arbiter-devel Mailing List for Arbiter
Status: Pre-Alpha
Brought to you by:
lastcraft
Menu
▾
▴
arbiter-devel — Technical discussion and project feedback
You can subscribe to this list here.
| 2004 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
(66) |
Oct
(20) |
Nov
(43) |
Dec
(2) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2005 |
Jan
(4) |
Feb
(9) |
Mar
(84) |
Apr
(6) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(3) |
Nov
(3) |
Dec
|
| 2006 |
Jan
(8) |
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(1) |
Nov
|
Dec
|
|
From: Marcus B. <ma...@la...> - 2006-10-11 16:33:35
|
Hi... Marcus Baker wrote: > It comes with some handy functions for string manipulation like > converting from CamelCase, unsdercored_words, pluralizing, singularizing. I am not sure I have already replied to you for this e-mail. It rather got lost in my mail system. Anyway, it sounds excellent! It's actually the sort of thing I need at work anyway, so I'll be taking a look. > Regards, > > Bermi > Where are you based? yours, Marcus -- Marcus Baker ma...@la... |
|
From: Marcus B. <ma...@la...> - 2006-01-21 02:41:26
|
Hi... Noel Darlow wrote: > Progress report: I'm up to my eyeballs at work and have only managed to > steal away for the odd hour here and there. I think I've figured out > enough about rtf to grab user stories. Should have something to check in > soon. Please feel free to take your time. You have already sorted out the web site, which is a good start. > > Noel yours, Marcus -- Marcus Baker, ma...@la... - http://www.lastcraft.com/ PHP London every first Thursday - http://www.phplondon.org/ |
|
From: Marcus B. <ma...@la...> - 2006-01-21 02:31:17
|
Hi...
Noel Darlow wrote:
> Should we create a precise definition of the Arbiter language? I think
> we'll need this for test generation.
We can evolve it as we go. The usability tests are the real decider for
the langauge.
>
> Some ideas in: spikes/arbiter_language/language.rtf.
Here are my guesses, but again this is armchair usability, and therefore
wrong for sure.
>> The vocabulary will have three main categories (?):
Macros are important. I find it's quite common to have to perform a
particular web test sequence, such as logging in.
>> go [to] "name of page" /(go)\s*to\s*"([^"]+)"/i
I think we should have a best guess without quoting. Quotes are fiddly
and could get used in places thet are not intended. Quoting should be
possible to remove ambiguity, but not necessary.
I think synonyms are important for readability (never mind this has to
one day work with multiple spoken languages):
Also "start at", "home" (predefined elsewhere as a macro?), "begin"?
>> click [on] "hyperlinked text" /(click)\s*[^"]*\s*"([^"]+)"/i
This should click on the first text found no matter what type it is.
There is the more specific "click link" when a link is needed. I have
just added this capability to SimpleTest (still a rough cut). I think
this will be the only command necessary unless the web tester cannot
handle it.
>> image names ought to be glossary terms? auto-added to glossary when
found?
Too confusing.
>> "field name" will of course be the visible field name (will there
always be one..?) not the html name
There is always the <label> tag and the <fieldset> tag. As a
requirements tool, we can "encourage" correct HTML. If they want to use
the name attribute or id then we will deny it initially and see how many
people complain. The beta testers will haveto decide this one.
>> see text "foo" /(see)\s*([^"]+)\s*"([^"]+)"/i
Synonyms "must see", "should see" and "expect". Also, the title should
count as a hit for a straight "see" (but not "see text", "see body text").
"see title" can double up as both heading and title bar. The exact
distinction will be up to graphic designers rather than project
managers. A wider range of possible hits (looser spec) should be the
default.
>> Some stories will require conditionals?
Why? Separate tests should be written (we can still use macros) and we
can invert the logic ("must not see").
>> Variables?
Yikes. Not on a first version. Even if users asked for this one, I would
fight it. Project specs should be clearly understood by everyone. This
is a step in the wrong direction for me.
>> A list of items in one line:
Possibly. Certainly we should make it as tolerant as possible.
