Re: [OpenFontLibrary] Phone conversation with Ed Trager
aiki 2010/3/8 Robert Martinez : > On 03/08/2010 04:47 PM, Ed Trager wrote: >> >> Hi, everyone, >> >> On Mon, Mar 8, 2010 at 9:18 AM, Alexandre Prokoudine >> wrote: >> >>> >>> On 3/8/10, Dave Crossland wrote: >>> >>> >> >> Ed thinks about replacing ccHost with a custom webapp. >> > > >> >> To set the record straight, like many of you I've been wondering all >> these months what the holdup was with the cchost prototype? I >> contributed a bunch of code to that, and naturally would like to see >> the fruits of my labor --as well as the fruits of the labors of so >> many others-- realized. >> >> So this puzzlement was really the genesis of my conversation with >> Dave. And my suspicion going into the conversation was: is cchost >> part of the problem? >> >> Also, I was thinking is cchost the right platform for what we want to do? >> >> So, I was thinking, let's explore an alternative pathway: What do we >> want to do? Write that down as an outline. If the most important >> parts of what we want to do are simple and straightforward, then let's >> just write a custom web app to do it. >> >> But note that such a custom web app would still capitalize on stuff >> that we already collectively know how to do well anyway: i.e., PHP >> with MySQL and, on the Javascript side, jQuery. >> >> Also, for the login and security aspects I was thinking about using >> OpenId (OpenId.net). >> >> Also, I've been working on a bunch of jQuery-based code for web >> applications in my day job, and have also been thinking of using >> OpenId for some web apps in my day job. So I was visualizing how I >> might be able to apply and reuse some of what I have been doing in my >> paid work toward the OFLB project. >> >> Now if Aiki is a really good solution to the problems, then I won't >> argue against it. But is it a really good solution or not? I don't >> yet know, because I only heard about it yesterday when talking with >> Dave, and I don't yet know anything else about it. >> >> In the end, I only argue for using the right tool(s) for the job. So >> that could be Aiki or something else. But it looks more and more like >> cchost was really not the right tool, and from what I understand, >> modifying and customizing cchost was very laborious and frought with >> bugs. >> >> > > Ed suggests just talking about OFLB v2 again for the 4th time at LGM? > >> >> Umm ... I wasn't even at LGM in the last couple of years ... >> >> Ed is offering to cut the code for this, so, don't be mean to him. :-) >>> >>> What's the point of being mean when I can be nasty? :) >>> >>> >> >> It's OK, Alexandre, I'll take it if you agree to let me dish it out too :-) >> >> >>> >>> Alexandre >>> >>> >> >> - Ed >> > > Was it ever considered to use drupal and enhance it? > Ed seems to know it, and I got the feeling there are plenty of PHP savvy > people here. > In terms of reusing a live community and popular code base it might be a good > idea. > > Just a thought! Please don't lynch me for bringing that up if it was a bad > idea :) > > > Robert > -- Jon Phillips http://rejon.org/ http://fabricatorz.com/ internet: @rejon + skype: kidproto +1.415.830.3884 (sf/global) +86.134.3957.2035 (china)
Re: [OpenFontLibrary] Phone conversation with Ed Trager
On 03/08/2010 04:47 PM, Ed Trager wrote: Hi, everyone, On Mon, Mar 8, 2010 at 9:18 AM, Alexandre Prokoudine wrote: On 3/8/10, Dave Crossland wrote: Ed thinks about replacing ccHost with a custom webapp. To set the record straight, like many of you I've been wondering all these months what the holdup was with the cchost prototype? I contributed a bunch of code to that, and naturally would like to see the fruits of my labor --as well as the fruits of the labors of so many others-- realized. So this puzzlement was really the genesis of my conversation with Dave. And my suspicion going into the conversation was: is cchost part of the problem? Also, I was thinking is cchost the right platform for what we want to do? So, I was thinking, let's explore an alternative pathway: What do we want to do? Write that down as an outline. If the most important parts of what we want to do are simple and straightforward, then let's just write a custom web app to do it. But note that such a custom web app would still capitalize on stuff that we already collectively know how to do well anyway: i.e., PHP with MySQL and, on the Javascript side, jQuery. Also, for the login and security aspects I was thinking about using OpenId (OpenId.net). Also, I've been working on a bunch of jQuery-based code for web applications in my day job, and have also been thinking of using OpenId for some web apps in my day job. So I was visualizing how I might be able to apply and reuse some