As more of a end user than a developer - looking from the outside in so-to-speak, and a relatively new user at that (9 months or so), my first impressions were that there were a Heinz variety of desktops (Puppy's) to choose from, all much the same, all doing much the same thing (launching programs that actually do stuff), remastered 101 different ways.
All I wanted was to get to the programs that actually did the stuff I needed, browse, write a document, edit a image, download burn video's from tube etc.
The immediate barrier was having to learn the core workings to make (what a end user might consider) small changes. Add a item to the tray bar - hmm! right click didn't work (jwm), off to boards/tech papers to find that you need to edit scripts (code).
Change the theme and that only applies to certain bits, not others - but I wanted that to be consistent across all - not a different menu bar set of colours to that of program menu colours.
Plug in/use a non supported device and you have to get pretty technical again to get that working.
And then there's a plethora of techie discussions about the desktop and menus that enable users to load programs, but when it comes to the actual programs often you'll find that there are bugs that put end users off. For me for instance Abiword (latest at the time) crashed after a lot of work/effort - I've since abandoned that altogether (and now purely use LibreOffice). Softmaker functional - but File, Open and I can't read the text comfortably, so again abandoned. Openshot for editing video and I've not come across one that fully works as intended - and you have to have the exact/right choice of Blender and Inkscape versions to go with a particular Openshot version.
Fine for those that enjoy learning the systems technicalities, not so good for end users that just want to do stuff.
By the time a user does get a decent working system that does the stuff they desire, so many changes have had to be made that moving on to another version becomes a daunting task as many of the forgotten changes would have to be reapplied to the new version.
If as much effort went into making things more user friendly, then more users rather than developers/coders would stick around. (Relatively) simple things like add a right click or drag/drop option to jwm panel to bring up configuration dialogs, interlink GTK/JWM etc themes (or at least provide the same theme for each).
Drag something to the desktop - then you need to add a icon for that item. Where are the icons? As a end user it appears all over the place and in cryptic sub directories - why aren't they all together in one single place (seems simple enough to drop them all into a single directory and then sym link other directories to that single directory).
Personally I now love my puppy and I've learned a lot over the last 9 months - and use it exclusively. Common colours for menus/themes, applications that actually work well and as intended. I don't want to have to continually learn and tweak however, nor should a end user have to put in the time/effort to get to a reasonable working 'tool' (system). Whilst I persisted, many I suspect wouldn't. What might seem simple/obvious to developers/coders is often double-dutch to users.
I'd guess that most of puppy users are developers/coders or those with a techie hobby type character. Limited audience/user-base - limited usage.
To me, core puppy should be a lightweight set of applications and where the desktop can be easily tweaked/changed. Basic tools such as Leafpad, galculator, mplayer, CD tools, network connect etc. Other programs - browser, word processor, video editior etc should all be separate add-ons such that the user can select which choices to add (or not). SFS's are ideal for that. Instead users have to wade through 101 different puppies, some with this program installed but not that one, another with that one installed but not this one ...etc. and as to which core puppy a user should initially choose for their particular hardware - its more often repeated trial and error to test out your patience/persistence - a trial by ordeal that many potential users I suspect fail to pass.
Standardization of Puppy Linux Desktop and App Platform
yes ..one of those boring standards.... still have choice just makes life easier. Was ages before there even was a script to mimic the standard handling of desktop files at all and the categories are still a bit of a one off.Is that how it's done in xfce, or other desktops?
Its also going to be much more efficient if whatever window/desktop manager is in use if a built in binary function does the job rather than an add on script. No point in a 'lightweight' window manager if you need a pile of scripts or binaries to fill in the blanks.
mike
Puppy is what it is. If you are "just" an end-user who wants to have the basic apps and a minimum of a futzing around to get things working, may I ask why you do not simply stick to one of the mainstream distros like Ubuntu, perhaps in a less radical version like Mint or a lighter version like Lubuntu? I run Puppy on 3 computers for playing around as a hobbyist. But for "end-user" mode I use LXLE (a Lubuntu spin).rufwoof wrote: All I wanted was to get to the programs that actually did the stuff I needed, browse, write a document, edit a image, download burn video's from tube etc.
