On Tue, Feb 6, 2024 at 10:53 AM Taher Alkhateeb <ta...@pythys.com.invalid>
wrote:
>
> When you have two code bases that share a common ancestry, then the two
> different lines of development are called forks,
Well, no. That's just not true. Do people say that JDK 7 and JDK 8 are
forks? No. The latter is not a fork, but just a continuation of development
on the same project.
I already explained the history of this. The FreeMarker 3 that I am working
on is the continuation of development on the trunk ("master" or "main" in
Git-speak) of the code from when the project was hosted on Sourceforge. A
continuation of development by the person who is, to all intents and
purposes, the original author. I daresay that nobody who understood the
situation would call this a "fork".
> it's not a bad thing and if you check github you'll notice that it's kind
> of a popular button over there.
Well, it's true that in git, the term "fork" does not have a negative
connotation, but I was pretty certain that the way you were using the term,
you did mean it negatively. That view is reinforced by the hostile tone of
your message.
In any case, if anything is a "fork", it is "Apache FreeMarker" because
Daniel chose to fork off from the main stream of development by taking an
older obsolete version of the code base, the FreeMarker 2.3 maintenance
branch and using that as the basis for "Apache FreeMarker". Once I resumed
my FreeCC work, which used the more advanced version of the code (the trunk
in the code repository), yes, there was effectively a fork, as in a
bifurcation, but the record is clear. The person who forked, i.e. caused a
bifurcation, was Daniel. My work is simply a continuation of work on the
main stream of development.
The history of all this is kind of convoluted in a way, but it's not really
that hard to understand either -- unless you very much don't want to
understand, which I suspect is your case...
> Sorry but I will refrain from nothing, especially when it's just _your_
> interpretation or mind-reading attempt.
>
Well, I do not think that I am imagining the hostility in your tone.
Anyway, I would re-iterate that you really ought to refrain from referring
to my work as a "fork" because it is not, and I already explained this. If
somebody was misspelling your last name, let's say with one 'e' instead of
two, and you corrected him, and he kept misspelling your name regardless,
would this not be some kind of passive aggression?
>
> I don't know you nor do I know Daniel beyond just interactions in here,
> but at this point and after everything I read, I don't care if your code
> quality is 10,000 times better. I just don't want to deal with you
> regardless of who you are or what your code is like.
Well, the "Apache FreeMarker" code was also largely written by me. I mean,
certainly the core parser/renderer part which is what FreeMarker mostly is.
"Apache FreeMarker" also includes the BeansWrapper written entirely by
Attila Szegedi, which was 12000 lines of rather grotesquely over-engineered
code. I ended up rewriting all of that in about 400 lines. So, as you could
imagine, it is a lot easier to work with!
But the problem with what you're saying is that if you get in there and
work on the "Apache FreeMarker" code, you're largely just working with an
older, obsolete version of the code by same author. Me!
But anyway, there's not much point in announcing loudly to the world that
you don't want to collaborate with me. Just don't collaborate with me. I
don't know what the point of this is. It's like you're trying to "virtue
signal" or something. Somehow making a show of this hostility towards me is
somehow virtuous or something... (SMH).
Check out new FreeMarker 3 features:
https://github.com/freemarker/freemarker3/wiki
Jonathan Revusky
>
> On Monday, February 05, 2024 15:44 +03, Jonathan Revusky <
> revu...@gmail.com> wrote:
> (Sigh.)
>
> Well, first of all, your characterization of the overall situation is
> pretty dubious. For one thing, you refer to *my* work as a "fork", which is
> quite loaded language, since a "fork" is usually taken to be a bad thing.
> But really, this is a very tenuous concept in this context. A "fork" is
> short for a bifurcation of effort, no? That would mean that I'm doing
> something and you guys are doing something, right? Except that is hardly a
> correct characterization. The cold, hard truth is that, over the last year
> in particular, I have been working on the thing (I mean in a fundamental,
> meaningful way, making significant, even revolutionary changes, to the code
> base, it's much better structured now) and this community has done
> basically NOTHING. That, my friends, is not a "bifurcation of effort". So I
> would suggest that you refrain from referring to my work as a "fork"
> because it isn't really, and to keep using this loaded language would be
> something of a provocation frankly.
>
> Now look, the very first thing anybody needs to understand about the
> situation is this: the only reason that there is such a thing as "Apache
> FreeMarker" is that I made a pretty massive donation of my work roughly 10
> years ago. At that time, I did not anticipate ever doing anything in
> FreeMarker again and it seemed like the existing community was enthusiastic
> about going to ASF -- and, more importantly, since I had not done anything
> in the project for about 5 years, I did not think I really had any right to
> block the move to ASF, even though I myself did not like the idea at all.
> Besides, at that point, one could not know what the results of all this
> would be. Maybe the move to ASF would give FreeMarker development a shot in
> the arm and wonderful things were going to happen. How do you know if you
> don't try it? But, obviously, that is most certainly not what happened.
>
> But anyway, what you are asking me, when you ask me why I don't try to
> merge my ongoing work with "Apache FreeMarker" is why, after making the
> substantial code donation (to which "Apache FreeMarker" owes its existence)
> I decline to even try to donate any further work. Well, why should I? If I
> feel (and I do!) that the initial code donation was a mistake on my part,
> wouldn't it stand to reason that I am not interested in repeating the
> mistake? I mean, let's be honest here (or try...). If you were in my shoes,
> would you want to donate any more work to ASF? Already, when one is treated
> with an incredible lack of graciousness after donating this much work, why
> would you ever donate any work again? Does that make any sense? Just think
> about that...