>
> Noel
Damn it, you've got me excited about the project again :).
yours, Marcus
--
Marcus Baker, ma...@la... - http://www.lastcraft.com/
PHP London every first Thursday - http://www.phplondon.org/
|
|
From: Mark R. <ma...@mc...> - 2006-01-20 08:01:15
|
> Some ideas in: spikes/arbiter_language/language.rtf. This is good... I'll have to read more closely, but here are some initial thoughts: "It should be tolerant of minor variations in user input" I particularly agree with this statement - this is one of the primary reasons for such a document based tool existing. "The vocabulary will have three main categories (?): actions, assertions, conditionals" I just checked in an old hack that demonstrates a proof of concept for executing basic test statements. There are some slight variations between how I constructed the test language, to what you have suggested, but they seem to be close enough to merge into something coherent. I described the vocabulary as made up of 'commands' and 'subjects'. 'commands' are verbs that perform an action or assertion eg: 'go','start','assert','should','expect','click','set' 'subjects' are elements or aspects of the target document to operate on eg: 'text', 'title', 'field', 'button', 'link', 'submit' For example, the following method phrases generate an identical assertion: - Expect text "Arbiter" - Expect to see text "Arbiter" - Should see text "Arbiter" - Should be text "Arbiter" - Assert text "Arbiter" Note that the test list I just checked in actually runs and passes green on my localhost install of Arbiter, but this is wildly different from the code currently in CVS. I need to go back in and pick it apart, and I doubt that many of my mods are even worth committing. Am just about to check out a clean CVS version and reinstall the whole thing. Regards, Mark |
|
From: Noel D. <ma...@mc...> - 2006-01-20 07:10:50
|
Should we create a precise definition of the Arbiter language? I think we'll need this for test generation. Some ideas in: spikes/arbiter_language/language.rtf. Noel |
|
From: Noel D. <ma...@mc...> - 2006-01-20 04:38:10
|
> I've been reading looking at Arbiter and found as you might find useful > a port of the Ruby on Rails Inflector Thanks. Will have a look at that. > Which of course reminds me that I still haven't checked in any of the > Arbiter stuff I did at the end of last year. Yes please. Ta. Progress report: I'm up to my eyeballs at work and have only managed to steal away for the odd hour here and there. I think I've figured out enough about rtf to grab user stories. Should have something to check in soon. Noel |
|
From: Marcus B. <ma...@la...> - 2006-01-19 23:41:29
|
Hi... Mark Rickerby wrote: > Which of course reminds me that I still haven't checked in any of the > Arbiter stuff I did at the end of last year. You should check in every few hours :(. If you don't like it, you can always rol back. This is a problem, because Noel will be learning code taht is out of date. Can you check in right away? >I got sidetracked with > learning Java and examining the differences between SimpleTest and > JWebUnit. What are the differences? > > Regards, > Mark yours, Marcus -- Marcus Baker, ma...@la... - http://www.lastcraft.com/ PHP London every first Thursday - http://www.phplondon.org/ |
|
From: Mark R. <co...@gm...> - 2006-01-19 23:20:49
|
Cool. Last week I noticed he added that link to my page on the vic university wik= i: http://www.elvis.ac.nz/brain?PluralizationMapping Which of course reminds me that I still haven't checked in any of the Arbiter stuff I did at the end of last year. I got sidetracked with learning Java and examining the differences between SimpleTest and JWebUnit. Sorry. Regards, Mark On 1/20/06, Marcus Baker <ma...@la...> wrote: > > > -------- Original Message -------- > Received: by 10.11.120.58 with HTTP; Wed, 18 Jan 2006 16:48:25 -0800 (PST= ) > Message-ID: <9c3...@ma...> > Date: Thu, 19 Jan 2006 01:48:25 +0100 > From: "Bermi Ferrer :: Akelos Media, S.L." <be...@gm...> > Reply-To: be...@ak... > To: ma...@la... > Subject: Arbiter: A text Inflector that you might find useful > MIME-Version: 1.0 > Content-Type: multipart/alternative; > boundary=3D"----=3D_Part_10835_31167361.1137631705422" > X-SpamCop-Checked: 192.168.1.103 195.177.192.5 66.249.82.199 10.11.98.47 > 10.11.120.58 > > > > Hi Marcus, > > I've been reading looking at Arbiter and found as you might find useful > a port of the Ruby on Rails Inflector that I've posted to > http://pear.php.net/pepr/pepr-proposal-show.php?id=3D342 > <http://pear.php.net/pepr/pepr-proposal-show.php?id=3D342> > It comes with some handy functions for string manipulation like > converting from CamelCase, unsdercored_words, pluralizing, singularizing. > > And of course it comes with unit tests I made using your wonderful > simpletest. > > Regards, > > Bermi > > -- > Marcus Baker, ma...@la... - http://www.lastcraft.com/ > PHP London every first Thursday - http://www.phplondon.org/ > > > > ------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi= les > for problems? Stop! Download the new AJAX search engine that makes > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat= =3D121642 > _______________________________________________ > Arbiter-devel mailing list > Arb...@li... > https://lists.sourceforge.net/lists/listinfo/arbiter-devel > -- ---------------------------------------------------------------------- mark rickerby http://maetl.coretxt.net.nz http://www.mcs.vuw.ac.nz/~maetl/ ---------------------------------------------------------------------- |
