @@
extern void region_warp(WRegion *reg); /* maybewarp TRUE */
extern void region_set_focus(WRegion *reg); /* maybewarp FALSE */
-extern void region_finalise_focusing(WRegion* reg, Window win, bool warp);
+extern void region_finalise_focusing(WRegion* reg, Window win, bool warp, Time
time);
DYNFUN void region_do_set_focus(WRegion *reg, bool warp);
extern void region_do_warp(WRegion *reg);
diff -rN -u old-ion-3/ioncore/group.c new-ion-3/ioncore/group.c
--- old-ion-3/ioncore/group.c 2009-02-09 00:35:07.272887653 +0200
+++ new-ion-3/ioncore/group.c 2009-02-09 00:35:07.460880061 +0200
@@ -241,7 +241,7 @@
if(st!=NULL st-reg!=NULL)
region_do_set_focus(st-reg, warp);
else
-region_finalise_focusing((WRegion*)ws, ws-dummywin, warp);
+region_finalise_focusing((WRegion*)ws, ws-dummywin, warp,
CurrentTime);
}
diff -rN -u old-ion-3/ioncore/window.c new-ion-3/ioncore/window.c
--- old-ion-3/ioncore/window.c 2009-02-09 00:35:07.332885230 +0200
+++ new-ion-3/ioncore/window.c 2009-02-09 00:35:07.480879252 +0200
@@ -182,7 +182,7 @@
void window_do_set_focus(WWindow *wwin, bool warp)
{
-region_finalise_focusing((WRegion*)wwin, wwin-win, warp);
+region_finalise_focusing((WRegion*)wwin, wwin-win, warp, CurrentTime);
}
diff -rN -u old-ion-3/mod_tiling/tiling.c new-ion-3/mod_tiling/tiling.c
--- old-ion-3/mod_tiling/tiling.c 2009-02-09 00:35:07.124893627 +0200
+++ new-ion-3/mod_tiling/tiling.c 2009-02-09 00:35:07.416881837 +0200
@@ -158,7 +158,7 @@
void tiling_fallback_focus(WTiling *ws, bool warp)
{
-region_finalise_focusing((WRegion*)ws, ws-dummywin, warp);
+region_finalise_focusing((WRegion*)ws, ws-dummywin, warp, CurrentTime);
}
Re: Multi-, different- monitors is multi-hopeless?
On 2009-01-18, ebik e...@drak.ucw.cz wrote: I don't use mod_xrandr, for me it is useless since it doesn't support xrandr 1.2. Xrandr 1.2 itself is useless shit-o-rama shit. *nix is dead. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and local vendors and artisans have had to yield to all under one roof big box hypermarkets.
Re: Multi-, different- monitors is multi-hopeless?
On 2009-01-18, ebik e...@drak.ucw.cz wrote: There is currently no option to multiple screens than xrandr today. Yes there is: it's called plain-old-X-multihead, or _the clean solution_ that doesn't smell of a rotten kludge to the other side of the globe. *Eugh*, adding more screens by pre-allocating a huge root window and then adding views to it, *eugh*, *puke*, *blaargh*. And that's the approach taken to everything these days in Futile/fascist Open Sore Software. (Futile, because it's destined to become just another incredibly poor bastard son of Windows and MacOS, instead of offering a real choice. Fascist, because it offers no practical choice anymore; only the _purely theoretical_ choice of forking the code yourself, wasting your life on it.) How about just creating new root windows dynamically? And using the same old APIs that are used for any other windows? -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: [SEMI-OT] Applications in a typical ion3 desktop environment
On 2009-01-17 11:47 +0100, Sylvain Abélard wrote: (The Modern Web counts as sheer lunacy. In a few years, given the trend, it will require triple-widescreen configurations for all the advertisements and other crap, yet still have a single 5cm text column in the middle.) The Web will not require a triple-wide screen since the trend seems precisely to get the content on a third of the width, two thirds being empty white (or styled) borders. The two thirds are mega-wide and force the text out of view in narrow browsers for sane content (such as my slightly smaller than portrait A4 or letter window on a 17 4:3 display) and viewing other stuff on the side. You need the tripe-wide screen configuration to not have to scroll horizontally to get to the text. In fact, an approx A4 sized screen [...] could be quite nice on a laptop. That would be great, readable, and not to mention easier to fit in normal bags. Of course, it would also be great if you could use it as a tablet in portrait mode... with an e-paper display mode. The current tablet PCs seem horribly kludgy, though, and, in fact, it might be better for the tablet mode to be the default, with the computer behind the screen and no keyboard except on a docking station that perhaps can be attached to the computer nicely for transportation. At least I don't think I'd miss the keyboard much when travelling lightly. The keyboard does work as a screen cover, though, and there needs to be something to replace that function. Surely it can be made lighter and less fragile than esp. tablet hinges. Or maybe the docking station can be made to fill this function, essentially making the computer of two detachable parts, so the tablet can be made less clunky. In fact, maybe you could also split the eletronics and battery between the two components, further reducing the tablet weight. E.g., a primary storage HD (or SSD) could go in the KB part, as well as any optical drives -- actually, it sucks that you have to again pay extra for a laptop that doesn't waste space for one; this is again something that belongs in a docking station or otherwise external device -- with a smaller SSD on the tablet part for the OS and some documents etc. The mechanical coupling between the parts of course has to be made reliable. -- Tuomo