>
> And again, I see no reason to make any bones about the fact that I consider
> the code donation I made back then to have been a terrible mistake. And,
> really, the results are pretty much a fiasco. There has been some work done
> on the thing, but I reckon that what has happened in ten years is quite a
> bit less than a single motivated person such as myself would do in a single
> month. The project was already pretty dead when it came to Apache, comatose
> at least, but now it's a full-blown nothingburger project. (I explain the
> "nothingburger" concept as best I can here:
> https://wiki.parsers.org/doku.php?id=nothingburger )
>
> But, look, there is another basic point to make about all of this. In this
> life, if you want to work on something with other people, you have to find
> people who share that common intersest. For example, if you're a guitarist
> or drummer and you want to be part of a musical group, you need to find
> other people who also (like you, presumably) genuinely want to make music!
> The same if you want to get involved in some local theater, or whatever.
> And this is no different in principle. If you do want to collaborate on a
> software development project, you need to find the right people who also
> want to do that. But regardless, if you really want to do something, it's
> just masochistic to try to get involved with people who don't really want
> to do anything. Why waste one's time?
>
> So this is a juncture where a person really has to be honest with himself,
> no? Do you really want to get involved in FreeMarker development? Or is
> this just some kind of weird posturing? Because if you do want to get
> involved, obviously you should want to get involved with people who
> actually want to do something. (Isn't that just common sense finally?) In
> terms of reviving a "nothingburger" project, which is what "Apache
> FreeMarker" is, granted, it's not entirely impossible, I suppose, but the
> prognosis is really very poor. And this basic problem, that "Apache
> FreeMarker" is a classic nothingburger, that's not something that can be
> laid at my doorstep.
>
> Basically, you have a choice between working on an earlier version of my
> work without me -- or the latest version, with the cleanest, best
> structured codebase... WITH my collaboration, the involvement of the
> original author. Based on my own values, it would be a very easy decision.
>
> Jon Revusky
>
>
>
>
>
> On Mon, Feb 5, 2024 at 7:22 AM Taher Alkhateeb <ta...@pythys.com.invalid>
> wrote:
>
> >
> > Hello Jonathan,
> >
> > Why yes if you recall I actually replied to you in that thread, and I was
> > asking you why not join hands in here instead of maintaining a fork and
> > confusing everyone as to what's going on not to mention fragmenting an
> > already small community?
> >
> > Regards,
> >
> > Taher Alkhateeb
> >
> > On Sunday, February 04, 2024 23:27 +03, Jonathan Revusky <
> > revu...@gmail.com> wrote:
> > Hi Taher (and everyone else).
> >
> > A couple of months ago, I announced the availability of a more advanced
> > FreeMarker 3 version here: https://github.com/freemarker/freemarker3
> >
> > Really, the bottom line is that if you do want to get involved in hacking
> > the FreeMarker code, this is the one you should get involved in. This is
> a
> > continuation of work by the original author (ME) and if you get in there
> > and have whatever questions about how the code works, you have the
> > collaboration of the original author (ME).
> >
> > If you work on Apache FreeMarker 2.x or 3.x you're working on a much more
> > primitive, older version of the code. For one thing, the FreeMarker 3
> that
> > I point to is rewritten to use a much more powerful parser generator,
> which
> > is CongoCC. And this really has allowed quite a streamlining of the code.
> > Just look at what the CongoCC grammar looks like:
> > https://github.com/freemarker/freemarker3/tree/master/src/parser And
> > compare that with what the legacy JavaCC grammar looks like for Apache
> > FreeMarker:
> >
> >
> https://github.com/apache/freemarker/blob/2.3-gae/freemarker-core/src/main/javacc/freemarker/core/FTL.jj
> >
> > Just eyeball the two and think about which one you would rather work
> with!
> > I can be quite objective because I am basically the author of both
> > versions!
> >
> > On Sat, Feb 3, 2024 at 9:20 AM Taher Alkhateeb <ta...@pythys.com.invalid
> >
> > wrote:
> >
> > > Hello, we were just having a discussion about this:
> > >
> > > https://lists.apache.org/thread/2p3521br9jnp9ww1f5vf80l90fntmfdf
> > >
> > > Essentially the way I understood it, it's better to focus on 2 and get
> > > things done as 3's future is not very clear and requires a lot of work
> > from
> > > developers intimate with the code base.
> > >
> >
> > Look, the real truth of the matter is that working with either Apache
> > FreeMarker 2 codebase or the 3, it's just an exercise in necrophilia.
> > Nothing meaningful has been done for ages and, at this point, there is
> just
> > about no prospect of anything happening. By all means, you could get in
> > there and try to clean it all up and so on, but frankly, your prospects
> of
> > ever catching up to the state of the FreeMarker 3 that I have pointed
> to...
> > it's quite bleak really.
> >
> > I mean, really, c'mon, even just reading between the lines in Daniel's
> > response to this question about FreeMarker development, you can get the
> > feeling that it's really just a waste of time. The thing is dead and
> Daniel
> > is not hardly even trying to hide this.
> >
> > But anyway, 'nuff said. I just would tell you to do your due diligence
> and
> > figure out which way is up! I would be delighted to have collaborators,
> and
> > you would be collaborating with the person who is, to all intents and
> > purposes, the original author of the tool.
> >
> > It really ought to be a very easy decision.
> >
> > Best Regards,
> >
> > Jonathan Revusky
> >
> >
> > >
> > > On February 3, 2024 10:51:15 AM GMT+03:00, Alon Ziv
> > > <nola...@google.com.INVALID> wrote:
> > > >Specifically - is there anything contributors can help with to get
> this
> > > >completed?
> > >
> >
>