of what I have been doing in my paid work toward the OFLB project. Now if Aiki is a really good solution to the problems, then I won't argue against it. But is it a really good solution or not? I don't yet know, because I only heard about it yesterday when talking with Dave, and I don't yet know anything else about it. In the end, I only argue for using the right tool(s) for the job. So that could be Aiki or something else. But it looks more and more like cchost was really not the right tool, and from what I understand, modifying and customizing cchost was very laborious and frought with bugs. Ed suggests just talking about OFLB v2 again for the 4th time at LGM? Umm ... I wasn't even at LGM in the last couple of years ... Ed is offering to cut the code for this, so, don't be mean to him. :-) What's the point of being mean when I can be nasty? :) It's OK, Alexandre, I'll take it if you agree to let me dish it out too :-) Alexandre - Ed Was it ever considered to use drupal and enhance it? Ed seems to know it, and I got the feeling there are plenty of PHP savvy people here. In terms of reusing a live community and popular code base it might be a good idea. Just a thought! Please don't lynch me for bringing that up if it was a bad idea :) Robert
Re: [OpenFontLibrary] Phone conversation with Ed Trager
On 3/8/10, Ed Trager wrote: > Also, for the login and security aspects I was thinking about using > OpenId (OpenId.net). Not everyone has one of those, so it'd rather be registration or openid > In the end, I only argue for using the right tool(s) for the job. Right on! :) Ed suggests just talking about OFLB v2 again for the 4th time at LGM? >>> > > Umm ... I wasn't even at LGM in the last couple of years ... You mean you've been there before and never told anyone? :) >>> Ed is offering to cut the code for this, so, don't be mean to him. :-) >> >> What's the point of being mean when I can be nasty? :) > > It's OK, Alexandre, I'll take it if you agree to let me dish it out too :-) Aye, cap'n :) Alexandre
Re: [OpenFontLibrary] Phone conversation with Ed Trager
Hi, everyone, On Mon, Mar 8, 2010 at 9:18 AM, Alexandre Prokoudine wrote: > On 3/8/10, Dave Crossland wrote: > Ed thinks about replacing ccHost with a custom webapp. >>> To set the record straight, like many of you I've been wondering all these months what the holdup was with the cchost prototype? I contributed a bunch of code to that, and naturally would like to see the fruits of my labor --as well as the fruits of the labors of so many others-- realized. So this puzzlement was really the genesis of my conversation with Dave. And my suspicion going into the conversation was: is cchost part of the problem? Also, I was thinking is cchost the right platform for what we want to do? So, I was thinking, let's explore an alternative pathway: What do we want to do? Write that down as an outline. If the most important parts of what we want to do are simple and straightforward, then let's just write a custom web app to do it. But note that such a custom web app would still capitalize on stuff that we already collectively know how to do well anyway: i.e., PHP with MySQL and, on the Javascript side, jQuery. Also, for the login and security aspects I was thinking about using OpenId (OpenId.net). Also, I've been working on a bunch of jQuery-based code for web applications in my day job, and have also been thinking of using OpenId for some web apps in my day job. So I was visualizing how I might be able to apply and reuse some of what I have been doing in my paid work toward the OFLB project. Now if Aiki is a really good solution to the problems, then I won't argue against it. But is it a really good solution or not? I don't yet know, because I only heard about it yesterday when talking with Dave, and I don't yet know anything else about it. In the end, I only argue for using the right tool(s) for the job. So that could be Aiki or something else. But it looks more and more like cchost was really not the right tool, and from what I understand, modifying and customizing cchost was very laborious and frought with bugs. >>> Ed suggests just talking about OFLB v2 again for the 4th time at LGM? >> Umm ... I wasn't even at LGM in the last couple of years ... >> Ed is offering to cut the code for this, so, don't be mean to him. :-) > > What's the point of being mean when I can be nasty? :) > It's OK, Alexandre, I'll take it if you agree to let me dish it out too :-) > Alexandre > - Ed
Re: [OpenFontLibrary] Phone conversation with Ed Trager
On 3/8/10, Dave Crossland wrote: >>> Ed thinks about replacing ccHost with a custom webapp. >> >> Ed suggests just talking about OFLB v2 again for the 4th time at LGM? > > Ed is offering to cut the code for this, so, don't be mean to him. :-) What's the point of being mean when I can be nasty? :) Alexandre