Re: Keyboard focus lost in Swing applications (ex: Netbeans) after changing to window
On 2009-01-16, Daniel Clemente dcl441-b...@yahoo.com wrote: I was using Ubuntu GNU/Linux 8.04.1 with these packages: ii sun-java6-bin 6-06-0ubuntu1 Sun Java(TM) Runtime Environment (JRE) 6 (architecture dependent fil Etch, sun-java5-bin. Also, just now whatever they've got on this FC4. ( $ java --version java version 1.4.2 gij (GNU libgcj) version 4.0.2 20051125 (Red Hat 4.0.2-8) ) -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: [SEMI-OT] Applications in a typical ion3 desktop environment
On 2009-01-16, Philip Snowberger psnow...@nd.edu wrote: On Sun, Jan 11, 2009 at 10:23 AM, Tuomo Valkonen tuo...@iki.fi wrote: bought almost an identical x61s about 18 months previous. That said, it's kinda sad that the only reason it could be considered 'affordable' is that Lenovo was blowing them out so that they can hurry up and end-of-life them (I guess?). One thing I've been wondering is, whether a widescreen display on a laptop (not a dragtop!) is actually a bad idea too. And it's the trend. For most _sane_ documents, the height of the screen is more important than width. (The Modern Web counts as sheer lunacy. In a few years, given the trend, it will require triple-widescreen configurations for all the advertisements and other crap, yet still have a single 5cm text column in the middle.) You'd need at least 15.4 WS (1.6 aspect) to fit a single A5-sized document (21cm tall) on the screen without noticeable downscaling. Such a laptop is already quite huge width-wise. In 4:3 aspect ratio 14 is enough, and width-wise it's about the same as 13.1 WS, which just about fits a keyboard without it looking crammed. (I don't have much direct experience with laptops.) And a 12 4:3 is approximately as tall as a 14.1 WS. (A5 is what you get when you print two pages on one A4 sheet, and anything noticeably smaller than that is too small for documents that you have not specifically prepared for printing small -- see earlier post. In fact, an approx A4 sized screen -- A4 has 14.5 diameter -- with sqrt(2)/1=~1.4 aspect ratio, which is wider than 4:3=~1.3 but taller than 1.6, could be quite nice on a laptop. The WS laptopss themselves actually have about the aspect ratio of A4, but tend to have huge uneven borders around the screen. [1] on the topic of mail clients, I haven't seen anybody mention 'sup' ( http://sup.rubyforge.org/ ), which didn't completely suck last time I tried it (early 2008). I've heard of it before and on principle it seemed almost what I've wanted. However, at least then it suffered from some limitations that I can't recall now, so I didn't bother trying it. I also read mail on a remote computer that I don't control, so installing programs written in one of the Popular Bloated Scripting Languages isn't very straightforward. it, I noticed on the mailing list lots of Architecture Astronaut talk about splitting it into sup the service and sup the client. YMMV, it may be shit by now. That sounds bad. [2] I used to use remind + wyrd with great success and much happiness until I started using google calendar+mail for work. The only thing I'd really need for remind to keep working for me is a better backup solution for ~/reminders than I had at the time. Every once in a while I think about going back to remind+wyrd. I use a combination of xmessages and the cell phone for reminders, depending on the importance of the reminder and whether I should be near a computer anyway to act on it. (Why can't computer calendar software be as simple as that in cell phones? Wait! Probably they've managed to ruin them in them too with increased screen size and CPU power.) ~$ cat bin/atxm #!/bin/sh if test $# -lt 1; then echo 'Usage: atxm time [message]' exit 0 fi TIME=$1 shift cd / ( echo '(iconv --to=latin1 | xmessage -display :0.0 -file -) EOF' if test $# -gt 0; then echo $@ else cat fi echo EOF ) | at $TIME -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Keyboard focus lost in Swing applications (ex: Netbeans) after changing to window
On 2009-01-15, Daniel Clemente dcl441-b...@yahoo.com wrote: In fact any Swing application has this problem. You can use this one to test: CBA going through all the trouble of installing java compilers and whatever dependencies there may be. Java on Linux is pain; already been too much trouble installing the VM back in the day. Why is it so? Probably because Swing is crap and sets the focus somewhere where it shouldn't be -- in violation of the ICCCM. X doesn't forward focus changes to the WM is requests, but rather obeys each and every one (unless the target window is hidden, etc.), so the WM should have to start fighting agianst them. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: [SEMI-OT] Applications in a typical ion3 desktop environment