|
From: Marcus B. <ma...@la...> - 2006-01-19 13:55:54
|
-------- Original Message -------- Received: by 10.11.120.58 with HTTP; Wed, 18 Jan 2006 16:48:25 -0800 (PST) Message-ID: <9c3...@ma...> Date: Thu, 19 Jan 2006 01:48:25 +0100 From: "Bermi Ferrer :: Akelos Media, S.L." <be...@gm...> Reply-To: be...@ak... To: ma...@la... Subject: Arbiter: A text Inflector that you might find useful MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_10835_31167361.1137631705422" X-SpamCop-Checked: 192.168.1.103 195.177.192.5 66.249.82.199 10.11.98.47 10.11.120.58 Hi Marcus, I've been reading looking at Arbiter and found as you might find useful a port of the Ruby on Rails Inflector that I've posted to http://pear.php.net/pepr/pepr-proposal-show.php?id=342 <http://pear.php.net/pepr/pepr-proposal-show.php?id=342> It comes with some handy functions for string manipulation like converting from CamelCase, unsdercored_words, pluralizing, singularizing. And of course it comes with unit tests I made using your wonderful simpletest. Regards, Bermi -- Marcus Baker, ma...@la... - http://www.lastcraft.com/ PHP London every first Thursday - http://www.phplondon.org/ |
|
From: Marcus B. <ma...@la...> - 2005-11-08 18:04:45
|
Hi... Mark Rickerby wrote: > Am assuming that you mean that a grammar will evolve as a reflection > of method names from a web testing API... In that sense, it may not > even need to be a grammar. No I mean that the language will evolve by usability. It's really a language for professionals, but not necessarily programmers. Although arbiter and the underlying test framework will be connected with each other, I think the language is king. Say we generate JWebUnit test cases and JWebUnit is missing some feature. I would rather write a JWebUnit extension than alter the test langauge. > > Anyway - I'm going to review all the stuff in CVS this week, and post > any tidy ups / questions / notes of relevance. The diagrams (UML) are horribly out of date. Feel free to dump them. > > Regards, > Mark yours, Marcus -- Marcus Baker http://www.lastcraft.com/ ma...@la... ma...@wo... |
|
From: Mark R. <ma...@mc...> - 2005-11-06 23:04:35
|
> > I've personally put a fair bit of effort in to this, and maybe made > > more than a few mistakes, but overall, think that the main problems > > have just been a lack of communication and consistent work being done > > / progress being made. > > Agreed, although I take most of the blame on my own shoulders. Well, it's not really about blame... something is better than nothing, and the roadmap provides a reasonably good direction for the project. Besides, we all have a lot to thank you for, in terms of the continuous evolution of SimpleTest over the past couple of years. Just gotta be wary of starting too many projects at once! > > I have come to understand that the core of Arbiter is not so > > much the document parser, as it is the test language, which should be > > independent of any particular markup format. > > I think such a grammer will evolve. Arbiter seems to be the only > project that is tackling this in relation to requirements. Selenium > is much more technically focussed. Yes, it's important to keep that in perspective. Theres no need to reinvent the wheel here. Am assuming that you mean that a grammar will evolve as a reflection of method names from a web testing API... In that sense, it may not even need to be a grammar. Anyway - I'm going to review all the stuff in CVS this week, and post any tidy ups / questions / notes of relevance. Regards, Mark |
|
From: Mark R. <ma...@mc...> - 2005-11-06 22:49:15
|
Hi! On 10/6/05, Jean-Marie Porchet <jm...@at...> wrote: > I haven't been active in this project for a long time, but starting from = the > 15th of November, I will have almost 8 hours a day to dedicate to > programming in general (and Arbiter in particular). Before that, I have t= o > pass exams, but I guess they will pose no problems, given that I dedicate > enough time. Sounds awesome, but be careful... I wouldn't jump in and commit too much time as yet - first get a feel for the flavour of the project, and what you would be most interested in doing, start with something small and easily achievable in a short timeframe (See the TODO file and the Roadmap: http://arbiter.sourceforge.net/roadmap.html). Then, just email the list, explaining which task you're undertaking, and get started. Focus on writing tests, and getting that particular