Re: Where is KControl?
Thank you, Ian. You saved me at least another two days of struggling. To return the favor - sorry for posting the info way outside of the list's subject - I have switched to Pathfinder exactly for the sake of the font and color control. It's no Krusader though :( Thank you once again ~Andrey -- View this message in context: http://mac-os-forge.2317878.n4.nabble.com/Where-is-KControl-tp218948p219077.html Sent from the MacPorts - Users mailing list archive at Nabble.com. ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
On May 27, 2013, at 21:41, Dan Aldrich wrote: I was thinking more of an interface like what sourceforge offers. Recommendations and number of d/l's a week. The comments make it sound like more trouble than it's worth. Sourceforge tracks downloads per file, so if you want that information shown with port information, then we would need to associate files with ports. If a port fetches a binary package, that's straightforward since the name is always based on the port name (and version and variants and OS). But if a port fetches distfiles, then it's not so simple. You might look at the distfile zlib-1.2.8.tar.xz and think that's easy to link back to the zlib port, but we can't generalize that; a port could fetch files of any name. jpegsrc.v9.tar.gz for example from the jpeg port. Some ports might fetch multiple distfiles. Some of those distfiles might not change each version (e.g. a game whose code changes but whose graphic and sound assets don't). Some distfiles might change multiple times per version (e.g. a stealth update). Some distfiles might even be used by multiple ports (e.g. py24-distribute, py25-distribute, py26-distribute, py27-distribute, py31-distribute, py32-distribute, py33-distribute which all use the distfile distribute-0.6.43.tar.gz; this is very common for language modules). So we could track download statistics, but it would be a lot of infrastructure work to get to that point. And although it might be interesting from an educational point of view, I don't think download statistics translate into recommendations. If I'm having problems with a port and I believe that reinstalling it will help, and I'm able to use MacPorts binaries, then reinstalling the port will cause MacPorts to redownload the binary, which would increase its download count, though I would not want other users to infer from that that I suddenly recommend the port more greatly. And there are a lot of ports that a download count would seem to suggest are very popular, yet might be something like pkgconfig or zlib that you probably need to have installed but is not something you will specifically use yourself. If you were thinking of showing a list of ports sorted by such a popularity indicator, with the hope of finding ports you might actually want to use, then it might be le ss useful than you think. ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: error message in Macports
Okay, did the sudo port -d sync again to get a verbose output, and it updated quite a few items. Here is the last of it: Total number of ports parsed: 8179 Ports successfully parsed: 8179 Ports failed: 0 Up-to-date ports skipped: 8927 The list is quite long on what was looked at and updated and added. To my eye, I cannot see anything wrong, so tried again with just: Sudo port sync And this time it just had: updating ports tree And went back to the prompt, no errors!! Am wondering if doing this in verbose mode forced it to add the items it needed? Any thoughts? Thanks, Lisa -- Lisa Madden, ACA Sr. Systems Administrator ADNET Systems, Inc. (SESDA 3 Contract) Terrestrial Information Systems Laboratory (619) Building 33, Room H-108 NASA Goddard Space Flight Center 301.614.6594 On 5/24/13 9:18 PM, Bradley Giesbrecht pixi...@macports.org wrote: On May 24, 2013, at 12:42 PM, Madden, Lisa E. (GSFC-619.0)[ADNET SYSTEMS INC] wrote: Hi all, Newbie to Macports so excuse if this is a dumb questionŠ. We cannot use rsync so am updating using svn. Anyway, I type port sync, and it comes back with this error: Synchronization of the local ports tree failed doing an svn update Port sync failed: synchronization of 1 source(s) failed WHAT does this mean and how do I find out which source failed? I'm not sure. Make port more verbose and see if you get any clues. $ sudo port -d sync Regards, Bradley Giesbrecht (pixilla) ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: error message in Macports
On May 28, 2013, at 13:25, Madden, Lisa E. (GSFC-619.0)[ADNET SYSTEMS INC] wrote: And went back to the prompt, no errors!! Am wondering if doing this in verbose mode forced it to add the items it needed? Verbose mode shouldn't behave any different, other than to show you more output. Perhaps there was a temporary network problem which has resolved itself. If it happens again, try again with verbose mode and see if there are errors shown. ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: error message in Macports