other stuff latex, rubber, ... dramatic pause ... bibtex, metapost, xdvi, xpdf, gv, psbind -3 [*], dvipdfm. And should we start listing latex packages too? Seriously though, this listing of all the software one uses and that is far removed from setting up a consistent ionic operating (no desktops here!) environment, could be seen as spamming the list. A wiki would be a better location for this sort of stuff but, alas, it's too much of a hassle to set up one that wouldn't start sucking. Maybe some site offers ad hoc wikis, a bit akin to pastebots? [*] With narrow text column classes such as amsart, three pages fit on one side of an A4 just fine with psbind, and saves a lot of paper -- together with duplex printing, of course. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: [SEMI-OT] Applications in a typical ion3 desktop environment
On 2009-01-11, Sylvain Abélard sylvain.abel...@gmail.com wrote:
I also heard you can find companies that replace MacBooks screens with
matte ones for $100.
FYI, the matte screen option is $100, only available on the biggest $2500
Mac.
Yeah, they only seem to sell matte displays on corporate-priced stuff
worth more than your monthly salary. To people who want quality and
are ready to pay for it. The typical idiot consumer is all wet over
a 19 (wow, big number!) dragtop with a glossy (shineee! my precious!)
finish.
Matte display finish is like green paint: worth more than the device
it is covered with. \end{army joke}
(In theory e.g. the el cheapo Thinkpad SL400 should be available in
matte, but all the shops only seem to carry glossy versions.)
--
Dreams of the past: Computers will automate menial tasks. Paperless office.
Reality: Paper forms replaced by a hundred times more electronic forms
required to be filled by human slaves for management departments.
ion-3-20090110
This is yet another maintenance release. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and local vendors and artisans have had to yield to all under one roof big box hypermarkets.
Re: path, alias with F3
On 2009-01-07, ebik e...@drak.ucw.cz wrote: Which scripts? If you set environment variable PATH for ion to contain ~/bin, it will search executables there also. Ion also has its own search path for its scripts. You can set it from the command line, or modifying the searchpath field of ioncore.set/get_paths. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Toggle horizontal/vertical maximize behaviour
On 2009-01-05, Chris Burkhardt ch...@mretc.net wrote: Ah, I see. But what I think is still a bug is that after performing a 'Mod1-K V' in the other top frame (as in step 5 in my example), then the originally maximized frame forgets its original height. 'Mod1-TAB'ing to it and trying to un-maximize it with 'Mod1-K V' has no effect, although even with the intended per-frame maximize I would expect it to (right?). The logic is that frames store their size when they're specifically shaded or maximised, and forget the stored size whenever they are otherwise resized -- including resizing is enforced in a tiled layout whenever some other frame is resized. Fixing these annoyances you've pointed out, involves a different kind of logic not based on maximizing individual frames, but altering the entire layout somehow. Thanks for everything, I hope your PhD defense/preparations went (are going) well for you. It was on the 22th of December already. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Toggle horizontal/vertical maximize behaviour
On 2009-01-02, Chris Burkhardt ch...@mretc.net wrote: 4) Again in the top frame do a 'Mod1-K S' to do a horizontal split. 5) In one of the top frames do a 'Mod1-K V' to collapse the bottom frame. Then 'Mod1-TAB' to the other top frame and 'Mod1-K V'... the bottom frame unexpectedly stays at 0! Ah... well I'd classify that as an (unintentional) feature then. Maximise is completely unaware of tiling/layout, it just works on a per-frame basis. It should be done somewhat differently (snapshotting entire layout instead of just saving the size of a single frame) for what you want to work. It'st hus will not be changed in the stable Ion3. For Ion3plus a good-quality patch would certainly be accepted; if I ever find the time, I may even do it myself. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: xkbion.lua
On 2008-12-19, Anton Yuzhaninov cit...@citrin.ru wrote: xkbion.lua from ion scripts collection works for me only with this patch: Please submit the fixes to the repository. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and local vendors and artisans have had to yield to all under one roof big box hypermarkets.