piece of functionality you've chosen working. It shouldn't take too much time to make some significant progress. > > The main problem with Arbiter is that I really have to dive in the code, = to > "feel it", and I foresee that it will take me a lot of time. > I guess I can do some things, if someone tells me what (and maybe even ho= w > :p ). Marcus has a couple of sketches of the proposed static structure (in CVS, under /design/ ), the direction the PHP code is heading should make a lot more sense after studying the diagrams. Also remember that the code in CVS is there because it is what works now, not because it is the best code in the world - it is in no way permanent, and if you need to rip stuff out and rewrite bits of it, don't be afraid to. Use the tests to guide this process. Any specific questions you have, just email the list. Thanks, Mark |
|
From: Marcus B. <ma...@la...> - 2005-10-28 18:27:05
|
Hi... Sorry about the delay. My truly lousy ISP (www.demon.net) stuffed up my email. Mark Rickerby wrote: > This project has been going for over a year now, and although it seems > to have been dead for a few months, there is actually some basic stuff > that works, and I still think that the idea is solid and has a huge > amount of potential. I just wanted to check in with the list and see > if there was still interest in contributing to Arbiter, and pushing > the project forward in coming months. I am feeling forever guilty that this has dropped out of the bottom of my stack. I still have it queued though, so you never know. > > I've personally put a fair bit of effort in to this, and maybe made > more than a few mistakes, but overall, think that the main problems > have just been a lack of communication and consistent work being done > / progress being made. Agreed, although I take most of the blame on my own shoulders. > > Secondly to Marcus, what do you see the current relationship between > Arbiter and Simpletest being? The extra work on SimpleTest was simply to add file upload support so thet web tests could be written for Arbiter. > Should we be waiting for anything (eg: > new mocks)? Should we be aiming for a specific version of Simpletest? No. > Do you still think we should base further development on the existing > TODO and BACKLOG? What do you want to see happen? I think the roadmap I checked into the documentation is a better guide. > I'm personally most interested in developing a grammar for web > testing. I have come to understand that the core of Arbiter is not so > much the document parser, as it is the test language, which should be > independent of any particular markup format. If anyone is interested, > I have some time in the next few weeks to develop and document this > further. I think such a grammer will evolve. Arbiter seems to be the only project that is tackling this in relation to requirements. Selenium is much more technically focussed. > > Regards, > Mark yours, Marcus -- Marcus Baker http://www.lastcraft.com/ ma...@la... ma...@wo... |
|
From: Jean-Marie P. <jm...@at...> - 2005-10-06 10:12:45
|
Hello, I haven't been active in this project for a long time, but starting from = the 15th of November, I will have almost 8 hours a day to dedicate to programming in general (and Arbiter in particular). Before that, I have = to pass exams, but I guess they will pose no problems, given that I = dedicate enough time. The main problem with Arbiter is that I really have to dive in the code, = to "feel it", and I foresee that it will take me a lot of time. I guess I can do some things, if someone tells me what (and maybe even = how :p ). In the mean time, I guess it would be a good idea to create such a = grammar. Maybe it would be a good opportunity to put my formal languages' class = to action. Let me know! Jean-Marie -----Message d'origine----- De : arb...@li... [mailto:arb...@li...] De la part de Mark Rickerby Envoy=E9 : lundi, 3. octobre 2005 03:30 =C0 : arb...@li... Objet : [Arbiter-devel] state of the project Hi all, This project has been going for over a year now, and although it seems = to have been dead for a few months, there is actually some basic stuff that works, and I still think that the idea is solid and has a huge amount of potential. I just wanted to check in with the list and see if there was still interest in contributing to Arbiter, and pushing the project = forward in coming months. I've personally put a fair bit of effort in to this, and maybe made more than a few mistakes, but overall, think that the main problems have just been a lack of communication and consistent work being done / progress = being made. So I have a couple of questions, firstly to anyone on the list - are you still interested in contributing? and what do you want to do? Secondly to Marcus, what do you see the current relationship between = Arbiter and Simpletest being? Should we be waiting for anything (eg: new mocks)? Should we be aiming for a specific version of Simpletest? Do you still