Re: [Factor-talk] User update magic in edit-profile
Sorry Slava, try the same repo again: g...@github.com:addrummond/factor.git Alex On 23 February 2010 02:43, Slava Pestov sl...@factorcode.org wrote: On Tue, Feb 23, 2010 at 5:25 PM, Alex Drummond a.d.drumm...@googlemail.com wrote: I'm a bit of a git newbie, so let me know if you'd rather I pushed the changes some other way. Hi Alex, You might want to create your repository again. You removed the work folder, added work.zip, and removed a file from io.encodings.japanese. Slava -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
Re: [Factor-talk] User update magic in edit-profile
Hi Slava, I've written some docs for the couchdb auth implementation and put it up here (in master): g...@github.com:addrummond/factor.git I'm a bit of a git newbie, so let me know if you'd rather I pushed the changes some other way. Currently, there is nothing to stop a user editing their profile such that they end up with the same registered email address as another user. Eventually, I may try extending the auth protocol a bit to allow errors of this sort to percolate. Alex On 21 February 2010 07:01, Slava Pestov sl...@factorcode.org wrote: On Sun, Feb 21, 2010 at 10:53 PM, Alex Drummond a.d.drumm...@googlemail.com wrote: What's the best way for me to send it to you? Set up a github account. I'll probably take a look at furnace.sessions at some point. Right now I am just using sqlite for sessions. In fact, this *probably* makes more sense than using couchdb, since couchdb has a write-only architecture which isn't ideal for making frequent updates to small documents. (Not that I have any data on this specific case -- just guessing.) But for sure there are plenty of non-SQL dbs out there that people might want to use for sessions. Or in-memory sessions for performance. Also, I've added a t:script tag to chloe which allows you to include scripts in the parent template either (i) inline between t:script/t:script, (ii) by specifying a local file to insert inline, or (iii) by setting 'src'. I'm also planning to update the t:style tag to be consistent with t:script. If you're interested in these updates, let me know. Sounds good. Slava -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
[Factor-talk] User update magic in edit-profile
Hi, I've written an implementation of the furnace authentication provider protocol for couchdb. I decided to do a direct implementation, rather than write a tuple db interface for couchdb, since when using couchdb you need to jump through various hoops in order to ensure the uniqueness of usernames and email addresses. It's all working fine, except that the edit-profile controller is not saving the modified user tuple back to the DB. I see that it sets the changed? flag of the tuple once its submit action finishes, but I'm having trouble working out how I should hook into whatever method is supposed to get called in order to commit the update to the DB. It works fine if I use the following nasty code to manually ensure that a user-saver is created, and its destructor called: edit-profile-action [ responder [ [ call( -- response ) logged-in-user get user-saver dispose ] curry ] change-submit drop ] [ auth-boilerplate edit-profile add-responder ] bi But this is pretty ugly, and I'd like to know what the Right Way is, if there is one. Any help appreciated. Alex -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
Re: [Factor-talk] User update magic in edit-profile
Oops, when I switched back to using allow-edit-profile instead of the hack in the previous email, it worked fine. I guess there must previously have been a bug in my update-user method which somehow led me to think that it wasn't being called at all. (I am just using login-realm, so there wouldn't have been an issue with call-next-method.) sorry for the false alarm... Alex On 21 February 2010 04:08, Slava Pestov sl...@factorcode.org wrote: Hi Alex, Does the call-responder* method of your authentication realm execute call-next-method? Because M: realm call-responder* calls save-user-after, which sets up a destructor that will call update-user when the request is done. If you implement the update-user generic word then editing the profile should work. Slava On Sun, Feb 21, 2010 at 9:59 PM, Alex Drummond a.d.drumm...@googlemail.com wrote: Hi, I've written an implementation of the furnace authentication provider protocol for couchdb. I decided to do a direct implementation, rather than write a tuple db interface for couchdb, since when using couchdb you need to jump through various hoops in order to ensure the uniqueness of usernames and email addresses. It's all working fine, except that the edit-profile controller is not saving the modified user tuple back to the DB. I see that it sets the changed? flag of the tuple once its submit action finishes, but I'm having trouble working out how I should hook into whatever method is supposed to get called in order to commit the update to the DB. It works fine if I use the following nasty code to manually ensure that a user-saver is created, and its destructor called: edit-profile-action [ responder [ [ call( -- response ) logged-in-user get user-saver dispose ] curry ] change-submit drop ] [ auth-boilerplate edit-profile add-responder ] bi But this is pretty ugly, and I'd like to know what the Right Way is, if there is one. Any help appreciated. Alex -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