As bark_bark_bark says, people and their needs are all different. The Linux world is full of options -- there's no need to be fighting with a distro that doesn't fit your requirements.
I still have a stack of around 50 DVD's that I burned in the early days (around 10 months ago) when first migrating from XP (in view of at that time XP support soon to be withdrawn).The CD burning was through ignorance of not knowing how to use a frugal HDD based puppy with Grub.may I ask why you do not simply stick to one of the mainstream distros like Ubuntu, perhaps in a less radical version like Mint or a lighter version like Lubuntu?
Many didn't boot to my hardware - typically either black screen or no wireless network. Of those that did some were much slower in operation. Slacko was one of the first that was OK and after trying 5.6 for a while I switched to using the thinned down 5.3.3 version that micko put out some time back.
Having reduced down that 5.3.3 thin even further - down around a 75MB size and found the main apps that I use that work well with that I'm content to stick with that now. Fixed unchanging version other than firefox portable that updates to the latest version whenever that's released. Runs almost entirely in ram (excepting firefox and the sym links to preserve config changes (I don't use a savefile)) and is grease lightening in speed. So now somewhat spoilt as after you get used to booting in seconds and everything running near instantaneously everything else seems slow. The Ubuntu repositories might be better/wider, but I have most of the apps I require working ok under Slacko. IIRC Ubuntu LiveCD was really really slow.
Yes I have copies of Knoppix and a talking based version of that, Mint, Ubuntu, Slitax, DSL etc etc etc. and numerous versions/remasters of Puppy. For me Puppy Slacko proved to be the best choice, but only after a extensive steep learning curve and a lot (for me) of tweaks to get the desktop/themes/display as I like and stable apps (SFS's) to go with that.
-
Moose On The Loose
- Posts: 965
- Joined: Thu 24 Feb 2011, 14:54
[quote="NeroVance"]
The only gripes I have are the loss of Tcl/Tk, other than that, having a good standard is useful, though I do admit I actually loathe the constant usage of Bash for graphical applications
[quote]
What is there not to like about Bash? It allows complete programs to be easily and efficiently written without adding another interpreter to the OS. I went to the bother of learning Tcl/Tk at one point. I never found a case where it was better suited to my purpose than bash driving gtkdialog3 has been.
Perhaps if I was porting things also onto Windows(spit) I may find use for Tcl/Tk. I have found that a lot of windows programs that allow scripting seem to use it rather that trying to make something that works in Visual Basic.
The only gripes I have are the loss of Tcl/Tk, other than that, having a good standard is useful, though I do admit I actually loathe the constant usage of Bash for graphical applications
[quote]
What is there not to like about Bash? It allows complete programs to be easily and efficiently written without adding another interpreter to the OS. I went to the bother of learning Tcl/Tk at one point. I never found a case where it was better suited to my purpose than bash driving gtkdialog3 has been.
Perhaps if I was porting things also onto Windows(spit) I may find use for Tcl/Tk. I have found that a lot of windows programs that allow scripting seem to use it rather that trying to make something that works in Visual Basic.
-
gcmartin
This is one of 2 words in Puppyland which evokes some real divisive responses. It a shame that it does because the real intent is to bring harmony. So the word and a perverse vision gets in the way of trying to see the benefit that the topic is trying to bring into focus. (kinda like the word "hacking")
YES, I fully agree that development would benefit from some sort of desktop harmony to reduce the pain they go thru when introducing new functionality into PUPPY Linux and community usage.
Should some harmony (really a standard) takes root, it should be made clear what and which timeslot of PCs it will address. We cannot continue to support 486s and the like in NEW future PUP development. Much is already ever present as it has already been done many times in our past.
We must accept the great things that has been done and lock those in. And we must be clear about where the landmark is to begin any effort where this kind of harmony is to exist. I am not so sure that old packages need be retooled in the new harmony (again,standard), but, for sure, harmony benefits development to produce such where, the exhibits produced, will work on all system distros built at or after the "landmark".