think we should base further development on the existing = TODO and BACKLOG? What do you want to see happen? If anyone has any questions about what's in CVS and what works and what doesn't, please let me know. I'm personally most interested in developing a grammar for web testing. = I have come to understand that the core of Arbiter is not so much the = document parser, as it is the test language, which should be independent of any particular markup format. If anyone is interested, I have some time in = the next few weeks to develop and document this further. Regards, Mark ------------------------------------------------------- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, = discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl _______________________________________________ Arbiter-devel mailing list Arb...@li... https://lists.sourceforge.net/lists/listinfo/arbiter-devel |
|
From: Mark R. <ma...@mc...> - 2005-10-03 01:29:50
|
Hi all, This project has been going for over a year now, and although it seems to have been dead for a few months, there is actually some basic stuff that works, and I still think that the idea is solid and has a huge amount of potential. I just wanted to check in with the list and see if there was still interest in contributing to Arbiter, and pushing the project forward in coming months. I've personally put a fair bit of effort in to this, and maybe made more than a few mistakes, but overall, think that the main problems have just been a lack of communication and consistent work being done / progress being made. So I have a couple of questions, firstly to anyone on the list - are you still interested in contributing? and what do you want to do? Secondly to Marcus, what do you see the current relationship between Arbiter and Simpletest being? Should we be waiting for anything (eg: new mocks)? Should we be aiming for a specific version of Simpletest? Do you still think we should base further development on the existing TODO and BACKLOG? What do you want to see happen? If anyone has any questions about what's in CVS and what works and what doesn't, please let me know. I'm personally most interested in developing a grammar for web testing. I have come to understand that the core of Arbiter is not so much the document parser, as it is the test language, which should be independent of any particular markup format. If anyone is interested, I have some time in the next few weeks to develop and document this further. Regards, Mark |
|
From: Marcus B. <ma...@la...> - 2005-04-06 01:52:20
|
Hi. Just a progress report: Hours volunteered = 40 Hours worked = 31 Ideal hours completed = 22 If people can write their likely available hours into the TODO and what they would most like to work on, I'll work out a possible iteration plan. Please take a look at the road map in the website folder in CVS before suggesting something taht is planned later. yours, Marcus -- Marcus Baker, ma...@la... - http://www.lastcraft.com/ PHP London every first Thursday - http://www.phplondon.org/ |
|
From: Marcus B. <ma...@la...> - 2005-04-06 01:45:41
|
Hi. This is just a tarball of the essential files in CVS. To run the tests... tar -zxf arbiter_iteration3.tar.gz cd arbiter chmod 777 temp chmod 777 configuration cd test ln -s /path/to/simpletest/ php all_tests.php yours, Marcus -- Marcus Baker, ma...@la... - http://www.lastcraft.com/ PHP London every first Thursday - http://www.phplondon.org/ |
|
From: Christian A. <ch...@mo...> - 2005-04-05 20:38:56
|
hey guys, just revisited the FIT website (http://fit.c2.com/) and found that they = made huge progress since my last visit. the documentation is more = complete and the added examples a really worth a read. cheers Christian |
|
From: Mark R. <co...@gm...> - 2005-04-03 04:54:16
|
> I am building the monthly tarball release and need to know which files > to include. Which of the samples are actually used in the test? Better > yet, if someone can just set them up in scripts/build_tarball.sh then I > can build teh release right after. Done. > yours, Marcus > -- Thanks, Mark |
|
From: Marcus B. <ma...@la...> - 2005-04-02 01:23:44
|
Hi... Mark Rickerby wrote: > I'm kind of embarassed because I feel like I've tried about 10 > different ways of extracting a title from an RTF document, but none of > them seemed to fit together quite right or work in all cases. Hey, that's normal :). > I've hooked this in to the Document/Repository, and added expectations > for the sample documents - all the tests are running green on my > system (PHP 4.3.8 / IIS/Win). Mine too. > > There are many many things wrong with the implementation. I think the > design is ok, but expect it to change heavily under refactoring and > evolution. Ok, I'll review it over the weekend prior to building a developer release. I'll do the statistics for the iteration as well. > Thanks, > Mark Good stuff. yours, Marcus -- Marcus Baker, ma...@la... - http://www.lastcraft.com/ PHP London every first Thursday - http://www.phplondon.org/ |