Re: [Factor-talk] Chloe call-next-template
Thinking about it, what I really want to be able to do is include templates by name inside other templates. I don't think that extending call-next-template is the best way to accomplish this, since one may wish to include one template file inside another without this having any semantic significance (i.e. without it making sense to have a nested form). To this end, I added the following crude extension to chloe: M: vocab vocabulary ( v -- v ) ; CHLOE: include [ vocab required-attr vocab ] [ template required-attr ] bi 2array resolve-template-path .xml append filexml body children compile-children ; I was really pleased by how easy it was to extend Chloe like this. I'll definitely be more open to adding custom tags in the future. (I'd assumed it would be more difficult.) Of course, this doesn't do any namespacing stuff. That doesn't bother me personally, but it probably is not suitable for inclusion as one of the standard chloe tags. Alex On 11 February 2010 14:19, Alex Drummond a.d.drumm...@googlemail.com wrote: Hey Slava, thanks for the helpful response. Yes, I might see if I can extend the code a little. I guess there are roughly two ways of doing it. Either (i) chloe templates have access to data from parent forms by default (i.e. you can write t:label t:name=parent.parent.foo /) or (ii) templates continue to operate in a flat namespace, but there is a means of accessing values from parent forms in factor controller code. In the case of (ii), you would manually copy those values of interest into the child template's namespace in the init quotation (or wherever). Do you have any view on which (if either) is the better approach? I guess there would also be issues regarding where validation errors live. I haven't made much use of the validations system yet so I don't have any ideas about how this would work. For the moment, I think I will just generate the list XML in factor using the child template, then bung the generated XML into the parent template. I hadn't thought of doing this when I originally posted, but it seems like an ok workaround. Alex On 11 February 2010 07:23, Slava Pestov sl...@factorcode.org wrote: Hi Alex, The reason you can't do what you want to do is that both the boilerplate and action responders call begin-form at the top of their call-responder* methods. Chloe forms are built on top of actions and the machinery in the html.forms vocabulary. Nesting forms is supported only to the extent that validation errors work properly for a page with multiple forms. For example, in the pastebin, a Chloe template renders the actual paste, and there is a sub-template with the form for adding an annotation. Validation errors in the annotation form appear on the annotation template's components, instead of on components of the same name in the paste template. To support what you want to do, the templating code needs to be extended to support nested templates in a more advanced way. In particular, Chloe tags for placing components need to be able to address values from parent forms as well as child forms, using some kind of namespacing syntax. If that's a project you feel like taking on, let me know ;-) Slava On Thu, Feb 11, 2010 at 7:09 PM, Alex Drummond a.d.drumm...@googlemail.com wrote: Hi, I've noticed that values from the parent template don't seem to be available to a child template rendered via t:call-next-template. I would like to render a list of records, with the html template for each record in a separate template file. Could anyone tell me the best way to do this with Chloe? The child template is also used to render individual records in other contexts, so I'd really like to keep it in its own file. thanks, Alex -- SOLARIS 10 is the OS for Data Centers - provides features such as DTrace, Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW http://p.sf.net/sfu/solaris-dev2dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk -- SOLARIS 10 is the OS for Data Centers - provides features such as DTrace, Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW http://p.sf.net/sfu/solaris-dev2dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk -- SOLARIS 10 is the OS for Data Centers - provides features such as DTrace, Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW http://p.sf.net/sfu/solaris-dev2dev ___ Factor-talk
Re: [Factor-talk] Chloe call-next-template