Re: [OpenFontLibrary] Phone conversation with Ed Trager
We are in a productive mode these days, and I think making conversion to aiki is going to be faster and better longer term than working on a dead codebase, with a system that requires limited access to our main server. The point with Aiki is to allow various levels of access from the web to enable people rather than lock out with arcane codebase and limited server access. Jon On Mon, Mar 8, 2010 at 6:36 PM, Schrijver wrote: > But I don’t understand what is holding back just launching the ccHost > version—why did I see a functioning prototype a year ago? > Are there some hurdles that are proving difficult, and that could be > overcome by scrapping it for a different system? > > Jon I admire your spirit and I think it is great you are willing to do > this; > (and congrats on the OCAL launch!) > the thing I am a bit scared of is that in shifting systems you will run > into unforeseen delays all over again; and right now the site just needs to > be up, no? > > Eric > > Op 8 mrt 2010, om 10:46 heeft Jon Phillips het volgende geschreven: > > Dudes, we are relaunching OCAL 2.0 monday PST, and we are getting huge > traffic and making great strides. Aiki is awesome, and we could easily port > current awesome design, and ccHost, since we know how to do it, and get the > same response IMO. > > My company has been doing the work, Fabricatorz.com, with Bassel running > the whole thing. > > So, IMO, cchost is over and done, long live aiki. > > At this point its about doing the work, and we are in a configuration to do > this, access ready, and system in place. > > S, the best thing we could get to make this real is support, filing > bugs, and compensation for ronaldo and bassel's time, if we want to > accelerate this. Of course, compensation is not a blocker, just nice icing. > > At the end of the day, i think we could turn around OFLB in 2 weeks, and > get the same work burst, same look, and all the features desired of the > current site. > > I would scrap any new new features for now, and push towards having OCAL > 2.0 and OFLB 2.0 done by LGM2010 so we can stop talking talking and doing > doing what we want, making great fonts and community. > > Anyway, let's get it doneof course, open to discussion about it. > > I don't really see any other viable timely options. ccHost dev is dead. > Writing a new webapp is a death sentence for OFLB. The best option is to > copy what we did for ocal, open up launchpad for oflb, and unite behind aiki > as the core, since we know how and built aiki to handle just this type of > content site. > > Cheers and I'm excited!!! > Jon > > On Mon, Mar 8, 2010 at 5:11 PM, Dave Crossland wrote: > >> On 8 March 2010 04:26, Alexandre Prokoudine >> wrote: >> > On 3/7/10, Dave Crossland wrote: >> > >> >> Ed thinks about replacing ccHost with a custom webapp. >> > >> > Ed suggests just talking about OFLB v2 again for the 4th time at LGM? >> >> Ed is offering to cut the code for this, so, don't be mean to him. :-) >> > > > > -- > Jon Phillips > http://rejon.org/ > http://fabricatorz.com/ > internet: @rejon + skype: kidproto > +1.415.830.3884 (sf/global) > +86.134.3957.2035 (china) > > > -- Jon Phillips http://rejon.org/ http://fabricatorz.com/ internet: @rejon + skype: kidproto +1.415.830.3884 (sf/global) +86.134.3957.2035 (china)
Re: [OpenFontLibrary] Phone conversation with Ed Trager
But I don’t understand what is holding back just launching the ccHost version—why did I see a functioning prototype a year ago? Are there some hurdles that are proving difficult, and that could be overcome by scrapping it for a different system? Jon I admire your spirit and I think it is great you are willing to do this; (and congrats on the OCAL launch!) the thing I am a bit scared of is that in shifting systems you will run into unforeseen delays all over again; and right now the site just needs to be up, no? Eric Op 8 mrt 2010, om 10:46 heeft Jon Phillips het volgende geschreven: > Dudes, we are relaunching OCAL 2.0 monday PST, and we are getting huge > traffic and making great strides. Aiki is awesome, and we could easily port > current awesome design, and ccHost, since we know how to do it, and get the > same response IMO. > > My company has been doing the work, Fabricatorz.com, with Bassel running the > whole thing. > > So, IMO, cchost is over and done, long live aiki. > > At this point its about doing the work, and we are in a configuration to do > this, access ready, and system in place. > > S, the best thing we could get to make this real is support, filing bugs, > and compensation for ronaldo and bassel's time, if we want to accelerate > this. Of