|
From: Marcus B. <ma...@la...> - 2005-04-02 01:20:39
|
Hi. I am building the monthly tarball release and need to know which files to include. Which of the samples are actually used in the test? Better yet, if someone can just set them up in scripts/build_tarball.sh then I can build teh release right after. yours, Marcus -- Marcus Baker, ma...@la... - http://www.lastcraft.com/ PHP London every first Thursday - http://www.phplondon.org/ |
|
From: Mark R. <co...@gm...> - 2005-03-31 05:16:54
|
I'm kind of embarassed because I feel like I've tried about 10 different ways of extracting a title from an RTF document, but none of them seemed to fit together quite right or work in all cases. Anyway - I cobbled together a new RTF parser based on Christian's tokenizer, but changed the processing loop to trigger event handlers rather than iterating through token objects... I've hooked this in to the Document/Repository, and added expectations for the sample documents - all the tests are running green on my system (PHP 4.3.8 / IIS/Win). There are many many things wrong with the implementation. I think the design is ok, but expect it to change heavily under refactoring and evolution. Rationale for doing it this way: 1) Scanner should have no concern with document structure, and should just be responsible for breaking down the RTF syntax into raw token events. 2) The role of the parser is to translate the raw token events into more structured document events. The parser is very light, and just ignores everything except for a small subset of useful controls. The idea is to keep up the pretence that we can read a nice hierachical set of nodes out of the document, regardless of how disordered the actual structure is. This might seem a weird way to parse RTF, it is quite different from Christian's original model for the rawdocument, but I think that might be folded back in as we go forward - just at the moment, I am finding the Sax based concept more familiar to work with, and easier to refactor and test. 3) The Reader is totally crap at the moment - but is supposed to be a simple callback handler to get wanted elements out of the parser. This should be the part that ends up doing the statistics and analysis on the document. Currently needs some more exploration, and verification of arguments to the callbacks - for example, it would be a good idea to pass the location of each token up from the scanner through the parser. API needs some thought, as to how much work the parser does in sorting out a structure for the document - whether the reader is expanded to cover more granular callbacks (eg: addFont(), addStyle()) or whether this work is done by the reader itself (addProperty()) Hope this is useful to you guys. Sorry if it is a mess, or too much of a braindump to make sense. I haven't had very much time to focus on this stuff, which I'm finding more than a little difficult, so my recent rantings may be quite confused. Thanks, Mark |
|
From: Mark R. <co...@gm...> - 2005-03-10 08:30:39
|
> Simple rule: get the project working end to end the siplest way possible > to get to alpha. Once that's done, if two people request it then do it. > Otherwise we will get feature creep at a time when the project is at > it's most vulnerable - the beginning. > you are right of course... I'm just trying to get my head around the ways that commercial products deviate in their use of RTF, the standard being what the software actually does, not what is written down in a spec. |
|
From: Marcus B. <ma...@la...> - 2005-03-10 02:17:33
|
Hi... Mark Rickerby wrote: > ...but don't want to miss any other > programs that are also relevant. Simple rule: get the project working end to end the siplest way possible to get to alpha. Once that's done, if two people request it then do it. Otherwise we will get feature creep at a time when the project is at it's most vulnerable - the beginning. > Thanks, > Mark yours, Marcus -- Marcus Baker, ma...@la... - http://www.lastcraft.com/ PHP London every first Thursday - http://www.phplondon.org/ |
|
From: Mark R. <co...@gm...> - 2005-03-10 00:25:03
|
Any other word processors we should be aiming for? I would be happy to settle with Word and Openoffice - but don't want to miss any other programs that are also relevant. RTF is a standard held by Microsoft, and Word is pretty much the benchmark that all other commercial tools measure themselves by, in terms of being able to import/export rtf/docs. I have decided to ignore WordPad (Win) and TextEdit (Mac OS/X) as the RTF they spit out is just too horrible to deal with, and neither supports highlighting words anyway. But there might be some other standards to look at... Lotus? Wordperfect? I don't even know if they still exist, but if anyone knows or uses them, let us know. Or else even better, check in a sample document with the filename reflecting the tool that it was saved in. That would be extraordinarily helpful at this stage. Thanks, Mark |