It's a pretty ordinary CRUD application. The idea is to present upcoming events in linguistics in an accessible way. Currently, most conferences etc. are announced on the Linguist List mailing list, which works great as a mailing list, but is not an ideal way to keep track of deadlines etc. I have some nasty Perl code which grovels emails from the mailing list for conference/job announcements and puts them into a CouchDB database. Now I'm writing a frontend to it in factor (also allowing events to be added directly, etc.) I started off writing the front end in Perl too, but I thought it would be more interesting to try it in factor. As you know, the actual code for this sort of thing tends to be incredibly dull, so it's nice to at least write it in a cool language :) It may never get finished, as I'm just doing it for fun and I have to prioritize my (ahem) real work as a grad student. Alex On 12 February 2010 06:56, Slava Pestov sl...@factorcode.org wrote: I'm curious, what are you building with Furnace and Factor? Slava On Fri, Feb 12, 2010 at 9:58 PM, Alex Drummond a.d.drumm...@googlemail.com wrote: Thinking about it, what I really want to be able to do is include templates by name inside other templates. I don't think that extending call-next-template is the best way to accomplish this, since one may wish to include one template file inside another without this having any semantic significance (i.e. without it making sense to have a nested form). To this end, I added the following crude extension to chloe: M: vocab vocabulary ( v -- v ) ; CHLOE: include [ vocab required-attr vocab ] [ template required-attr ] bi 2array resolve-template-path .xml append filexml body children compile-children ; I was really pleased by how easy it was to extend Chloe like this. I'll definitely be more open to adding custom tags in the future. (I'd assumed it would be more difficult.) Of course, this doesn't do any namespacing stuff. That doesn't bother me personally, but it probably is not suitable for inclusion as one of the standard chloe tags. Alex On 11 February 2010 14:19, Alex Drummond a.d.drumm...@googlemail.com wrote: Hey Slava, thanks for the helpful response. Yes, I might see if I can extend the code a little. I guess there are roughly two ways of doing it. Either (i) chloe templates have access to data from parent forms by default (i.e. you can write t:label t:name=parent.parent.foo /) or (ii) templates continue to operate in a flat namespace, but there is a means of accessing values from parent forms in factor controller code. In the case of (ii), you would manually copy those values of interest into the child template's namespace in the init quotation (or wherever). Do you have any view on which (if either) is the better approach? I guess there would also be issues regarding where validation errors live. I haven't made much use of the validations system yet so I don't have any ideas about how this would work. For the moment, I think I will just generate the list XML in factor using the child template, then bung the generated XML into the parent template. I hadn't thought of doing this when I originally posted, but it seems like an ok workaround. Alex On 11 February 2010 07:23, Slava Pestov sl...@factorcode.org wrote: Hi Alex, The reason you can't do what you want to do is that both the boilerplate and action responders call begin-form at the top of their call-responder* methods. Chloe forms are built on top of actions and the machinery in the html.forms vocabulary. Nesting forms is supported only to the extent that validation errors work properly for a page with multiple forms. For example, in the pastebin, a Chloe template renders the actual paste, and there is a sub-template with the form for adding an annotation. Validation errors in the annotation form appear on the annotation template's components, instead of on components of the same name in the paste template. To support what you want to do, the templating code needs to be extended to support nested templates in a more advanced way. In particular, Chloe tags for placing components need to be able to address values from parent forms as well as child forms, using some kind of namespacing syntax. If that's a project you feel like taking on, let me know ;-) Slava On Thu, Feb 11, 2010 at 7:09 PM, Alex Drummond a.d.drumm...@googlemail.com wrote: Hi, I've noticed that values from the parent template don't seem to be available to a child template rendered via t:call-next-template. I would like to render a list of records, with the html template for each record in a separate template file. Could anyone tell me the best way to do this with Chloe? The child template is also used to render individual records in other contexts, so I'd really like to keep it in its own file. thanks, Alex
Re: [Factor-talk] Chloe call-next-template
Hey Slava, thanks for the helpful response. Yes, I might see if I can extend the code a little. I guess there are roughly two ways of doing it. Either (i) chloe templates have access to data from parent forms by default (i.e. you can write t:label t:name=parent.parent.foo /) or (ii) templates continue to operate in a flat namespace, but there is a means of accessing values from parent forms in factor controller code. In the case of (ii), you would manually copy those values of interest into the child template's namespace in the init quotation (or wherever). Do you have any view on which (if either) is the better approach? I guess there would also be issues regarding where validation errors live. I haven't made much use of the validations system yet so I don't have any ideas about how this would work. For the moment, I think I will just generate the list XML in factor using the child template, then bung the generated XML into the parent template. I hadn't thought of doing this when I originally posted, but it seems like an ok workaround. Alex On 11 February 2010 07:23, Slava Pestov sl...