Re: ion scripts location
On 2008-12-19, Anton Yuzhaninov cit...@citrin.ru wrote: darcs get http://iki.fi/tuomov/repos/ion-scripts-3/ Works for me. However, http://modeemi.fi/~tuomov/repos/ion-scripts-3/ may be faster; at least darcs1 used to handle redirects stupidly. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Toggle horizontal/vertical maximize behaviour
On 2008-12-15, Ilya Schurov ilya.schu...@noo.ru wrote: resize commands. (E.g. I see only tab heading, but if I switch to that heading and try to do ALT+K-V or even ALT+R-N, there's no reaction.) The only way to unshrink it is to resize main frame. Is it a bug or a feature? :) For some reason, Alt+R Mod1+N does work, although it's the move binding. I may look into this more closely next year; now I have other things to concentrate on. -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: Tab width in 'shaped' bar style
On 2008-11-29, Oskar Nordquist [EMAIL PROTECTED] wrote: However, the tab width depends on the maximum tab width in their containing frame. Is there any way for the individual tabs to only stretch as far as their own title width? Other than touching the code for a custom hack? No. Tabs have uniform widths. For a good reason. (I find varying widths confusing.) -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. By 2008, Linux has bloated into a gas-guzzler, and the cycle paths have been replaced with polluted motorways.
Re: statusd_mpd-socket.lua
On 2008-11-19, Marc Hartstein [EMAIL PROTECTED] wrote: A drop-in replacement for statusd_mpd.lua which uses sockets instead of popen to communicate with mpd. Suggest renaming to replace the other script. I suggest sending it to the scripts repository for easy availability. Also, on a cursory glance, it appeared to be a timer hack. If it's at all possible to get the file descriptor from Lua, I suggest adding it to those selected on for event-based processing. This will require a little bit of C code to add the desciptor, since popen_bgread is currently the only interface and unlikely to work correctly on sockets. While to ion3plus, certainly new interfaces could be added to support selecting on arbitrary fds from the Lua side (and the primary reason the support isn't there, is that fds are not well available from Lua), this can be done with e.g. the TCC/alien/c/invoke hacks disgussed e.g. at http://www.mail-archive.com/ion-general@lists.berlios.de/msg02559.html -- In 1995, Linux was almost a bicycle; an alternative way of live to the Windows petrol beasts that had to be taken to the dealer for service. In 2008, Linux has bloated into a gas-guzzler, and cycle paths have been replaced with polluted motorways.
Re: Can't type queries
On 2008-11-16, sughosh ganu [EMAIL PROTECTED] wrote: The problem was that ion did not recognize the locale on my system (en_IN). Changing it to en_US.UTF-8 solved the problem. Thanks for the help. It's not Ion. It's Xlib. en_IN.UTF-8 should work, if it's been compiled. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Can't type queries
On 2008-11-16, Tuomo Valkonen [EMAIL PROTECTED] wrote: On 2008-11-16, sughosh ganu [EMAIL PROTECTED] wrote: The problem was that ion did not recognize the locale on my system (en_IN). Changing it to en_US.UTF-8 solved the problem. Thanks for the help. It's not Ion. It's Xlib. en_IN.UTF-8 should work, if it's been compiled. I'm not exactly sure why Xlib handles LC_CTYPEs without the encoding part specified in a fucked up manner, but it does have a locale.alias database of its own, /usr/share/X11/locale/locale.alias, which does e.g. list en_IN: en_IN.ISO8859-1 If Xlib uses this database instead of the information from libc, nl_langinfo(CODESET), and libc has been compiled with a different character set for en_IN, it could be why it's fscked up. The modern (read: shit) toolkits afaik pretty much reinvent all this, being utf-8 monoculturistic [ which also results in nobody bothering to fix Xlib, the generic multibyte routines of which handle utf-8 locales very badly ], and xterm also uses the Xutf8 routines instead of Xmb routines when in an utf-8 locale. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: two quick misc. questions - labeling workspace and changing title bar height...