Puppy is a very small community with none of the funding of the large distros. As such, though, it is complete and a bit of assistance in some developer standards could not deter the creativity they bring to the table.
Over the last year, I can count more than 5 productions in Puppyland that have produced a harmony of methods which should be called standards. The individual offerings and the community support which rallyed around them constituted a new and better way of approaches. This bring a new standard (thru the awareness that surfaced).
The term, standard, really has never been bad. It does have benefit that the choir/band is singing the same song in the production, even though some/many members have differing roles in the choir.
I hope we see some good value in what the OP presents in any effort for an approach to community harmony supporting developer and user subsystem-application developments and remasterings.
YES, I fully agree that development would benefit from some sort of desktop harmony to reduce the pain they go thru when introducing new functionality into PUPPY Linux and community usage.
Should some harmony (really a standard) takes root, it should be made clear what and which timeslot of PCs it will address. We cannot continue to support 486s and the like in NEW future PUP development. Much is already ever present as it has already been done many times in our past.
We must accept the great things that has been done and lock those in. And we must be clear about where the landmark is to begin any effort where this kind of harmony is to exist. I am not so sure that old packages need be retooled in the new harmony (again,standard), but, for sure, harmony benefits development to produce such where, the exhibits produced, will work on all system distros built at or after the "landmark".
Puppy is a very small community with none of the funding of the large distros. As such, though, it is complete and a bit of assistance in some developer standards could not deter the creativity they bring to the table.
Over the last year, I can count more than 5 productions in Puppyland that have produced a harmony of methods which should be called standards. The individual offerings and the community support which rallyed around them constituted a new and better way of approaches. This bring a new standard (thru the awareness that surfaced).
The term, standard, really has never been bad. It does have benefit that the choir/band is singing the same song in the production, even though some/many members have differing roles in the choir.
I hope we see some good value in what the OP presents in any effort for an approach to community harmony supporting developer and user subsystem-application developments and remasterings.
-
bark_bark_bark
- Posts: 1885
- Joined: Tue 05 Jun 2012, 12:17
- Location: Wisconsin USA
Forcing users onto one DE is wrong in my opinion. People come to linux because there is choice. You take that away and you have nothing and you just have another windows.
EDIT: I, for one, would NOT be happy to be forced into having to use the bloated and unstable mess known as GNOME 3.
Such a move would only turn me away from ever supporting the use of linux ever again. There are certainly many people who would agree.
EDIT: I, for one, would NOT be happy to be forced into having to use the bloated and unstable mess known as GNOME 3.
Such a move would only turn me away from ever supporting the use of linux ever again. There are certainly many people who would agree.
....
In my sad experience, lack of standards restricts freedom of choice, not the other way round. I have a personal list of the best Puppy apps, I'd like to have on DebianDog and proper Debian/Ubuntu and their spins. Not a huge list, about 4 or 6 apps and they are all gtkdialog based. OOB, they won't work, because the above distros default shell is dash. To make them work, you have to reconfigure bash to be the default shell. In other words, you gotta bend bigger standards to satisfy your non-standard desires. By which I mean, that gtkdialog itself has issues with standards, namely, it doesn't get along very well with bashisms. Technosaurus has a better explanation here: http://murga-linux.com/puppy/viewtopic. ... 026#814026.
----------
"Any customer can have a car painted any color he wants so long as it is black."
- Henry Ford
----------
"Any customer can have a car painted any color he wants so long as it is black."
- Henry Ford
"(Relatively) simple things like add a right click or drag/drop option to jwm panel to bring up configuration dialogs, interlink GTK/JWM etc themes (or at least provide the same theme for each). "
What you might find for desires like this (note, I don't know if it's true in this particular case) is that that functionality requires a switch to another WM that has it coded in, and that WM isn't substantially bigger than the original choice, but brings in a lot of dependencies that ARE, even if the dependencies aren't needed to support the functionality ... so that looks pretty unappealing to someone trying to get their ISO down as close to possible to 100MB.