course, compensation is not a blocker, just nice icing. > > At the end of the day, i think we could turn around OFLB in 2 weeks, and get > the same work burst, same look, and all the features desired of the current > site. > > I would scrap any new new features for now, and push towards having OCAL 2.0 > and OFLB 2.0 done by LGM2010 so we can stop talking talking and doing doing > what we want, making great fonts and community. > > Anyway, let's get it doneof course, open to discussion about it. > > I don't really see any other viable timely options. ccHost dev is dead. > Writing a new webapp is a death sentence for OFLB. The best option is to copy > what we did for ocal, open up launchpad for oflb, and unite behind aiki as > the core, since we know how and built aiki to handle just this type of > content site. > > Cheers and I'm excited!!! > Jon > > On Mon, Mar 8, 2010 at 5:11 PM, Dave Crossland wrote: > On 8 March 2010 04:26, Alexandre Prokoudine > wrote: > > On 3/7/10, Dave Crossland wrote: > > > >> Ed thinks about replacing ccHost with a custom webapp. > > > > Ed suggests just talking about OFLB v2 again for the 4th time at LGM? > > Ed is offering to cut the code for this, so, don't be mean to him. :-) > > > > -- > Jon Phillips > http://rejon.org/ > http://fabricatorz.com/ > internet: @rejon + skype: kidproto > +1.415.830.3884 (sf/global) > +86.134.3957.2035 (china)
Re: [OpenFontLibrary] Phone conversation with Ed Trager
Dudes, we are relaunching OCAL 2.0 monday PST, and we are getting huge traffic and making great strides. Aiki is awesome, and we could easily port current awesome design, and ccHost, since we know how to do it, and get the same response IMO. My company has been doing the work, Fabricatorz.com, with Bassel running the whole thing. So, IMO, cchost is over and done, long live aiki. At this point its about doing the work, and we are in a configuration to do this, access ready, and system in place. S, the best thing we could get to make this real is support, filing bugs, and compensation for ronaldo and bassel's time, if we want to accelerate this. Of course, compensation is not a blocker, just nice icing. At the end of the day, i think we could turn around OFLB in 2 weeks, and get the same work burst, same look, and all the features desired of the current site. I would scrap any new new features for now, and push towards having OCAL 2.0 and OFLB 2.0 done by LGM2010 so we can stop talking talking and doing doing what we want, making great fonts and community. Anyway, let's get it doneof course, open to discussion about it. I don't really see any other viable timely options. ccHost dev is dead. Writing a new webapp is a death sentence for OFLB. The best option is to copy what we did for ocal, open up launchpad for oflb, and unite behind aiki as the core, since we know how and built aiki to handle just this type of content site. Cheers and I'm excited!!! Jon On Mon, Mar 8, 2010 at 5:11 PM, Dave Crossland wrote: > On 8 March 2010 04:26, Alexandre Prokoudine > wrote: > > On 3/7/10, Dave Crossland wrote: > > > >> Ed thinks about replacing ccHost with a custom webapp. > > > > Ed suggests just talking about OFLB v2 again for the 4th time at LGM? > > Ed is offering to cut the code for this, so, don't be mean to him. :-) > -- Jon Phillips http://rejon.org/ http://fabricatorz.com/ internet: @rejon + skype: kidproto +1.415.830.3884 (sf/global) +86.134.3957.2035 (china)
Re: [OpenFontLibrary] Phone conversation with Ed Trager
On 8 March 2010 04:26, Alexandre Prokoudine wrote: > On 3/7/10, Dave Crossland wrote: > >> Ed thinks about replacing ccHost with a custom webapp. > > Ed suggests just talking about OFLB v2 again for the 4th time at LGM? Ed is offering to cut the code for this, so, don't be mean to him. :-)
Re: [OpenFontLibrary] Phone conversation with Ed Trager
On 3/7/10, Dave Crossland wrote: > Ed thinks about replacing ccHost with a custom webapp. Ed suggests just talking about OFLB v2 again for the 4th time at LGM? Alexandre
Re: [OpenFontLibrary] Phone conversation with Ed Trager
On Sun, 2010-03-07 at 22:31 +0100, Schrijver wrote: > thanks, +1 :-) > > my thoughts: > > 1. > > new architecture is not #1 priority right now— > lets get this one up! +1 again Lots of good ideas for future direction, got to be able to walk before we wear shoes. Or something like that. Liam -- Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/ Pictures from old books: http://fromoldbooks.org/ Ankh: irc.sorcery.net irc.gnome.org www.advogato.org
Re: [OpenFontLibrary] Phone conversation with Ed Trager