On 2008-11-14, Juri Mianovich [EMAIL PROTECTED] wrote: I did this, and the workspace still has: xterm5 in the title. It's not workspace title. It's the window title. Workspace titles aren't shown. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Can't type queries
On 2008-11-15, Tuomo Valkonen [EMAIL PROTECTED] wrote: On 2008-11-14, sughosh ganu [EMAIL PROTECTED] wrote: So pressing F3 prompts for a command, but typing does not produce any string. Same is the case with other keys like F4 and F5. IIRC it's a locale setup fuckup. You have to have the encoding specified as part of LC_CTYPE for X, e.g. LC_CTYPE=fi_FI.ISO-8859-1, and you have to have the locale compiled for libc. (Debian: config /etc/locale.gen, generate: locale-gen.) Another thing you might look for is, if the startup errors (non-logged; you'll have to dig into the console or .xsession-errors of you run a DM and it supports it) has complaints about input methods or input contexts, in the case that you're running in a strange locale (as your name might imply). Ion only has very basic support for X input managers, so pre-edit windows etc. might not work. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Ion rewriting
On 2008-11-13, Roy Lanek [EMAIL PROTECTED] wrote: etc. A bit like with ice hockey, which maybe (Finland) you know better than scuba diving. I don't. That approach doesn't scale to big projects Well, I have not proposed to rewrite Linux or PostgreSQL in Haskell yet ... Even Ion is far too much work to rewrite, with little practical benefit. Rewrites of working code just to conform to fads are pure insanity. I rather concentrate my academic wanking on the far more interesting mathematics... Indeed, one of the biggest obstacles I have in producing new code is that I'd need to be able to take a lot of time to properly concentrate on it. But I don't have that time and energy, and it's always a lot of effort to start working for a few spare hours on something you only ever look at for a few hours every couple of weeks. Maintaining an existing codebase that you've previously had more time and energy to concentrate on, so it's burned into your neural patterns, is far easier. One should be able to get into The Zone for an extended period to properly get going with a new project, but with projects bigger than a single-file hack, it's difficult to find the time and energy. Indeed, I found The Zone to write Riot in, immediately after having lost the shitjob I had back then. (One of the best days of my life.) -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Ion rewriting
On 2008-11-13, Roy Lanek [EMAIL PROTECTED] wrote: Which The Zone?! ... The Zone, the flow, super-concentration/immersion on/in a problem. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: two quick misc. questions - labeling workspace and changing title bar height...
On 2008-11-13, Juri Mianovich [EMAIL PROTECTED] wrote: Hello, When I create a workspace, it is named: empty frame That's not the workspace name; it's just indicator of an empty frame. my favorite xterms (or whatever) and have that name/title stick ? Mod1+M workspace/rename Second, I'd like to change the height of the title bar slightly and have that new height apply to all future title bars. How can I do that ? Change the style file. PS. Turn on line in your mail program. Max. ~76 characters/line. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: quickly cloning an existing statusbar script (date)?
On 2008-11-13, Juri Mianovich [EMAIL PROTECTED] wrote:
If that is correct, would you tell me what/where the monitor
is ?
Where ever you found the following line:
statusd.inform('date', os.date(settings.date_format, tm))
to this:
statusd.inform('date -v +7H', os.date(settings.date_format, tm))
does that seem reasonable ?
The monitor name must be parseable, and should be based on the
file name. RTFM:
http://modeemi.fi/~tuomov/ion-doc-3/ionconf/node6.html#SECTION0064
--
[Fashion] is usually a form of ugliness so intolerable that we have
to alter it every six months. -- Oscar Wilde
The computer industry is the only industry that is more fashion-driven
than women's fashion. -- RMS
Re: Ion rewriting
On 2008-11-13, Philip Snowberger [EMAIL PROTECTED] wrote: On Tue, Nov 11, 2008 at 12:53 AM, Tuomo Valkonen [EMAIL PROTECTED] wrote: On 2008-11-10, Tuomo Valkonen [EMAIL PROTECTED] wrote: And Haskell just sucks for the moderately object-oriented approach that is obvious for certain things. OTOH, Haskell and Parsec [1] are simply wonderful for writing parsers. Do you find that's the case because of pattern-matching function definition syntax ? In case of Parsec, I think it's more about how monads can be used to create domain specific languages (DSLs), and the monad syntax. Of course, one can write primitive parsers with the language's pattern matcher too, but Parsec is even more convenient than for anything except the most trivial parsers. That said, pattern matching and algebraic data types / tagged unions are a very nice thing. If there are two features I'd most like C to have without significantly altering its nature, it would be tagged unions and subtyping (i.e. the most basic component of inheritance). -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Ion rewriting