That is what I would have thought, but I first ran just plain port sync, got the error, then did port -d sync and it ran thru a ton of updates/adds, then port sync again, and no error. This is a virtual Monday for a lot of us who had yesterday off on holiday I guessŠ.we'll blame that. Thanks, Lisa -- Lisa Madden, ACA Sr. Systems Administrator ADNET Systems, Inc. (SESDA 3 Contract) Terrestrial Information Systems Laboratory (619) Building 33, Room H-108 NASA Goddard Space Flight Center 301.614.6594 On 5/28/13 2:26 PM, Ryan Schmidt ryandes...@macports.org wrote: On May 28, 2013, at 13:25, Madden, Lisa E. (GSFC-619.0)[ADNET SYSTEMS INC] wrote: And went back to the prompt, no errors!! Am wondering if doing this in verbose mode forced it to add the items it needed? Verbose mode shouldn't behave any different, other than to show you more output. Perhaps there was a temporary network problem which has resolved itself. If it happens again, try again with verbose mode and see if there are errors shown. ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
While download statistics might not be a good system, I do concur that MacPorts very much would benefit from having a discovery mechanism by which users find out about useful ports. Searching is nice, but it's not discovery. Some kind of top ports list (however implemented) would be useful, imho. I do like the idea of opt-inable call-home reporting about which ports are installed. Even if a minority opts-in, it would be useful. Ubuntu does a similar thing IIRC. For non-end-users, it would also inform port maintainers about how many people depend on their work. Jean-François Begin forwarded message: Re: Suggestion ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
Hi, On 28 May 2013, at 08:12 PM, Jean-François Caron jfca...@phas.ubc.ca wrote: While download statistics might not be a good system, I do concur that MacPorts very much would benefit from having a discovery mechanism by which users find out about useful ports. Searching is nice, but it's not discovery. Some kind of top ports list (however implemented) would be useful, imho. Personally, I fail to see how a 'top ports' list would tell me much. The ports i find essential are likely very different from others, so i don't see how using some sort of a list showing the most used ports would help me in any way in choosing new ones to install. Some ports likely have a low user base, but never less are critical to those that need them, such as more esoteric ports from the science section. Maybe such a list might be of some limited interest to developers, so we know which ports are most used, and perhaps more usefully those without users, but beyond that i doubt such a list would tell us much or help users. What would help users i think would be to have a much better search/browsing interface, to allow them to browse available ports. Chris I do like the idea of opt-inable call-home reporting about which ports are installed. Even if a minority opts-in, it would be useful. Ubuntu does a similar thing IIRC. For non-end-users, it would also inform port maintainers about how many people depend on their work. Jean-François Begin forwarded message: Re: Suggestion ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
On Tue, May 28, 2013 at 1:36 PM, Chris Jones jon...@hep.phy.cam.ac.ukwrote: Hi, On 28 May 2013, at 08:12 PM, Jean-François Caron jfca...@phas.ubc.ca wrote: While download statistics might not be a good system, I do concur that MacPorts very much would benefit from having a discovery mechanism by which users find out about useful ports. Searching is nice, but it's not discovery. Some kind of top ports list (however implemented) would be useful, imho. Personally, I fail to see how a 'top ports' list would tell me much. The ports i find essential are likely very different from others, so i don't see how using some sort of a list showing the most used ports would help me in any way in choosing new ones to install. Some ports likely have a low user base, but never less are critical to those that need them, such as more esoteric ports from the science section. Maybe such a list might be of some limited interest to developers, so we know which ports are most used, and perhaps more usefully those without users, but beyond that i doubt such a list would tell us much or help users. What would help users i think would be to have a much better search/browsing interface, to allow them to browse available ports. Chris I think the discovery mechanism argument is valid. The search mechanism only works if I know what I'm looking for. I don't know what I don't know but if 90% of MacPorts users install a specific port I would check it out even if I wasn't searching for it in particular. A better search interface would be an improvement too but I usually just install what I know I want. ...Stephen ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