So those are tough choices and people seem to go both ways. But I find, usually choices have a reason behind them if you can find them out.
My personal perfect Puppy? Hmm ... a good looking blue-grey-tinted icon set that 'goes with everything', a beautiful landscape picture for wallpaper, tint2 panel, a slick font choice in nearly-black, ubuntu-repository-compatible, Firefox browser (latest version), lots of wireless drivers, and very few other apps unless the user wants to install them. The applications to include is difficult; I prefer just utilities (they tend to be both small and useful, unlike many 'desktop applications', and the other apps are easily available for addition). But you know what? My desired Puppy could be a lot different than many other people's. Tahrpup takes twice as long to load as Lucid 5287, and while I will most definitely not categorize it as bloat, it's most likely simply stuff I don't need.
What you might find for desires like this (note, I don't know if it's true in this particular case) is that that functionality requires a switch to another WM that has it coded in, and that WM isn't substantially bigger than the original choice, but brings in a lot of dependencies that ARE, even if the dependencies aren't needed to support the functionality ... so that looks pretty unappealing to someone trying to get their ISO down as close to possible to 100MB.
So those are tough choices and people seem to go both ways. But I find, usually choices have a reason behind them if you can find them out.
My personal perfect Puppy? Hmm ... a good looking blue-grey-tinted icon set that 'goes with everything', a beautiful landscape picture for wallpaper, tint2 panel, a slick font choice in nearly-black, ubuntu-repository-compatible, Firefox browser (latest version), lots of wireless drivers, and very few other apps unless the user wants to install them. The applications to include is difficult; I prefer just utilities (they tend to be both small and useful, unlike many 'desktop applications', and the other apps are easily available for addition). But you know what? My desired Puppy could be a lot different than many other people's. Tahrpup takes twice as long to load as Lucid 5287, and while I will most definitely not categorize it as bloat, it's most likely simply stuff I don't need.
I have a 105mb slax including xfce4...also several pups that size and smaller with older xfce4...there are other wm/de that are small but respect desktop standards.
Cars have standards... tyre sizes, machine screw threads etc etc... its normal in the engineering world and software is a form of engineering.
Package management seems to have an 'each to their own ' methods too ...not sure why since the linux/posix file structure is the same for all. Packages are bunches of files in the pretty much standard places.... ship, unpack and go surely. Result is handling of several formats needed and there will always be some RPM joke funny to defeat you
Interestingly enough the windows world has a plethora of packaging methods.. a pain in the neck really, the worst of all being install sheild. Such diversity causes minor chaos and breaks itself or something else all too easily.
How many video formats out there...and many are really the same underneath..eg h264 codecs and all doing basically the same thing. Result...huge video players bundling piles of libraries or demanding them from the system. And such size and complexity multiplies the potential bugs that result.
hmm a can of worms opening.
Its not just puppy that needs to get a grip on harmonization but it could take steps in the right direction.
mike
Cars have standards... tyre sizes, machine screw threads etc etc... its normal in the engineering world and software is a form of engineering.
Package management seems to have an 'each to their own ' methods too ...not sure why since the linux/posix file structure is the same for all. Packages are bunches of files in the pretty much standard places.... ship, unpack and go surely. Result is handling of several formats needed and there will always be some RPM joke funny to defeat you
Interestingly enough the windows world has a plethora of packaging methods.. a pain in the neck really, the worst of all being install sheild. Such diversity causes minor chaos and breaks itself or something else all too easily.
How many video formats out there...and many are really the same underneath..eg h264 codecs and all doing basically the same thing. Result...huge video players bundling piles of libraries or demanding them from the system. And such size and complexity multiplies the potential bugs that result.
hmm a can of worms opening.