On 2008-11-10, Tuomo Valkonen [EMAIL PROTECTED] wrote: And Haskell just sucks for the moderately object-oriented approach that is obvious for certain things. OTOH, Haskell and Parsec [1] are simply wonderful for writing parsers. In fact, I use Haskell in the build process of Ion for parsing translatable strings from Lua source with lua-xgettext [2]. With Parsec, the parser [3] can be pretty much written in a DSL monad as the language is specified, without crummy parser generators. [1]: http://www.cs.uu.nl/~daan/parsec.html [2]: darcs get http://modeemi.fi/~tuomov/repos/lua-xgettext/ [3]: http://modeemi.fi/~tuomov/repos/lua-xgettext/Kuu/Parser.hs -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Ion rewriting
On 2008-11-10, Roy Lanek [EMAIL PROTECTED] wrote: Haskell language-of-the-day ... why?, Umm... increasing popularity? Oh dear, hello Moses- (the illiterate who brought back the astonishing writings) alike, or Muhammad- (ditto [illiterate]) alike. And who did write your frameworks ages ago? I have written object crap etc. for C. I can use them comfortably enough. The language just isn't meant for that kind of stuff. Can't tell now. On the other hand: Roll Your Own Window Manager: Tracking Focus with a Zipper They have a adaptive data structure for O(1) access to currently focussed window! How incredible! How about all the other state updates triggered by client software? A window manager depends on a lot of state, and pure FP makes it difficult/laboursome to maintain it. Pointers are comfortable. You do have a good, comfortable life, there in Finland, eh? :) Finland is a shithole. Bullshit. It's very very laboursome to make fast Haskell code, and especially such that doesn't eat gigabytes of memory for breakfast. I have, e.g., *met* reference counting in the *real world* for the first time It's not ref. counting/GC/shit I'm talking about. I'm talking about all the boxing and shit. The basic data structures incredibly inefficient, so you have to use other more hacky ones for efficiency, and they are again more difficult to use generically, may demand messing in the IO monad, etc. (The uniqueness typing of Clean seems more comfortable..) -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Ion rewriting
On 2008-11-10, Roy Lanek [EMAIL PROTECTED] wrote: instead of their heads.) In the hands of a great programmer (like James Clark), it's astounding what can be done in a few lines of Haskel, and it's interesting to compare the amount of Java code it takes the same programmer to implement the same algorithm. Yep, you can do incredible things in 10 lines. But to think how to write those few lines takes ten times as long as the writing the corresponding 1000 lines of C. That approach doesn't scale to big projects, where you have some other objective than an academic exercise in the language in question. And Haskell just sucks for the moderately object-oriented approach that is obvious for certain things. You have to build workarounds upon workarounds, because the obvious approach can't be sanely done.[*] And it gets very tiresome, if you have to try to come up with a non-obvious model for everything in a big project. The massive threading approach that I proposed might be workable and functional (as well as actually message passing OO), perhaps even scalable, but it also would need better language support to be convenient to use. I am actually very very slowly working on something in Haskell, dunno if I'll ever finish, and in all the time I've spent fighting with the language, I'd already have finished the project in C. Actually many things that would be a single line in C are tens of lines of Haskell... But the entire project is really an academic exercise. [*] That said, a lot of time should be spent on designing the API of important core libraries etc. that everyone gets to suffer from, to avoid the crummy obvious approach, and maybe come up with something nicer. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Ion rewriting
On 2008-11-10, Leslie P. Polzer [EMAIL PROTECTED] wrote: Ever tried Common Lisp? I'm not a fan of Lisp syntax. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Ion rewriting
On 2008-11-10, Roy Lanek [EMAIL PROTECTED] wrote: What--and I am serious--about rewriting Ion in ... Haskell? Rewrites in The Language of the Day are lame (see also the .signature), for people with a lot of time in their hands and no original ideas. Why so?!, because Ion is written--shudder--in the US-dollar of the programming languages anno 2008 still; hence, Tuomo (but not only he) has, to ... I am paraphrasing Torvalds, play at the masturbating monke^H^H^H^H^Hreindeer every time he needs to change something in the code that goes beyond cosmetics, and be it as as an experiment only, quite likely. ( C's *expressivity* oblige.) C works quite well, especially combined with Lua for the tasks that are not nicely done in C. Working on Ion's code, I don't often run into the limitations of C, because I have my frameworks written ages ago. What would make sense is gradually moving some of auxiliary C code to the Lua side. I really don't like writing the kind of stuff a window manager is in Haskell. It's too painful. The language just isn't meant