@factorcode.org wrote: Hi Alex, The reason you can't do what you want to do is that both the boilerplate and action responders call begin-form at the top of their call-responder* methods. Chloe forms are built on top of actions and the machinery in the html.forms vocabulary. Nesting forms is supported only to the extent that validation errors work properly for a page with multiple forms. For example, in the pastebin, a Chloe template renders the actual paste, and there is a sub-template with the form for adding an annotation. Validation errors in the annotation form appear on the annotation template's components, instead of on components of the same name in the paste template. To support what you want to do, the templating code needs to be extended to support nested templates in a more advanced way. In particular, Chloe tags for placing components need to be able to address values from parent forms as well as child forms, using some kind of namespacing syntax. If that's a project you feel like taking on, let me know ;-) Slava On Thu, Feb 11, 2010 at 7:09 PM, Alex Drummond a.d.drumm...@googlemail.com wrote: Hi, I've noticed that values from the parent template don't seem to be available to a child template rendered via t:call-next-template. I would like to render a list of records, with the html template for each record in a separate template file. Could anyone tell me the best way to do this with Chloe? The child template is also used to render individual records in other contexts, so I'd really like to keep it in its own file. thanks, Alex -- SOLARIS 10 is the OS for Data Centers - provides features such as DTrace, Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW http://p.sf.net/sfu/solaris-dev2dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk -- SOLARIS 10 is the OS for Data Centers - provides features such as DTrace, Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW http://p.sf.net/sfu/solaris-dev2dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk -- SOLARIS 10 is the OS for Data Centers - provides features such as DTrace, Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW http://p.sf.net/sfu/solaris-dev2dev ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
Re: [Factor-talk] can somebody test a factor branch on macosx for me please?
It compiles ok, but I get the same bootstrap error for both images: *** Stage 2 early init... You have triggered a bug in Factor. Please report. critical_error: Bad rel type: c009 This was using make macosx-x86-32' / make macosx-x86-64, and then ./factor -i=IMAGE to bootstrap. It's been a while since I've done a build without the factor.sh script, so sorry if I was just executing the wrong commands. Alex 2009/8/22 Phil Dawes p...@phildawes.net: Hi Factor list, I'm still working toward a kernel threaded factor vm and the finish line is in sight! The only problem is that I don't have a macosx platform to test on. If there's anybody reading this on osx with a build setup would you mind testing this for me please?: http://github.com/phildawes/factor/tree/reentrantvm-dev http://phildawes.net/tmp/images/boot.x86.32.image http://phildawes.net/tmp/images/boot.unix-x86.64.image I'm hoping it should compile and bootstrap fine (win32 and linux64 are working). Many thanks, Phil -- Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk -- Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
[Factor-talk] Division by zero
Hi all, The new 1/0 0/0 -1/0 syntax for the special floating point values seems to raise a division by zero error whenever it's used. For example, in evaluating the following expression: 1/0 0.0 Is there any way of using these values in calculations that doesn't cause an error to be raised? I couldn't find anything in the documentation on floats. thank you, Alex -- Register Now Save for Velocity, the Web Performance Operations Conference from O'Reilly Media. Velocity features a full day of expert-led, hands-on workshops and two days of sessions from industry leaders in dedicated Performance Operations tracks. Use code vel09scf and Save an extra 15% before 5/3. http://p.sf.net/sfu/velocityconf ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
Re: [Factor-talk] Division by zero
Ah, but if you actually evaluate [ 1.0 0.0 / ], it seems that you get the value 1/0 and no error is raised in subsequent calculations using this value. Alex 2009/5/2 Alex Drummond a.d.drumm...@googlemail.com: Hi all, The new 1/0 0/0 -1/0 syntax for the special floating point values seems to raise a division by zero error whenever it's used. For example, in evaluating the following expression: 1/0 0.0 Is there any way of using these values in calculations that doesn't cause an error to be raised? I couldn't find anything in the documentation on floats. thank you, Alex -- Register Now Save for Velocity, the Web Performance Operations Conference from O'Reilly Media. Velocity features a full day of expert-led, hands-on workshops and two days of sessions from industry leaders in dedicated Performance Operations tracks. Use code vel09scf and Save an extra 15% before 5/3. http://p.sf.net/sfu/velocityconf ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