Its not just puppy that needs to get a grip on harmonization but it could take steps in the right direction.
mike
Closer GTK/jwm theme's given that there are utilities (PET's) to create themes for both, could be just a merging of two utilities into one.wboz wrote:What you might find for desires like this (note, I don't know if it's true in this particular case) is that that functionality requires a switch to another WM that has it coded in, and that WM isn't substantially bigger than the original choice, but brings in a lot of dependencies that ARE, even if the dependencies aren't needed to support the functionality ... so that looks pretty unappealing to someone trying to get their ISO down as close to possible to 100MB.(Relatively) simple things like add a right click or drag/drop option to jwm panel to bring up configuration dialogs, interlink GTK/JWM etc themes (or at least provide the same theme for each).
Personally I struggled some with Rox when I first started with Puppy last February, but have since come to like it a lot. With Rox taking care of the desktop (Pin) and jwm the menu's, combined with Rox Apps and SFS's and Puppy is one lean mean machine.
My own core puppy is lightweight, just a notepad (leafpad), calculator, the usual puppy desktop utilities etc. i.e. takes care of the interface (network, display, GUI, CD/DVD/USB etc.) and weighs in at a 75MB puppy sfs. I run with no savefile, ram booted and that flies. The rest I load using sfs's wrapped inside rox-apps i.e the rox-app takes care of repointing the programs configuration file from under /root/.config to a copy in the rox-app directory, so that any program config changes are preserved across reboots. The only exception is firefox, which I run using Shinobar's portable firefox - so always using the latest version and again any changes (bookmarks etc.) are preserved across reboots.
Generally the core puppy changes little (rarely) and when change are made I just remaster using no compression and it takes seconds. I then drop the puppy sfs inside initrd and compress that - so only vmlinuz and initrd.gz required to boot.
Running 90%+ in ram and it really flies. The main reason I've stayed with puppy - as in comparison everything else is slow.
I know its not, but to me the core puppy is Rox + jwm, with most things coded as scripts for ease of change/customisation. That's unique and what makes puppy different from the rest. Whilst not complying with those others standards being on the outside with a unique and highly customisable/fast/simple alternative is a definite plus. Foregoing that to fall in more alignment with others and Puppy becomes much more of just another alternative/similar choice to the others rather than something unusual on the outside.
"To me the core puppy is Rox + jwm, with most things coded as scripts for ease of change/customisation."
Agree with this although it's sometimes not immediately clear to me if I am using a JWM or Openbox pup. Not sure how different they are.
Any idea what is the minimal WM/DE most likely to be able to adopt touch controls at some point? Not that I'm a touch person but I do think it's a key piece of future-proofing.
Agree with this although it's sometimes not immediately clear to me if I am using a JWM or Openbox pup. Not sure how different they are.
Any idea what is the minimal WM/DE most likely to be able to adopt touch controls at some point? Not that I'm a touch person but I do think it's a key piece of future-proofing.
using non standard desktop and menu generation does not make it special...it makes it awkward. JWM does not handle menus...add on scripts have to hack that one...same for ROX.
You can make fast and light versions of both that do follow standards...
The puppy features are ability to run from anything including the Out house sink, loading into ram, sfs usage...that sort of core structures that make it stand out and perform .... plus some neat scripts that seem to do what the big boys fails to do...eg wireless connectivity.
Note there are plenty of puppy derivatives that abandon JWM and ROX yet still are very much puppy and are slick into the bargain..
Just want to make that distinction there.
Ok flies in the wind of barry's definition which is a pup is only a pup IF its built with woof and used JWM and ROX ...seems a very narrow view to me and that sounds more like restricting choices than following certain standard methods.
I am not totally biased...I keep ROX and its pinboard on me pups but have xfce4 do the rest...those handy standards are one reason.
grizzly adams...
mike
You can make fast and light versions of both that do follow standards...
The puppy features are ability to run from anything including the Out house sink, loading into ram, sfs usage...that sort of core structures that make it stand out and perform .... plus some neat scripts that seem to do what the big boys fails to do...eg wireless connectivity.
Note there are plenty of puppy derivatives that abandon JWM and ROX yet still are very much puppy and are slick into the bargain..
Just want to make that distinction there.