Hi, On 28 May 2013, at 10:14 PM, Stephen Rasku macpo...@srasku.net wrote: On Tue, May 28, 2013 at 1:36 PM, Chris Jones jon...@hep.phy.cam.ac.uk wrote: Hi, On 28 May 2013, at 08:12 PM, Jean-François Caron jfca...@phas.ubc.ca wrote: While download statistics might not be a good system, I do concur that MacPorts very much would benefit from having a discovery mechanism by which users find out about useful ports. Searching is nice, but it's not discovery. Some kind of top ports list (however implemented) would be useful, imho. Personally, I fail to see how a 'top ports' list would tell me much. The ports i find essential are likely very different from others, so i don't see how using some sort of a list showing the most used ports would help me in any way in choosing new ones to install. Some ports likely have a low user base, but never less are critical to those that need them, such as more esoteric ports from the science section. Maybe such a list might be of some limited interest to developers, so we know which ports are most used, and perhaps more usefully those without users, but beyond that i doubt such a list would tell us much or help users. What would help users i think would be to have a much better search/browsing interface, to allow them to browse available ports. Chris I think the discovery mechanism argument is valid. The search mechanism only works if I know what I'm looking for. I don't know what I don't know but if 90% of MacPorts users install a specific port I would check it out even if I wasn't searching for it in particular. A better search interface would be an improvement too but I usually just install what I know I want. I can see how for something that has a massive user base, such as the Apple 'App Store', a user based popularity poll would help to tell me something. I still maintain though that in general just because a lot of other users install something, does not mean i need to do the same, but if the statistics are large enough so you can start to discriminate within a specific category, it might be useful. This works for places like the App store, that has the massive user base needed, but i suspect MacPorts is no where near that so it really doesn't help. I still think a better documentation system would help more, so given a few keywords a user can search macports to find what ports are available and provide what they need, is more useful to MacPorts than some popularity based polling system. Chris ...Stephen ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
Please excuse typos--sent from my iPhone. On 2013-05-28, at 5:35 PM, Chris Jones jon...@hep.phy.cam.ac.uk wrote: Hi, On 28 May 2013, at 10:14 PM, Stephen Rasku macpo...@srasku.net wrote: On Tue, May 28, 2013 at 1:36 PM, Chris Jones jon...@hep.phy.cam.ac.uk wrote: Hi, On 28 May 2013, at 08:12 PM, Jean-François Caron jfca...@phas.ubc.ca wrote: While download statistics might not be a good system, I do concur that MacPorts very much would benefit from having a discovery mechanism by which users find out about useful ports. Searching is nice, but it's not discovery. Some kind of top ports list (however implemented) would be useful, imho. Personally, I fail to see how a 'top ports' list would tell me much. The ports i find essential are likely very different from others, so i don't see how using some sort of a list showing the most used ports would help me in any way in choosing new ones to install. Some ports likely have a low user base, but never less are critical to those that need them, such as more esoteric ports from the science section. Maybe such a list might be of some limited interest to developers, so we know which ports are most used, and perhaps more usefully those without users, but beyond that i doubt such a list would tell us much or help users. What would help users i think would be to have a much better search/browsing interface, to allow them to browse available ports. Chris I think the discovery mechanism argument is valid. The search mechanism only works if I know what I'm looking for. I don't know what I don't know but if 90% of MacPorts users install a specific port I would check it out even if I wasn't searching for it in particular. A better search interface would be an improvement too but I usually just install what I know I want. I can see how for something that has a massive user base, such as the Apple 'App Store', a user based popularity poll would help to tell me something. I still maintain though that in general just because a lot of other users install something, does not mean i need to do the same, but if the statistics are large enough so you can start to discriminate within a specific category, it might be useful. This works for places like the App store, that has the massive user base needed, but i suspect MacPorts is no where near that so it really doesn't help. I still think a better documentation system would help more, so given a few keywords a user can search macports to find what ports are available and provide what they need, is more useful to MacPorts than some popularity based polling system. Chris ...Stephen __ I'd like to know that a port has been successfully installed, recently. IOW that it hasn't suffered bitrot! Craig___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