thanks, my thoughts: 1. new architecture is not #1 priority right now— lets get this one up! the demo site has been looking so great. i love the design, I love the features, the work all of you have put into it! and it seemed perfectly usable for day to day use. users don’t really care what it is hosted on… 2. when it’s up, can we asap host the code on git or mercurial so anyone can fork and submit patches? the development proces is closed now which is a shame for an open project— with the source available bugs can be filed and patches submitted! 3. after which, the backend… I agree the fonts should be in dvcs. Dvcs is great! I use git but apparently mercurial is awesome as well. The merging/branching thing is what dvcs is all about, it would be a shame to reinvent the weel with something homegrown. these systems are by default not very friendly to novice users since they use the command line (though they are not that difficult—i got taught the basics of git in half an hour). But the web app could provide a nice graphical interface to them! I am sure there is work being done in this area, and if not, well it would be all the more useful to work on it :) I imagine more tech-savvy users could still use the clone url’s to access the repositories directly. Then you would really have a rock-solid collaboration platform. For the app, django or web.py seem like a good idea. I do not know the aiki framework, so i am not sure if it would be suited to integration with the dvcses? An online font editor is a lofty but lovely goal. I imagine it could plug in to the same architecture of dvc. Maybe the repositories need abstraction layer then, so they can easily be acccessed through a restful api (though that would be hard to code i imagine). 4. As a sidenote, i remember dave we wrote about generating otf from ufo’s. i could not get fontforge installed at the time, because my shared hosting did not allow it. but i will be getting a linux vps shortly, so i can try it out. And it is funny that you mention web.py, because I was planning on trying to use that to make a little web service that generated the font files. The idea: -users commits updated ufo to dvcs -post commit hook sends ufo to service -service generates otf, which gets send, to, eeuhm, well some specified location, so that there always is a font file to use for the most recent changes Hey wait that is like a compiler Maybe something nice to tinker at for LGM Best Eric Op 7 mrt 2010, om 21:02 heeft Dave Crossland het volgende geschreven: > Hi! > > Today I made a phonecall to Ed Trager to catch up and discuss the > project. Here are my notes from the talk: > > Dave explained why the site is being held up - that there are 4 people > with SSH access to the new server (me, Ben Weiner, Jon Philips, and Ed > Trager) and of them, I am the most free, and I haven't done the work > of getting ccHost 5 running with the theme Ben and Aaron made and the > font preview program Ed made. Once this is done, the next step will be > to consider what to replace ccHost with. Jon has been involved with > the creation of aikiframework.org (NOT a wiki framework!) that has > just gone live for openclipart.org and this could well be a contender. > > Ed thinks about replacing ccHost with a custom webapp. The most needed > feature is managing the fonts; thats 1 sql table for where the fonts > are on disk, and their metadata. Then another feature for font > authors/editors, with an OpenID backend so its just tracking who is an > editor for what files. The web design and the important system > architecture can be reused from Ben Weiner's work on ccHost. Ed is > using Drupal for other projects. > > Dave thought that whatever it is, it must be RESTful for integration > with FF and so on. The feature he would think of when looking at > suitability is, how easy it will be to implement a user viewing a > font, downloading it, modifying on their computer, and then uploading > v1.1 with some glyphs changed and some new ones added, as a fork of > the original. ccHost supports this, albeit not very well. And then, > crucially, the feature would be extended to support the original's > author to merge these patches into their trunk. This could be with > Aiki or anything custom with a DVCS file store - web.py and django > were my immediate thoughts. > > Ed is very happy with his latest "Font Playground" program, which is a > jQuery plugin. He recently met a Nigerian guy in Michigan who has > difficulty with the keyboard input for the Nigerian language (a latin > variant). With the jQuery framework it is easy for Ed to implement > this layout. > > The next feature Ed plans to add to the playground is using browser > feature detection to transparently switch from PNG font rendering to > @font-face rendering as used in http://www.oep-h.com/LOF/ (also qith > jQuery!). I am happy to support this. > > Dave suggested the end result of Open Font Library ought to be an