for that kind of stuff. Dynamic data structures (e.g. objects) are _pain_; perhaps the best way to get around this is _heavy_ threading, every object being its own thread... it's an academic exercise that I have no time for, and of little practical benefit over a tried C implementation that works today. I skip mentioning other advantages in general, moreover I am sure Tuomo perfectly knows that Compiled Haskell is as fast than C Bullshit. It's very very laboursome to make fast Haskell code, and especially such that doesn't eat gigabytes of memory for breakfast. Lie, statistic, benchmark. Liar, politician, software advocate.. Lua may be fun, but Haskell (and other modern functional languages) are no match ... on standard environments at the least. Haskell is fun for academic exercises. Real-world programs are pain. Plus the Haskell build environments are shit. GHC is infinite pain to install anything under. All these language fundies (which includes the majority of FP people) are always reinventing the OS for their language only, instead of using the operating system's tools (such as the file system). dwm and all its clone and derivatives look like nano (the editor), or pico of the window managers pretty much to me, where ion could be vi; Ion is joe http://joe-editor.sf.net/. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Ion rewriting
On 2008-11-10, Christian Walther [EMAIL PROTECTED] wrote: I'll never forget one of my teachers telling me that I used java to store my functions. I'm still not sure that I understand what he meant. ;-) One example of Pure Uglyness in programming languages are the OO first class function hacks. i.e. writing some kind of Function1Int,Int template class, overloading the call operator, etc. That's why I think that this discussion is kind of... pointless. Of course it is, as are rewrites just because the language used isn't in fashion anymore. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Disable automatic saving of workspace (frame layout)
On 2008-11-06, Sylvain Abélard [EMAIL PROTECTED] wrote: How do I save workspace and frame layout? Ion does it automatically for you when you exit it cleanly. Can ion be configured not to do this? A long time ago, my hack was to remove the write rights to the saved-layout file. If there isn't a better way now, I hope it helps Exit with ioncore.resign() instead of ioncore.shutdown() -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Disable automatic saving of workspace (frame layout)
On 2008-11-06, Daniel Clemente [EMAIL PROTECTED] wrote: The restart will save the session unless the source code is modified. Or maybe it can be overriden in Lua? No. And it doesn't make much sense to override; saving is needed for sane restart behaviour that will place the windows where they were. (Of course, one might use some X root window property too in this case.) The primary reason for the existence of resign() is also being able to quit the WM under a session manager (that supports remote exit) without shutting down and saving the entire session. If one wants a static setup, it's perhaps best to just clean up files in .xsession. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: Disable automatic saving of workspace (frame layout)
On 2008-11-06, Tuomo Valkonen [EMAIL PROTECTED] wrote: If one wants a static setup, it's perhaps best to just clean up files in .xsession. Perhaps this needs a clarification: you can put a saved_layout.lua in ~/.ion3/. If one doesn't exist in the session directory (which one can explicitly specify with -session), this one is loaded. If you remove the one in the session directory in .xsession, restarts should work fine, as the state gets saved over restarts. (It's also possible to do some search path mangling in Lua and from the command line, and maybe even set X properties to detect restarts. But it's more kludgy. And the session directory is always the first directory looked into, despite other search path orders, which excludes certain options.) -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Moving tabbar to bottom of the screen
On 2008-10-31, Eider Oliveira [EMAIL PROTECTED] wrote: Is there a way for me to place the tab below the frame, instead of on top of it? Other than writing a patch, no. It shouldn't be too difficult. There's already the 'bar' parameter of frame styles, that supports the modes inside, outside, and shaped. Just add -below variants, and wherever FRAME_BAR_* is referred to, add extra cases for the below geometry. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: OpenOffice focus
On 2008-10-13, John Harrigan [EMAIL PROTECTED] wrote: Both Acrobat and OpenOffice open in the 'Aux' frame as desired, but starting OpenOffice moves focus to the 'Aux' frame. Acrobat does not change the focus. Changing switchto from true to false doesn't seem to change this behavior. It is possible that OO is crap [of course it is, and utter and total at that] and forces the focus, in violation of the ICCCM [1]. Focus change is not a request, but a command that the X server follows. ICCCM establishes protocols on when it may be used. [1]: http://www.tronche.com/gui/x/icccm/sec-4.html#s-4.1.7 -- Be an early adopter! Beat the herd! Choose Windows today!