On May 28, 2013, at 16:56, Craig Treleaven wrote: I'd like to know that a port has been successfully installed, recently. IOW that it hasn't suffered bitrot! To some degree, you can answer that already, by checking if there is a recent package on the packages server. To answer it with statistics, we would not only have to phone home to track successful port installs, but also keep track of install dates, not just a count of installs. Presumably we could keep just the latest install date. ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
Hi, On 28 May 2013, at 10:59 PM, Ryan Schmidt ryandes...@macports.org wrote: On May 28, 2013, at 16:56, Craig Treleaven wrote: I'd like to know that a port has been successfully installed, recently. IOW that it hasn't suffered bitrot! To some degree, you can answer that already, by checking if there is a recent package on the packages server. I was about to say the same. To my mind, this question is much better/simply answered by just checking to see if the build bots successfully built the port question. I don't see what you gain by getting user stats as well. Chris To answer it with statistics, we would not only have to phone home to track successful port installs, but also keep track of install dates, not just a count of installs. Presumably we could keep just the latest install date. ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
On May 28, 2013, at 17:17, Chris Jones wrote: On 28 May 2013, at 10:59 PM, Ryan Schmidt wrote: On May 28, 2013, at 16:56, Craig Treleaven wrote: I'd like to know that a port has been successfully installed, recently. IOW that it hasn't suffered bitrot! To some degree, you can answer that already, by checking if there is a recent package on the packages server. I was about to say the same. To my mind, this question is much better/simply answered by just checking to see if the build bots successfully built the port question. I don't see what you gain by getting user stats as well. Just because the buildbots could build the port when it was last updated doesn't mean that anybody can still build it today. For example, one of its dependencies could have been updated in a way that is incompatible. ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Upgrading perl from 5.12 to 5.16?
On May 27, 2013, at 1:37 PM, Ryan Schmidt ryandes...@macports.org wrote: On May 27, 2013, at 14:11, Johannes Kastl wrote: 3. How to get intltool to use p5.16 instead of 5.12 for its dependencies? Reinstalling gimp2 and inkscape leads to some p5.12-* modules being installed. There is intentionally no way for you to do so. Ports like intltool that use specific perl modules must depend on a specific perl version thereof. Currently the default perl in MacPorts is 5.12. I'm sure we'd be willing to entertain the idea of changing what the default perl is, but it should then be done uniformly in all ports. Can't we have perl variants for ports that use perl like we have python variants for ports that use python? -Frank ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
On 29/05/2013, at 6:36 AM, Chris Jones wrote: What would help users i think would be to have a much better search/browsing interface, to allow them to browse available ports. Yes indeed. It is hard to see the wood for the trees. I did some investigation on this recently with a view to designing a search facility in a GUI app for Macports. Firstly, there are nearly 17,000 ports and the distribution is quite skewed. For example, more than 10,000 of the ports have names beginning with the letter p, because so many ports are written in Python, Perl or PHP and even in specific versions of same (names like py27*, py26*, etc.). In among that 10,000 is a legion of specific functionality of different kinds, but how do you find it, unless you already know the name or were recommended to it by a friend? Macports has categories but IMHO they are not always rigorously applied and re-cataloging thousands of ports is a huge exercise. Also some categories are quite broad, which limits their usefulness in searching (e.g. the science, math and games categories). I looked also at extracting a list of top-level ports. A large proportion of ports are dependencies of something else. However, those that are not