[Factor-talk] init-gadget
Hi, I just noticed that the init-gadget word was removed at some point. Can I just remove calls to this in my UI code (this seems to work), or is no explicit initialization necessary any more? thanks, Alex -- Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
[Factor-talk] FFI question
Hi, I was wondering if there is any way to pass a Factor object to C callback that was constructed using alien-callback. thanks, Alex -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
[Factor-talk] Strange 'let' behavior
Hi, Not sure if I'm misunderstanding how let is supposed to work, but I'm getting some strange behavior when using it to bind symbols to multiple values. Specifically, the 'good' word works as expected, but the 'bad' word doesn't: : good ( -- v ) [let | x [ 1 2 ] | x drop ] ; : bad ( -- v ) [let | x [ 1 2 ] | x 2drop x drop ] ; 'good' returns 1, as expected, whereas 'bad' leads to a compile error and adds nothing to the stack. Alex -- This SF.net email is sponsored by: SourcForge Community SourceForge wants to tell your story. http://p.sf.net/sfu/sf-spreadtheword ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
Re: [Factor-talk] Locals Usage
Dan This is my philosophy too. I do use locals sometimes, but Dan typically only one or two locals words per vocabulary. The Dan version without locals is usually cleaner, I find. Same here, mostly. I think a good knowledge of the libraries really helps for writing clean stack code, because this way most of the local state ends up stored in objects (e.g. your current position in a sequence is stored in a slice object, rather than being another variable that you have to carry around on the stack). Alex -- This SF.net email is sponsored by: SourcForge Community SourceForge wants to tell your story. http://p.sf.net/sfu/sf-spreadtheword ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
Re: [Factor-talk] Which Emacs to use on Mac?
I use Carbon Emacs which you can download from apple's site: http://www.apple.com/downloads/macosx/unix_open_source/carbonemacspackage.html I think it's more or less the same as Emacs.app from the CVS. Alex 2009/1/12 Jon Kleiser jon.klei...@usit.uio.no: Hi, Which Emacs do you recommend for use with Factor/FUEL on Mac OS X (Intel)? Is Aquamacs a good one? /Jon -- Check out the new SourceForge.net Marketplace. It is the best place to buy or sell services for just about anything Open Source. http://p.sf.net/sfu/Xq1LFB ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk -- This SF.net email is sponsored by: SourcForge Community SourceForge wants to tell your story. http://p.sf.net/sfu/sf-spreadtheword ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
[Factor-talk] sequences.lib
Hi, I was just wondering what's going on with the removal of the sequences.lib vocabulary. Some useful words like reduce* seem to have been lost. Is there any particular reason for removing these rather than moving them to another vocab? thanks, Alex -- ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
Re: [Factor-talk] sequences.lib
That all makes sense, I would just suggest reduce* (the reduce that requires the list to be of at least length one) as a useful word to have in the sequences library. This is mostly my own personal bias, as I've found it useful in my own code, but FWIW Haskell has an equivalent in its standard List library (foldl1). Alex 2008/12/19 Slava Pestov sl...@factorcode.org: The existence of sequences.lib, combinators.lib and so on always bothered me, and I was opposed to the idea from the very start, because instead of thinking about the right place to put a utility word, people would just dump them in there. So we ended up with an ad-hoc grab-bag of utilities with no consistency, documentation or tests. Many of them were not used, and didn't even work correctly. Over time the useful and clean words were split off into their own vocabularies or moved into core, and code in basis/ and extra/ that used the *.lib vocabularies was refactored to use the new words, or in some cases, not need them. Now that nothing in basis/ and extra/ used them anymore, I put them in unmaintained/, and this is where they will stay. If you see some words in there that are useful, and there is no clean way to express your code without them, and you don't want to just have a copy of these words in your own vocabulary as a private utility, then we can consider moving them into a new vocabulary, or even the core. But I don't want to return to having these vocabularies of random, poorly-documented and poorly-tested utilities that end up cluttering the language only to save a few tokens here and there. Slava On Fri, Dec 19, 2008 at 7:28 AM, Alex Drummond a.d.drumm...@googlemail.com wrote: Hi, I was just wondering what's going on with the removal of the sequences.lib vocabulary. Some useful words like reduce* seem to have been lost. Is there any particular reason for removing these rather than moving them to another vocab? -- ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk -- ___ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