Ok flies in the wind of barry's definition which is a pup is only a pup IF its built with woof and used JWM and ROX ...seems a very narrow view to me and that sounds more like restricting choices than following certain standard methods.
I am not totally biased...I keep ROX and its pinboard on me pups but have xfce4 do the rest...those handy standards are one reason.
grizzly adams...
mike
-
bark_bark_bark
- Posts: 1885
- Joined: Tue 05 Jun 2012, 12:17
- Location: Wisconsin USA
It allows complete programs to be easily and efficiently writtenMoose On The Loose wrote:What is there not to like about Bash? It allows complete programs to be easily and efficiently written without adding another interpreter to the OS. I went to the bother of learning Tcl/Tk at one point. I never found a case where it was better suited to my purpose than bash driving gtkdialog3 has been.NeroVance wrote: The only gripes I have are the loss of Tcl/Tk, other than that, having a good standard is useful, though I do admit I actually loathe the constant usage of Bash for graphical applications
Perhaps if I was porting things also onto Windows(spit) I may find use for Tcl/Tk. I have found that a lot of windows programs that allow scripting seem to use it rather that trying to make something that works in Visual Basic.
Though I've always found Bash to look extremely ugly by itself, if suffices for being a shell, but I like my scripting languages to be full of curly-braces, though that might just be my preference.
And I mean code, not the UIs themselves, though I do find it jarring to have some windows become basically grayed out because I move a window for a dialog box, but that I do believe is possible to fix in code.
Personally, I'd like to see how pMusic and the like would fare against other software in non-BASH in terms of speed and performance.
Though I will admit, I am kinda a Tcl fanboy
-
LazY Puppy
- Posts: 1934
- Joined: Fri 21 Nov 2014, 18:14
- Location: Germany
I had stumbled lately over a posting -I think made by technosaurus- about the mess in GTK (especially of deprecating things and to replace them with NOTHING). He mentioned the use of Tk/Tcl instead of GTK/Bash (iIrc).NeroVance wrote:Though I will admit, I am kinda a Tcl fanboy
Is there a Tk/Tcl development package that will run in Lucid and/or Precise based Puppies?
Have a Link?
Some documentation and/or manuals also?
RSH
"you only wanted to work your Puppies in German", "you are a separatist in that you want Germany to secede from Europe" (musher0) :lol:
No, but I gave my old drum kit away for free to a music store collecting instruments for refugees! :wink:
"you only wanted to work your Puppies in German", "you are a separatist in that you want Germany to secede from Europe" (musher0) :lol:
No, but I gave my old drum kit away for free to a music store collecting instruments for refugees! :wink:
Ah, nice to see some others willing to learn the lost art of Tcl. I wonder if this is how most gurus in the *nix world feel when they see people willing to learn something you support?LazY Puppy wrote:I had stumbled lately over a posting -I think made by technosaurus- about the mess in GTK (especially of deprecating things and to replace them with NOTHING). He mentioned the use of Tk/Tcl instead of GTK/Bash (iIrc).NeroVance wrote:Though I will admit, I am kinda a Tcl fanboy
Is there a Tk/Tcl development package that will run in Lucid and/or Precise based Puppies?
Have a Link?
Some documentation and/or manuals also?
Well Tcl/Tk is pretty easy to get setup and working, And I did create a couple packages for Precise of 8.6 here http://www.murga-linux.com/puppy/viewto ... 486ac8e8ae
As for documentation, I can offer a couple superb places to not only get information, but where I first started Tcl. http://www.bin-co.com/tcl/tutorial/
I always suggest bin-co for newcomers, and those who wish to learn Tcl/Tk. And here is a major tip for making Tcl scripts run on a variety of platforms.
Code: Select all
#!/bin/sh
#The next line executes wish - wherever it is \
exec wish "$0" "$@"
).But since it is open source you can contribute to fix the poorly written software that you said and you cannot suffer from severe hopelessness.bark_bark_bark wrote:Standardization lead to two options:
1) Use it and suffer from poorly written software, but you can't use something else because there are no other options.
2) Don't use it at all and suffer from severe hoplessness.