dependencies of anything else are (to me) a surprising proportion of the total --- about 30 to 40% IIRC. In among those thousands of top-level ports are major packages such as Inkscape and toys such as 2Pong. In the end, I found the most powerful simple search was much like port's own search command, i.e. look for a substring anywhere in Name, Description or Full Description. If I picked a fairly specific string, such as quantum or jigsaw, I got quite good results, as long as I searched the Full Description as well as the Description (which apparently port search does not do by default). Beyond this, I am looking at complex searches involving predicates (Cocoa NSPredicate), such as all KDE apps but excluding translation packages for languages other than English, but it is not easy to come up with simple user interfaces. More ideas on searching anyone? All the best, Ian W. ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
On May 28, 2013, at 8:08 PM, Ian Wadham iandw...@gmail.com wrote: If I picked a fairly specific string, such as quantum or jigsaw, I got quite good results, as long as I searched the Full Description as well as the Description (which apparently port search does not do by default). Searching is highly overrated. Conceptually, it is useful, but in reality it fails miserably. a fairly specific string, such as quantum or jigsaw, are not strings but rather tokens. (And it's hard to get more specific than that.) Searching by strings is almost impossible. ... even when you quote them: quantum jigsaw The number of search engines which will ONLY return hits on your requested search are few and far between (if any) They are ALL committed to the idea that people will not use us again if we don't return thousands and thousands of hits. And we all know that the reason for that is 100% financial -- that's how those search engines get paid. And it gets even more aggravating when you know that half of the hits on any given page are PAID to be there, they are not even the result of the companies search algorithm. And they are NEVER highlighted. ... because the search algorithm's are being gamed. However, searching is far better than being handed a oh...mee-tooo list. Just because some piece of Malware has been downloaded by hundreds or thousands of people does not mean that I want, or should, download it also. Johnny or Suzzie does it, why can't I... that's the attitude of a 7 year old. But folks seem to fall for it day after day after day at sites like Amazon, Twitter, iTunes, Facebook, and the like. There is a difference between searching and browsing. I don't know how many Mac Ports users remember visiting a major Library full of books not looking for a specific title, but looking for serendipity. Granted it was just a different generation's version of attention grabbers -- a particular title, or a specific jacket illustration. You picked a room -- a category -- and started wandering the shelves. Maybe you looked at the display the Librarian put together but usually you just wandered through the stacks. (I know I'm REALLY dating myself now.) Personally, I find a search engine which does not tokenize every bloody word in the search string just so that it can return me something really annoying.--- Returning nothing is JUST AS VALID, it has just as much, and maybe more, meaning than a bunch of random hits. Just my 2bits. T.T.F.N. William H. Magill # iMac11,3 Core i7 [2.93GHz - 8 GB 1067MHz] OS X 10.8.3 # MacBook Pro4.1 Core 2 Duo [2.5GHz - 4GB 667] OS X 10.6.8 # Macmini6,1 Intel Core i5 [2.5 Ghz - 4GB 1600MHz] OS X 10.8.3 mag...@icloud.com mag...@mac.com whmag...@gmail.com ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
Re: Suggestion
At 5:26 PM -0500 5/28/13, Ryan Schmidt wrote: On May 28, 2013, at 17:17, Chris Jones wrote: On 28 May 2013, at 10:59 PM, Ryan Schmidt wrote: On May 28, 2013, at 16:56, Craig Treleaven wrote: I'd like to know that a port has been successfully installed, recently. IOW that it hasn't suffered bitrot! To some degree, you can answer that already, by checking if there is a recent package on the packages server. True but a non-trivial fraction of ports not able to be distributed as binaries. I was about to say the same. To my mind, this question is much better/simply answered by just checking to see if the build bots successfully built the port question. I don't see what you gain by getting user stats as well. Just because the buildbots could build the port when it was last updated doesn't mean that anybody can still build it today. For example, one of its dependencies could have been updated in a way that is incompatible. Am I missing something; how does one look up whether/when the buildbots built a specific port? Craig ___ macports-users mailing list macports-users@lists.macosforge.org https://lists.macosforge.org/mailman/listinfo/macports-users