Re: maximizing
On 2008-10-08, panman [EMAIL PROTECTED] wrote: I've been trying to figure out where the problem is with maximizing any of the flash video players out there, google, youtube, etc to the frame they're in. They all try and then return to the normal embedded player mode. Does anyone have any input into that? They're crap that try to be their own WMs. Flash in particular doesn't think it should be in FS mode if it doesn't have the focus. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: maximizing
On 2008-10-08, Sylvain Abélard [EMAIL PROTECTED] wrote: Flash in particular doesn't think it should be in FS mode if it doesn't have the focus. Flash has many defects, but is this one so bad ? It is. Most importantly, there's a race condition: if the flash window doesn't get the focus fast enough, when it goes into FS mode, it will close. But this trying to outsmart the WM also means you can't open the scratchpad while flash is FS, can't switch back to another WS while flash is FS, or the window will un-FS. I wouldn't like the ugly Flash ads on a web-page to make itself fullscreen without asking, This feature won't stop that, if the flash crap can avoid the above-mentioned race condition. Flash is made for dummies. All Web technology invented after HTML 2.0, with the exception of CSS, is for dummies. (And even CSS is too complicated. I suspect it could be simplified a lot, if would only have to cater to document layout, not application layout. And applications would benefit from something not suited to documents. *sigh* again trying to do everything with one language.) Be an early adopter! Beat the herd! Choose Windows today! Your troll-power is exceptional :D Just stating where things seem to be going. PS: would it be easy to fix this ML's reply-to? Is it possible, do we want to avoid it? There's a problem? It should not be set, and there's a school of thought that thinks this is how it should be. I have no particular opinion (!) about the matter, but tend to slightly side with not setting it. I just wish MUAs were more intelligent about List-Post and headers. Even mutt -- the mail user agent that sucks the least -- is stupid with regard to mailing lists. It sucks having to configure the lists in .procmailrc, include a subscribe line in .muttrc, include a mailbox line in .muttrc for checking for new mail (or alternatively, if not procmailed into separate folders, write save-hooks), and then remember to press L instead of r to reply. The default reply key should ask what to do when there are multiple options (list, group, or just the poster, etc.). Setting Reply-To is a workaround to retarded MUAs. I just use slrn and nntp://news.gmane.org whenever possible. -- [Fashion] is usually a form of ugliness so intolerable that we have to alter it every six months. -- Oscar Wilde The computer industry is the only industry that is more fashion-driven than women's fashion. -- RMS
Re: pidgin add a buddy transient not moveable
On 2008-10-07, A K [EMAIL PROTECTED] wrote:
I have my pidgin (.im) buddy list nice and narrow on the right of my
screen. When someone wants to add themselves as a buddy of mine a
transient window pops up but half of it is hidden and when I move the
buddy list to another (bigger) frame the transient doesn't move with
it.
How is the window hidden? Is it clearly inside the frame of the main
window, or could it be just partially outside the screen?
These days Ion automatically detaches (or unsqueezes) big transients
that don't properly fit in the frame, and they don't follow their main
windows around. You can disable this behaviour with
ioncore.set{ unsqueeze = false }
--
[Fashion] is usually a form of ugliness so intolerable that we have
to alter it every six months. -- Oscar Wilde
The computer industry is the only industry that is more fashion-driven
than women's fashion. -- RMS
Re: Move frames to and from scratchpad
On 2008-10-05, Daniel Clemente [EMAIL PROTECTED] wrote: I have been using the function you sent and it works very well; thanks. With Tuomo's improvements, I think it would be very helpful for the scratchpad feature of Ion. Could it be included? Just send it to the scripts repo. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: unfocused floating windows
On 2008-10-05, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
My question is, is it possible to force a
window not to have focus even if the pointer is on it (something
similar to what happens to the rest of the windows when the scratchpad
is visible)?
ioncore.set{mousefocus='disabled'} will disable pointer-based
focusing entirely. Restore with 'sloppy'. Currently there are
no other options, as nobody's been arsed to write them.
Also, it's possible to make floating windows 'passive'
(setting passive = true in WMPlex.attach[_new] parameters),
but there's no winprop hack for it; using it would demand
some custom scripting.
--
Be an early adopter! Beat the herd! Choose Windows today!
ion-3-20081002
Some minor fixes again, including a few that should've already been in the previous release. -- Be an early adopter! Beat the herd! Choose Windows today!
Re: Move frames to and from scratchpad
On 2008-09-25, Canaan Hadley-Voth [EMAIL PROTECTED] wrote: After trying several combinations, I found one that correctly brings a frame from scratchpad to the background frame: In Lua console: _:screen_of():mx_current():current():current():attach(_sub) In a key binding: kpress(META..Shift+k, _:screen_of():mx_current():current():current():attach(_)), I have been using this function to do what you are trying to do. You both are failing to simply recursively apply WRegion.current, until something interesting is found. Something like: reg=cwin:screen_of() function scan(reg) if not reg then return nil elseif obj_is(reg, WFrame) then return reg:attach(cwin) elseif ... then ... else return scan(reg:current()) end end scan(cwin:screen_of()) Maybe it indeed makes sense to use WTiling attach if discovered, as it can better find something useful, if the current isn't actually a frame (unlikely). -- Tuomo
